In the sound card market a real battle has been unleashed, what is at stake is to impose a "definitive" professional standard; often, however, the weapons used end up by slightly differing: the resolution and sampling frequency, the format of the inputs and outputs, the presence or absence of preamplifiers, the methods of connection to the computer's “mother” card. How to orient yourself in this chaos of brands and numbers? In this issue I wanted to focus on two leading houses in the sound card sector: MOTU and RME.
Prologue
As mentioned in the foreword, this comparison test involves two leading boards in the professional sector. Having been the owner of one for more than a year RME Fireface 800, I preferred to make this article looking for a comparison with another professional sound engineer, user of a Motu 896 HD; In short, I sought the greatest possible objectivity in this complicated and demanding task. When we talk about the “top” of audio interfaces, the choice is often determined by the “name” of the product and by one's own sound taste rather than by the actual possibilities of using the functions offered by the card. I want to clarify that the two cards have been tested for about five months in a professional recording studio, where everything is recorded, from the pop group, to the string quartet, from the 30-piece gospel choir to the singer-songwriter, always looking for sophisticated sounds. . A hard test to testify their stability and reliability over time. Before writing this article I read various English and American magazines to decide how to set up the comparison, which points to highlight and which ones to leave out as useless and dispersive. I realized how difficult it is today for the end user to choose one product over another; all cards often appear perfect, the only drawback may be some small, perhaps deliberate, forgetfulness in software or hardware design, justified by the good Q / P ratio. "Poor", therefore, the buyers who, before taking the fateful step, must rely on magazines, forums or the advice of friends. I understand them, I am one of them too, perhaps a little luckier as today I can try some products before buying them, but know that I am not fasting on the classics I was prey when I was more inexperienced. Unfortunately, there are no shops that allow the simultaneous testing of multiple cards in the same environment and with the same monitoring boxes. I will give a hint on the components and on the design used for the realization of these two products, as my intent is not to push the reader to buy one model over another, but to bring pros and cons of individual products and the intended use estimated by the respective manufacturers.
Audio chain
I used two computers for this test, a MAC G4 with 876 mHz / 1.5 giga ram processor, with Audio logic vers. 6.0 and one Pc MusicDaw 2, (Intel 775 3,4 Ghz processor and 2 Mb cache) with XP and Cubase SX3. For listening, I used both my TBS2-A PMCs and Dynaudio BM6A and Yamaha NS10M that were already in the studio, so that both the second sound engineer and I had a reference point for listening that we already knew well. The mixer is a Soundcraft 3200 and as a recorder (with I / O Adat) we used an Alesis HD 24, instead for the Tascam “TDIF” format we got a Tascam. Microphones: Neuman U 87, Neuman KM 184, Akg C414, Akg C451, Audio-Technica AT4041, Rode NT2, a “Spectralab” spectrum analyzer from Sound Technology and a lot of patience.
RME Fireface 800
The RME Fireface 800 (Fig.1) is a FireWire sound card, “made in Germany”, a professional product that allows you to manage up to 28 I / O simultaneously. The card has been studied down to the smallest details and is the result of two years of development. It integrates the best of modern technology, the integrated 32-bit RISC processor, the use of latest generation reprogrammable logic components (FPGA, 90 nanometer Field Programmable Gate Arrays), the latest generation communication protocol FireWire at 800 Mb / s , AKM 5385AVF converters supporting up to 24bit 192kHz. The internal technology is all in SMD. But let's go in order ...
How it looks
The Fireface 800 is presented in a cardboard box, together with a photocopied manual in English and German, a pair of Adat cables, a cable firewire 400, 4 rubber pads to put under the card to prevent it from being scratched, a CD with the software and a warranty sheet from the Italian importer. The machine is in standard rack format with one unit in light gray and purple color. From left to right we find: channel 1 (Fig.2) which reproduces a "DI BOX" to allow you to directly record a bass / guitar signal and all those instruments that have a very low output gain. At the top there are three LEDs: DRIVE yellow, LIM (Limiter) red and SIG (signal input control) green, while at the bottom an Instr / Line “Speaker Emulation” input with a Gain control next to it by potentiometer. Moving to the right we find four entrances, from 7 to 10, (Fig.3) with INSERT (LINE) connection on standard jacks in TRS format, or MIC on XLR individually controlled by a potentiometer for the gain of the preamplifiers (GAIN) which have an excursion from + 10 to +80 dB. At the top we find for each of the inputs a LED that indicates the activation of the 48 V (yellow color) the CLIP signal (red color) and the green SIG (signal input control). We note with regret that no “switch” has been provided to activate the phantom power from the RME panel. Activation can only take place via software and it will eventually be possible to save it in the machine's routing settings for "stand-alone" use. Continuing on the right we find an ANALOG / DIGITAL STATE LED control (Fig.4) which is none other than the panel where we can view the status of the input and output signal "LEVEL IN / OUT" and the status of the clock for the various digital sources. The last connection on the front panel is that of the headphones, controlled by a volume potentiometer to be transformed if necessary into in / out 9 and 10 (unbalanced).
Back panel
From left to right we find (Fig.5): connector for power supply, MIDI IN / OUT, a Firewire 400 port and two Firewire 800 ports with a metal hook to hook the Firewire cable, WORD OUT on BNC, 2 digital I / O in ADAT format (total 16 I / O independent), I / OS / PDIF, WORD IN on BNC with 75 ohm impedance selector, at the bottom 8 analog audio outputs in TRS format (Fig.6). The last stage, from left to right starting from the top, we find: Optional time code, while at the bottom there are 8 outputs in TRS format. Already from the description we can understand that the board was designed more for studio use than for "live" shooting, as for live use it would have been appropriate to have at least 8 analog inputs, with the possibility of supporting a "phantom" power supply selectable from the panel and not via software. Furthermore, the level indicator LEDs on the front panel are missing. I believe that these "shortcomings" have been well considered in order not to jeopardize the original design of a "light" system for a rack unit.
First impressions
We install the card using the Firewire 400 port cable of the PC mod. MusicDaw 2, we load the software for the management of the machine and we begin to discover flaws and virtues of this much publicized card. The card's routing management software is very basic and lean. Although the “minimalist design” has very small fonts, functionality is not impaired in any way. The use of the "Fireface Setting" panel is easily understood regardless of the manuals (Fig.7). From here we can decide which inputs to control, analog, Adat, S / PDIF, and the relative frequency from 32 to 192 kHz. We can decide if the XLR inputs must power microphones that require the Phantom, or simple cardiodes; all parameter changes can be performed in real time without ever locking the system. Analog and digital inputs and outputs can be combined simultaneously. Latency can drop to very low values, equal to about one millisecond, but I strongly advise against it, it would only mean making the computer's CPU work harder without having any kind of advantage. The card has ASIO2.0, ASIO Multiclient, MME, WDM, Multiclient, GSFI drivers and is compatible with ADM (ASIO Direct Monitoring). If you use the FireWire 800 connection, in theory you could connect up to a maximum of 3 RMEs for a total of 84 24-bit / 48 KHz I / O.
RME Pressed
The first tests I did were those relating to virtual software, from the heavier ones like "Symphonic Orchestra Platinum" to those compiled "maluccio" like Tritone Digital Hydratone. In none of the cases have I ever had problems of any kind. A few months ago a friend made me try the same card under LINUX with "sequencers" and "plugs" which I did not know existed, to my amazement the card worked at values of one millisecond without the CPU going into TILT. The real test of fire, however, I did it in the studio. Starting from existing material on ALESIS HD24, my friend Rino Morra and I tried to redo a mix; we simultaneously take the 10 analog inputs with the 16 in ADAT format and transfer everything to the computer, finally the result is transferred to an analog mixer, using the analog outputs of the RME and using the ALESIS HD 24 as a D / A converter to transform the outputs optics of the sound card in analog. Not satisfied, we also try to import some tracks from the analog sources of the card and transfer them directly in ADAT format. Card routing allowed us to do this without first having to record audio content to the computer. I wanted to make a clarification on the Fireface inputs: they are 12 (4 XLR and 8 TRS to Jack) but they can work at most 10 at the same time, as 2 inputs are just repetitions. One downside, worthily overshadowed by "dubious" advertising phrases is that the card is advertised as "Firewire 56-channel audio interface - 192 kHz / 24 bit". But where are these 56 channels? In reality it is possible to have a maximum of 28 I / O inputs at 24 bit / 44kHz or 48kHz simultaneously using all the I / O (analog-digital-S / PDIF). If instead we wanted to record at 24 bit 96 kHz we could reach a maximum of 20 recording inputs (10 analog - 8 digital - 2 S / PDIF). Working at 24 bit / 192 kHz we would reach a maximum of 12 recording inputs (10analog - 2 S / PDIF). In this case it is not possible to use the Adat channels. The same test was done on the MAC G4 and although it is not a latest generation computer, we have not had any kind of problem. At this point the question arises: how does the instrument play? We understood that the system is very stable on both Mac and PC and that it can also work under Linux, guaranteeing total audio-video integration. Anyway, what exactly happens when the signal comes in and out? How do the preamps and the “Speaker Emulation” input sound? Section A / D RME The first test I perform concerns input 1 “Speaker Emulation”. In a corner of the studio there is an old Rhodes Mark 2 piano, slightly out of tune and with a few keys not perfectly working. With the advent of Virtual Instruments, the owner of the studio relegated it to a piece of furniture rather than a musical instrument. I persuade him to move him to our post. I don't have the necessary tools to tune it and I honestly don't think I'm capable of it; however I want to hear what happens by plugging it directly into the “Speaker Emulation” input. This connection has the characteristics of a DI box, the input level is higher and gritty than the line inputs. By exasperating the Gain a little, you get a drive effect very similar to that of a real cabinet (Fig.8). The Rhodes recorded in this way took back all the original breath and brilliance. Percussion can be heard perfectly without being overwhelmed by background noise. Long last! With all its "detuning" it is a pleasure to play and record this instrument, to hear it "alive" in its harmonic nuances. The test was also carried out on guitar and bass, always with excellent results, but the most difficult test was the relative one. to the microphone channels.
The PRE section, according to many magazines and forums, is not a strong point of the card; I read that it would have been better to opt for 2 PRE of superior quality compared to the 4 chosen supplied. In this case, I want to break a lance in favor of the Fireface. The PREs are of excellent quality and, as with everything, you have to understand what use they were designed for. If our purpose is to record voices we should not expect optimal results, the converters are too transparent and do not have a strong sound, too little colored. This is not a flaw, indeed in many cases it can be a real strength when it comes to miking a classical guitar, a violin or a piano, since we would be sure that the sound will never be mixed. If we wanted to record a voice instead, (unless your “performer” is Barry White), it would be advisable to add a good more characterizing PRE in order to guarantee a more amalgamated sound. Perhaps the only flaw concerns the potentiometers that regulate the input level (Gain), these do not respond smoothly to the impressed movements: not very sensitive for almost three quarters of their stroke, rapid surges in the residual phase making it really difficult to manage the volume. It is likely that in later versions the problem has been solved with a better calibration, but at the moment I do not have the possibility to verify. The last test I made was to insert an external Firewire accelerator card on the PC, to be precise a TC Electronics. Unfortunately I have had conflict problems which have been confirmed to me in the various forums I usually frequent. If instead we use PCI accelerator cards no problem, whether it is TC or Universal Audio.
RME conclusions
There is still a lot to say about this card, its routing and all the many applications, perhaps I have highlighted more defects than merits, but I assure you that so far I have never found anything absolutely perfect over the years. One of the factors common to all sound cards, even professional ones, is that as the value of the sampling frequency increases, the sound becomes harsher and less natural: this problem does not concern RME, the sound remains very linear even at varying frequencies. The signal in general is very homogeneous and pleasant, well cared for and natural at the extremes of the band. The bass is slightly full and round and acquires that right amount of color without ever being false; the highs are transparent and silky. Even after many hours of work the sound is not tired and its “right” dose of neutrality allows to obtain good results even in mixes crowded with tracks. Of course I would have preferred a "hardware" button for the selection of the phantom, more balanced volume potentiometers, we hope that these defects will be a stimulus for a future two-channel project, specific for use as a Mastering or for small "Home recording" projects. . It would be nice to expect a 2 rack unit with all inputs equipped with phantom enabling, not orphans of their own level indicator, it would be an ideal solution for live or studio use: we would benefit from a good input stage and comfort of a mix all using DAW.
Motu 2408 MK3 as it presents itself
The Motu 2408 MK3 is presented in a cardboard box, together with 2 manuals, PCI-424 and AUDIO DESK vers. 2.0 for MAC in English, a power cable, a PCI connection cable, a CD with the software and a registration card from the American distributor. The machine is in standard rack format with one unit in dark gray color (Fig.9).
From left to right we find: the headphone input with its VOL potentiometer and the MAIN OUT VOL. A DIGITAL I / O control, divided into three BANK A, B and C (Fig.10). Moving further to the left we find the display that manages the volume of the 8 ANALOG IN analog inputs and the one that manages the 8 ANALOG OUT analog outputs. Finally we find a FORMAT CONVERSION button selector and at the bottom a SET button and lastly the classic ON / OFF switch.
Back panel
From left to right we find a fairly crowded panel (Fig.11): the connector for power supply, WORD / VIDEO IN / OUT, SMPTE (LTC), S / PDIF, 3 ports for signals in TDIF format (Tascam), a main out, OUT to connect the card to the PCI-424, 3 optical I / O in Adat format, for a total of other 24 I / O, 8 ANALOG OUT and 8 ANALOG IN in TRS format and finally MAIN OUT, right and left.
Finally the real heart of this card is the PCI 424, (Fig.12) which houses the DSP that manages all the routing of inputs and outputs, relieving the computer CPU and making the system more reliable and stable, with latency close to zero.
First impressions of the MOTU
I plugged the Motu PCI controller into the motherboard slot, booted the computer, and Windows XP promptly detected the presence of new hardware. I inserted the required disk drive and started the installation. The procedure was successful without any hitch. After rebooting the system I connected the audio interface to the relative PCI card, set the "drivers" and launched Cubase SX 3 which immediately recognized the card and all its functions. The control panel is similar to that of the model 896 HD that I mounted on the Mac in order to perceive the differences in sound (Fig.13). As this is a card designed more as a controller and system heart rather than as a support for digital inputs and outputs, I ran into an initial difficulty for my audio tests; in this case too, however, ingenuity won, in fact, in addition to the expected tests I was able to invent new interesting tests, even if not standard.
Motu 2408 Under Press
The first test I made was to put a Focus ISA 8 MK430 in one of the 2 analog inputs of the Motu, not so much to test the quality of the recording sound, but to test if the analog input signal of the card went into saturation. I inserted the signal of a bass slap directly into the Focus, and tried to intelligently dose the PRE input and output; the card perfectly handled the fast transient changes without ever going into digital distortion. This demonstrates the great quality of the converters and instrument design. I tried to record and play 8 audio signals using the analog inputs and outputs. The result was excellent in both sound and system stability. The microphone test (Voice) also went well. For this test I used a Neumann U 87 microphone powered by the mixer Soundcraft 3200, a nice analog match: no problem using both the MusicDaw 2 and the G4, which can be defined as less powerful in terms of processor and RAM. One of the most demanding tests I have carried out in these 5 months is to stress the system with a huge amount of data. The studio where I carried out the test was commissioned to record a classical music concert made in a theater: for the occasion all 24 tracks of the Alesis HD 24 recorder were occupied (Fig.14). Subsequently, for reasons of “editing” of the single audio tracks, the need arose to import the entire recording onto the computer. The entire concert was poured into a single passage using the 24 optical outputs of the Alesis HD 24 and setting the Motu to the 24 Adat (optical) inputs, all synchronizing the recorder and the interface via ADAT SYNC. The recording (24 bit / 48kHz) lasted 70 minutes and during this time the card did not present any kind of problems. The signal never went off hook and there were no drops in the recording. The problems started when, due to business needs, it was necessary to transfer the work from the Adat to the Tascam. Not having other ADAT cables available, I connected the HD24 outputs to ADAT on the board and from the 2408 MK3, using TDIF cables, I went to the TASCAM MX2448 multi-track. Bitter surprise! If you set the card to receive 24 signals in “optical adat” format, real-time playback can only take place via ADAT. Unfortunately, the fact of not being able to transfer from one format to another in real time is a major limitation. I want to list a strong point that instead I found only in this tab: while the play scrolls, if you enable only one track in recording, the transition takes place in a lightning-fast way, without a minimum of latency, just as if you were working with a Alesis HD 24 type recorder or a classic Fostex D 160. The same performance is obtained by recording up to four tracks simultaneously. From five tracks onwards the delay begins to be felt, even if minimal. All this is possible both because the management software is very well written, and we had no doubts about this, and because, as far as we can tell, PCI cards always work better than Firewire, receive more power and are more stable. The last test I made was to insert an external accelerator card on the PC, to be precise TC Electronics: no conflicts.
Motu conclusions
Even on this card there is still a lot to say, starting with the enormous power of his CueMix DSP, to all possible applications with the audio and video world.
It belongs to a specific professional sector. It does not want to be addressed to all the small household “Home studios” but to an audience of professionals who adopt hardware supports for data transfer. Think of the advantages for those who record live and who still find themselves using "workstations" or recorders such as Tascam MX 2424, or the old Alesis Adat, etc. Some will turn up their noses calling them technologically "outdated" products, especially after the advent of computers, but nothing can be said about their reliability. I have to say that even though I personally own a RME Fireface 800 and am used to a slightly warmer sound, I don't mind the aroma of this card. In my opinion,
what differentiates the Motu house from many others is the harshness, which does not make the high frequencies shine or emphasize the low ones. From the converters, said in a very simple language, “only what comes in comes out”, this allows in the mixing phase a great workability of the sound without having to calculate and balance the intrinsic characteristics of the converters. We can add in this regard that compared to the little sister 896 HD, the 2408 MK3 is slightly more colorful in the high frequencies (it shines a little more).
General consideration
The two cards both sound good and the incoming signal is recorded in a very balanced way. Each one has a little emphasis in a part of the sound spectrum, however we talk about small things; I am sure that if they had to make me listen to one of the two cards in another environment and blindfolded, I would not be able to distinguish them. They are both pleasant to listen to and in the long run do not tire the ear of the engineer on duty. Certainly with such a neutral and well balanced signal, it will be very difficult to miss a mix in either model. A common problem with the two cards is the lack of analog inserts. Both in a home studio and in a professional one it would be interesting to use a tube compressor to be combined with the internal microphone preamps, unfortunately the two manufacturers wanted to deny us this small but very effective "gem". If I had to add some consideration I would say that the RME is more versatile and more comfortable due to its portability, the Motu, on the other hand, which adopts a PCI bus, denies this convenience but provides great value performances. After being an impartial judge, I defer to your chosen sums which will surely be dictated by the future use of the product and the expense that you will be willing to face.
In the sound card market a real battle has been unleashed, what is at stake is to impose a "definitive" professional standard; often, however, the weapons used end up by slightly differing: the resolution and sampling frequency, the format of the inputs and outputs, the presence or absence of preamplifiers, the methods of connection to the computer's “mother” card. How to orient yourself in this chaos of brands and numbers? In this issue I wanted to focus on two leading houses in the sound card sector: MOTU and RME.
Prologue:
As mentioned in the introduction, this comparison test involves two leading professional audio interfaces. Having owned an RME Fireface 800 for over a year, which I used for all of CM2 magazine's tests, I chose to conduct this article by comparing it with another professional sound engineer who uses a MOTU 896 HD. In short, I sought to be as objective as possible in this complicated and demanding task. When it comes to top-of-the-line audio interfaces, the choice is often determined by the product's name and one's personal sonic taste rather than by the actual usability of the interface's features. I should point out that both interfaces were tested for approximately five months in a professional recording studio, where everything from pop groups and string quartets to 30-piece gospel choirs and singer-songwriters are recorded, always seeking sophisticated sounds. This was a tough test to demonstrate their stability and reliability over time. Before writing this article, I read various English and American magazines to decide how to frame the comparison, which points to highlight and which to leave out as unnecessary and dispersive. I realized how difficult it is for the end user today to choose one product over another; all the cards often appear perfect, the only flaw being some small, perhaps deliberate, oversight in the software or hardware design, justified by the good Q/P ratio. Poor, then, those buyers who, before taking the fateful step, have to rely on magazines, forums, or the advice of friends. I understand them—I'm one of them, too—perhaps a little luckier in that these days I can try some products before buying them, but know that I'm not unfamiliar with the classics I fell prey to when I was more inexperienced. Unfortunately, there are no stores that allow you to test multiple cards simultaneously in the same environment and with the same monitoring speakers. I will give a brief overview of the components and design used to make these two products, as my aim is not to push the reader to purchase one model over another, but to report, as clearly as possible, the pros and cons of the individual products and the intended use envisaged by the respective manufacturers.
Audio chain:
For this test I used two computers, a MAC G4 with an 876 MHz processor/1.5 GB of RAM, with Logic Audio version 6.0 and a MusicDaw 2 PC (Intel 775 processor at 3,4 GHz and 2 MB cache) with XP and Cubase SX3. For listening, I used both my PMC TBS2-A, the Dynaudio BM6A and the Yamaha NS10M that were already in the studio, so that both I and the second sound engineer had a listening reference point that we already knew well. The mixer is a Soundcraft 3200 and as a recorder (with Adat I/O) we used an Alesis HD 24, while for the Tascam “TDIF” format we got a Tascam. Microphones: Neuman U 87, Neuman KM 184, Akg C414, Akg C451, Audio-Technica AT4041, Rode NT2, a “Spectralab” spectrum analyzer from Sound Technology and a lot of patience.
RME Fireface 800
The RME Fireface 800 (Fig.1) is a FireWire sound card, “made in Germany”, a professional product that allows you to manage up to 28 I / O simultaneously. The card has been studied down to the smallest details and is the result of two years of development. It integrates the best of modern technology, the integrated 32-bit RISC processor, the use of latest generation reprogrammable logic components (FPGA, 90 nanometer Field Programmable Gate Arrays), the latest generation communication protocol FireWire at 800 Mb / s , AKM 5385AVF converters supporting up to 24bit 192kHz. The internal technology is all in SMD. But let's go in order ...
What is presented:
The Fireface 800 comes in a cardboard box, along with a photocopied manual in English and German, a pair of Adat cables, a FireWire 400 cable, four rubber pads to place under the card to prevent it from being scratched, a software CD, and a warranty sheet from the Italian importer. The machine is a standard 1U rack unit, and is light gray and purple in color. From left to right, we find: channel 4 (Fig.2) which reproduces a "DI BOX" to allow you to directly record a bass / guitar signal and all those instruments that have a very low output gain. At the top there are three LEDs: DRIVE yellow, LIM (Limiter) red and SIG (signal input control) green, while at the bottom an Instr / Line “Speaker Emulation” input with a Gain control next to it by potentiometer. Moving to the right we find four entrances, from 7 to 10, (Fig.3) with INSERT (LINE) connection on standard jacks in TRS format, or MIC on XLR individually controlled by a potentiometer for the gain of the preamplifiers (GAIN) which have an excursion from + 10 to +80 dB. At the top we find for each of the inputs a LED that indicates the activation of the 48 V (yellow color) the CLIP signal (red color) and the green SIG (signal input control). We note with regret that no “switch” has been provided to activate the phantom power from the RME panel. Activation can only take place via software and it will eventually be possible to save it in the machine's routing settings for "stand-alone" use. Continuing on the right we find an ANALOG / DIGITAL STATE LED control (Fig.4) which is none other than the panel where we can view the status of the input and output signal "LEVEL IN / OUT" and the status of the clock for the various digital sources. The last connection on the front panel is that of the headphones, controlled by a volume potentiometer to be transformed if necessary into in / out 9 and 10 (unbalanced).
Rear Panel:
From left to right we find (Fig.5): connector for power supply, MIDI IN / OUT, a Firewire 400 port and two Firewire 800 ports with a metal hook to hook the Firewire cable, WORD OUT on BNC, 2 digital I / O in ADAT format (total 16 I / O independent), I / OS / PDIF, WORD IN on BNC with 75 ohm impedance selector, at the bottom 8 analog audio outputs in TRS format (Fig.6). The last stage, from left to right starting from the top, we find: Optional time code, while at the bottom there are 8 outputs in TRS format. Already from the description we can understand that the board was designed more for studio use than for "live" shooting, as for live use it would have been appropriate to have at least 8 analog inputs, with the possibility of supporting a "phantom" power supply selectable from the panel and not via software. Furthermore, the level indicator LEDs on the front panel are missing. I believe that these "shortcomings" have been well considered in order not to jeopardize the original design of a "light" system for a rack unit.
First impressions:
We install the card using the Firewire 400 port cable of the PC mod. MusicDaw 2, we load the software for the management of the machine and we begin to discover flaws and virtues of this much publicized card. The card's routing management software is very basic and lean. Although the “minimalist design” has very small fonts, functionality is not impaired in any way. The use of the "Fireface Setting" panel is easily understood regardless of the manuals (Fig.7). From here we can decide which inputs to control, analog, Adat, S / PDIF, and the relative frequency from 32 to 192 kHz. We can decide if the XLR inputs must power microphones that require the Phantom, or simple cardiodes; all parameter changes can be performed in real time without ever locking the system. Analog and digital inputs and outputs can be combined simultaneously. Latency can drop to very low values, equal to about one millisecond, but I strongly advise against it, it would only mean making the computer's CPU work harder without having any kind of advantage. The card has ASIO2.0, ASIO Multiclient, MME, WDM, Multiclient, GSFI drivers and is compatible with ADM (ASIO Direct Monitoring). If you use the FireWire 800 connection, in theory you could connect up to a maximum of 3 RMEs for a total of 84 24-bit / 48 KHz I / O.
RME Under the Press:
The first tests I did were those relating to virtual software, from the heaviest ones like “Symphonic Orchestra Platinum” to the “badly” compiled ones like Tritone Digital Hydratone. In none of the cases have I ever had any problems of any kind. A few months ago a friend let me try the same card under LINUX with “sequencers” and “plugs” that I didn't know existed. To my amazement, the card ran at millisecond values without the CPU going into TILT. But the real test of fire was in the studio. Starting from material already existing on the ALESIS HD24, my friend Rino Morra and I attempted to redo a mix; we simultaneously took the 10 analog inputs with the 16 in ADAT format and transferred everything to the computer. Finally, the result was transferred to an analog mixer, using the RME's analog outputs and using the ALESIS HD 24 as a D/A converter to transform the optical outputs of the sound card into analog. Not content with that, we also try to import some tracks from the card's analog sources and transfer them directly into ADAT format. The card's routing allowed us to do this without having to first record the audio content into the computer. I wanted to make a clarification about the Fireface inputs: there are 12 of them (4 XLR and 8 TRS Jack) but a maximum of 10 can function simultaneously, as 2 inputs are just repetitions. One downside, which is worthily obscured by “equivocal” advertising phrases, is that the card is advertised as a “56-channel Firewire audio interface – 192 kHz/24 bit”. But where are these 56 channels? In reality, it is possible to have a maximum of 28 I/O inputs at 24 bit / 44kHz or 48 kHz simultaneously using all the I/Os (analog-digital-S/PDIF). If instead we wanted to record at 24 bit 96 kHz we could reach a maximum of 20 recording inputs (10 analog – 8 digital – 2 S/PDIF). Working at 24 bit/192 kHz we would reach a maximum of 12 recording inputs (10 analog – 2 S/PDIF). In this case it is not possible to use Adat channels. The same test was done on the MAC G4 and although it is not a latest generation computer, we did not have any kind of problem. At this point the question arises spontaneously: how does the instrument sound? We found that the system is extremely stable on both Mac and PC and that it can also work under Linux, ensuring complete audio-video integration. Anyway, what exactly happens when the signal goes in and out? How do the preamps and “Speaker Emulation” input sound? RME A/D Section The first test I carry out concerns input 1 “Speaker Emulation”. In one corner of the studio is an old Rhodes Mark 2 piano, slightly out of tune and with some keys not working perfectly. With the advent of Virtual Instruments, the studio owner relegated it to a decorative object rather than a musical instrument. I convince him to move it towards our station. I don't have the equipment to tune it, and I honestly don't think I'm capable of it; however, I'd like to hear what happens when I plug it directly into the Speaker Emulation input. This connection has the characteristics of a DI box, the input level is higher and more punchy than the line inputs. By increasing the Gain a little you get a drive effect very similar to that of a real cabinet (Fig.8). The Rhodes recorded in this way took back all the original breath and brilliance. Percussion can be heard perfectly without being overwhelmed by background noise. Long last! With all its "detuning" it is a pleasure to play and record this instrument, to hear it "alive" in its harmonic nuances. The test was also carried out on guitar and bass, always with excellent results, but the most difficult test was the relative one. to the microphone channels.
The preamp section, according to many magazines and forums, is not a strong point of the card; I've read that it would have been better to opt for two higher-quality preamps rather than the four supplied. In this case, I want to break a lance in favor of the Fireface. The preamps are of excellent quality and, as with everything, you need to understand what they were designed for. If our goal is to record vocals, we shouldn't expect optimal results; the converters are too transparent and don't have a clear sound, too little color. This isn't a flaw; in fact, in many cases, it can be a real strength when it comes to mic'ing a classical guitar, violin, or piano, as we can be sure that the sound will never sound muddy. If we wanted to record vocals, however (unless your "performer" is Barry White), it would be advisable to add a good preamp with more character to ensure a more balanced sound. Perhaps the only flaw concerns the potentiometers that adjust the input level (Gain), which don't respond smoothly.
to the movements they impart: insensitive for almost three-quarters of their travel, rapid surges in the remaining phase, making it really difficult to manage the volume. It's likely that in later versions the problem was solved with better calibration, but I don't have the opportunity to verify this at the moment. The last test I performed was to insert an external FireWire accelerator card into the PC, to be precise, a TC Electronics one. Unfortunately, I had conflict problems that were confirmed to me in the various forums I regularly frequent. If we use PCI accelerator cards, however, no problem, whether they are TC or Universal Audio.
RME CONCLUSIONS:
There's much more to say about this card, its routing, and all its many applications. Perhaps I've highlighted more flaws than strengths, but I assure you that over the years I've never found anything absolutely perfect. One of the factors common to all sound cards, even professional ones, is that as the frequency increases,
As the sampling rate increases, the sound becomes increasingly harsh and less natural: this problem doesn't affect the RME, the sound remains very linear even when varying frequencies. The overall signal is very homogeneous and pleasant, well-defined and natural at the frequency extremes. The bass is slightly full and rounded and acquires just the right amount of color without ever sounding false; the highs are transparent and silky. Even after many hours of work, the sound doesn't tire, and its "just right" amount of neutrality allows for good results even in crowded mixes. Of course, I would have preferred a "hardware" button for selecting phantom power and more balanced volume pots. Let's hope these small flaws inspire a future two-channel project, specifically for mastering use or small home recording projects. It would be nice to expect a 2 rack unit with all inputs equipped with phantom power, not without their own level indicator, it would be an ideal solution for live or studio use: we would benefit from a good input stage and the convenience of mixing entirely via DAW.
Motu 2408 MK3 what it looks like:
The Motu 2408 MK3 is presented in a cardboard box, together with 2 manuals, PCI-424 and AUDIO DESK vers. 2.0 for MAC in English, a power cable, a PCI connection cable, a CD with the software and a registration card from the American distributor. The machine is in standard rack format with one unit in dark gray color (Fig.9). From left to right we find: the headphone input with its VOL potentiometer and the MAIN OUT VOL potentiometer. A DIGITAL I/O control, divided into three BANKS A, B and C (Fig.10). Moving further to the left we find the display that manages the volume of the 8 ANALOG IN analog inputs and the one that manages the 8 ANALOG OUT analog outputs. Finally we find a FORMAT CONVERSION button selector and at the bottom a SET button and lastly the classic ON / OFF switch.
Rear Panel:
From left to right we find a fairly crowded panel (Fig.11): the power connector, WORD/VIDEO IN/OUT, SMPTE (LTC), S/PDIF, 3 ports for signals in TDIF format (Tascam), a main out, OUT to connect the card to the PCI-424, 3 optical I/O in Adat format, for a total of 24 other I/O, 8 ANALOG OUT and 8 ANALOG IN in TRS format and finally MAIN OUT, right and left. Finally the real heart of this card is the PCI 424, (Fig.12) which houses the DSP that manages all the routing of inputs and outputs, relieving the computer CPU and making the system more reliable and stable, with latency close to zero.
First impressions of MOTU:
I plugged the Motu PCI controller into the motherboard slot, booted the computer, and Windows XP promptly detected the presence of new hardware. I inserted the required disk drive and started the installation. The procedure was successful without any hitch. After rebooting the system I connected the audio interface to the relative PCI card, set the "drivers" and launched Cubase SX 3 which immediately recognized the card and all its functions. The control panel is similar to that of the model
896 HD that I mounted on the Mac in order to perceive the differences in sound (Fig.13). As this is a card designed more as a controller and system heart rather than as a support for digital inputs and outputs, I ran into an initial difficulty for my audio tests; in this case too, however, ingenuity won, in fact, in addition to the expected tests I was able to invent new interesting tests, even if not standard.
Motu 2408 Under Press:
The first test I carried out was to plug a Focus ISA 430 MK2 into one of the 8 analog inputs of the Motu, not so much to test the sound quality during recording, but to test whether the analog input signal of the card would saturate. I directly inserted the signal of a slap bass into the Focus, and tried to dose it in a way
The PRE's input and output were intelligent; the card handled fast transient changes perfectly without ever going into digital distortion. This demonstrates the high quality of the converters and the instrument's design. I tried recording and playing back 8 audio signals using the analog inputs and outputs. The result was excellent both in terms of sound and system stability. The microphone test (voice) also went well. For this test I used a Neumann U 87 microphone powered by the
Soundcraft 3200 mixer, a nice analog pairing: no problems using either the MusicDaw 2 or the G4, which can be defined as less powerful in terms of processor and RAM. One of the most demanding tests I've carried out in these 5 months is to stress the system with a huge amount of data. The studio where I carried out the test was commissioned to record a classical music concert held in a theater: for the occasion, all 24 tracks of the Alesis HD 24 recorder were used (Fig.14). Subsequently, for "editing" reasons of the individual audio tracks, the need arose to import the entire recording into the computer. The entire concert was transferred in a single pass using the 24 optical outputs of the Alesis HD 24 and setting the Motu to the 24 Adat (optical) inputs, all synchronizing the recorder and the interface via ADAT SYNC. The recording (24 bit/48kHz) lasted 70 minutes and during this time the card did not present any problems of any kind. The signal never
unhooked and there were no drops in the recording. The problems began when, for work reasons, it was necessary to transfer the work from the Adat to the Tascam. No
Having some other ADAT cables available, I connected the HD24 outputs to ADAT on the card and, from the 2408 MK3, I connected them to the TASCAM MX2448 multitrack player using TDIF cables. What a surprise! If you set the card to receive 24 signals in "optical adat" format, real-time playback can only occur via ADAT. Unfortunately, the inability to transfer between formats in real time is a major limitation. I'd like to list a strength that I've only found with this card:
While playback is running, if you enable recording for just one track, the transition happens lightning-fast, with no latency whatsoever, just like working with a recorder like the Alesis HD 24 or a classic Fostex D 160. The same performance is achieved by recording up to four tracks simultaneously. From five tracks onwards, the delay begins to be noticeable, even if minimal. All this is possible both because the management software is very well written, and we had no doubts about this, and because, as far as I can tell, PCI cards always work better than Firewire ones, receive more power, and are more stable. The last test I carried out was
that of inserting an external accelerator card into the PC, to be precise TC Electronics: no kind of conflict.
Conclusions Motu:
Even on this card there is still a lot to say, starting with the enormous power of his CueMix DSP, to all possible applications with the audio and video world.
It belongs to a specific professional sector. It does not want to be addressed to all the small household “Home studios” but to an audience of professionals who adopt hardware supports for data transfer. Think of the advantages for those who record live and who still find themselves using "workstations" or recorders such as Tascam MX 2424, or the old Alesis Adat, etc. Some will turn up their noses calling them technologically "outdated" products, especially after the advent of computers, but nothing can be said about their reliability. I have to say that even though I personally own a RME Fireface 800 and am used to a slightly warmer sound, I don't mind the aroma of this card. In my opinion,
what differentiates the Motu house from many others is the harshness, which does not make the high frequencies shine or emphasize the low ones. From the converters, said in a very simple language, “only what comes in comes out”, this allows in the mixing phase a great workability of the sound without having to calculate and balance the intrinsic characteristics of the converters. We can add in this regard that compared to the little sister 896 HD, the 2408 MK3 is slightly more colorful in the high frequencies (it shines a little more).
General consideration:
Both cards sound good, and the incoming signal is recorded very balanced. Each has a slight emphasis on one part of the sound spectrum, but these are minor details. I'm sure that if I were to listen to one of the two cards in another room blindfolded, I wouldn't be able to tell them apart. They're both pleasant to listen to, and in the long run, they don't tire the engineer's ears. Certainly, with such a neutral and well-balanced signal, it will be very difficult to mess up a mix with either model. A common problem with both cards is the lack of analog inserts. In both a home studio and a professional one, it would be interesting to use a tube compressor to combine with the internal microphone preamps, but unfortunately both manufacturers chose to deny us this small but very effective "gem." If I had to add any additional considerations, I would say that the RME is more versatile and more convenient due to its portability, while the Motu, which uses a PCI bus, negates this convenience but provides high-quality performance. After being an impartial judge, I defer to your chosen sums, which will certainly be dictated by the future use of the product and the expense you are willing to face.
In the sound card market a real battle has been unleashed, what is at stake is to impose a "definitive" professional standard; often, however, the weapons used end up by slightly differing: the resolution and sampling frequency, the format of the inputs and outputs, the presence or absence of preamplifiers, the methods of connection to the computer's “mother” card. How to orient yourself in this chaos of brands and numbers? In this issue I wanted to focus on two leading houses in the sound card sector: MOTU and RME.
Prologue:
As mentioned in the introduction, this comparison test involves two leading professional audio interfaces. Having owned an RME Fireface 800 for over a year, which I used for all of CM2 magazine's tests, I chose to conduct this article by comparing it with another professional sound engineer who uses a MOTU 896 HD. In short, I sought to be as objective as possible in this complicated and demanding task. When it comes to top-of-the-line audio interfaces, the choice is often determined by the product's name and one's personal sonic taste rather than by the actual usability of the interface's features. I should point out that both interfaces were tested for approximately five months in a professional recording studio, where everything from pop groups and string quartets to 30-piece gospel choirs and singer-songwriters are recorded, always seeking sophisticated sounds. This was a tough test to demonstrate their stability and reliability over time. Before writing this article, I read various English and American magazines to decide how to frame the comparison, which points to highlight and which to leave out as unnecessary and dispersive. I realized how difficult it is for the end user today to choose one product over another; all the cards often appear perfect, the only flaw being some small, perhaps deliberate, oversight in the software or hardware design, justified by the good Q/P ratio. Poor, then, those buyers who, before taking the fateful step, have to rely on magazines, forums, or the advice of friends. I understand them—I'm one of them, too—perhaps a little luckier in that these days I can try some products before buying them, but know that I'm not unfamiliar with the classics I fell prey to when I was more inexperienced. Unfortunately, there are no stores that allow you to test multiple cards simultaneously in the same environment and with the same monitoring speakers. I will give a brief overview of the components and design used to make these two products, as my aim is not to push the reader to purchase one model over another, but to report, as clearly as possible, the pros and cons of the individual products and the intended use envisaged by the respective manufacturers.
Audio chain:
For this test I used two computers, a MAC G4 with an 876 MHz processor/1.5 GB of RAM, with Logic Audio version 6.0 and a MusicDaw 2 PC (Intel 775 processor at 3,4 GHz and 2 MB cache) with XP and Cubase SX3. For listening, I used both my PMC TBS2-A, the Dynaudio BM6A and the Yamaha NS10M that were already in the studio, so that both I and the second sound engineer had a listening reference point that we already knew well. The mixer is a Soundcraft 3200 and as a recorder (with Adat I/O) we used an Alesis HD 24, while for the Tascam “TDIF” format we got a Tascam. Microphones: Neuman U 87, Neuman KM 184, Akg C414, Akg C451, Audio-Technica AT4041, Rode NT2, a “Spectralab” spectrum analyzer from Sound Technology and a lot of patience.
RME Fireface 800
The RME Fireface 800 (Fig.1) is a FireWire sound card, “made in Germany”, a professional product that allows you to manage up to 28 I / O simultaneously. The card has been studied down to the smallest details and is the result of two years of development. It integrates the best of modern technology, the integrated 32-bit RISC processor, the use of latest generation reprogrammable logic components (FPGA, 90 nanometer Field Programmable Gate Arrays), the latest generation communication protocol FireWire at 800 Mb / s , AKM 5385AVF converters supporting up to 24bit 192kHz. The internal technology is all in SMD. But let's go in order ...
What is presented:
The Fireface 800 comes in a cardboard box, along with a photocopied manual in English and German, a pair of Adat cables, a FireWire 400 cable, four rubber pads to place under the card to prevent it from being scratched, a software CD, and a warranty sheet from the Italian importer. The machine is a standard 1U rack unit, and is light gray and purple in color. From left to right, we find: channel 4 (Fig.2) which reproduces a "DI BOX" to allow you to directly record a bass / guitar signal and all those instruments that have a very low output gain. At the top there are three LEDs: DRIVE yellow, LIM (Limiter) red and SIG (signal input control) green, while at the bottom an Instr / Line “Speaker Emulation” input with a Gain control next to it by potentiometer. Moving to the right we find four entrances, from 7 to 10, (Fig.3) with INSERT (LINE) connection on standard jacks in TRS format, or MIC on XLR individually controlled by a potentiometer for the gain of the preamplifiers (GAIN) which have an excursion from + 10 to +80 dB. At the top we find for each of the inputs a LED that indicates the activation of the 48 V (yellow color) the CLIP signal (red color) and the green SIG (signal input control). We note with regret that no “switch” has been provided to activate the phantom power from the RME panel. Activation can only take place via software and it will eventually be possible to save it in the machine's routing settings for "stand-alone" use. Continuing on the right we find an ANALOG / DIGITAL STATE LED control (Fig.4) which is none other than the panel where we can view the status of the input and output signal "LEVEL IN / OUT" and the status of the clock for the various digital sources. The last connection on the front panel is that of the headphones, controlled by a volume potentiometer to be transformed if necessary into in / out 9 and 10 (unbalanced).
Rear Panel:
From left to right we find (Fig.5): connector for power supply, MIDI IN / OUT, a Firewire 400 port and two Firewire 800 ports with a metal hook to hook the Firewire cable, WORD OUT on BNC, 2 digital I / O in ADAT format (total 16 I / O independent), I / OS / PDIF, WORD IN on BNC with 75 ohm impedance selector, at the bottom 8 analog audio outputs in TRS format (Fig.6). The last stage, from left to right starting from the top, we find: Optional time code, while at the bottom there are 8 outputs in TRS format. Already from the description we can understand that the board was designed more for studio use than for "live" shooting, as for live use it would have been appropriate to have at least 8 analog inputs, with the possibility of supporting a "phantom" power supply selectable from the panel and not via software. Furthermore, the level indicator LEDs on the front panel are missing. I believe that these "shortcomings" have been well considered in order not to jeopardize the original design of a "light" system for a rack unit.
First impressions:
We install the card using the Firewire 400 port cable of the PC mod. MusicDaw 2, we load the software for the management of the machine and we begin to discover flaws and virtues of this much publicized card. The card's routing management software is very basic and lean. Although the “minimalist design” has very small fonts, functionality is not impaired in any way. The use of the "Fireface Setting" panel is easily understood regardless of the manuals (Fig.7). From here we can decide which inputs to control, analog, Adat, S / PDIF, and the relative frequency from 32 to 192 kHz. We can decide if the XLR inputs must power microphones that require the Phantom, or simple cardiodes; all parameter changes can be performed in real time without ever locking the system. Analog and digital inputs and outputs can be combined simultaneously. Latency can drop to very low values, equal to about one millisecond, but I strongly advise against it, it would only mean making the computer's CPU work harder without having any kind of advantage. The card has ASIO2.0, ASIO Multiclient, MME, WDM, Multiclient, GSFI drivers and is compatible with ADM (ASIO Direct Monitoring). If you use the FireWire 800 connection, in theory you could connect up to a maximum of 3 RMEs for a total of 84 24-bit / 48 KHz I / O.
RME Under the Press:
The first tests I did were those relating to virtual software, from the heaviest ones like “Symphonic Orchestra Platinum” to the “badly” compiled ones like Tritone Digital Hydratone. In none of the cases have I ever had any problems of any kind. A few months ago a friend let me try the same card under LINUX with “sequencers” and “plugs” that I didn't know existed. To my amazement, the card ran at millisecond values without the CPU going into TILT. But the real test of fire was in the studio. Starting from material already existing on the ALESIS HD24, my friend Rino Morra and I attempted to redo a mix; we simultaneously took the 10 analog inputs with the 16 in ADAT format and transferred everything to the computer. Finally, the result was transferred to an analog mixer, using the RME's analog outputs and using the ALESIS HD 24 as a D/A converter to transform the optical outputs of the sound card into analog. Not content with that, we also try to import some tracks from the card's analog sources and transfer them directly into ADAT format. The card's routing allowed us to do this without having to first record the audio content into the computer. I wanted to make a clarification about the Fireface inputs: there are 12 of them (4 XLR and 8 TRS Jack) but a maximum of 10 can function simultaneously, as 2 inputs are just repetitions. One downside, which is worthily obscured by “equivocal” advertising phrases, is that the card is advertised as a “56-channel Firewire audio interface – 192 kHz/24 bit”. But where are these 56 channels? In reality, it is possible to have a maximum of 28 I/O inputs at 24 bit / 44kHz or 48 kHz simultaneously using all the I/Os (analog-digital-S/PDIF). If instead we wanted to record at 24 bit 96 kHz we could reach a maximum of 20 recording inputs (10 analog – 8 digital – 2 S/PDIF). Working at 24 bit/192 kHz we would reach a maximum of 12 recording inputs (10 analog – 2 S/PDIF). In this case it is not possible to use Adat channels. The same test was done on the MAC G4 and although it is not a latest generation computer, we did not have any kind of problem. At this point the question arises spontaneously: how does the instrument sound? We found that the system is extremely stable on both Mac and PC and that it can also work under Linux, ensuring complete audio-video integration. Anyway, what exactly happens when the signal goes in and out? How do the preamps and “Speaker Emulation” input sound? RME A/D Section The first test I carry out concerns input 1 “Speaker Emulation”. In one corner of the studio is an old Rhodes Mark 2 piano, slightly out of tune and with some keys not working perfectly. With the advent of Virtual Instruments, the studio owner relegated it to a decorative object rather than a musical instrument. I convince him to move it towards our station. I don't have the equipment to tune it, and I honestly don't think I'm capable of it; however, I'd like to hear what happens when I plug it directly into the Speaker Emulation input. This connection has the characteristics of a DI box, the input level is higher and more punchy than the line inputs. By increasing the Gain a little you get a drive effect very similar to that of a real cabinet (Fig.8). The Rhodes recorded in this way took back all the original breath and brilliance. Percussion can be heard perfectly without being overwhelmed by background noise. Long last! With all its "detuning" it is a pleasure to play and record this instrument, to hear it "alive" in its harmonic nuances. The test was also carried out on guitar and bass, always with excellent results, but the most difficult test was the relative one. to the microphone channels.
The preamp section, according to many magazines and forums, is not a strong point of the card; I've read that it would have been better to opt for two higher-quality preamps rather than the four supplied. In this case, I want to break a lance in favor of the Fireface. The preamps are of excellent quality and, as with everything, you need to understand what they were designed for. If our goal is to record vocals, we shouldn't expect optimal results; the converters are too transparent and don't have a clear sound, too little color. This isn't a flaw; in fact, in many cases, it can be a real strength when it comes to mic'ing a classical guitar, violin, or piano, as we can be sure that the sound will never sound muddy. If we wanted to record vocals, however (unless your "performer" is Barry White), it would be advisable to add a good preamp with more character to ensure a more balanced sound. Perhaps the only flaw concerns the potentiometers that adjust the input level (Gain), which don't respond smoothly.
to the movements they impart: insensitive for almost three-quarters of their travel, rapid surges in the remaining phase, making it really difficult to manage the volume. It's likely that in later versions the problem was solved with better calibration, but I don't have the opportunity to verify this at the moment. The last test I performed was to insert an external FireWire accelerator card into the PC, to be precise, a TC Electronics one. Unfortunately, I had conflict problems that were confirmed to me in the various forums I regularly frequent. If we use PCI accelerator cards, however, no problem, whether they are TC or Universal Audio.
RME CONCLUSIONS:
There's much more to say about this card, its routing, and all its many applications. Perhaps I've highlighted more flaws than strengths, but I assure you that over the years I've never found anything absolutely perfect. One of the factors common to all sound cards, even professional ones, is that as the frequency increases,
As the sampling rate increases, the sound becomes increasingly harsh and less natural: this problem doesn't affect the RME, the sound remains very linear even when varying frequencies. The overall signal is very homogeneous and pleasant, well-defined and natural at the frequency extremes. The bass is slightly full and rounded and acquires just the right amount of color without ever sounding false; the highs are transparent and silky. Even after many hours of work, the sound doesn't tire, and its "just right" amount of neutrality allows for good results even in crowded mixes. Of course, I would have preferred a "hardware" button for selecting phantom power and more balanced volume pots. Let's hope these small flaws inspire a future two-channel project, specifically for mastering use or small home recording projects. It would be nice to expect a 2 rack unit with all inputs equipped with phantom power, not without their own level indicator, it would be an ideal solution for live or studio use: we would benefit from a good input stage and the convenience of mixing entirely via DAW.
Motu 2408 MK3 what it looks like:
The Motu 2408 MK3 is presented in a cardboard box, together with 2 manuals, PCI-424 and AUDIO DESK vers. 2.0 for MAC in English, a power cable, a PCI connection cable, a CD with the software and a registration card from the American distributor. The machine is in standard rack format with one unit in dark gray color (Fig.9). From left to right we find: the headphone input with its VOL potentiometer and the MAIN OUT VOL potentiometer. A DIGITAL I/O control, divided into three BANKS A, B and C (Fig.10). Moving further to the left we find the display that manages the volume of the 8 ANALOG IN analog inputs and the one that manages the 8 ANALOG OUT analog outputs. Finally we find a FORMAT CONVERSION button selector and at the bottom a SET button and lastly the classic ON / OFF switch.
Rear Panel:
From left to right we find a fairly crowded panel (Fig.11): the power connector, WORD/VIDEO IN/OUT, SMPTE (LTC), S/PDIF, 3 ports for signals in TDIF format (Tascam), a main out, OUT to connect the card to the PCI-424, 3 optical I/O in Adat format, for a total of 24 other I/O, 8 ANALOG OUT and 8 ANALOG IN in TRS format and finally MAIN OUT, right and left. Finally the real heart of this card is the PCI 424, (Fig.12) which houses the DSP that manages all the routing of inputs and outputs, relieving the computer CPU and making the system more reliable and stable, with latency close to zero.
First impressions of MOTU:
I plugged the Motu PCI controller into the motherboard slot, booted the computer, and Windows XP promptly detected the presence of new hardware. I inserted the required disk drive and started the installation. The procedure was successful without any hitch. After rebooting the system I connected the audio interface to the relative PCI card, set the "drivers" and launched Cubase SX 3 which immediately recognized the card and all its functions. The control panel is similar to that of the model
896 HD that I mounted on the Mac in order to perceive the differences in sound (Fig.13). As this is a card designed more as a controller and system heart rather than as a support for digital inputs and outputs, I ran into an initial difficulty for my audio tests; in this case too, however, ingenuity won, in fact, in addition to the expected tests I was able to invent new interesting tests, even if not standard.
Motu 2408 Under Press:
The first test I carried out was to plug a Focus ISA 430 MK2 into one of the 8 analog inputs of the Motu, not so much to test the sound quality during recording, but to test whether the analog input signal of the card would saturate. I directly inserted the signal of a slap bass into the Focus, and tried to dose it in a way
The PRE's input and output were intelligent; the card handled fast transient changes perfectly without ever going into digital distortion. This demonstrates the high quality of the converters and the instrument's design. I tried recording and playing back 8 audio signals using the analog inputs and outputs. The result was excellent both in terms of sound and system stability. The microphone test (voice) also went well. For this test I used a Neumann U 87 microphone powered by the
Soundcraft 3200 mixer, a nice analog pairing: no problems using either the MusicDaw 2 or the G4, which can be defined as less powerful in terms of processor and RAM. One of the most demanding tests I've carried out in these 5 months is to stress the system with a huge amount of data. The studio where I carried out the test was commissioned to record a classical music concert held in a theater: for the occasion, all 24 tracks of the Alesis HD 24 recorder were used (Fig.14). Subsequently, for "editing" reasons of the individual audio tracks, the need arose to import the entire recording into the computer. The entire concert was transferred in a single pass using the 24 optical outputs of the Alesis HD 24 and setting the Motu to the 24 Adat (optical) inputs, all synchronizing the recorder and the interface via ADAT SYNC. The recording (24 bit/48kHz) lasted 70 minutes and during this time the card did not present any problems of any kind. The signal never
unhooked and there were no drops in the recording. The problems began when, for work reasons, it was necessary to transfer the work from the Adat to the Tascam. No
Having some other ADAT cables available, I connected the HD24 outputs to ADAT on the card and, from the 2408 MK3, I connected them to the TASCAM MX2448 multitrack player using TDIF cables. What a surprise! If you set the card to receive 24 signals in "optical adat" format, real-time playback can only occur via ADAT. Unfortunately, the inability to transfer between formats in real time is a major limitation. I'd like to list a strength that I've only found with this card:
While playback is running, if you enable recording for just one track, the transition happens lightning-fast, with no latency whatsoever, just like working with a recorder like the Alesis HD 24 or a classic Fostex D 160. The same performance is achieved by recording up to four tracks simultaneously. From five tracks onwards, the delay begins to be noticeable, even if minimal. All this is possible both because the management software is very well written, and we had no doubts about this, and because, as far as I can tell, PCI cards always work better than Firewire ones, receive more power, and are more stable. The last test I carried out was
that of inserting an external accelerator card into the PC, to be precise TC Electronics: no kind of conflict.
Conclusions Motu:
Even on this card there is still a lot to say, starting with the enormous power of his CueMix DSP, to all possible applications with the audio and video world.
It belongs to a specific professional sector. It does not want to be addressed to all the small household “Home studios” but to an audience of professionals who adopt hardware supports for data transfer. Think of the advantages for those who record live and who still find themselves using "workstations" or recorders such as Tascam MX 2424, or the old Alesis Adat, etc. Some will turn up their noses calling them technologically "outdated" products, especially after the advent of computers, but nothing can be said about their reliability. I have to say that even though I personally own a RME Fireface 800 and am used to a slightly warmer sound, I don't mind the aroma of this card. In my opinion,
what differentiates the Motu house from many others is the harshness, which does not make the high frequencies shine or emphasize the low ones. From the converters, said in a very simple language, “only what comes in comes out”, this allows in the mixing phase a great workability of the sound without having to calculate and balance the intrinsic characteristics of the converters. We can add in this regard that compared to the little sister 896 HD, the 2408 MK3 is slightly more colorful in the high frequencies (it shines a little more).
General consideration:
Both cards sound good, and the incoming signal is recorded very balanced. Each has a slight emphasis on one part of the sound spectrum, but these are minor details. I'm sure that if I were to listen to one of the two cards in another room blindfolded, I wouldn't be able to tell them apart. They're both pleasant to listen to, and in the long run, they don't tire the engineer's ears. Certainly, with such a neutral and well-balanced signal, it will be very difficult to mess up a mix with either model. A common problem with both cards is the lack of analog inserts. In both a home studio and a professional one, it would be interesting to use a tube compressor to combine with the internal microphone preamps, but unfortunately both manufacturers chose to deny us this small but very effective "gem." If I had to add any additional considerations, I would say that the RME is more versatile and more convenient due to its portability, while the Motu, which uses a PCI bus, negates this convenience but provides high-quality performance. After being an impartial judge, I defer to your chosen sums, which will certainly be dictated by the future use of the product and the expense you are willing to face.
In the sound card market a real battle has been unleashed, what is at stake is to impose a "definitive" professional standard; often, however, the weapons used end up by slightly differing: the resolution and sampling frequency, the format of the inputs and outputs, the presence or absence of preamplifiers, the methods of connection to the computer's “mother” card. How to orient yourself in this chaos of brands and numbers? In this issue I wanted to focus on two leading houses in the sound card sector: MOTU and RME.
Prologue:
As mentioned in the introduction, this comparison test involves two leading professional audio interfaces. Having owned an RME Fireface 800 for over a year, which I used for all of CM2 magazine's tests, I chose to conduct this article by comparing it with another professional sound engineer who uses a MOTU 896 HD. In short, I sought to be as objective as possible in this complicated and demanding task. When it comes to top-of-the-line audio interfaces, the choice is often determined by the product's name and one's personal sonic taste rather than by the actual usability of the interface's features. I should point out that both interfaces were tested for approximately five months in a professional recording studio, where everything from pop groups and string quartets to 30-piece gospel choirs and singer-songwriters are recorded, always seeking sophisticated sounds. This was a tough test to demonstrate their stability and reliability over time. Before writing this article, I read various English and American magazines to decide how to frame the comparison, which points to highlight and which to leave out as unnecessary and dispersive. I realized how difficult it is for the end user today to choose one product over another; all the cards often appear perfect, the only flaw being some small, perhaps deliberate, oversight in the software or hardware design, justified by the good Q/P ratio. Poor, then, those buyers who, before taking the fateful step, have to rely on magazines, forums, or the advice of friends. I understand them—I'm one of them, too—perhaps a little luckier in that these days I can try some products before buying them, but know that I'm not unfamiliar with the classics I fell prey to when I was more inexperienced. Unfortunately, there are no stores that allow you to test multiple cards simultaneously in the same environment and with the same monitoring speakers. I will give a brief overview of the components and design used to make these two products, as my aim is not to push the reader to purchase one model over another, but to report, as clearly as possible, the pros and cons of the individual products and the intended use envisaged by the respective manufacturers.
Audio chain:
For this test I used two computers, a MAC G4 with an 876 MHz processor/1.5 GB of RAM, with Logic Audio version 6.0 and a MusicDaw 2 PC (Intel 775 processor at 3,4 GHz and 2 MB cache) with XP and Cubase SX3. For listening, I used both my PMC TBS2-A, the Dynaudio BM6A and the Yamaha NS10M that were already in the studio, so that both I and the second sound engineer had a listening reference point that we already knew well. The mixer is a Soundcraft 3200 and as a recorder (with Adat I/O) we used an Alesis HD 24, while for the Tascam “TDIF” format we got a Tascam. Microphones: Neuman U 87, Neuman KM 184, Akg C414, Akg C451, Audio-Technica AT4041, Rode NT2, a “Spectralab” spectrum analyzer from Sound Technology and a lot of patience.
RME Fireface 800
The RME Fireface 800 (Fig.1) is a FireWire sound card, “made in Germany”, a professional product that allows you to manage up to 28 I / O simultaneously. The card has been studied down to the smallest details and is the result of two years of development. It integrates the best of modern technology, the integrated 32-bit RISC processor, the use of latest generation reprogrammable logic components (FPGA, 90 nanometer Field Programmable Gate Arrays), the latest generation communication protocol FireWire at 800 Mb / s , AKM 5385AVF converters supporting up to 24bit 192kHz. The internal technology is all in SMD. But let's go in order ...
What is presented:
The Fireface 800 comes in a cardboard box, along with a photocopied manual in English and German, a pair of Adat cables, a FireWire 400 cable, four rubber pads to place under the card to prevent it from being scratched, a software CD, and a warranty sheet from the Italian importer. The machine is a standard 1U rack unit, and is light gray and purple in color. From left to right, we find: channel 4 (Fig.2) which reproduces a "DI BOX" to allow you to directly record a bass / guitar signal and all those instruments that have a very low output gain. At the top there are three LEDs: DRIVE yellow, LIM (Limiter) red and SIG (signal input control) green, while at the bottom an Instr / Line “Speaker Emulation” input with a Gain control next to it by potentiometer. Moving to the right we find four entrances, from 7 to 10, (Fig.3) with INSERT (LINE) connection on standard jacks in TRS format, or MIC on XLR individually controlled by a potentiometer for the gain of the preamplifiers (GAIN) which have an excursion from + 10 to +80 dB. At the top we find for each of the inputs a LED that indicates the activation of the 48 V (yellow color) the CLIP signal (red color) and the green SIG (signal input control). We note with regret that no “switch” has been provided to activate the phantom power from the RME panel. Activation can only take place via software and it will eventually be possible to save it in the machine's routing settings for "stand-alone" use. Continuing on the right we find an ANALOG / DIGITAL STATE LED control (Fig.4) which is none other than the panel where we can view the status of the input and output signal "LEVEL IN / OUT" and the status of the clock for the various digital sources. The last connection on the front panel is that of the headphones, controlled by a volume potentiometer to be transformed if necessary into in / out 9 and 10 (unbalanced).
Rear Panel:
From left to right we find (Fig.5): connector for power supply, MIDI IN / OUT, a Firewire 400 port and two Firewire 800 ports with a metal hook to hook the Firewire cable, WORD OUT on BNC, 2 digital I / O in ADAT format (total 16 I / O independent), I / OS / PDIF, WORD IN on BNC with 75 ohm impedance selector, at the bottom 8 analog audio outputs in TRS format (Fig.6). The last stage, from left to right starting from the top, we find: Optional time code, while at the bottom there are 8 outputs in TRS format. Already from the description we can understand that the board was designed more for studio use than for "live" shooting, as for live use it would have been appropriate to have at least 8 analog inputs, with the possibility of supporting a "phantom" power supply selectable from the panel and not via software. Furthermore, the level indicator LEDs on the front panel are missing. I believe that these "shortcomings" have been well considered in order not to jeopardize the original design of a "light" system for a rack unit.
First impressions:
We install the card using the Firewire 400 port cable of the PC mod. MusicDaw 2, we load the software for the management of the machine and we begin to discover flaws and virtues of this much publicized card. The card's routing management software is very basic and lean. Although the “minimalist design” has very small fonts, functionality is not impaired in any way. The use of the "Fireface Setting" panel is easily understood regardless of the manuals (Fig.7). From here we can decide which inputs to control, analog, Adat, S / PDIF, and the relative frequency from 32 to 192 kHz. We can decide if the XLR inputs must power microphones that require the Phantom, or simple cardiodes; all parameter changes can be performed in real time without ever locking the system. Analog and digital inputs and outputs can be combined simultaneously. Latency can drop to very low values, equal to about one millisecond, but I strongly advise against it, it would only mean making the computer's CPU work harder without having any kind of advantage. The card has ASIO2.0, ASIO Multiclient, MME, WDM, Multiclient, GSFI drivers and is compatible with ADM (ASIO Direct Monitoring). If you use the FireWire 800 connection, in theory you could connect up to a maximum of 3 RMEs for a total of 84 24-bit / 48 KHz I / O.
RME Under the Press:
The first tests I did were those relating to virtual software, from the heaviest ones like “Symphonic Orchestra Platinum” to the “badly” compiled ones like Tritone Digital Hydratone. In none of the cases have I ever had any problems of any kind. A few months ago a friend let me try the same card under LINUX with “sequencers” and “plugs” that I didn't know existed. To my amazement, the card ran at millisecond values without the CPU going into TILT. But the real test of fire was in the studio. Starting from material already existing on the ALESIS HD24, my friend Rino Morra and I attempted to redo a mix; we simultaneously took the 10 analog inputs with the 16 in ADAT format and transferred everything to the computer. Finally, the result was transferred to an analog mixer, using the RME's analog outputs and using the ALESIS HD 24 as a D/A converter to transform the optical outputs of the sound card into analog. Not content with that, we also try to import some tracks from the card's analog sources and transfer them directly into ADAT format. The card's routing allowed us to do this without having to first record the audio content into the computer. I wanted to make a clarification about the Fireface inputs: there are 12 of them (4 XLR and 8 TRS Jack) but a maximum of 10 can function simultaneously, as 2 inputs are just repetitions. One downside, which is worthily obscured by “equivocal” advertising phrases, is that the card is advertised as a “56-channel Firewire audio interface – 192 kHz/24 bit”. But where are these 56 channels? In reality, it is possible to have a maximum of 28 I/O inputs at 24 bit / 44kHz or 48 kHz simultaneously using all the I/Os (analog-digital-S/PDIF). If instead we wanted to record at 24 bit 96 kHz we could reach a maximum of 20 recording inputs (10 analog – 8 digital – 2 S/PDIF). Working at 24 bit/192 kHz we would reach a maximum of 12 recording inputs (10 analog – 2 S/PDIF). In this case it is not possible to use Adat channels. The same test was done on the MAC G4 and although it is not a latest generation computer, we did not have any kind of problem. At this point the question arises spontaneously: how does the instrument sound? We found that the system is extremely stable on both Mac and PC and that it can also work under Linux, ensuring complete audio-video integration. Anyway, what exactly happens when the signal goes in and out? How do the preamps and “Speaker Emulation” input sound? RME A/D Section The first test I carry out concerns input 1 “Speaker Emulation”. In one corner of the studio is an old Rhodes Mark 2 piano, slightly out of tune and with some keys not working perfectly. With the advent of Virtual Instruments, the studio owner relegated it to a decorative object rather than a musical instrument. I convince him to move it towards our station. I don't have the equipment to tune it, and I honestly don't think I'm capable of it; however, I'd like to hear what happens when I plug it directly into the Speaker Emulation input. This connection has the characteristics of a DI box, the input level is higher and more punchy than the line inputs. By increasing the Gain a little you get a drive effect very similar to that of a real cabinet (Fig.8). The Rhodes recorded in this way took back all the original breath and brilliance. Percussion can be heard perfectly without being overwhelmed by background noise. Long last! With all its "detuning" it is a pleasure to play and record this instrument, to hear it "alive" in its harmonic nuances. The test was also carried out on guitar and bass, always with excellent results, but the most difficult test was the relative one. to the microphone channels.
The preamp section, according to many magazines and forums, is not a strong point of the card; I've read that it would have been better to opt for two higher-quality preamps rather than the four supplied. In this case, I want to break a lance in favor of the Fireface. The preamps are of excellent quality and, as with everything, you need to understand what they were designed for. If our goal is to record vocals, we shouldn't expect optimal results; the converters are too transparent and don't have a clear sound, too little color. This isn't a flaw; in fact, in many cases, it can be a real strength when it comes to mic'ing a classical guitar, violin, or piano, as we can be sure that the sound will never sound muddy. If we wanted to record vocals, however (unless your "performer" is Barry White), it would be advisable to add a good preamp with more character to ensure a more balanced sound. Perhaps the only flaw concerns the potentiometers that adjust the input level (Gain), which don't respond smoothly.
to the movements they impart: insensitive for almost three-quarters of their travel, rapid surges in the remaining phase, making it really difficult to manage the volume. It's likely that in later versions the problem was solved with better calibration, but I don't have the opportunity to verify this at the moment. The last test I performed was to insert an external FireWire accelerator card into the PC, to be precise, a TC Electronics one. Unfortunately, I had conflict problems that were confirmed to me in the various forums I regularly frequent. If we use PCI accelerator cards, however, no problem, whether they are TC or Universal Audio.
RME CONCLUSIONS:
There's much more to say about this card, its routing, and all its many applications. Perhaps I've highlighted more flaws than strengths, but I assure you that over the years I've never found anything absolutely perfect. One of the factors common to all sound cards, even professional ones, is that as the frequency increases,
As the sampling rate increases, the sound becomes increasingly harsh and less natural: this problem doesn't affect the RME, the sound remains very linear even when varying frequencies. The overall signal is very homogeneous and pleasant, well-defined and natural at the frequency extremes. The bass is slightly full and rounded and acquires just the right amount of color without ever sounding false; the highs are transparent and silky. Even after many hours of work, the sound doesn't tire, and its "just right" amount of neutrality allows for good results even in crowded mixes. Of course, I would have preferred a "hardware" button for selecting phantom power and more balanced volume pots. Let's hope these small flaws inspire a future two-channel project, specifically for mastering use or small home recording projects. It would be nice to expect a 2 rack unit with all inputs equipped with phantom power, not without their own level indicator, it would be an ideal solution for live or studio use: we would benefit from a good input stage and the convenience of mixing entirely via DAW.
Motu 2408 MK3 what it looks like:
The Motu 2408 MK3 is presented in a cardboard box, together with 2 manuals, PCI-424 and AUDIO DESK vers. 2.0 for MAC in English, a power cable, a PCI connection cable, a CD with the software and a registration card from the American distributor. The machine is in standard rack format with one unit in dark gray color (Fig.9). From left to right we find: the headphone input with its VOL potentiometer and the MAIN OUT VOL potentiometer. A DIGITAL I/O control, divided into three BANKS A, B and C (Fig.10). Moving further to the left we find the display that manages the volume of the 8 ANALOG IN analog inputs and the one that manages the 8 ANALOG OUT analog outputs. Finally we find a FORMAT CONVERSION button selector and at the bottom a SET button and lastly the classic ON / OFF switch.
Rear Panel:
From left to right we find a fairly crowded panel (Fig.11): the power connector, WORD/VIDEO IN/OUT, SMPTE (LTC), S/PDIF, 3 ports for signals in TDIF format (Tascam), a main out, OUT to connect the card to the PCI-424, 3 optical I/O in Adat format, for a total of 24 other I/O, 8 ANALOG OUT and 8 ANALOG IN in TRS format and finally MAIN OUT, right and left. Finally the real heart of this card is the PCI 424, (Fig.12) which houses the DSP that manages all the routing of inputs and outputs, relieving the computer CPU and making the system more reliable and stable, with latency close to zero.
First impressions of MOTU:
I plugged the Motu PCI controller into the motherboard slot, booted the computer, and Windows XP promptly detected the presence of new hardware. I inserted the required disk drive and started the installation. The procedure was successful without any hitch. After rebooting the system I connected the audio interface to the relative PCI card, set the "drivers" and launched Cubase SX 3 which immediately recognized the card and all its functions. The control panel is similar to that of the model
896 HD that I mounted on the Mac in order to perceive the differences in sound (Fig.13). As this is a card designed more as a controller and system heart rather than as a support for digital inputs and outputs, I ran into an initial difficulty for my audio tests; in this case too, however, ingenuity won, in fact, in addition to the expected tests I was able to invent new interesting tests, even if not standard.
Motu 2408 Under Press:
The first test I carried out was to plug a Focus ISA 430 MK2 into one of the 8 analog inputs of the Motu, not so much to test the sound quality during recording, but to test whether the analog input signal of the card would saturate. I directly inserted the signal of a slap bass into the Focus, and tried to dose it in a way
The PRE's input and output were intelligent; the card handled fast transient changes perfectly without ever going into digital distortion. This demonstrates the high quality of the converters and the instrument's design. I tried recording and playing back 8 audio signals using the analog inputs and outputs. The result was excellent both in terms of sound and system stability. The microphone test (voice) also went well. For this test I used a Neumann U 87 microphone powered by the
Soundcraft 3200 mixer, a nice analog pairing: no problems using either the MusicDaw 2 or the G4, which can be defined as less powerful in terms of processor and RAM. One of the most demanding tests I've carried out in these 5 months is to stress the system with a huge amount of data. The studio where I carried out the test was commissioned to record a classical music concert held in a theater: for the occasion, all 24 tracks of the Alesis HD 24 recorder were used (Fig.14). Subsequently, for "editing" reasons of the individual audio tracks, the need arose to import the entire recording into the computer. The entire concert was transferred in a single pass using the 24 optical outputs of the Alesis HD 24 and setting the Motu to the 24 Adat (optical) inputs, all synchronizing the recorder and the interface via ADAT SYNC. The recording (24 bit/48kHz) lasted 70 minutes and during this time the card did not present any problems of any kind. The signal never
unhooked and there were no drops in the recording. The problems began when, for work reasons, it was necessary to transfer the work from the Adat to the Tascam. No
Having some other ADAT cables available, I connected the HD24 outputs to ADAT on the card and, from the 2408 MK3, I connected them to the TASCAM MX2448 multitrack player using TDIF cables. What a surprise! If you set the card to receive 24 signals in "optical adat" format, real-time playback can only occur via ADAT. Unfortunately, the inability to transfer between formats in real time is a major limitation. I'd like to list a strength that I've only found with this card:
While playback is running, if you enable recording for just one track, the transition happens lightning-fast, with no latency whatsoever, just like working with a recorder like the Alesis HD 24 or a classic Fostex D 160. The same performance is achieved by recording up to four tracks simultaneously. From five tracks onwards, the delay begins to be noticeable, even if minimal. All this is possible both because the management software is very well written, and we had no doubts about this, and because, as far as I can tell, PCI cards always work better than Firewire ones, receive more power, and are more stable. The last test I carried out was
that of inserting an external accelerator card into the PC, to be precise TC Electronics: no kind of conflict.
Conclusions Motu:
Even on this card there is still a lot to say, starting with the enormous power of his CueMix DSP, to all possible applications with the audio and video world.
It belongs to a specific professional sector. It does not want to be addressed to all the small household “Home studios” but to an audience of professionals who adopt hardware supports for data transfer. Think of the advantages for those who record live and who still find themselves using "workstations" or recorders such as Tascam MX 2424, or the old Alesis Adat, etc. Some will turn up their noses calling them technologically "outdated" products, especially after the advent of computers, but nothing can be said about their reliability. I have to say that even though I personally own a RME Fireface 800 and am used to a slightly warmer sound, I don't mind the aroma of this card. In my opinion,
what differentiates the Motu house from many others is the harshness, which does not make the high frequencies shine or emphasize the low ones. From the converters, said in a very simple language, “only what comes in comes out”, this allows in the mixing phase a great workability of the sound without having to calculate and balance the intrinsic characteristics of the converters. We can add in this regard that compared to the little sister 896 HD, the 2408 MK3 is slightly more colorful in the high frequencies (it shines a little more).
General consideration:
Both cards sound good, and the incoming signal is recorded very balanced. Each has a slight emphasis on one part of the sound spectrum, but these are minor details. I'm sure that if I were to listen to one of the two cards in another room blindfolded, I wouldn't be able to tell them apart. They're both pleasant to listen to, and in the long run, they don't tire the engineer's ears. Certainly, with such a neutral and well-balanced signal, it will be very difficult to mess up a mix with either model. A common problem with both cards is the lack of analog inserts. In both a home studio and a professional one, it would be interesting to use a tube compressor to combine with the internal microphone preamps, but unfortunately both manufacturers chose to deny us this small but very effective "gem." If I had to add any additional considerations, I would say that the RME is more versatile and more convenient due to its portability, while the Motu, which uses a PCI bus, negates this convenience but provides high-quality performance. After being an impartial judge, I defer to your chosen sums, which will certainly be dictated by the future use of the product and the expense you are willing to face.
In the sound card market a real battle has been unleashed, what is at stake is to impose a "definitive" professional standard; often, however, the weapons used end up by slightly differing: the resolution and sampling frequency, the format of the inputs and outputs, the presence or absence of preamplifiers, the methods of connection to the computer's “mother” card. How to orient yourself in this chaos of brands and numbers? In this issue I wanted to focus on two leading houses in the sound card sector: MOTU and RME.
Prologue:
As mentioned in the introduction, this comparison test involves two leading professional audio interfaces. Having owned an RME Fireface 800 for over a year, which I used for all of CM2 magazine's tests, I chose to conduct this article by comparing it with another professional sound engineer who uses a MOTU 896 HD. In short, I sought to be as objective as possible in this complicated and demanding task. When it comes to top-of-the-line audio interfaces, the choice is often determined by the product's name and one's personal sonic taste rather than by the actual usability of the interface's features. I should point out that both interfaces were tested for approximately five months in a professional recording studio, where everything from pop groups and string quartets to 30-piece gospel choirs and singer-songwriters are recorded, always seeking sophisticated sounds. This was a tough test to demonstrate their stability and reliability over time. Before writing this article, I read various English and American magazines to decide how to frame the comparison, which points to highlight and which to leave out as unnecessary and dispersive. I realized how difficult it is for the end user today to choose one product over another; all the cards often appear perfect, the only flaw being some small, perhaps deliberate, oversight in the software or hardware design, justified by the good Q/P ratio. Poor, then, those buyers who, before taking the fateful step, have to rely on magazines, forums, or the advice of friends. I understand them—I'm one of them, too—perhaps a little luckier in that these days I can try some products before buying them, but know that I'm not unfamiliar with the classics I fell prey to when I was more inexperienced. Unfortunately, there are no stores that allow you to test multiple cards simultaneously in the same environment and with the same monitoring speakers. I will give a brief overview of the components and design used to make these two products, as my aim is not to push the reader to purchase one model over another, but to report, as clearly as possible, the pros and cons of the individual products and the intended use envisaged by the respective manufacturers.
Audio chain:
For this test I used two computers, a MAC G4 with an 876 MHz processor/1.5 GB of RAM, with Logic Audio version 6.0 and a MusicDaw 2 PC (Intel 775 processor at 3,4 GHz and 2 MB cache) with XP and Cubase SX3. For listening, I used both my PMC TBS2-A, the Dynaudio BM6A and the Yamaha NS10M that were already in the studio, so that both I and the second sound engineer had a listening reference point that we already knew well. The mixer is a Soundcraft 3200 and as a recorder (with Adat I/O) we used an Alesis HD 24, while for the Tascam “TDIF” format we got a Tascam. Microphones: Neuman U 87, Neuman KM 184, Akg C414, Akg C451, Audio-Technica AT4041, Rode NT2, a “Spectralab” spectrum analyzer from Sound Technology and a lot of patience.
RME Fireface 800
The RME Fireface 800 (Fig.1) is a FireWire sound card, “made in Germany”, a professional product that allows you to manage up to 28 I / O simultaneously. The card has been studied down to the smallest details and is the result of two years of development. It integrates the best of modern technology, the integrated 32-bit RISC processor, the use of latest generation reprogrammable logic components (FPGA, 90 nanometer Field Programmable Gate Arrays), the latest generation communication protocol FireWire at 800 Mb / s , AKM 5385AVF converters supporting up to 24bit 192kHz. The internal technology is all in SMD. But let's go in order ...
What is presented:
The Fireface 800 comes in a cardboard box, along with a photocopied manual in English and German, a pair of Adat cables, a FireWire 400 cable, four rubber pads to place under the card to prevent it from being scratched, a software CD, and a warranty sheet from the Italian importer. The machine is a standard 1U rack unit, and is light gray and purple in color. From left to right, we find: channel 4 (Fig.2) which reproduces a "DI BOX" to allow you to directly record a bass / guitar signal and all those instruments that have a very low output gain. At the top there are three LEDs: DRIVE yellow, LIM (Limiter) red and SIG (signal input control) green, while at the bottom an Instr / Line “Speaker Emulation” input with a Gain control next to it by potentiometer. Moving to the right we find four entrances, from 7 to 10, (Fig.3) with INSERT (LINE) connection on standard jacks in TRS format, or MIC on XLR individually controlled by a potentiometer for the gain of the preamplifiers (GAIN) which have an excursion from + 10 to +80 dB. At the top we find for each of the inputs a LED that indicates the activation of the 48 V (yellow color) the CLIP signal (red color) and the green SIG (signal input control). We note with regret that no “switch” has been provided to activate the phantom power from the RME panel. Activation can only take place via software and it will eventually be possible to save it in the machine's routing settings for "stand-alone" use. Continuing on the right we find an ANALOG / DIGITAL STATE LED control (Fig.4) which is none other than the panel where we can view the status of the input and output signal "LEVEL IN / OUT" and the status of the clock for the various digital sources. The last connection on the front panel is that of the headphones, controlled by a volume potentiometer to be transformed if necessary into in / out 9 and 10 (unbalanced).
Rear Panel:
From left to right we find (Fig.5): connector for power supply, MIDI IN / OUT, a Firewire 400 port and two Firewire 800 ports with a metal hook to hook the Firewire cable, WORD OUT on BNC, 2 digital I / O in ADAT format (total 16 I / O independent), I / OS / PDIF, WORD IN on BNC with 75 ohm impedance selector, at the bottom 8 analog audio outputs in TRS format (Fig.6). The last stage, from left to right starting from the top, we find: Optional time code, while at the bottom there are 8 outputs in TRS format. Already from the description we can understand that the board was designed more for studio use than for "live" shooting, as for live use it would have been appropriate to have at least 8 analog inputs, with the possibility of supporting a "phantom" power supply selectable from the panel and not via software. Furthermore, the level indicator LEDs on the front panel are missing. I believe that these "shortcomings" have been well considered in order not to jeopardize the original design of a "light" system for a rack unit.
First impressions:
We install the card using the Firewire 400 port cable of the PC mod. MusicDaw 2, we load the software for the management of the machine and we begin to discover flaws and virtues of this much publicized card. The card's routing management software is very basic and lean. Although the “minimalist design” has very small fonts, functionality is not impaired in any way. The use of the "Fireface Setting" panel is easily understood regardless of the manuals (Fig.7). From here we can decide which inputs to control, analog, Adat, S / PDIF, and the relative frequency from 32 to 192 kHz. We can decide if the XLR inputs must power microphones that require the Phantom, or simple cardiodes; all parameter changes can be performed in real time without ever locking the system. Analog and digital inputs and outputs can be combined simultaneously. Latency can drop to very low values, equal to about one millisecond, but I strongly advise against it, it would only mean making the computer's CPU work harder without having any kind of advantage. The card has ASIO2.0, ASIO Multiclient, MME, WDM, Multiclient, GSFI drivers and is compatible with ADM (ASIO Direct Monitoring). If you use the FireWire 800 connection, in theory you could connect up to a maximum of 3 RMEs for a total of 84 24-bit / 48 KHz I / O.
RME Under the Press:
The first tests I did were those relating to virtual software, from the heaviest ones like “Symphonic Orchestra Platinum” to the “badly” compiled ones like Tritone Digital Hydratone. In none of the cases have I ever had any problems of any kind. A few months ago a friend let me try the same card under LINUX with “sequencers” and “plugs” that I didn't know existed. To my amazement, the card ran at millisecond values without the CPU going into TILT. But the real test of fire was in the studio. Starting from material already existing on the ALESIS HD24, my friend Rino Morra and I attempted to redo a mix; we simultaneously took the 10 analog inputs with the 16 in ADAT format and transferred everything to the computer. Finally, the result was transferred to an analog mixer, using the RME's analog outputs and using the ALESIS HD 24 as a D/A converter to transform the optical outputs of the sound card into analog. Not content with that, we also try to import some tracks from the card's analog sources and transfer them directly into ADAT format. The card's routing allowed us to do this without having to first record the audio content into the computer. I wanted to make a clarification about the Fireface inputs: there are 12 of them (4 XLR and 8 TRS Jack) but a maximum of 10 can function simultaneously, as 2 inputs are just repetitions. One downside, which is worthily obscured by “equivocal” advertising phrases, is that the card is advertised as a “56-channel Firewire audio interface – 192 kHz/24 bit”. But where are these 56 channels? In reality, it is possible to have a maximum of 28 I/O inputs at 24 bit / 44kHz or 48 kHz simultaneously using all the I/Os (analog-digital-S/PDIF). If instead we wanted to record at 24 bit 96 kHz we could reach a maximum of 20 recording inputs (10 analog – 8 digital – 2 S/PDIF). Working at 24 bit/192 kHz we would reach a maximum of 12 recording inputs (10 analog – 2 S/PDIF). In this case it is not possible to use Adat channels. The same test was done on the MAC G4 and although it is not a latest generation computer, we did not have any kind of problem. At this point the question arises spontaneously: how does the instrument sound? We found that the system is extremely stable on both Mac and PC and that it can also work under Linux, ensuring complete audio-video integration. Anyway, what exactly happens when the signal goes in and out? How do the preamps and “Speaker Emulation” input sound? RME A/D Section The first test I carry out concerns input 1 “Speaker Emulation”. In one corner of the studio is an old Rhodes Mark 2 piano, slightly out of tune and with some keys not working perfectly. With the advent of Virtual Instruments, the studio owner relegated it to a decorative object rather than a musical instrument. I convince him to move it towards our station. I don't have the equipment to tune it, and I honestly don't think I'm capable of it; however, I'd like to hear what happens when I plug it directly into the Speaker Emulation input. This connection has the characteristics of a DI box, the input level is higher and more punchy than the line inputs. By increasing the Gain a little you get a drive effect very similar to that of a real cabinet (Fig.8). The Rhodes recorded in this way took back all the original breath and brilliance. Percussion can be heard perfectly without being overwhelmed by background noise. Long last! With all its "detuning" it is a pleasure to play and record this instrument, to hear it "alive" in its harmonic nuances. The test was also carried out on guitar and bass, always with excellent results, but the most difficult test was the relative one. to the microphone channels.
The preamp section, according to many magazines and forums, is not a strong point of the card; I've read that it would have been better to opt for two higher-quality preamps rather than the four supplied. In this case, I want to break a lance in favor of the Fireface. The preamps are of excellent quality and, as with everything, you need to understand what they were designed for. If our goal is to record vocals, we shouldn't expect optimal results; the converters are too transparent and don't have a clear sound, too little color. This isn't a flaw; in fact, in many cases, it can be a real strength when it comes to mic'ing a classical guitar, violin, or piano, as we can be sure that the sound will never sound muddy. If we wanted to record vocals, however (unless your "performer" is Barry White), it would be advisable to add a good preamp with more character to ensure a more balanced sound. Perhaps the only flaw concerns the potentiometers that adjust the input level (Gain), which don't respond smoothly.
to the movements they impart: insensitive for almost three-quarters of their travel, rapid surges in the remaining phase, making it really difficult to manage the volume. It's likely that in later versions the problem was solved with better calibration, but I don't have the opportunity to verify this at the moment. The last test I performed was to insert an external FireWire accelerator card into the PC, to be precise, a TC Electronics one. Unfortunately, I had conflict problems that were confirmed to me in the various forums I regularly frequent. If we use PCI accelerator cards, however, no problem, whether they are TC or Universal Audio.
RME CONCLUSIONS:
There's much more to say about this card, its routing, and all its many applications. Perhaps I've highlighted more flaws than strengths, but I assure you that over the years I've never found anything absolutely perfect. One of the factors common to all sound cards, even professional ones, is that as the frequency increases,
As the sampling rate increases, the sound becomes increasingly harsh and less natural: this problem doesn't affect the RME, the sound remains very linear even when varying frequencies. The overall signal is very homogeneous and pleasant, well-defined and natural at the frequency extremes. The bass is slightly full and rounded and acquires just the right amount of color without ever sounding false; the highs are transparent and silky. Even after many hours of work, the sound doesn't tire, and its "just right" amount of neutrality allows for good results even in crowded mixes. Of course, I would have preferred a "hardware" button for selecting phantom power and more balanced volume pots. Let's hope these small flaws inspire a future two-channel project, specifically for mastering use or small home recording projects. It would be nice to expect a 2 rack unit with all inputs equipped with phantom power, not without their own level indicator, it would be an ideal solution for live or studio use: we would benefit from a good input stage and the convenience of mixing entirely via DAW.
Motu 2408 MK3 what it looks like:
The Motu 2408 MK3 is presented in a cardboard box, together with 2 manuals, PCI-424 and AUDIO DESK vers. 2.0 for MAC in English, a power cable, a PCI connection cable, a CD with the software and a registration card from the American distributor. The machine is in standard rack format with one unit in dark gray color (Fig.9). From left to right we find: the headphone input with its VOL potentiometer and the MAIN OUT VOL potentiometer. A DIGITAL I/O control, divided into three BANKS A, B and C (Fig.10). Moving further to the left we find the display that manages the volume of the 8 ANALOG IN analog inputs and the one that manages the 8 ANALOG OUT analog outputs. Finally we find a FORMAT CONVERSION button selector and at the bottom a SET button and lastly the classic ON / OFF switch.
Rear Panel:
From left to right we find a fairly crowded panel (Fig.11): the power connector, WORD/VIDEO IN/OUT, SMPTE (LTC), S/PDIF, 3 ports for signals in TDIF format (Tascam), a main out, OUT to connect the card to the PCI-424, 3 optical I/O in Adat format, for a total of 24 other I/O, 8 ANALOG OUT and 8 ANALOG IN in TRS format and finally MAIN OUT, right and left. Finally the real heart of this card is the PCI 424, (Fig.12) which houses the DSP that manages all the routing of inputs and outputs, relieving the computer CPU and making the system more reliable and stable, with latency close to zero.
First impressions of MOTU:
I plugged the Motu PCI controller into the motherboard slot, booted the computer, and Windows XP promptly detected the presence of new hardware. I inserted the required disk drive and started the installation. The procedure was successful without any hitch. After rebooting the system I connected the audio interface to the relative PCI card, set the "drivers" and launched Cubase SX 3 which immediately recognized the card and all its functions. The control panel is similar to that of the model
896 HD that I mounted on the Mac in order to perceive the differences in sound (Fig.13). As this is a card designed more as a controller and system heart rather than as a support for digital inputs and outputs, I ran into an initial difficulty for my audio tests; in this case too, however, ingenuity won, in fact, in addition to the expected tests I was able to invent new interesting tests, even if not standard.
Motu 2408 Under Press:
The first test I carried out was to plug a Focus ISA 430 MK2 into one of the 8 analog inputs of the Motu, not so much to test the sound quality during recording, but to test whether the analog input signal of the card would saturate. I directly inserted the signal of a slap bass into the Focus, and tried to dose it in a way
The PRE's input and output were intelligent; the card handled fast transient changes perfectly without ever going into digital distortion. This demonstrates the high quality of the converters and the instrument's design. I tried recording and playing back 8 audio signals using the analog inputs and outputs. The result was excellent both in terms of sound and system stability. The microphone test (voice) also went well. For this test I used a Neumann U 87 microphone powered by the
Soundcraft 3200 mixer, a nice analog pairing: no problems using either the MusicDaw 2 or the G4, which can be defined as less powerful in terms of processor and RAM. One of the most demanding tests I've carried out in these 5 months is to stress the system with a huge amount of data. The studio where I carried out the test was commissioned to record a classical music concert held in a theater: for the occasion, all 24 tracks of the Alesis HD 24 recorder were used (Fig.14). Subsequently, for "editing" reasons of the individual audio tracks, the need arose to import the entire recording into the computer. The entire concert was transferred in a single pass using the 24 optical outputs of the Alesis HD 24 and setting the Motu to the 24 Adat (optical) inputs, all synchronizing the recorder and the interface via ADAT SYNC. The recording (24 bit/48kHz) lasted 70 minutes and during this time the card did not present any problems of any kind. The signal never
unhooked and there were no drops in the recording. The problems began when, for work reasons, it was necessary to transfer the work from the Adat to the Tascam. No
Having some other ADAT cables available, I connected the HD24 outputs to ADAT on the card and, from the 2408 MK3, I connected them to the TASCAM MX2448 multitrack player using TDIF cables. What a surprise! If you set the card to receive 24 signals in "optical adat" format, real-time playback can only occur via ADAT. Unfortunately, the inability to transfer between formats in real time is a major limitation. I'd like to list a strength that I've only found with this card:
While playback is running, if you enable recording for just one track, the transition happens lightning-fast, with no latency whatsoever, just like working with a recorder like the Alesis HD 24 or a classic Fostex D 160. The same performance is achieved by recording up to four tracks simultaneously. From five tracks onwards, the delay begins to be noticeable, even if minimal. All this is possible both because the management software is very well written, and we had no doubts about this, and because, as far as I can tell, PCI cards always work better than Firewire ones, receive more power, and are more stable. The last test I carried out was
that of inserting an external accelerator card into the PC, to be precise TC Electronics: no kind of conflict.
Conclusions Motu:
Even on this card there is still a lot to say, starting with the enormous power of his CueMix DSP, to all possible applications with the audio and video world.
It belongs to a specific professional sector. It does not want to be addressed to all the small household “Home studios” but to an audience of professionals who adopt hardware supports for data transfer. Think of the advantages for those who record live and who still find themselves using "workstations" or recorders such as Tascam MX 2424, or the old Alesis Adat, etc. Some will turn up their noses calling them technologically "outdated" products, especially after the advent of computers, but nothing can be said about their reliability. I have to say that even though I personally own a RME Fireface 800 and am used to a slightly warmer sound, I don't mind the aroma of this card. In my opinion,
what differentiates the Motu house from many others is the harshness, which does not make the high frequencies shine or emphasize the low ones. From the converters, said in a very simple language, “only what comes in comes out”, this allows in the mixing phase a great workability of the sound without having to calculate and balance the intrinsic characteristics of the converters. We can add in this regard that compared to the little sister 896 HD, the 2408 MK3 is slightly more colorful in the high frequencies (it shines a little more).
General consideration:
Both cards sound good, and the incoming signal is recorded very balanced. Each has a slight emphasis on one part of the sound spectrum, but these are minor details. I'm sure that if I were to listen to one of the two cards in another room blindfolded, I wouldn't be able to tell them apart. They're both pleasant to listen to, and in the long run, they don't tire the engineer's ears. Certainly, with such a neutral and well-balanced signal, it will be very difficult to mess up a mix with either model. A common problem with both cards is the lack of analog inserts. In both a home studio and a professional one, it would be interesting to use a tube compressor to combine with the internal microphone preamps, but unfortunately both manufacturers chose to deny us this small but very effective "gem." If I had to add any additional considerations, I would say that the RME is more versatile and more convenient due to its portability, while the Motu, which uses a PCI bus, negates this convenience but provides high-quality performance. After being an impartial judge, I defer to your chosen sums, which will certainly be dictated by the future use of the product and the expense you are willing to face.
