MACKIE ONYX 400F (audio interface)

[MikeRivers]

Originally posted by harryjames For me effects are part of the performance. Going back after the fact and editing/adding effect to say a dry vocal is not how I wish to record.

It sounds like you'd be a good candidate for an outboard effect processor. To try to stay on topic here, this is what the Inserts on the 400F are for. If you have a device that gives you the vocal sound that you want to record, you can connect it in line between the mic preamp and the A/D converter via the Insert so that your vocal will be recorded with the effect.

The disadvantage to this is that you're stuck with what you recorded. If your projects are fairly simple and you have them well thought out, you can work that way - we did for years and years. But if you figure that you'll add 30 more parts and then start mixing, you might decide that the processed vocal that you recorded no longer fits.

When I first saw the Mackie 400F, I got the distinct impression that it would be a great I/O to use with Tracktion2 (which I was already pleased with its easy of use including ReWire) and that I could monitor the DAW mix, with effects, while recording, through the 400F with "no latency."

You could, if you used the Inserts and an outboard processor. But once you send the audio as data out to the computer, you're at the mercy of the computer as to what you hear in the monitor. If it takes 100 msec to crunch the numbers that gives you the effect you wan tto record, that's how long (plus other delays) it will take for the processed signal to get back to your ears.

When you have a dedicated piece of hardware that has nothing to do but process your vocal in real time on that pass, and you can route its output both to the recording software and to the monitor, you get very little latency in the monitor path. If the DSP in the 400F was pressed into service creating effects as well as mixing and routing audio, there would almost certainly be some delay - probalby not as much as the route through the computer and back, but enough to be annoying.

As to the comment inferring I am some consumer wanting $100 grand equipment for $1000 bucks bucks; (not an exact quote) this is not the case. I am very appreciative that one now can purchase a quality 24 track recorder for $2grand.

This is indeed amazing. And for the same $2K, you also get a digital editor, a mixer, effect processor, and even some musical instruments. But when you try to use all of that stuff simultaneoulsly, you suffer a speed hit. If you were to use the computer only as a recorder and take care of your effects and monitoring in the classic way, you'd get rid of latency. But that costs more money. Maybe not $100,000 (not everyone needs an SSL 900 console to control their DAW) but it does expand your $2K investment.

I just have found it very challenging [if not impossible in my area] to physically get a hands on demo of any of these devices. Which makes it harder to grasp how these things work.

This is a real problem. A few major cities have shops where they have a dedicated showroom for DAWs and a few knowledgeable people on staff who will spend an hour with you giving you a good tour. But your normal music store clerk knows little beyond brand names other than perhaps the particular softwar that he himself uses (if any). But this is really no different than when you needed to choose between an Ampex, Otari, or Studer recorder and an API, Neve, or maybe Soundcraft console. You couldn't really go into a shop and demo them effectively, you bought based on research, knowledge of others, budget, and to a certan extent, gut feeling. If you were in an "Ampex" market, you probalby wouldn't buy a Studer recorder because potential clients would ask for an Ampex. Today the equivalent is not buying Tracktion because the clients want ProTools. But if you have no clients, your choices are far more flexible - however that makes it harder because you have so many options.

Do some netwoking. Find other musicians in your area who are recording themselves, ask what they use, and offer to bring over a pizza, a six pack, and your guitar, and get a a demonstration of whatever he or she is using. It's not like booking studio time - most musicians are firendly. And there are SO many of them.

By the way, latency bothers me terribly. Thiis is why I use a dedicated recorder and console, and if I want to monitor with effects, I can easily patch those into just the monitor path and make final decisions later. With a digital recorder, there's still the time it takes for A/D and D/A conversion, but that's on the order of 1 msec round trip, and is usually tolerable. On the other hand, I've been accumulating gear for 40 years now, and I can assure you that I've invested well over $2K.

If I were to start out fresh today, I'd still use a mixing console (rather than a hardware control surface) and I'd use a multi-channel interface (as many channels as I wanted to mix tracks - 24 is enough for me, thank you) and keep my outboard processors and mic preamps. I'd use the DAW as a recorder, and occasionally an editor while tracking to clean up stray noises that are distracting, or to edit in the best take if we've done multiple takes. When mixing, I'd take advantage of plug-ins on the DAW if they're effective, or use outboards if that's what I want in the mix. But that's just me.

[DemoKing]

"and that I could monitor the DAW mix, with effects, while recording, through the 400F with "no latency."

You CAN monitor the DAW mix through the 400F with no latency. Playback is NOT the problem, there is no latency there because you dont hear or know it. Your latency will be ONLY the track you are recording, during the recording process. It will be fine on playback. While playing back your mix (which you'll hear no latency on) use the 400F to monitor any incoming signals with no latency. Problem solved. If you want the effects there already, like Mike said, use an outboard, but thats no way to go these days. If you get the best take ever, and the effect {censored}s it up later, too bad. And, any effect will cause latency someway or another, no matter how or where you put it.

[MikeRivers]

Originally posted by MikeRivers (that's me!) The really important specification is the one that manufacturers don't publish - how much noise comes out under realistic operating conditions. Actually, this is test data and not a specification - a spec is something that you start with and then design and build to meet it, or figure out a way to measure it so that it looks like you're meeting it.

A disgruntled former 400F owner over on the Mackie forum replaced it with an RME Fireface 800 and pointed me to the web page to compare features. While there, I found the data that I said was really useful that practically nobody publishes. Now I know someone who does, at least for the line inputs:

Input/Output level for 0 dBFS @ Hi Gain: +19 dBu
Input/Output level for 0 dBFS @ +4 dBu: +13 dBu
Input/Output level for 0 dBFS @ -10 dBV: +2 dBV

I don't know why they call the lowest sensitivity "Hi Gain" - it must mean something different in German.

[harryjames]

pardon my short term memory; but thank you to both King and Mike for your input and confirmations of what I *was* actually thinking. Maybe I initially posed my quest incorrectly; I am aware of no latency playback; when you spoke of using outboard gear the bells rang true for me. I already do have some outboard gear and I do wish to dabble and perhaps by dabbling with a software DAW I shall grow an addiction to the zillions of RTAS and VSTs that can be used in computer recording. As of now I have a much clearer picture as to what I am going put together; thanks to your *confirmations* which IS what I was thinking I just dont have the hands-on experience with ProTools and other other DAWs using plugins, and freezing tracks and all that to deal with CPU load and latency issues. I imagine it is actually a science in itself to get good at doing that. To me it's fair to say just as a good engineer on a mixing console can make all the difference in the world to someone's already recorded material, the same applies to those who dive into software DAWs. Even though you may have unlimited tracks, and tons of RAM, and resources are low, and you got tracks frozen and say you have a mix of 6 tracks to monitor while you record you next mono or stereo track, it takes thought and planning. I am sure you can just cut and paste your way into latency and CPU overload hell very easy when using plugins and soft synths. I've done a lot of soundclip editing using Soundforge and now that I wish to finally pour my songwriting into concrete I know I would prefer using outboard gear, a dedicated recorder and then use a DAW to create whatever I cannot create utilizing the other equipment. I already have an Alesis MIDIverbIII and MIDIVerb4 so I've decided to buy the Mackie 400F, a Yammie MOTIF ES and the AW2400 because 24 tracks is enough for me too. This should avoid latency pains and still give me the editing flexibility when I choose to utilize it. As per your comments about "thinking ahead", "planning" "making changes later" and "effects that are in concrete that dont work later in the entire mix"... I understand all of that. Actually I am an excellent arranger and have an uncanny sense of putting these thing together without ending up those problems during mixdown automation and mastering a burn. I know my personal story has veered off the forum topic which is, "the Mackie 400F" but then again, I am going to purchase one. No doubt it is a wonderful I/O unit and Mackie brilliantly constructed a quality piece of pro audio gear at a pricepoint that makes musicians and songwriters jump; or at least it sure did me. So again, I thank all of you and this forum for helping me to finalize my initial hardware purchases to set up my home studio which is actually in my 26ftRV. So my initial setup is going to be Mackie400F wired and racked with my 2 other RUs, Yamaha's AW2400, a pair of HS50M's, HS10Wsub, my Korgi3 and MOTIF ES-6, and even though Tracktion2 comes with the 400F, I'm gonna buy the standalone bundle version so I can doodle with all the plugins. Whereas the AW2400 has easy connectivity with my PC, I have a feeling I will be using the software DAW for some things and working mostly with my hardware and outboard gear. Funny thing; I found it was just easier to add $1000 into my budget than it was to try and decide which method to record would satisfy my needs the best. Thanks to AmericanMusicalSupply, I can buy all my stuff cash and get the AW on 5 payments of $200. So sometimes some creative financing helps a whole bunch. For now I will use my AudioTechnica phantom head mic and my new AT4040 and see if will actually really need a tube preamp for the AW's mic pre's. You guys are great. Thanks again!

[Pleasant]

Originally posted by MikeRivers It sounds like you'd be a good candidate for an outboard effect processor. To try to stay on topic here, this is what the Inserts on the 400F are for. If you have a device that gives you the vocal sound that you want to record, you can connect it in line between the mic preamp and the A/D converter via the Insert so that your vocal will be recorded with the effect. The disadvantage to this is that you're stuck with what you recorded.

There is an alternative. One could use the Inserts as a direct out by only inserting the patch cord (to the effect processor) partway into the Insert....you can read about this in the Mackie manual. This effectively splits the signal allowing it to pass both into the effect processor AND to the DAW. One could then route the output of the effect processor back into an input of the 400f, and monitor this signal with the heaphone monitor.

So, you get a wet monitor signal as well as a dry signal being recorded at the DAW.

I plan to do this with the 400f -- when I finally get around to buying it.

Have a good,

- P

[Anderton]

For this section, I spent some time checking out comments in other forums (Mackie, Gearslutz) about the 400F because I wanted to get some more background on some of the negatives that were touched on here. It was quite revealing.

Obviously, Mackie has already sold quite a few of these yet the number of people having problems seems relatively small. Of course, those with issues are most likely to post in forums, either because they’re upset or they’re looking for solutions (or both). Those who don’t have problems just go on their merry way making music. So let’s address the main problems that were mentioned, and my take on them.

Low input gain on Line Ins 5-8. I tested several electronic synths, and was able to kick the meters into the -10dB range on peaks. That’s not really all that bad – you’re basically throwing away less than 2 bits of resolution – but then again, that’s why the 400F has two instrument inputs for devices with high output impedances, and two additional inputs with preamps. Using these, I was able to peg the meters with just about anything, including a Zen portable player. Clearly, the line ins are optimized for typical pro studio gear (DAT machines, mixers, etc.) as opposed to stage gear, although I was able to pin the meters easily with the DigiTech GNX4.

Is the line in gain a problem? Maybe if you have four mics on drums and also want to run an electric piano and guitar processor designed to plug into guitar amps. But overall, I don’t find this is a big deal, given the inclusion of the four mic pres.

Problems with long headphone extension cords. I wasn’t able to duplicate this, nor find additional confirmation from others. The only explanation I can think of is that the headphone extension cord used by the person who mentioned this had so much cable capacitance that it somehow loaded down the output stage and created some kind of oscillation, but that’s a long shot. I wouldn’t worry about this one.

”Brittle” sound. Those hearing a negatively bright tonal quality seemed to be in a distinct minority. Although many people, myself included, have commented on the bright, clean high end, this is generally seen as a positive, not a negative. Personally, I like the 400F’s high end, as it’s very appropriate for most of the music I do. I can understand that some might like a “warmer” sound, but hey, that’s why different products exist.

High frequency whine. This is the big mystery. Some people hear an audible, annoying whine when using the 400F; most don’t, but that doesn’t help those who do. Even more frustratingly, there are reports of whines with other interfaces – again, not consistent – with Apple G4 PowerBooks mentioned frequently, but with other laptops as well. Some people might hear a whine with the 400F but not a different interface; others might hear a whine from another interface, but not the 400F. The one interface that seems to have none of these problems is the RME FireFace 800.

So what does the RME do that the others don’t? Beats me. But I suspect the whine issue reported with some 400F (and other) interfaces has less to do with the interface and more to do with the computer. I also feel this might be an instance where RME’s tendency to “overengineer,” and charge accordingly, makes up for marginal operation at the computer itself.

Over the years I’ve often speculated on why I have so few problems with my Windows software and peripherals. I install a lot of beta software, do a lot of uninstalls, and my registry probably looks like the inside of a serial killer’s brain. Yet it just keeps running. The only real problem I’ve had recently (although it was maddeningly hard to track down) was due to installing a new DVD drive that turned out to be defective.

But I always use computers that were integrated for music applications. My last two computers were designed by ace computer guru Pete Leoni and they have done yeoman service. Pete always said the most important part of the computer for good audio performance was the motherboard, and almost equally important was the way you handled graphics (e.g., AGP over PCI, simple board instead of hot rod game board). I’ve done most of the standard recommended software tweaks, but not even turned off all the unnecessary services and such. (I’m also still on SP1, which may be important: If you have SP2, you need to install a fix from the Microsoft web site for proper FireWire operation.)

I’d almost be willing to bet that some of the people experiencing “whine” problems are using on-board FireWire rather than from a plug-in card. And I’d also almost be willing to bet that the company making the computer tested the FireWire port with some digital camcorders and left it at that. I get the general feeling that on some PCs, FireWire is more of an afterthought than USB. Then again, it seems a lot of whine problems occur with G4 PowerBooks, so who knows – if anyone should know the ins and outs of FireWire, it would be Apple.

What’s even weirder is you’ll see someone say “Oh yeah, I had a whine like that but I disabled wireless networking and it went away.” So someone with the whine disables wireless networking, and it doesn’t go away. Welcome to the bleeding edge.

Bottom line: None of my research has indicated that anyone has a definitive answer as to why some devices whine and some don’t. Although it doesn’t seem that any whine shows up in the record path, I would suggest that if you’re concerned about the possibility of a whine situation, try before you buy and order from a company with a solid return policy. I never experienced any kind of whine under any of the circumstances people suggested, so obviously, I couldn’t try different things to get rid of it (although I did try different things to try to get it to happen, without success).

That’s pretty much it for negatives. Now let’s cover some other aspects.

[Anderton]

One aspect that doesn’t excite me too much is the stand-alone mixer option. Yes, the 400F is light, small, and solid; I could definitely see it as a keyboard mixer or other utility mixer (and there’s nothing wrong with a 64-bit audio path, either), but I’d get real nervous about not having some kind of real-time hardware control. That’s why it was interesting to see Mackie’s Dan Steinberg mention that Mackie is considering enabling operation with the Mackie Control Unit (MCU). Whether that actually happens or not is up in the air, but MCU control would add another element to the 400F and open it up for more possible applications.

 

Now let’s go to the undisputed positives. Clearly, the mic pres are exceptional, and even more so at this price point. They really do sound great – or rather, don’t sound great, because they really don’t “sound” at all. If the goal of a mic pre is to be a “straight wire with gain,” this is probably as close as you’ll get without spending a whole lot more. Prefer a creamy tube sound with a somewhat attenuated high end, coupled with a little warm midrange ringing from a quality transformer? That’s not the 400F. We’re talking clean, quiet, clear, transparent, detailed…great stuff.

 

And as I mentioned previously in this review, when used as guitar instrument inputs the preamps are within spitting distance of the Radial Labs’ JDL class A direct box, which to my ears is about as good as it gets.

 

But good mic pres aren’t worth much without good A/D and signal path. In one forum, in a blind shootout of two boutique A/Ds and Onyx A/Ds, a surprising number of people preferred the Onyx (which uses the 24-bit/192kHz AKM 5385 and 4358 phase-accurate audio converters) compared to far more expensive A/Ds. And the internal processing is 64-bit, floating point. As Cakewalk has been more than happy to point out since the release Sonar 5 with its 64-bit audio engine, this does make a difference.

 

I’m not sure that 192kHz is a huge selling point, but it’s there if you want it. I did test out 96kHz, and yes, it is a bit better than 44.1kHz. But the sound quality is good enough at the lower sample rates that in some cases you may stick with 44.1/48kHz to save a little space and let your computer work less hard.

 

The console software does the job, looks good (well yeah, that does matter) and the ASIO drivers are excellent. The WDM drivers work superbly with Adobe Audition and other apps that don’t try to tap into the depths of the WDM protocol, like the way Sonar does. In fact, I feel it’s essential to use Cakewalk’s Sonar and Project5 with ASIO instead of WDM, given that you have the choice to use either one.

 

The construction and overall build quality also get very high marks. Even the knobs are nice, and the drop-in power supply and FireWire boards simplify service for those elements most likely to cause any problems in the long term.

 

The icing on the cake is cross-platform operation, and the inclusion of Tracktion 2 – a fine program in its own right, along with a useful set of plug-ins – as part of the package. Couple that with a friendly yet reasonably detailed manual, a company forum that offers assistance and tolerates dissent, and the fact that updated drivers have already appeared, and I’d have to say it’s clear that Mackie is behind this product. Their participation in this Pro Review is, I believe, another indication that they sincerely want to connect with their customer base.

[Anderton]

The genesis of this review was at the Fall 2005 AES in New York, when I first saw the 400F. I was explaining the Pro Review concept to Mackie’s Kyle Ritland, and he immediately “got it.” He suggested doing the 400F. I was a bit skeptical that it would get a huge response; the previous Pro Reviews covered fairly complex products, whereas hey, what’s the deal with an interface? It’s just ins and outs.

Wrong. Clearly, the 400F got serious attention for what it offered given the price point. Participation in this review was strong from the very beginning, and got another shot in the arm when Mackie sent out an email to their user base touting the review. Credit to Mackie for that: The fun, edgy, and somewhat scary thing about Pro Reviews is that no one knows how they’re going to turn out. Mackie might have sent thousands of people here just as someone discovered that the 400F self-destructed if you turned all the knobs up to maximum. But I guess they had confidence in their product, and based on the reactions here, that confidence seems justified.

So what about the RME FireFace? Clearly, it sets the current standard against which FireWire interfaces are measured. I’ve reviewed it, and it’s an exceptional piece of gear. But it’s also very pricey. What the 400F does is come in with some serious sound quality, feature set, and construction chops at half the price. Granted, no one reported whines with the FireFace, so the extra bucks do buy you some added bullet-proofing. But I suspect that for most studios, the 400F will do everything they and need at a much lower entry price.

Finally, a Pro Review is nothing without the people who participate in it. I’d like to thank all of you for your comments, the high level and civility of discourse, and the useful insights. Kudos to Dan Steinberg for restraining his enthusiasm enough to avoid sounding like a marketing guy (I know it was tough, Dan!), and Kyle Ritland for getting behind and supporting the Pro Review concept in the first place. Also, a huge thank-you to Mike Rivers for adding his considerable expertise. I’ve always appreciated that he uses his knowledge to enlighten rather than intimidate.

This completes my official “review,” but does not mean the end of the thread: Feel free to keep discussing the 400F. As long as the thread has some steam, we’ll keep it going. After a while we’ll un-sticky it, but you’ll always be able to access it from the thread Links to All Existing Pro Reviews.

The PreSonus ADL 600 preamp will be the subject of our next Pro Review, which starts next week. It’ll be interesting, because we’re taking a 180 degree turn: This is a preamp that’s intended to color the sound rather than be transparent, but do so in a subjectively pleasing way. Have they succeeded? We’ll find out.

[Ninjä]

First of all. Thanks to all of you for participating in what must be the greatest review I've ever witnessed in my life.

I just recieved my unit and have been very pleased with it overall. The preamps are maybe the best I've heard on a "soundcard" and I definetly fall into the category of people who find the DA converter very crisp and clean indeed. Definetly not "too bright" for my taste.

I do however fall into the category of people falling victim to the high pitched whine from my firewire port. I've tried connecting the unit to both the FW400 and FW800 port on my mac and the problem persists both places. I've tried removing just about everything without the sound going away. I've also tried all the FW cables I own (some of which should be pretty high quality). I might add that the building I live in is fairly new and modern, and that such the power wiring is pretty state of the art for a normal apartmentbuilding. I also live in Norway where (and I say this without any form of prejudice) I believe that the quality of power seems to be higher quality than over seas (less spikes, and what have you).

I've tried a bunch of different firewire units and the 400F is the only one with this problem.

What strikes me though is that my mac is fairly "standard" (Apple G5 DP 1.8, 2GB ram, two uad-1 cards), and so any problems I might run into are probable to occur elsewhere. I also don't want to point fingers, but it does seem strange if this isn't a problem related to the Mackie interface as, as you all know Apple designed firewire and for them to not implement it correctly is highly unlikely. I do hope that Mackie will check into this scenario and that it might be possible to remedy it with a software update.

Thanks again.

[harryjames]

Please excuse me for being "vague" but with an M-Audio firewire product my PC's built-in 4pin firewire did the "whine" thing. i bought a PCMCIA 6pin firewire "with AC power" and using AC power the whine dissapeared. using the same PCMCIA firewire w/o the AC assist, the whine remained. apparently must have something to do with buss power or something. i am no electronics engineer by any means. As for the Mackie 400F I will be ordering mine from American Musical Supply next week. This forum review has been an excellent source of information. I applaud the creators and every one who has contributed here. I can't wait to sing through those clean Mackie mic pre's! ;;Merry Christmas;;

[MikeRivers]

Originally posted by Pleasant One could use the Inserts as a direct out by only inserting the patch cord (to the effect processor) partway into the Insert....you can read about this in the Mackie manual. This effectively splits the signal allowing it to pass both into the effect processor AND to the DAW. One could then route the output of the effect processor back into an input of the 400f, and monitor this signal with the heaphone monitor.

There are almost always are ways to work around limitations, as long as you don't run into other limitations - like suppose you're using all the inputs as real inputs?

 

By the way, while I know it says in the manual that you can do that, I don't think the people who wrote the manual ever tried it with the actual hardware. The jacks that Mackie used in the older compact mixer line (CR-1604 up through the VLZ Pro series) would do a pretty good job of holding a plug that was pushed in only as far as the ring contact. The jacks that they use on the new Onyx mixers will not. A "halfway-inserted" plug flops around in the jack and the ring contact doesn't quite reach down to the bottom of the "V" in the plug. It's a very unreliable connection. I suspect (but don't know) that the same jacks are used on the 400F.

 

The way to solve this problem is to build some cables with a TRS plug on the mixer end and connect the tip adn ring together. That way you can insert the plug all the way wihout breaking the channel signal path.

[MikeRivers]

Good wrap-up. Your observation (and I'll bet this isn't the first time you've recognized this) is very true - that "we, the people" hear more from users with problems than users who are happy (or don't realize yet that they have the same problems). Also, Mackie sells a lot more units than RME, so the number of RME participants are fewer - which means fewer reported problems as well as praise. Valid statistics on the "test group" are pretty hard to obtain.

Originally posted by Anderton Low input gain on Line Ins 5-8. I tested several electronic synths, and was able to kick the meters into the -10dB range on peaks. That’s not really all that bad – you’re basically throwing away less than 2 bits of resolution – but then again, that’s why the 400F has two instrument inputs for devices with high output impedances, and two additional inputs with preamps. Using these, I was able to peg the meters with just about anything

The sensitivity that you reported is about what I expected. But DAW people rarely havet VU meters that represent loudness fairly well, they watch peak reading meters to avoid digital clipping. A -10 dBFS eyeball average recording produces a waveform display that's barely a squiggle down the middle of the "track." This is perceived by many as "not a very hot recording" when compared to a graphic-filling ripped commercial CD that's been heavily limited, pushed to full scale peaks and has a much lower peak-to-average ratio than a raw studio track. And if one sets the playback volume based on that full scale track, a track that's 10 dB lower on peaks and has a lower average level than a commercial recording will sound pretty wimpy by comparison. Unfortunately too many people judge the quality of an audio device by how far they have to turn up the volume control. This is a perception that goes away with experience.

Putting this rant aside, I really think that, based on your experiments, the line inputs need more sensitivity. A low impedance line level output on a synth should be able to drive an audio interface to full scale. It's OK if you need to use maximum gain (or minimum attenuation) to do this, but you should be able to do it. On stage, we usually use a DI bewtween the instrument and the mixer and connect it to a mic input, but the DI's primary purpose in this application (as opposed ot connecting directly to a guitar pickup) is to provide a balanced run back to the mixer rather than an unbalancd run from the typical electornic instrument output.

Today many synths have balanced outputs, but we still use DIs on stage rather than using a balanced line level run to the mixer simply because our snake goes to mic inputs rather than line inputs. There's no need to do this in the studio, however, where you can run a balanced cable at line level, or even an unbalanced cable if it's an older synth, for any reasonable studio length. By using a DI and a mic input, you're providing the gain that Mackie should have provided on the line inputs, given the expected application of those inputs.

Interesting that the instrument inputs on the 400F having more gain than the line inputs. Since analog circuitry is something that Mackie designers do well, I would have expected that this would be common with the Onyx mixers. The instrument inputs on the mixers are unbalanced and have about a 1 megohm input impedance (an excellent match for instrument pickups) but they have practically the same gain as the line inputs. Although I've found that they have plenty of gain for the few electric instruments that I have around here, perhaps those less heavy-handed than I have wished for more gain so they provided it in the 400F. I think the instrument inputs on the Onyx mixers sound great, maybe even better than on my Great River preamp. It sounds like you were also impressed with their performance.

But . . . they're unbalanced. If you have a synth in the studio that has a balanced output, you lose the benefits of a balanced connection. For a short run and with no grounding problems, this won't matter. Still, they're so close to getting it right. Too bad they didn't go a little further.

I’d almost be willing to bet that some of the people experiencing “whine” problems are using on-board FireWire rather than from a plug-in card. And I’d also almost be willing to bet that the company making the computer tested the FireWire port with some digital camcorders and left it at that.

I think you're right about that. I think the only people who test computers with audio interfaces are guys like Pete Leoni who custom build computers for audio applications. Still, I think the whine problem is real and I'm glad to hear that Mackie was able to reproduce it and will give it some thought. It really does have the characteristics of a grounding problem, which can affect some external devices and not others.

I'm not excited about 192 kHz operation either. I haven't looked at the specs of the 400F, but I did notice that the noise performance of the Onyx Mixer+Firewire noise and distortion specs were slightly poorer at 96 kHz than at 48. We got over the "marketing bits" scare but now I think we're getting into a "marketing sample rate" scare. While there's no disputing that the sample rate is really 192 kHz (as opposed more than a bit or two beyond 15 being significant for the early 24-bit converters) there's no clear evidence that there's any sonic improvement (though there are some theoretical arguments, both pro and con). It sure sells more disk space, however, and more DVDs for backup.

[amplayer]

Originally posted by MikeRivers You pretty much answered your own question - they're two different measurements. EIN is actually a computed number and represents the ideal noise level contributed by the input before you add any gain. S/N is the difference between the maximum possible output level and the noise present under (presumably) known conditions. EIN is the measured noise output minus the gain, It's really only significant for the first input stage, but, for marketing reasons, seems to be getting attached to mic preamps (the whole box) because it's a nice low number. If the EIN is -129 dBu and you have 60 dB of gain (as measured by conventional means), you have 129-60 or -69 dBu noise just sitting there at the output with no input. That doesn't look very impressive, so they have to come up with a different way of saying it that doesn't look so dreadful. Add 20 dB of headroom to that and you have: a noise floor of -69 dBu and a maximum output of +20 dBu, giving a S/N ratio of 89 dB. Close enough? It really isn't all that simple because you're really comparing apples and oranges, but that's sort of the way it works. You can make the S/N ratio look better in most cases by simply reducing the gain because you're amplifiying the input noise less and the noise output isn't swamped by input noise. The really important specification is the one that manufacturers don't publish - how much noise comes out under realistic operating conditions. Actually, this is test data and not a specification - a spec is something that you start with and then design and build to meet it, or figure out a way to measure it so that it looks like you're meeting it.

Ok. This sort of makes sense.
Now a slightly different (possibly stupid) question: Since a whopping -129 dB EIN can only get you an 89 dB S/N because of the preamp gain, then what does it take to make 24-bit conversion necessary? That is, with a S/N of 89 dB, why bother with 24-bit conversion when clearly a 16 bit converter would be fine?

[MikeRivers]

Originally posted by amplayer Since a whopping -129 dB EIN can only get you an 89 dB S/N because of the preamp gain, then what does it take to make 24-bit conversion necessary? That is, with a S/N of 89 dB, why bother with 24-bit conversion when clearly a 16 bit converter would be fine?

That's a pretty good question, actually. I do almost all of my live recording, 2-track or multitrack, at 16-bit resolution and I don't seem to be suffering any great hardships from it, but I use 24-bit for more critical work.

First off, you don't always use all the gain of the preamp, so it's often possible to achieve a S/N ratio of better than 89 dB. You'll never get to the theoretical 144 dB or so, but decent analog circuitry ahead of good converters can achieve a usable dynamic range exceeding the 96 dB theoretical limit of 16-bit recording. I say "usable dynamic range" rather than S/N ratio becaue our ears and brains are pretty good at extracting intelligence below the noise floor. Of course you have to turn the listening volume up so that you can't listen to anything other than reverb tails and noise without blowing your speakers, but it's nice to know that there's something down there.

Another advantage to using 24-bit resolution is that you can allow more headroom without losing anything but volume (which you can make up later). It's no big deal (as Craig ponted out in his discussion of the 400F line input sensitivity) to record 10 dB below full scale. This means that when someone unexpectedly bellows into a mic, you have another 10 dB to go before your A/D converter clips.

Another advantage of increasing resolution is that when you go through the "make it louder" exercise, compressing the signal, and then boosting it, the low level information that you're boosting is quieter.

Or course when the final product has less than 5 dB of dynamic range and you listen in the car with the windows rolled down, there isn't a whole lot of point to high resolution recording.

[sekim]

Again, has anyone conducted a multi generation loopback recording test on this box? I'm interested in buying, but I'd like to see something that can QUANTIFY whether or not the D/A is misrepresenting anything. This test will prove whether the alleged brightness of the D/A is perception or reality, since *if misrepresentation is occuring* then the error will be *cumulative* accross the generations. It will become both obvious and *measureable* if the alleged problem actually exists. I just don't want to buy one and have to do this and possibly do a return if others own this box and can do it and let us all know how it came out.

[blue2blue]

Great write-up, Craig -- although I'm still absorbing it.

On the whine issue -- if it's at all similiar to the whine I experience under certian circumstances with my MOTU 828 mkII, I'd strongly recommend others so afflicted to the same kind of isolation testing that one might use to approach a more common audio ground loop issue.

In my case I get the impression -- but I'm no circuit whiz -- that some kind of ground loop issue seems to be causing 'contamination' of audio signals with sounds associated with activity on screen (mouse actions, browser windows changing.) But it only occurs with certain combinations of equipment connection -- just like a conventional ground loop.

My specific problem, which I outline somewhere above, is a whirring, whining noise when there is video/mouse activity -- but ONLY when both my MOTU and the laptop's internal speaker/headphone out are both hooked up to the same 'prosumer' amp/tuner (which has a bunch of other stuff plugged in -- including the output of a DVD player AND a digital cable box. The all DO resolve to the same outlet -- but it's a complex set of interconnections.

Anyhow, simply breaking the audio connection from the laptop headphone out OR the MOTU main outs to the amp fixes the whining/whirring which shows up (apparently) in the output of BOTH units... (maybe I should switch over to another sound source and see if the whirring contams ALL the signals on the amp...)

Anyhow, in my case, on my MOTU, it can be avoided. Others may not be so lucky

[mandoman]

So I've had more time to play with the unit on the road and in the studio.

I took the 400f over to my friends house and set it up to record some acoustic bluegrass in stereo. Unit is fairly portable: powerbook, external firewire hdd, 400f, cables, mics, stands, mandolin, and a six pack - just barely able to make it with only one trip from the car. Anymore gear would have been two trips.

Setup was a breeze, though I think I would prefer to get a faster 7200 internal drive and a bus powered unit to cut down on cables for this type of recording.

I typically use a pair of LDC's either omni spaced or M/S. This time I went for omni spaced plunked right in the middle of the jam. Recording into digital performer was a breeze. No hickups at all, at least I hope not, I only recorded two channels!

The resulting recording sounds pretty incredible. I compared this directly to a similarly recorded setup from a year or two ago recorded on the same mics into a presonus digimax/828 rig. The 400f recording is noticeably more clean, open, and clear sounding. Not to dog the presonus recording, that's good too, but there is definately an improvement with the 400f.

Oh, almost forgot to mention. At the end of the recording, I realized that I hadn't noticed any 'firewire whine'. At this point the fan on my powerbook was blasting away after a 3 hour recording session, but even so, I never once noticed a firewire whine. So, my conclusion with that is, the firewire whine is definately ground related. So, my suggestion to Mackie is for the next production revision, please work on your power supply grounding...

I also have played a bit with the unit in the studio. One of the first things I tried to do was hookup a Pod XT Live to the line ins. I barely got any signal hooking it up this way. I had to hook it up to one of the mic pres to get enough gain out of it. I also tried an hhb burnit - both through spdif and rca outs (9db out) - got full 0db from the spdif and only -10db from the rca. I have to say, I'm pretty dissapointed in the sensitivity on the line ins 5-8 - Mackie could have provided a little more gain on these ins. This probably won't be an issue for me since I usually never use more than 4 channels, but anyone out there that has any non pro (ie non 4db) gear to hookup and is already utilizing the 4 mics pres may be a little dissapointed. I know MOTU's products allow for 4db and -10db devices...

I tried guitar rig 2 with the 400f and was not able to get the best latency with even minimal effects (~60% cpu). I only got about 18ms. I can't knock Mackie though as my computer is pretty dated, a 667Mhz g4 10.3.9, and guitar rig is a cpu hog. I desperately need a new computer to get that latency down! Regardless, I hope they continue to improve their drivers to achieve lower latencies. I will say the front panal Hi-Z guitar in is a nice feature and sounds great with guitar rig despite the high latency.

With my cd player hooked up to the line ins, the eq really didn't sound quite right. I thought this might be the bright DA issue that others are complaining about so I decided to do a comparison. The four setting were:

1) CD track wav ripped on computer and played from quicktime over firewire
Same CD track played on HHB burnit and:
2) Monitored directly from HHB
3) Connected via 400f Spdif (clock set to spdif) and monitored through control mixer
4) Connected via line-in 5-6 and monitored through control mixer.

The best sound was 1 - clip from computer. I don't think the sound was too bright. It was very clear and 'crisp' sounding, but not too bright.

Next best sound was 2 - monitored directly from the HHB. This bit sounded a little darker than 1, but still pretty clean. Sound is pretty good considering these are pretty old DA converters.

Number 3s, spdif, sounded horrible at first - sounded like some digital clipping was going on. I had to turn down the software control mixer down about 5db to get rid of the clipping. Mackie, you need adjust the software control panel because spdif in should not clip period. With the gain turned down, the sound was ok, but not great. I thought it was a little muddy sounding.

The worst of the lot was 4, monitoring from line-in 5-6. Not too surprsing since you got DA (hhb) ADA (400f) going on. For casual listening the sound is ok, but for critical listening, it may not cut the mustered. Muddy with rough edges.

Conclusion? The DA itself sounds good - crisp and clean. However, the 64 bit digital mixer needs some work - it should NOT clip with a spdif input and it could sound better. Fortunately, this is the type of thing Mackie should be able to easily fix in software.

Control Panel software is as straightforward as it gets. Not sure what I'd improve except maybe two meters per channel, one showing the raw input level and the other showing the the output level. The lack of metering and ability to change settings on the unit itself is a bit of a bummer, but than again, I wouldn't want to clutter the unit too much. I think an ideal compromise would be to allow 2 to 4 different flash settings that could be saved and recalled without a computer. Some extra memory, a button, and couple of leds could have accomplished this.

I'll briefly mention the Tracktion 2 software. I opened it up, and it really seems alien at first. I like the concept - everything in plain view on a single page and low cpu overhead. I think the only thing keeping from using it is lack of Audio Unit support. I think this is on Mackie's list to eventually include. With AU and VST support, it has some serious pluggin potential. Get it to work with Garageband loops too (doubtful) and I would not hesitate to give Tracktion some serious attention.

Us Mac users will soon be going through a rough time with the transition to Intel processors. If Mackie comes out with native x86 OSX tracktion binaries early, that's another thing that could get me to start using tracktion, because I know MOTU is going to lag with Digital Performer in this regard.

Finally, I have to mention the overall feel of this unit. I love it! It feels nice and sturdy. All steel and aluminum baby! Two headphone jacks that sound great and are loud too, plus a seperate control room out. Very well thought out. This feels more like a nice analog unit, not something digital.

Overall conclusion? Well, pretty much the same as before I bought the unit.

The good:
Excellent sounding Mic Pre's and AD
Crisp and clean DA
Tracktion has potential
Great build quality
2 headphone outs and seperate C/R volume

The bad:
Grounding issues and firewire whine
Not enough gain on line ins 5-8

The ugly:
64 bit mixer needs more tweaking

I may still return this unit as the firewire whine really bugs the heck out me. I am getting a new computer soon and hope that will correct the issue, but probably after my return window closes with my dealer. So, I may have to return the unit. When I get my new computer, maybe demo a Traveler and 400f side by side to make my final decision? Another option is to just get a Traveler for portability/bus power/swiss army knife connectivity and pair it with an 800R for when I want the sweet pres. But that's an expensive dream...

I think Mackie has a real contender on their hands. Just replace that lame 'Spike' with a 200f - 2 onyx mic pres, 2 line ins, 2 spdif, 2 headphone out, compact bus powered firewire 6x6 interface, and they will have the whole market covered. I'd buy a 200f forsure, nod, wink, nudge ;-)

Thanks Craig for your wonderful review and to all involved for such a wonderful forum.

[MikeRivers]

Originally posted by sekim Again, has anyone conducted a multi generation loopback recording test on this box? I'm interested in buying, but I'd like to see something that can QUANTIFY whether or not the D/A is misrepresenting anything. This test will prove whether the alleged brightness of the D/A is perception or reality, since *if misrepresentation is occuring* then the error will be *cumulative* accross the generations.

While this is a relatively simple test to do at home, it doesn't represent anything that we do in real life. Multi-generation testing was improtant back when we had to bounce tracks, but that's no longer important.

 

I think that it's possible that someone does have a legitimate beef about the D/A cound of the 400F (or he might just be an anti-Mackie crank) and it's certainly reasonable that it sounds different from some other D/A converter that he has. They all sound different. What would be useful is not just to confirm that it sounds different than something else (no big surprise there) but that there's something about how it sounds that's unpleasant and that might (or might not) be able to be fixed.

 

It's easy to measure frequency response. If it has a high frequency rise or an irregularity (microphone manufacturers do this intentionaly to give their mic a characteristic sound) it would be easy to tell. Harmonic distortion is also easy to measure, and it's typically absurdly low with digital eqipment. So while these things can be measured, it almost doesn't matter unless there's something dreadfully wrong. I would expect that the analog output stage gets a bit ragged at very high and very low levels, but this isn't where we normally listen.

 

Mostly what makes digital devices sound unpleasant is the presence of things in the output that aren't related to the input. Intermodulation distortion can sound pretty nasty. In analog systems, we put in two frequencies and look at the output for frequencies that are related to the sum or difference of the input frequencies. But in a digital system, we don't have to put in two frequencies to have intermodulation - there's already lots of frequency-related stuff floating around that can interact with the intended audio. So you have to look at the full spectrum of what's coming out when you know what's going in.

 

Still, we're talking about pretty low ampltudes so it takes more than casual test equipment or procedures to detect, and more important, analyze what's going on. This is why you rarely see this sort of data in product reviews, particularly with MI products. Recycling the signal may make it more apparent that something's wrong, but it probalby won't point out what's wrong.

 

Just proving that it sounds different from another D/A converter is easy. Deciding if the difference is going to ruin your day (or your project) is personal and subjective. Because one person doesn't like it doesn't mean that it will be a problem for you. Someone on another forum said that the great George Massenburg "got screwed" (the poster's words, surely not George's) because he mixed a project monitoring through a Lavry D/A converter. Yet Lavry is one of the most respected manufacturers in the business and George would know if there was something wrong with his unit or setup.

 

Is this guy who loves his RME converter ever going to book your studio? Will George Massenburg? (George loves good music and will work with whatever meets the client's budget - if he has the time and inclination) What's most important is whether you're getting something you can work with and that you can afford. If the nature of the reported brittleness is such that you can't stand to listen for more than for short periods of time, then obviously it's not for you. But if the top end sounds a little brighter than what you're used to, you can get accustomed to that fairly quickly.

 

Remember, if you wait for the ideal product to come along and be validated by everyone, you'll never buy anything. This isn't going to be your last purchase. If you have a fair amount of experience already and are looking to expand your system, you can probably make a fair judgement with a couple of hours of listening.

 

It's up to you.

[snaildarter]

First - great review by all.
I'm currently torn between an Onyx Mixer with FW and the 400F.
Can anyone compare spec's and fundamental differences between the two interfaces and associated AD/DA chipsets?

Does it already have - or is Mackie considering upgrading the FW interface available for the Onyx mixers to the chipset used in the 400F?

[mandoman]

Originally posted by snaildarter First - great review by all. I'm currently torn between an Onyx Mixer with FW and the 400F. Can anyone compare spec's and fundamental differences between the two interfaces and associated AD/DA chipsets? Does it already have - or is Mackie considering upgrading the FW interface available for the Onyx mixers to the chipset used in the 400F?

 

 

The functional differences are detailed earlier in this thread, but not sure what the exact converters are used in the firewire mixer card. Would be nice to know.

 

I do have both 400f and 1640/onyx card and I'd say on the recording side, the results are very similar @ 24/44.1k, thanks in part to the the mic pres. Not so sure on the output side.

 

Mackie's recent free firewire card promotion on their 1220 mixers could be a sign they are clearing out inventory in time to bring out a new card. Winter NAMM is just around the corner...

[sekim]

Originally posted by MikeRivers
While this is a relatively simple test to do at home, it doesn't represent anything that we do in real life.

Processing with outboard gear and external summing, both of which require round trips thru the A/D/A, which is what this test checks, certainly represent "real life" to me, but to each his own I suppose... A definitive answer to the question is easily obtainable, too bad nobody is doing it to put the matter to bed once and for all...

[MikeRivers]

Originally posted by sekim Processing with outboard gear and external summing, both of which require round trips thru the A/D/A, which is what this test checks, certainly represent "real life" to me, but to each his own I suppose...

Well, OK, one trip, but the gear you're processing through is almost certainly worse than a round trip thorugh the converters, so why worry?

 

As far as external summing goes, you're only making a single pass out of the D/A converters (unless you're mixing back through the unit's A/D converter). Besides, most DAW external summing boxes cost more than a 400F, so you probably want to use a really, really good set of D/A converters (which the 400F's might actually be - I dunno) with it rather than a sub-$1K do-it-all box.

 

You can exaggerate any small fault to the point where it becomes objectionable, but as long as you don't do that, it's not a problem.

A definitive answer to the question is easily obtainable, too bad nobody is doing it to put the matter to bed once and for all...

What's easily obtainable? The result of a multi-pass test? What will that put to bed?

 

A good quantitative evaluation of the converters requires test equipment that apparently none of us has available. A couple of trade shows back, I was talking with the Prism folks, who make a pretty sophisticated and fairly expensive PC-based digital and analog test set. I remarked that I'd love to have one, but that I'd never make enough money with it to pay for it. The guy from Prism suggested that they might consider some sort of a long term loan program for reviewers, but I never heard back from him. Maybe I should bug them next time I see them. I'd be happy to test it if I had the unit to test and the equipment to test it with. Until then, I'll just have to trust those with the real world experience.

[MikeRivers]

Originally posted by snaildarter I'm currently torn between an Onyx Mixer with FW and the 400F. Can anyone compare spec's and fundamental differences between the two interfaces and associated AD/DA chipsets?

Forget specs, at least at the chip level, at least for now. Look at the functional differences. They're SO different that if you need one, you wouldn't have much use for the other.

 

If you're mixing live sound and want to bring home multitrack recordings of your shows that you can mix in your computer later, the Onyx mixer is the one to get. If you're working in a home project studio, the 400F is the one to get. If you do both, consider getting both, or the Onyx mixer without the Firewire card, and and some other interfacie for your computer that will give you more I/O routing flexibility than the Onyx mixer setup.

 

It's highly unlikely that Mackie will update the Onyx mixer Firewire card with the 400F converter chips because they're totally different. And for the purpuse to which it's most suited, there would be no great advantage to improviing it from the pretty darn good that it already is. At least that's my opinion.

[sekim]

Originally posted by MikeRivers What's easily obtainable? The result of a multi-pass test? What will that put to bed?

 

 

Yes, the results of a multi generation test.

 

If someone tells me "I couldn't measure or hear a significant difference after 6 generations", this puts the whole issue to bed as far as I'm concerned since this is typical performance for today's technology. But if someone said "after 6 generations it was +4dB @ 10K and +8dB @20K" then I'd know there is a problem.

 

Given the nature and interest level concerning this thread and the number of folks that own the box, I thought it was worth asking the question. If this isn't info that would help you in making a purchase decision that's totally fine, but it definitely helps me.

[MikeRivers]

Originally posted by sekim If someone tells me "I couldn't measure or hear a significant difference after 6 generations", this puts the whole issue to bed as far as I'm concerned since this is typical performance for today's technology. But if someone said "after 6 generations it was +4dB @ 10K and +8dB @20K" then I'd know there is a problem.

Yeah, so would I. On the other hand, if there was that sort of error build-up after six generations, it should be measurable at the first generation, and that's what I'd like to be able to measure. First, by doing that, you get a more accurate assessment of what the actual problem is, and scond, while you may be able to come up with some hypothetical cases, most people wouldn't be making six passes through the same path. While knowing the result of this test might make you comfortable, I don't think it would really matter to most other potential users,

 

If you read "Frequency response at 44.1 kHz sample rate was flat within 0.1 dB from 20 Hz to 5 kHz, rose to +0.3 dB at 10 kHz and 1 dB at 20 kHz before dropping to -36 dB at 21 kHz." would you be suspicious? Most people would be completely happy to know that the respnse was flat within 0.3 dB over the normal audio range. Now if the response continued to rise at higher sample rates, and was up 2 dB at 35 kHz, that might be reason for concern by some, but not all of us.

 

What would concern me more is if I found significant sidebands around a single frequency, or unexplained frequencies generated by the device in an intermodulation test.

Given the nature and interest level concerning this thread and the number of folks that own the box, I thought it was worth asking the question. If this isn't info that would help you in making a purchase decision that's totally fine, but it definitely helps me.

I hope you get a volunteer. Anyone?