MOTU Audio interfaces, any good?

[Leh173]

I'm looking at buying a MOTU 8pre or bigger interface than my Firestudio Mobile as I need more mic channels for recordings. I've been looking at MOTU for a while and seem good value and well built. Does anybody here use them and if so how do you find them?

[Griffinator]

MOTU makes excellent interfaces. They're not top of the line, but they don't charge top of the line prices for them either.

[NewImprov]

I use an 828 MK3 as my main studio interface, and the 8Pre as an expander, running on a Mac Pro and Logic 9. I also use the 8Pre as an interface when doing remote recording with my Windows laptop running Reaper. I'm very happy with both interfaces. My studio has a kind of arcane routing setup, and with all the inputs and outputs available on the 828, I can get everything going where I need it to. I totally got rid of my mixer about 6 months ago.

 

The 8Pre has surprisingly good mic preamps for the price, and they have enough gain for my ribbon mics. IMHO, the MOTU mic preamps are better than anything else in the price range, at least that I have tried, I have compared them to ART, M-Audio, and Presonus, and I like them better.

 

the only complaint I have is that the MOTU drivers do not play nicely with Windows XP. The drivers work fine with Windows if you don't intend to use any other interfaces, but I switch back and forth between the 8Pre for recording and an M-Audio interface for live gigs, and I have to uninstall the MOTU drivers or I get an error message every time I launch an audio application with the M-Audio attached. If I click through the error message, the other interface works fine, but it is an annoying extra step to take.

[zephonic]

Ya, my experience with MOTU on Windows is similarly frustrating.

 

But on Macs they reign supreme. They may or may not be the latest and greatest in conversion, but their drivers are second to none.

 

 

[Mark Schmieder]

Always good value, and rock solid on Macs. My bandmate uses them on Windows but the multiple-driver problem seems more an issue with Sonar (he has an old version) than the hardware per se. Sounds like the Windows guys here feel the blame lies with MOTU, so who knows?

 

You can also do the Black Lion mod to dramatically tighten up the clocking of your MOTU interfaces and improve their op amps.

 

I chose instead to upgrade to RME a few years ago and have never looked back, but it's all relative as RME is twice as expensive. I feel that MOTU offers the best quality and reliability for conversion (and even mic pre's, depending on the model) in their budget-to-mid-priced range. That have a quality "24 I/O" board at the higher end as well, that is truly pro-level. PCI-based.

[Leh173]

Thanks for the replies guys. I'm thinking about the 8 Pre and then maybe adding a 828 or a FF400 later on, as I need mic pres right now. I use my MBP with OSX 10.6.6 and Logic Pro. Do you guys get any noise across the audio outs from the computer down the firewire cable? I use my mac in Clamshell mode a lot and the presonus firestudio mobile picks up data noise it seems. I'm actually using the Fantom G and Firestudio as an Aggregate device and using the Fantom Gs audio outs to my powered monitors to get around it. Clean as a whistle, but the FG isn't bus powered. Even using the Presonus power supple I get the noise. Hopefully the MOTU and RMEs don't have this problem.

[Ashville.Guru]

I vaguely recall reading that the Windows driver issues were firewire-related, and that is why Motu came out with the 'hybrid' (USB/firewire) versions, which play well on Windows. Does anyone know if this is true?

[Mark Schmieder]

You should go USB at this point; depending on vendor. Not sure about MOTU, but RME's USB is now better than their Firewire in terms of latency etc. But that's because they do their own proprietary chip designs so have full control over the implementation, architecture, and protocol. They are geniuses. MOTU would have had to independently solve the problem of USB bandwidth and non-isochronicity their own way, but I wouldn't be surprised if they are also now better on USB vs. FW.

[Ashville.Guru]

You should go USB at this point; depending on vendor. Not sure about MOTU, but RME's USB is now better than their Firewire in terms of latency etc.

I don't know where you got that info from, but the Dawbench test results say exactly the opposite. Within the RME UFX, FW gives marginally better performance than USB. Both the UFX and the Fireface are way better than the USB-only Babyface. In fact, latency-wise, the Babyface is the laggard of the entire RME family.

 

Contrary to popular opinion, the Avid MBox Pro 3 (firewire) outperforms the much-glorified RME Babyface. Note that I'm strictly referring to latency here. There are many other yardsticks for interface performance. I only bring up the Dawbench results because you specifically mentioned latency, for which Dawbench is the only objective test (AFAIK).

 

The MOTU driver issues I was referring to in my post had nothing to do with latency. As I recall, it was something about conflict with other interfaces. In simple setups, the drivers worked fine on Windows 7, if I recall correctly.

[Mark Schmieder]

That's interesting; thanks for posting that chart.

 

I have spoken to RME reps at trade shows and they insist the USB is faster than FW (on a Mac) using their latest chip designs, during overdub sessions where one is going to be most concerned about latency (although this is also critical when recording original tracks one at a time using computer-based MIDI-triggered sounds, but I consider that to be an Overdub situation where there are no pre-existing tracks).

 

I'm not sure how many variables are different vs. that statement. The tests were done on Windows. I couldn't tell how many simultaneous tracks of playback and record; it's probably there if I take a lot longer to read the details (not during the work day!). RME folk are some of the most direct and honest in the business, so there is probably a difference in the context of their claim than the DAWbench test.

 

At any rate, the numbers are VERY close, and that itself shows that the advantage of FW is minimal these days and shrinking. Companies are not going to put continued resources into Firewire vs. other protocols. Also, Thunderbolt is likely to make the whole issue moot before long, as well as another protocol based on Ethernet. I can't really say more about that without doing a lot of checking first to verify what is public knowledge and what is under non-disclosure.

 

The point is that someone buying a NEW interface will get minimal benefit from FW vs. USB if they buy the right gear. And it is better not to have the same bus used for external data drives and audio interfaces. The inherent advantages of FW are FAR more critical in external hard drive applications than in audio input/output paths, so it is better to use FW for the drives and USB for the audio.

 

I now have a MacPro so don't have to worry about double-purposing my buses, but still noticed an improvement when I switched from FW to USB for audio. It was minor, but most noticeable on overdubs, where I care the most.

[Mark Schmieder]

Interesting; the older FF800 is about the same in the chart as the UFX; whereas RME tell me the UFX has the same latency as their PCIe solutions -- which is why I went the UFX route vs. continuing to fret over WHICH PCIe approach to day (RAYdat, HDSP, MADI, etc.) or whether to combine RME converters with Lynx I/O.

 

Amazing that MOTU's 24 I/O is still so near the top of the list! It's ancient by now, but has long been MOTU's best piece of gear.

[Mark Schmieder]

OK, I found enough work-related discussion over there to delve deeper, and drew some clear conclusions:

 

1. The measurements were carefully constructed to be real-world

2. During the period of the testing, RME improved their drivers so that FW is now slightly better than USB (not the case when the UFX was first released)

3. PCIe still has the edge due to not needing to over-buffer

4. Mac results are harder to separate factors but close to Windows results

 

The chart is thus very useful for first-time buyers, upgraders, and those adding additional channel count to their studios.

 

I am most shocked to see what a poor performer Focusrite is. I thought I was doing something "wrong", that my additional tracks (usually when recording drums with up to sixteen mics) brought in via ADAT from the OctoPre mk II, required some nudging to line up properly with the Fireface UFX tracks and/or had timing anomalies now and then.

 

I already hated that interface from the get go due to its dirty and grainy high-noise mic pre's and how it isn't even clean when going in with minimal gain just to use the ADAT converters for additional track count on the RME. I just haven't found a good replacement yet (I had the Behringer ADDA before and it was a total piece of crap sound quality-wise).

 

All the same, I can't currently justify the price differential of my current setup vs. either the MOTU 24 I/O route or the RME HDSPe-ADI8-QS route -- that's a LOT of serious dough!

 

For those on a budget, the charts clearly show the benefits of the mid-range and budget MOTU units, if the drivers are up to snuff on your platform of choice.

 

 

[Ashville.Guru]

Mark, it's possible that the RME reps were being honest with you, if a bit naive and uninformed. I've been struggling over the past couple of weeks to wrap my head around the Dawbench numbers and their interpretation, and I'm still learning things each day. The stuff is rather tricky, and it's possible for a team of clever, competent engineers take their eyes off the ball (giving them the benefit of the doubt here).

 

They are right about this - with RME devices, USB has lesser RTL (round trip latency) than firewire! The real interesting info is in the tables below the charts:

@128 buffer size, UFX FW:8.548ms, UFX USB: 7.959ms; Fireface: 9.002ms, Babyface: 8.481ms.

Similar trend at lower buffer sizes.

 

However, RTL alone gives an incomplete picture of performance. How much audio processing can be supported without pops and crackles is also crucial. The Dawbench test measures this ability by adding more and more plugins until pops and crackles start appearing. The RXC, CV and NCV columns refer to different categories of plugins. For the RME interfaces, check out the rows corresponding to 128 buffer size. Not much difference between firewire and USB. Now check out the 64 buffer size rows.

Huge difference! @64 buffer size, UFX firewire can handle ~30% more plugins than USB, without breaking a sweat. Similar trend with Fireface Vs. Babyface.

 

The Dawbench ratings take this into account, and return an average % for all the columns in those tables, as measured against the benchmark. But does it matter?

 

In many cases, yes. In the UFX case, firewire is much more likely to allow you to go from 128 to 64 samples, effectively bringing down RTL from 7.959ms (USB@128) to 5.646ms (FW@64)! The Dawbench ratings therefore are correct in bumping up FW against USB.

 

Unfortunately, there are still many interfaces which have not been tested by Dawbench standards, such as the new MOTU microbook II. It's important therefore to create awareness about the tests, and put pressure on manufacturers to submit their interfaces for testing. Here's a big thumbs-up to Vin Curigliano (may his tribe increase!) who's designed the Dawbench tests and also the Dawbench measurement utility. Without his considerable efforts, we would be utterly clueless as to the real latency performance of interfaces. And left to the mercy of manufacturer claims and subjective anecdotal forum reports.

[steadyb]

(I work for PreSonus)

 

The PreSonus AudioBox 1818VSL should give you the additional mic inputs you're looking to add, plus: latency of only 2.8 ms - 6 ms, analog in to out (round trip) -

24-bit / 96 kHz

 

Input/Output -

1 USB 2.0 port

2 combo XLR / ¼ mic/instrument inputs

6 combo XLR / ¼ balanced mic/line inputs

2 balanced ¼ main (L/R) line outputs

6 balanced ¼ auxiliary (L/R) line outputs

8 channels ADAT optical input and output (at 44.1 or 48 kHz; 4 channels at 88.2 or 96 kHz)

2 channels (stereo) S/PDIF coaxial input and output

1 stereo ¼ headphone output

1 BNC word-clock output

MIDI In/Out

[zephonic]

(I work for PreSonus)

 

The PreSonus AudioBox 1818VSL should give you the additional mic inputs you're looking to add, plus: latency of only 2.8 ms - 6 ms, analog in to out (round trip) -

24-bit / 96 kHz

 

At what buffer settings?

 

[Ashville.Guru]

Zephonic, I must thank you for being the first to mention the Dawbench standard on this forum...! Your post on the Workstation VST thread was the first time I heard about it. Totally changed the way I evaluated audio performance on computers, not just interfaces.

 

Steadyb, there's no easy way to say this - I'm going to be blunt, and I really hope you won't take this personally. You present a single metric as indication of the Presonus' latency performance. Totally unhelpful, and I'm just stopping short of calling it misleading.

 

There's RTL - both absolute value, and how it 'falls off' with decreasing buffer size - 256, 128, 64, (perhaps) 32 samples. At a given sampling frequency, as Zeph points out. Then there's the audio load handling capability - again, both absolute value, and how it falls off with decreasing buffer size. The latency performance of a given interface is a complex interplay of all these factors.. One simply cannot represent it by a single measurement. It needs to be an aggregate of several measurements, performed in a standardized manner, so that it can be compared objectively across different devices.

 

The Presonus Audiobox has been subject to the above rigorous standards. The verdict - it ranks third... in the bottom quantile. 20th overall out of 27 tested. ~44% of the top contender. If Presonus has improved performance with newer models/drivers, and is truly sincere about regaining lost credibility in the latency department, there's only one way to go about it. You need to submit a demo piece to http://www.dawbench.com for testing.

 

If you sense a tone of harshness in my post, it is born out of indignation and frustration. As consumers, we're kept in the dark regarding one of the most important quality yardsticks - latency . Totally impossible to gauge from manufacturer spec sheets. Even the bottom-feeders provide info on number of ins and outs, balanced vs. unbalanced, impedence, frequency response, etc.

 

Thats perhaps because analog audio has been around long enough for the industry to adopt audio specification standards. No one would want to buy from a manufacturer who didn't reveal the specs. Sadly, that hasn't happened with audio interfaces. And it's party time for the manufacturers! Without standards, every Tom, Dick and Harry can claim ultra-low latency on their product. And most do! Point out a manufacturer who doesn't use 'low latency' on their blurb, at least on their flagship models.

 

Naturally, many manufacturers aren't too enthusiastic about Dawbench. To quote Vin:

I have to admit I was surprised by the level of resistance I received from various angles during the last 12 months, a quick overview would be that some manufacturers / distros were more than happy to send me product to test and report on , others were not prepared to unless I assured them a positive result ???, while others simply didn't deliver on the promised test units after reading the reports.

 

Steadyb, I do appreciate your coming clean regarding your affiliation. I also appreciate that you've at least taken the trouble to provide some measurement (useless as it may be). But as a consumer, I have the right to demand that you put your money where your mouth is. Again, please don't take this personally.

 

Cheers,

 

Guru

 

P.S. Just wanted to reiterate that yes, latency is merely one of the yardsticks for evaluating an interface. It's also the one that is hardest for a consumer to gauge a priori.

[steadyb]

Nothing taken personally... at all. All good here on my end, Guru.

 

(I am not a software/hardware engineer, nor do I write driver code)

 

I'll see if I can come up with/find out a specific answer to you question(s). I'll ask around.

 

When you mention the AudioBox, is that the original or the VSL models?

 

 

[Ashville.Guru]

Thanks, man.

 

The unit that was tested was a Presonus Audiobox 44VSL with driver version 1.47 (as reported here). There was also a StudioLive 16.4.2 with driver version 3.5.5, which (understandably) fared much worse.

[Tom Fiala]

Happy with my 896HD. Working fine with Cubase 6.5 and Windows

[Mark Schmieder]

That's all useful info -- I have only gotten crackles or hiccups OCCASIONALLY with the UFX, when doing overdubs at lowest buffer setting against HUGE track counts that I haven't pre-bounced to a reference submix or stems yet.

 

Now I have valuable information towards possibly switching back to FW -- at least for overdubs. I was using FW800 with the FF800, which is partly why I went to USB with the UFX as they no longer support FW800 (RME folk have told me many times that FW800 didn't pan out so well reliability-wise -- I think because the industry never quite got behind it so it lagged in development).

 

Maybe the moral of this story though, is that people in the market for a new interface might be best off buying a hybrid interface like one of the new ones from RME or MORU, or any other manufacturer that has come out with a hybrid interface (I haven't been watching M-Audio, PreSonus, etc. of late).

 

And yes, it is possible that the RME folk have a certain set of criteria and metrics they use to determine things that may not tell as complete of a story as that Dawbench study. But I can guarantee you they are 100% honest. Some of my co-workers occasionally deal with them also, due to MADI (where they are still fairly unique in the industry) and have likewise found them 100% transparent compared to most other companies.

 

You'd be surprised how many differences of opinion there are in the audio industry. And many of the players are eager to improve their test methodologies and extend their measurement criteria, as after all it helps towards building better and more competitive products.

 

It's too bad Lynx haven't stayed on the ball a much as RME and MOTU, as they make great gear but compatibility (even including cable connectivity!) is an ongoing concern and can slow people down in setting up a properly working system (look at all the GS threads on trying to find compatible pin-outs for connecting to other gear via the D-Subs; not to mention driver out-of-date-ness).

 

I am hoping with the release of the excellent Lynx HiLo recently, that they will use what they did for that model to retrofit their entire line to be more up-to-date so that more people can take advantage of their excellent interface technology.

[Throbert]

aurora & ASE-16e pinouts Redco makes custom cables for lynx products

[Mark Schmieder]

Good to know; thanks. The few people on GS who tried to take care of this on their own ended up with unexpected problems. Looks like there's a trouble-free way to deal with the D-Sub connectivity now between Yamaha-compliant and Tascam-compliant devices.

[Throbert]

Lynx has the same digital pin out as some of yamahas products such as... Their analog pinout is the same as the standard Tascam/yamaha pinouts http://sphotos.xx.fbcdn.net/hphotos-ash3/560607_146632132128990_100003466126272_169532_1274743358_n.jpg