A/D/A converters - help me cut through the BS

well admittedly this is not really a DIY-type post and I'm not sure which sub-forum it belongs in, but I value the input of the folks here, so here it goes.  I'm shopping for a  new multi-channel A/D/A converter for my DAW.  As I'm sure you know, the web is full of confusing, contradictory, vague, and subjective information about converters and all the voodoo they contain. 

What I'm wondering, from the folks who have more intimate knowledge of converter design thank I do, is what specs are worth paying attention to, and what meaningful measurements can I do on my own (using modest test equipment) to test the "quality" of a unit.  Is it as simple as doing a loopback test using something like RMAA to analyze the results?  Is there much else I could do without some more heavy duty test equipment?

Any advice here would be appreciated, it's becoming very easy to overthink this decision!

Matt C said:

well admittedly this is not really a DIY-type post and I'm not sure which sub-forum it belongs in, but I value the input of the folks here, so here it goes.  I'm shopping for a  new multi-channel A/D/A converter for my DAW.  As I'm sure you know, the web is full of confusing, contradictory, vague, and subjective information about converters and all the voodoo they contain. 

Click to expand...

I would suggest you read the Dan Lavry papers, in particular
http://www.lavryengineering.com/pdfs/lavry-white-paper-the_optimal_sample_rate_for_quality_audio.pdf

  Is it as simple as doing a loopback test using something like RMAA to analyze the results?

Click to expand...

In most cases, you wouldn't see much. The typical levels of frequency response error, distortion and noise are so low they don't constitute an element of choice. Very recently, I found a sound card that had respectable published data turned out to have unforgivable frequency response errors, though, but it's really an exception. The amplitude and phase response in the transition band (0.45 to 0.55Fs) is a very important factor, particularly if you consider working at Single-Speed (44.1/48k).  Check how a 23kHz tone is filtered, because it will produce an audible image frequency.
As you mentioned, there is a lot of talk, and in particular a lot of expectations about the possible differences, when in fact most of the times they are not clearly audible.

IMO published specs are all but meaningless with the exception of a handful of manufacturers.

As for measuring yourself I agree with Abbey, I've got a pretty decent bench and have spent time measuring quite a few commercial AD/DA plus many evaluation boards.  Getting good measurements is not trivial.

What is your budget and channel requirement?

ruairioflaherty said:

What is your budget and channel requirement?

Click to expand...

I need 24 channels of I/O. Budget is sort of nebulous, I could spend a few thousand dollars if I honestly thought it was justified, but I'd rather not.  The unit I'm most likely going to get is the MOTU 24 i/o.  In terms of features it's exactly what I want, and is very affordable.  But there's always that voice in the back of my mind wondering whether I'm making a serious compromise in audio quality compared to the newer, "fancier", way more expensive units made today.

I'll read the Lavry paper mentioned above. 

It seems like I should probably just stop worrying about it and buy the MOTU unit.  I've used a few different converters over the years and always thought they all sounded fine, but never had the chance to do any side-by-side comparisons.  I was never able to credit/blame any converter for a particularly good/bad sound in the past, which makes me think I've got much bigger fish to fry before I start worrying about the minute differences between different units.

I'd be interested in hearing opinions too.

I was enlightened on one thread about the sound of different tubes and how it's really a matter of the shortcomings of a design that is revealing what these differences are more or less.

As of now, I'm a huge fan of the Lavry sound. The Myteks are really nice too. But, are they pulling some trickery among some poor design to make me think they sound as good as they do and I'm just a fan of this because it's all I've listened to in the context of a bunch of units that are more budget friendly?

Matt C said:

I've used a few different converters over the years and always thought they all sounded fine, but never had the chance to do any side-by-side comparisons.  I was never able to credit/blame any converter for a particularly good/bad sound in the past, which makes me think I've got much bigger fish to fry before I start worrying about the minute differences between different units.

Click to expand...

You really should compare if you can....

Although I'm not familiar with the newer technology and the size of the quality gap between units costing less than the pricier ones.....

It was a large gap not too many years ago

Just my opinion....

My favorite test is to listen to tape hiss through an AD/DA loop. Its a torture test. The best sounding hiss always wins with music. I find it is also important to use an analog source that hasn't been through an A/D conversion. The nyquist filters all but eliminate the difference between converters if you are just looping stuff back.

scott2000 said:

As of now, I'm a huge fan of the Lavry sound. The Myteks are really nice too. But, are they pulling some trickery among some poor design to make me think they sound as good as they do ?

Click to expand...

This may have been more or less the case years ago, when getting actual performance was hard and costly, but today, conversion chips offer objectively excellent performance, requiring only thorough implementation, which is IMO what makes the difference between brands.

Another thing to consider is the interface to the computer,  USB,  pcie, thunderbolt,  etc.  Having something reliable with solid drivers,  that works with your software  is just as important as the converters themselves.

On the converter front I'm currently using two Lynx Aurora 16. I tested a few in that price range and the Lynx had the lowest distortion,  which for me was a plus since it can double as a front end for a test rig.  The new version recently came out,  which I would imagine is even better,  but it also means your can get good deals right now on the older ones.

abbey road d enfer said:

This may have been more or less the case years ago, when getting actual performance was hard and costly, but today, conversion chips offer objectively excellent performance, requiring only thorough implementation, which is IMO what makes the difference between brands.

Click to expand...

I was somewhat thinking this. And I think the topic points to this in some way. So, are there any opinions on units that have proper implementation that are flying under the radar or any that are phoolishly being used?

I'm guessing schematics will be non existent as these are newer devices we are talking about so details of the design won't be readily available.

Interesting to hear any experiences.....

I have heard many positives on  Lynx even from it's earlier offerings so, it seems like there is some truth to the hype surrounding this company with the distortion testing mentioned above being superior to others in it's price range????

abbey road d enfer said:

This may have been more or less the case years ago, when getting actual performance was hard and costly, but today, conversion chips offer objectively excellent performance, requiring only thorough implementation, which is IMO what makes the difference between brands.

Click to expand...

I guess this is why I'm a little hesitant to buy the MOTU unit - it's pretty old (came out over ten years ago I think), so it might belong to an era in which converter technology just wasn't as mature as it is now.

Matt C said:

I guess this is why I'm a little hesitant to buy the MOTU unit - it's pretty old (came out over ten years ago I think), so it might belong to an era in which converter technology just wasn't as mature as it is now.

Click to expand...

I  agree. Unless the sound of inferior conversion is a sound you are after or becomes fashionable as an effect, I would look for something that won't change what you are supposed to hear or what you are trying to record. But, it all comes down to results at the end of the day I suppose....

But I can't see how you would ever go back to inferior conversion once you've compared it to better..... It's really just an effect or a necessary convenience at that point//////

Like Bob Katz said....Good digital ain't cheap....And cheap digital ain't good.....or something like that...... But that was a while ago and that MOTU was a while ago.....??? Or maybe not....... The Rosetta 200 has a pretty nice sound (effect??) imo and it's pretty old...... The Lavry's and Myteks have units that old that sound really fantastic......

Guess you'd need to find out by comparing..... differences in implementation then too I'm sure....

scott2000 said:

But, it all comes down to results at the end of the day I suppose....

The Lavry's and Myteks have units that old that sound really fantastic......

Guess you'd need to find out by comparing..... differences in implementation then too I'm sure....

Click to expand...

Yeah this is the whole gray area I'm trying to navigate through by asking about meaningful specs, etc. It's easy for me as a relatively naive consumer to look at something like the Lynx units and think "well it's newer and it's 10x more expensive so it must be WAY better".  But obviously that's foolish.  I've actually used both the MOTU 24io and the Lynx Aurora before and both sounded fine, it's possible the Lynx sounded better but it was never, like "wow everything used to sound awful and now everything sounds amazing". 

Side by side listening tests would be great but it's not always easy to arrange that.  Hence me wondering about any concrete specs that would justify the huge price tag of newer high-end converters. So I don't mean to ask for specific model recommendations so much as ideas for criteria on which to judge these things. 

Paul your tip about the tape hiss is good, I'll make sure to try that.

Matt C said:

Yeah this is the whole gray area I'm trying to navigate through by asking about meaningful specs, etc.

Click to expand...

Yeah....

Matt C said:

  I've actually used both the MOTU 24io and the Lynx Aurora before and both sounded fine, it's possible the Lynx sounded better but it was never, like "wow everything used to sound awful and now everything sounds amazing". 

Click to expand...

Very interesting and it "may" point to the close in the quality gap being mentioned.

Never heard either .....

Matt C said:

So I don't mean to ask for specific model recommendations so much as ideas for criteria on which to judge these things. 

Click to expand...

I would love to hear more of this too. 

Tightly level matched comparisons are the only way, there is no spec sheet that will tell you how the distortion characteristics of a given converter will sound.

I've done a lot of those comparisons on high end converters, less in the range you are talking about but I have still heard most things out there at some point or another.  If you are recording music for fun I would not worry about it too much.

Connectivity, a modicum of future proofing, stability all matter as some one else mentioned.

I have strong opinions on certain brands and units but I don't share as I still do some work for Prism Sound.

I’ll just say that when I had my 192s for 10-years, I longed for better converters, but still did the best I could with them. Nowadays with the new Avid converters (along with the Iikes of Lynx and  plenty others), it make me careless about better converters. Not to say that there aren’t better ones I’d love to have, but honestly I’d rather more great gear, a bunch more cool accessories, even better acoustics, and simply way more space. If that was ever satisfied, sure I’d look at better converters.

I did get to hear direct Burl and Avid comparisons at Fantasy Studios. A difference? Yes. And I still feel the same way.

I wouldn't overthink this... Modern convertors are good enough that they can deliver excellent analog specifications.. If they deliver good analog performance the digital specs are superfluous.

JR 

Hi all

I'm still using the old Motu PCIe ring, 24io + 2408(mk3).
Nobody never complain about the sound coming out of the studio, nor grin looking at the old convertor
I mix all analog, meaning there is always extra conversion, mutli out to stereo in.

So for the OP, if you can put hand on 24io for low price (which is common today), just try it...

They also have an interesting spec for those using syth and CV, all motu product are DC coupled 

Best
Zam

zamproject said:

They also have an interesting spec for those using syth and CV, all motu product are DC coupled 

Click to expand...

Really? Aren't they using sigma/delta conversion?

abbey road d enfer said:

Really? Aren't they using sigma/delta conversion?

Click to expand...

Hi Abbey
I don't know, I'm not expert at all with converter technologies and underlying consequences.
To my mind it's just that they don't put any caps in the signal chain from DA chips to jack out, but maybe I'm wrong ?
I just know that I can drive my juno60 arpegiator clock directly from my motu converter.
By hand drawing a wave form (in waveform editor) which is a 0dbfs DC signal, then apply a pattern gate plugin or DAW fader automation (square), triggering the arpeggio steps. I do this since years... later Motu offer a dedicated plugin to shape CV, LFO or any DC signal directly from they DA converter to any analog syth.
Best
Zam