Quantcast
D/A/D Copy : How close should it be? - Gearspace.com
The No.1 Website for Pro Audio
D/A/D Copy : How close should it be?
Old 28th January 2009
  #1
Gear Addict
 
NotVeryLoud's Avatar
 
🎧 10 years
D/A/D Copy : How close should it be?

Maybe this is a really dumb question, but . . .

I just installed a DA*824 card in my Lavry Blue, giving me 2ch of A/D and D/A. After calibrating everything, I played a music track out of my DAW into the Lavry and recorded it back into my DAW.

DAW-->Ensemble SPDIF-->DA824-->AD824-->Ensemble SPDIF-->DAW.

Because there's a stage of A/D in there, I know I'm no longer getting an exact copy coming back in, but I wanted to hear how different the copy was, so I inverted its phase and summed it with the original.

Here comes the potentially dumb part . . .
I was a little surprised at how little cancellation there was when I summed everything. Taking latency into account, I even spent some time nudging the copy by individual samples, to see if I could get it to line up with the original and cause greater cancellation. It never happened. I got some great phasing effects, but no severe broad spectrum cancellation.

Is this what one would expect? Have I simply discovered the obvious?
Is the A/D stage changing things that much? Even in a unit as transparent as the Lavry?
Old 28th January 2009
  #2
Lives for gear
 
🎧 10 years
The coloration may as well be from the DA if there is one.

Did you calibrate the levels?

When you loop a signal as you did there may be level difference of a tenth or a couple of tenths of a dB on the resulting file and that will effect the nulling.

Quote:
Even in a unit as transparent as the Lavry?
How do you know that it is transparent?


/Peter
Old 28th January 2009
  #3
Gear Guru
 
Ethan Winer's Avatar
 
🎧 15 years
Quote:
Originally Posted by NotVeryLoud ➑️
I wanted to hear how different the copy was, so I inverted its phase and summed it with the original.
That doesn't always work, as you have discovered. Phase shift, which is otherwise inaudible, "rotates" the wave form as viewed on an oscilloscope or in a DAW. So the frequency content and distortion after A/D/A may be extremely close to what you started with, but nulling is no longer possible. I have heard that the DiffMaker program (freeware, Google it) can take this into account, but I can't confirm this.

--Ethan
Old 28th January 2009 | Show parent
  #4
Lives for gear
 
🎧 10 years
You can always perform a loop measurement which will show you not only phase but also frequency response, noise and distortion.

Please do that as I would be very intrested to see this performance. RMAA is very easy to use and automatically put the results in a HTML report.

Also it would be interesting if you could loop some short pieces (30s or so) music and post original + loop so one can listen to the effect of the DA + AD and compare to the original.


/Peter
Old 28th January 2009 | Show parent
  #5
Gear Guru
 
Ethan Winer's Avatar
 
🎧 15 years
Quote:
Originally Posted by Audiop ➑️
it would be interesting if you could loop some short pieces (30s or so) music and post original + loop so one can listen to the effect of the DA + AD and compare to the original.
Agreed. And doing multiple passes lets you quickly assess the degradation. I once did that 20 times with a cheap sound card, and by pass 5 you could hear the change. One pass was pretty good though! A high-end converter should be able to do many more passes with minimal degradation.

--Ethan
Old 28th January 2009
  #6
Gear Addict
 
The Reel Thing's Avatar
 
🎧 15 years
in an A/B situation you can hear the degradation of the signal in the first copy generation, no matter which converter.
if you can't, you need different monitors, or you need to train your ears.

in the 90s when DAT came out, everybody said an SP/DIF or AES copy wouldn't change the signal. it was a lie back then, and it is a lie today.
it would be against the laws of physics.
stupid thing is, good ears heard it all along, just nobody wanted to hear it, because digital was the new golden calf.
nothing has changed.

tom

analoghaus :: studio label verlag - home
Old 28th January 2009 | Show parent
  #7
Gear Guru
 
theblue1's Avatar
 
🎧 15 years
Quote:
Originally Posted by The Reel Thing ➑️
in an A/B situation you can hear the degradation of the signal in the first copy generation, no matter which converter.
if you can't, you need different monitors, or you need to train your ears.


in the 90s when DAT came out, everybody said an SP/DIF or AES copy wouldn't change the signal. it was a lie back then, and it is a lie today.
it would be against the laws of physics.
stupid thing is, good ears heard it all along, just nobody wanted to hear it, because digital was the new golden calf.
nothing has changed.


tom

analoghaus :: studio label verlag - home
DAT, like CD-Audio, is engineered with an orientation to correction and recovery. If there are uncorrectable errors, recovery is the next option, followed by momentary failure.

By contrast, storage and copying of data-oriented storage formats like AIFF and WAV are processes designed with no tolerance for data inaccuracy.

If your Mac or PC can't copy a file with 100% accuracy, it is supposed to fail the process and throw an error message.

But your CD player or DAT is designed to continue playing back at all costs... as a consequence, you would tend to have a lot of errors before you would begin to hear them -- unless you keep track of your DAT's error statistics. (Consumer DAT machines didn't tend to make that display 'publicly' accessible. But any serious 'pro' DAT would have some form of error counting.)


The problem was not so much either the S/PDIF or AES/EBU transmission protocols but the accuracy and durability of the storage format/medium.

Let's face it, the heliscan head was a thing of its time.

VCRs, DATs and a few other high bandwidth analog and/or digital formats that used heliscan head tape drives helped define the technology of a decade or two of the late 20th century -- but like any technology which has been exteded to the point of breaking -- the heliscan head had as many weaknesses as it had strengths.
Old 29th January 2009 | Show parent
  #8
Gear Addict
 
NotVeryLoud's Avatar
 
🎧 10 years
Great responses, thank you.

Quote:
Originally Posted by Audiop ➑️
How do you know that it is transparent?
I was just referring to the Lavry's reputation as a very "uncolored" converter, unlike some other brands that may have a certain sonic character.

Quote:
Originally Posted by Audiop ➑️
You can always perform a loop measurement
Also it would be interesting if you could loop some short pieces (30s or so) music and post original + loop so one can listen to the effect of the DA + AD and compare to the original.
I double checked my level calibrations and ran another test. I've posted two clips. One is the original imported CD track, and the other is the D/A/D copy, using the chain described in my initial post. I kept the transfer at 16/44.1 throughout, locked to the Lavry's word clock.

I'll also check out RMAA and DiffMaker.
Attached Files

ItsMyLife.wav (5.22 MB, 583 views)

ItsMyLife_DAD_Copy.wav (5.22 MB, 136 views)

Old 29th January 2009 | Show parent
  #9
Lives for gear
 
🎧 10 years
Thanks for posting the files.

I thought the looped file was a tiny bit less resolved and maybe a tiny bit rolled off in the high range but I'm not sure. I only scored 7/10 in ABX which is to close to chance to be really significant. I'll take another listen later.

There is a 0.07dB level difference that possibly could affect what I hear.

Obviously very good performance and I'm looking forward to measurements and possibly some more test material.




/Peter
Old 30th January 2009 | Show parent
  #10
Gear Addict
 
NotVeryLoud's Avatar
 
🎧 10 years
Quote:
Originally Posted by Audiop ➑️
RMAA is very easy to use and automatically put the results in a HTML report.
Unfortunately, RMAA is PC only, and my audio hardware is connected to my MAC. I'd be interested in finding equivalent software for the MAC. Any suggestions?

Quote:
Originally Posted by Audiop ➑️
I thought the looped file was a tiny bit less resolved and maybe a tiny bit rolled off in the high range but I'm not sure. I only scored 7/10 in ABX which is to close to chance to be really significant. I'll take another listen later.
I can hear a very subtle difference between the two files as well. I prefer the original version, but it's really splitting hairs when comparing the two.
Old 30th January 2009 | Show parent
  #11
Lives for gear
 
🎧 10 years
Sorry I don't know much about software for Mac.


/Peter
Old 10th February 2009 | Show parent
  #12
Lives for gear
 
🎧 10 years
Hi!

I made a loop of the same tune with a Lynx Aurora 8 and Line Audio 8MP. I guess it could be interesting for some people to compare these tests.


Test Line Audio 8MP + Aurora 8



/Peter
Old 8th April 2009 | Show parent
  #13
Gear Addict
 
NotVeryLoud's Avatar
 
🎧 10 years
Very interesting test. The differences are so subtle. At first it sounded like there was slightly less bass in the Aurora clip, but then it all just started to sound very similar.

Thanks for posting!
πŸ“ Reply

Similar Threads

Thread / Thread Starter Replies / Views Last Post
replies: 1788 views: 732029
Avatar for tourtelot
tourtelot 2 weeks ago
replies: 2 views: 679
Avatar for Almy
Almy 23rd January 2019
Post Reply

Welcome to the Gearspace Pro Audio Community!

Registration benefits include:
  • The ability to reply to and create new discussions
  • Access to members-only giveaways & competitions
  • Interact with VIP industry experts in our guest Q&As
  • Access to members-only sub forum discussions
  • Access to members-only Chat Room
  • Get INSTANT ACCESS to the world's best private pro audio Classifieds for only USD $20/year
  • Promote your eBay auctions and Reverb.com listings for free
  • Remove this message!
You need an account to post a reply. Create a username and password below and an account will be created and your post entered.


 
 
Slide to join now Processing…

Forum Jump
Forum Jump