Become a Site Supporter and Never see Ads again!

Author Topic: ratshack digital signal converter testing :)  (Read 22719 times)

0 Members and 1 Guest are viewing this topic.

Offline bhtoque

  • Trade Count: (3)
  • Needs to get out more...
  • *****
  • Posts: 3308
  • Gender: Male
  • So much music, so little time.
Re: ratshack digital signal converter testing :)
« Reply #45 on: September 07, 2004, 12:14:37 AM »
Ok, I figured I'd do the testing on this, and I've opened another can of worms.

Here's what I did.

I recorded a cd track from my Sony C701ES>optical cable>jb3 (did this twice) and
Sony C701ES>optical cable>ratshack converter>optical cable>jb3

Synchro was on on the jb3 so the files should all have the same starting point.

I then used cool edit pro to make all 3 files the same length. The file size matches too. If I invert-mix any of the 3 they cancel themselves out like they should, but the inverse of one does not cancel out either of the other 2.

If I put the two straight cd>jb3 files in multitrack view they match up.

I was willing to accept that the rs box was not bit perfect, but shouldn't the two cd>jb3 files match? Does this mean the jb3 is not bit-perfect? ???

I can upload the files if someone has space.

JAson
MG 200/210/270
AKG c422>s42>Hydra silver interconnects
AKG 391/92/93>MK 90/3 actives
>AM Hyper-Conductors
Studio Projects LSD2>MiAGi II
>Edirol R-4 (Oade T Mod)

db.etree.org/bhtoque

Offline Brian Skalinder

  • Complaint Dept.
  • Trade Count: (28)
  • Needs to get out more...
  • *****
  • Posts: 18868
  • Gender: Male
Re: ratshack digital signal converter testing :)
« Reply #46 on: September 07, 2004, 07:46:00 AM »
Try EACs WAV Compare function with the files you have.

Theoretically the two CD > JB3 files should match, but that assumes [1] the C701ES doesn't have a resampling output and [2] provides a signal to the JB3 that the JB3 likes, so to speak.

Also, I don't see in your chain where you're actually converting the digital signal with the RS box:  C701ES > optical > RS box > optical > JB3.  Can you run C701ES > coax > RS box > optical > JB3?
Milab VM-44 Links > Fostex FR-2LE or
Naiant IPA (tinybox format) >
Roland R-05

Offline bhtoque

  • Trade Count: (3)
  • Needs to get out more...
  • *****
  • Posts: 3308
  • Gender: Male
  • So much music, so little time.
Re: ratshack digital signal converter testing :)
« Reply #47 on: September 07, 2004, 02:21:27 PM »
Try EACs WAV Compare function with the files you have.

Theoretically the two CD > JB3 files should match, but that assumes [1] the C701ES doesn't have a resampling output and [2] provides a signal to the JB3 that the JB3 likes, so to speak.

Also, I don't see in your chain where you're actually converting the digital signal with the RS box:  C701ES > optical > RS box > optical > JB3.  Can you run C701ES > coax > RS box > optical > JB3?

The rs box takes in either and puts out both, so I doubt it is just a pass through for the incoming format. Unfortunately I don't have anything that will put out coax besides my ua-5.

To make things worse I can't run EAC on my xp system. EAC reboots the pc as soon as I try to launch it. I've spent several hours troubleshooting it, changing aspi layers etc and have had no luck.

JAson
MG 200/210/270
AKG c422>s42>Hydra silver interconnects
AKG 391/92/93>MK 90/3 actives
>AM Hyper-Conductors
Studio Projects LSD2>MiAGi II
>Edirol R-4 (Oade T Mod)

db.etree.org/bhtoque

Offline rustoleum

  • Trade Count: (0)
  • Taperssection All-Star
  • ****
  • Posts: 1015
  • Gender: Male
  • AKG 481s->MiaGi IIs->MiniMe->MTII
Re: ratshack digital signal converter testing :)
« Reply #48 on: September 15, 2004, 10:41:54 AM »
I picked one of these up and have begun testing it and wanted to share my results so far.  After jumping through some hoops to get a reliable test setup (see this thread for the gory details: http://www.taperssection.com/yabbse/index.php?topic=26297.msg329324#msg329324) I have managed to have some success with this box.

The final phase 1 test setup consisted of a one minute long sine curve generated by WaveLab, output to a Sony D100 via a Zoltrix Nightingale  coax output.  The sine curve ran from 44.1 Hz to about 20,000 Hz and back down.  I then ran the signal back out of the D100 into both the Coax-In of the zoltrix (as a control source) and then into the Rat Shack digital convertor's coax input -> toslink output -> toslink cable -> toslink digital input on the zoltrix.  The samples once trimmed to the same start and end sample matched perfectly in both WaveLab4's file comparison and EAC's Wav compare tools.

Pretty simple test, so far, but good results indicating the convertor is indeed bit accurate.

My next test is conisting of more Sine curves, however I've added a lot more variation.  Ex - I'm using one sine curve spanning 16hz to 20,000 and back down in the left channel and in the right channel I'm using a curve that begins at about 700 Hz, crawls up to 20,000 Hz, and then sweeps down from 20Khz to 44 Hz.  After 10 seconds the curves repeat but change channels.  The 20 second pattern then repeats for an hour.

I'm dumping the new hour long test signal to DAT as I type and will then run it through the convertor and see what happens.  more to come...

Rusty


Offline MattD

  • Taper Emeritus
  • Trade Count: (1)
  • Needs to get out more...
  • *****
  • Posts: 4634
  • Gender: Male
Re: ratshack digital signal converter testing :)
« Reply #49 on: September 15, 2004, 11:14:00 AM »
Why the sines? White noise will tell you what you need to know here.
Out of the game … for now?

Offline rustoleum

  • Trade Count: (0)
  • Taperssection All-Star
  • ****
  • Posts: 1015
  • Gender: Male
  • AKG 481s->MiaGi IIs->MiniMe->MTII
Re: ratshack digital signal converter testing :)
« Reply #50 on: September 15, 2004, 11:18:15 AM »
I wanted to force the test signal to utilize the majority of the frequency spectrum available to DAT.  Additionally, I've seen threads here involving channel swap in similar situations so that's why I forced the channels to alternate... I want to know if that type of problem is happening and a constant signal that does not fluctuate in some way would not allow me to notice a change like that.

Is there a reason I should use white noise instead of the sine curves or are you just mentioning it since it would be less work, yet still give us an idea of the accuracy?  Thanks!

Offline MattD

  • Taper Emeritus
  • Trade Count: (1)
  • Needs to get out more...
  • *****
  • Posts: 4634
  • Gender: Male
Re: ratshack digital signal converter testing :)
« Reply #51 on: September 15, 2004, 01:36:18 PM »
I mentioned mostly from the "less work" perspective. It covers your entire frequency spectrum and provides channel separation (if you use a different signal for each channel). Generate a good length of noise in a file. Play back the file through the device to DAT, transfer DAT, chop leader/trailer on DAT, compare files.
Out of the game … for now?

Offline rustoleum

  • Trade Count: (0)
  • Taperssection All-Star
  • ****
  • Posts: 1015
  • Gender: Male
  • AKG 481s->MiaGi IIs->MiniMe->MTII
Re: ratshack digital signal converter testing :)
« Reply #52 on: September 17, 2004, 11:50:24 AM »
I'm about ready to call this thing bit-accurate.  I've put about 2.5 hours of audio through it and gotten 100% of the same bits put through the coax-in of the zoltrix.

Unfortunately, I have also gotten about 2 hours that were 99.9544% accurate.  However, although I can't be 100% certain at  this point,  I am willing to attribute this to the Zoltrix Nightingale and/or the DAT deck as several transfers of the same 'offending' WAV through just the coax-in also resulted in a similar accuracy.

I still have yet to hook it up to the MiniMe and do a comparison feeding it that way, but once I do that I will presumably have a better idea if the DAT deck was the culprit.  For the moment, I would venture to say that this thing is field ready.

Also - please note that in the .0466% of the time where problems arose, they were inaudible - The samples were not dropped, but rather slightly altered (I can post a pic or two if desired).  Listening did not reveal a problem, but rather bit-for-bit file comparisons.  So, you guys using the Hosa ODLs and saying "I haven't heard any problems" should know that just because you can't hear it, doesn't mean it's bit-perfect.  That said, I do realize testing similar to what I am doing has been performed on the Hosas... I'm just saying this because in addition to that there seems to be a lot of stock put into what our ears can pick up which doesn't necessarily proove a thing.



Offline Brian Skalinder

  • Complaint Dept.
  • Trade Count: (28)
  • Needs to get out more...
  • *****
  • Posts: 18868
  • Gender: Male
Re: ratshack digital signal converter testing :)
« Reply #53 on: September 17, 2004, 12:03:28 PM »
Thanks for your testing efforts!

I vaguely recall Jamie Lutch testing the C02 and sometimes having to run through 10-20 hrs before encountering problems.  Just suggesting that a couple hours testing may not be enough to decide one way or the other.

And if your problems appear on both the control WAV and the DFC WAV then it sounds like you're right - the Zoltrix and/or DAT may have issues.  I'm real curious to see the results once you fire up the MMe and get some additional testing hours completed.

FWIW, I tested the ODL-312 ~100+ hrs and it was flawless, totally bit-perfect - after ruling out a seemingly random channel-swapping problem I seem to have in my EgoSys Waveterminal 2496 and/or recording software.  I don't know of anyone who's tested the -276.  And you're absolutely right - just because you can't hear it doesn't mean it's bit-perfect.  Though for some people that's good enough.  :)
Milab VM-44 Links > Fostex FR-2LE or
Naiant IPA (tinybox format) >
Roland R-05

Offline rustoleum

  • Trade Count: (0)
  • Taperssection All-Star
  • ****
  • Posts: 1015
  • Gender: Male
  • AKG 481s->MiaGi IIs->MiniMe->MTII
Re: ratshack digital signal converter testing :)
« Reply #54 on: September 17, 2004, 03:17:20 PM »
I definately want to do more testing but unfortunately need to change up my process to determine if the zoltrix or DAT deck was causing me troubles.

The next round of testing will consist of sending audio into the MiniMe out to my lappy via USB and out to my standalone CD-burner via Coax -> Convertor -> Optical In.  If I'm able to get several successful hours out of this I'll probably go back to using the Zoltrix in an effort to determine where my original flaws were coming from. 

And, as long as I'm posting, here's an example of one of the .0456%  of my transfers that failed verification.  The difference in samples is pretty minor, but it ain't an exact copy.


Offline dklein

  • Trade Count: (0)
  • Taperssection All-Star
  • ****
  • Posts: 1184
  • Gender: Male
Re: ratshack digital signal converter testing :)
« Reply #55 on: September 17, 2004, 06:08:49 PM »
fwiw, I think I did about 20 or 30 hours of testing to verify UA-5 > JB3 bit accuracy http://ca.geocities.com/dkleined@rogers.com/audio/bit/BitAccuracy.htm
« Last Edit: March 04, 2005, 11:01:34 PM by Brian Skalinder »
KM 184 > V2 > R4
older recording gear: UA-5  / emagic A62 / laptop / JB3 / CSB / AD20 / Sharp MT-90 / Sony MDS-JE510
Playback: Pioneer DV-578 > Lucid DA 9624 >many funny little british boxes > Linn Isobarik PMS

Offline ChrisV

  • Trade Count: (0)
  • Taperssection All-Star
  • ****
  • Posts: 1371
  • Gender: Male
  • Proud member of team 'Lux!
    • My Taping Webpage
Re: ratshack digital signal converter testing :)
« Reply #56 on: September 23, 2004, 12:32:09 AM »
I just picked one of these up :)
Hope I can help out somewhat on the testing.   I've got a UA-5 and 2 JB3's...would a viable test be to run one JB3 optical out of the UA-5 and another JB3 out of the coaxial with the Ratshack box converted to optical into the other JB3.   Then compare the 2 files from the same starting point?   Has it been verified that the UA-5 has exactly the same bits coming out of the optical and coax jacks?   If so, then I think this should be a good test.
http://taping.4t.com
http://db.etree.org/CV
Superlux CM-H8K/C/H/O -> digimod UA-5 -> JB3

Offline rustoleum

  • Trade Count: (0)
  • Taperssection All-Star
  • ****
  • Posts: 1015
  • Gender: Male
  • AKG 481s->MiaGi IIs->MiniMe->MTII
Re: ratshack digital signal converter testing :)
« Reply #57 on: September 24, 2004, 01:29:20 PM »
I can't comment on the UA5 outputs, but if they are indeed the same bits then this would be a great test.

Offline bhtoque

  • Trade Count: (3)
  • Needs to get out more...
  • *****
  • Posts: 3308
  • Gender: Male
  • So much music, so little time.
Re: ratshack digital signal converter testing :)
« Reply #58 on: September 24, 2004, 04:10:48 PM »
I just picked one of these up :)
Hope I can help out somewhat on the testing.   I've got a UA-5 and 2 JB3's...would a viable test be to run one JB3 optical out of the UA-5 and another JB3 out of the coaxial with the Ratshack box converted to optical into the other JB3.   Then compare the 2 files from the same starting point?   Has it been verified that the UA-5 has exactly the same bits coming out of the optical and coax jacks?   If so, then I think this should be a good test.

I doubt he has run a test on it, but Doug Oade probably knows the signal path in that thing better than anybody else in the community. Shoot him and e-mail, I'm sure he'd help.

JAson
MG 200/210/270
AKG c422>s42>Hydra silver interconnects
AKG 391/92/93>MK 90/3 actives
>AM Hyper-Conductors
Studio Projects LSD2>MiAGi II
>Edirol R-4 (Oade T Mod)

db.etree.org/bhtoque

Offline dklein

  • Trade Count: (0)
  • Taperssection All-Star
  • ****
  • Posts: 1184
  • Gender: Male
Re: ratshack digital signal converter testing :)
« Reply #59 on: September 24, 2004, 05:44:15 PM »
 I've got a UA-5 and 2 JB3's...would a viable test be to run one JB3 optical out of the UA-5 and another JB3 out of the coaxial with the Ratshack box converted to optical into the other JB3.   Then compare the 2 files from the same starting point?   Has it been verified that the UA-5 has exactly the same bits coming out of the optical and coax jacks?   If so, then I think this should be a good test.
the coax and optical are separately buffered but driven from the same source - they are the same.  You can also compare the JB3 to a USB recording if you've got that going glitch free.
KM 184 > V2 > R4
older recording gear: UA-5  / emagic A62 / laptop / JB3 / CSB / AD20 / Sharp MT-90 / Sony MDS-JE510
Playback: Pioneer DV-578 > Lucid DA 9624 >many funny little british boxes > Linn Isobarik PMS

 

RSS | Mobile
Page created in 0.137 seconds with 43 queries.
© 2002-2024 Taperssection.com
Powered by SMF