R/E/P Community

Please login or register.

Login with username, password and session length
Advanced search  

Pages: 1 [2]  All   Go Down

Author Topic: Another ADDA test  (Read 7218 times)

danlavry

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 997
Re: Another ADDA test
« Reply #15 on: March 09, 2005, 04:51:38 PM »

Here is a plot of what happens when one adds the same signal with 1 sample delay (red), 2 sample delay (blue), 3 sample delay (black) and 4 sample delay (purple).

index.php/fa/815/0/

The more delay the more "dips" in the comb filter.

The comb effect does not require that both signals are the same. It will effect will apply to the portion of the signal that is common to both signals.

This is about summing of signals with ELECTRICAL delay. not acoustical delay.

A little delay in adding musical material that is common to 2 channels can really screw things up   Sad  

Regards
Dan Lavry
www.lavryengineering.com
Logged

PookyNMR

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1991
Re: Another ADDA test
« Reply #16 on: March 09, 2005, 10:05:47 PM »

danlavry wrote on Wed, 09 March 2005 14:51

This is about summing of signals with ELECTRICAL delay. not acoustical delay.

A little delay in adding musical material that is common to 2 channels can really screw things up   Sad  



Especially when summed to mono...
Logged
Nathan Rousu

cerberus

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 2651
Re: Another ADDA test
« Reply #17 on: March 10, 2005, 02:45:26 AM »

 Shocked


cerberus

Joe Crawford

  • Full Member
  • ***
  • Offline Offline
  • Posts: 107
Re: Another ADDA test
« Reply #18 on: March 10, 2005, 01:21:52 PM »

Yannick, et. al.

Well I’ve isolated my sample swap/delay problem.  After getting the same failures in both Wavelab and Nuendo, I moved my attack down a notch to the ADDA box.  Just as a wildass guess I reconfigured the clocking on the ADI-8 DS ( the ADDA box) and the problem disappeared.  When it was failing, I had the RME ADI-8 DS externally clocked from the word clock output of the RME Hammerfall Digiface via 3ft. of 75 Ohm terminated coax.  Simply changing the clocking on the ADI-8 DS from ‘EXT’ to ‘INT’ cleared up the problems (all of them).  On restoring it back to external clocking, the problems reoccurred.

Well, Dan always said it would work better on internal clock, right Dan?  I’ve e-mailed REM and they are looking at it.  Maybe I can get some work done now....  

Joe Crawford
Logged

danlavry

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 997
Re: Another ADDA test
« Reply #19 on: March 10, 2005, 03:08:14 PM »

Joe Crawford wrote on Thu, 10 March 2005 18:21

Yannick, et. al.

Well I’ve isolated my sample swap/delay problem.  After getting the same failures in both Wavelab and Nuendo, I moved my attack down a notch to the ADDA box.  Just as a wildass guess I reconfigured the clocking on the ADI-8 DS ( the ADDA box) and the problem disappeared.  When it was failing, I had the RME ADI-8 DS externally clocked from the word clock output of the RME Hammerfall Digiface via 3ft. of 75 Ohm terminated coax.  Simply changing the clocking on the ADI-8 DS from ‘EXT’ to ‘INT’ cleared up the problems (all of them).  On restoring it back to external clocking, the problems reoccurred.

Well, Dan always said it would work better on internal clock, right Dan?  I’ve e-mailed REM and they are looking at it.  Maybe I can get some work done now....  

Joe Crawford



Great job!!!  

Dan Lavry
www.lavryengineering.com
Logged

Yannick Willox

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 264
Re: Another ADDA test
« Reply #20 on: March 11, 2005, 01:57:07 AM »

One problem down.

But all my tests were with all AD converters on internal clock...
So I still have some finding out to do  Confused
Logged
Yannick Willox
Acoustic Recording Service

Yannick Willox

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 264
Re: Another ADDA test
« Reply #21 on: March 11, 2005, 08:49:52 AM »

There is a big chance that the issue lies in a non-standard tdif protocol.

If this is the case, this would mean the offending AD converters are OK, and it is in the digital interface of my DAW (but not with the converters of my brand ...)

Keep you posted.
Logged
Yannick Willox
Acoustic Recording Service

Joe Crawford

  • Full Member
  • ***
  • Offline Offline
  • Posts: 107
Re: Another ADDA test
« Reply #22 on: March 11, 2005, 08:50:41 AM »

Yannick – How were you connecting the converters?  Give us a description of the system and the data paths you are using, and especially the clocking.  Next step is to try to isolate it down to either the output side (D/A) or the input side (A/D) of the converters (or both).  You might have to get a friend to generate a test CD for you, one that you know is clean, with a few seconds of 4kHz stereo sine wave on it.  This way you can read it in with different sample rates (44.1 & 88.2) and check the results. If you can get the failure simplified it will be easier to shoot.  George at RME also suggested downloading a trial version of Samplitude (or another application that will run on your system and use the same converter interface) to eliminate the application from the picture.  Give these a try and let us know what you have found out.

Joe Crawford
Logged
Pages: 1 [2]  All   Go Up
 

Site Hosted By Ashdown Technologies, Inc.

Page created in 0.029 seconds with 20 queries.