Upgrading convertors

  • Thread starter Thread starter kenoflife
  • Start date Start date
mshilarious said:
Yes, at the individual sample level. So the difference is more than just latency.
Click to expand...

It looks like you have a gain loss in there. Try using a reference tone at -3dB and calibrate your input until the reference tone is also at -3dB. You shouldn't have anywhere -near- that much signal left after a null test unless the level is way off.

Edit: scratch that. It looks like your high frequencies are nulling out a good bit, but you're getting a massive low frequency roll-off. Use a low frequency reference tone (500 Hz, maybe) and adjust that level to be the same. Then see what happens.
 
dgatwood said:
It looks like you have a gain loss in there. Try using a reference tone at -3dB and calibrate your input until the reference tone is also at -3dB. You shouldn't have anywhere -near- that much signal left after a null test unless the level is way off.
Click to expand...

I spent a lot of time calibrating levels, and repeated the test today. After long thought, I think I've come to the conclusion that it is perhaps not a valid test, because levels cannot be matched closely enough. I can only normalize to .001dB resolution, which equates to something like a -77dB difference file.

But even using the most careful procedures I have, I could get an average -31dB difference on program material (with most of the improvement from yesterday in high frequencies), and on a 200Hz sine wave, -41dB.

Without more converters to test, I can't draw any conclusions. Is there that much distortion (0.008V, or .006% THD if my math is right) in a D/A/D roundtrip with a midgrade converter? Maybe, I don't know. Food for thought for the OTB crowd ;)


The filter behavior and the sine wave distortion test are still quite interesting, as neither is dependent upon matching levels closely.
 
You must log in or register to reply here.
Back
Top