Latency Experiments

[jjhuntfox]

I've started doing some latency experiments with my 001 / Octopre.

I wanted to see if all ya all... ... think I'm on the right track. Scientifically...

I've started with the click plug on a track. I sent that click signal on a bus and recorded it to another track. I checked for any latency but didn't hear any and didn't see any. I'm a little concerned there might be a little bit of latency in this process even though it is internal. I cant see a way to check the recorded click against the original click.... but it sounds dead on.

My next step has been to record the click output from different sources, monitors and headphones, and compare the latency for each. (Also low latency monitoring on and off...)

I just expand the waveform and see if they line up, vertically. I then use the shift and experiment with different values to get them to line up.

What do you think?

I'll have some detailed results tomorrow, unless you think I'm full o crap.

Thanks,

John

[jjhuntfox]

Bump, thought there might be a bit more interest in this....

[Chris Cavell]

I have a feeling that the buffer and converter delay compensation that pro tools LE actually does have (it just doesn't have plugin delay compensation) will thwart or confuse your efforts.

[technician]

I'm interested! I'd say still go for it, there are people who will appreciate!

[da BaSsTaRd!]

been done already: http://pages.sbcglobal.net/ccash/latency.htm

[Chris Cavell]

dB, it would appear that the newer versions of pro tools also compensate for the converter latency in some cases...which has evidently changed since that experiment.

[da BaSsTaRd!]

are you sure about that? i think even ADC only compensates for buffer latency. maybe rail knows.

[Chris Cavell]

I just checked it, it compensates for the converters in the 002...in most cases...I have found a way to defeat the compensation alltogether, defeating both buffer and converter compensation. I think it won't compensate for any non-digi converters though.

[jjhuntfox]

And by the way...

I saw no difference with low latency monitoring. The figures were the same with it on or off. I did have a small session and I only did my experiments at 128 buffer.

But time after time the results were exactly the same. Try it is you don't believe me. I'd welcome it if someone would prove me wrong.

Just trying to understand this issue.