Re: X32 Discussion - Latency
So I went and tried this on my Pro Tools 8 LE rig with a 003R. I had two mono audio tracks and the click track. I output the click track directly to one audio track, and bussed from the click to a 3-4 stereo headphone mix output (separate from main out 1-2). Had a mic plugged into the other audio track, and put the headphones up to the mic. First I recorded at 1024 buffer rate, so hearing the click back through the mic in real time had some serious delay, as expected. I recorded for 5 seconds to each track, then switched the buffer rate to 128 and did the same to thing for comparison.
Although lowering the buffer rate greatly decreased latency in monitoring, it made no difference to the outcome of recorded alignment. In both cases, the clicks were almost completely lined up. Only when I zoomed in as far as I could did I see the recorded track was just slightly off. Played back and switched between and could hear no delay what so ever.
I know we're getting off topic of the X32. I apologize, but we already have this conversation going so why not.
edit: Crap, I don't think I did it right. It needs to be a prerecorded click and not streaming at the same time maybe? I'll have to try that again.
FWIW - did these rough tests a few weeks ago
"out of interest just did some rudimentary latency tests using reaper as the in and out - and these should not be taken as fact - as simply reading off the offsets by eye
Recorded a sine wave to a reaper track and then played back through x32 and recorded on new track in reaper ( note output from x32 was via the P16's to allow x32 processing and inserts to be included in the rerecorded track - don't know if this routing would make any significant difference suspect not).
First With the reaper automatic audio driver reported latency adjustment turned OFF ( out of interest and to see how good the reaper auto latency adjustment is). the rerecorded sine wave was offset by a approximately 57.4ms - this could be card / usb / driver etc adding to the X32 processing time -see below.
Measuring relative to this offset:
With standard dynamics and EQ the latency was the same as without these ( as expected).
With an insert effect - waves - the latency increased by approximately 1.5Ms ( but see below)
With an insert GEQ the latency was less about 0.8ms.
As expected with two inserts in line - i.e. via a bus ( in this case two waves) offset doubled.
Repeated the above with the reaper using the reaper reported driver latency adj turned ON - to see how this impacted
With no processing - the offset of the recorded sine wave was 0.6 ms ( this would ally with Behringers stated 0.7ms as the accuracy I could read off needs a little give and take).
And with standard dynamics and EQ was the same.
With the waves insert I got a slightly better figure than the first test about 1.2 ms but again the accuracy of my reading would need to be better and maybe do again.
With the Geq - 0.8ms.
Have not looked at any of the other inserts but suppose there will be some variation in the processing overhead and latency.
Finally doing a test with direct output then i.e Card in and card direct out - i.e. recording back to reaper with no x32 eq/ dynamic processing/effects
without reaper latency adjustment for drivers offset is 57.1 ms
So the drivers etc have a latency of about 57ms
With driver reported latency adjustment turned ON no offset"
