Timing reference not detected in import sweep recordings (beta build)

[Bowkey]

Hi John,

Importing sine sweeps recordings with timing reference doesn't seem to work in the beta build, it works fine here in the release build though.

A simple way to demonstrate this is to re-import an unaltered REW generated measurement sweep (from the same build). This does work fine in the release build.

1. Export measurement sweep:

2. Try to import using 'import sweep recordings':

Would love to know if this is an error on my side or wether this is a bug that could be changed.

Thanks,

Bowkey

 

[John Mulcahy]

The timing signal search on the captured sweep could terminate prematurely if it found a block of zero samples. I wouldn't expect that to happen for an actual recording, though it could for a synthesised file. I have put in a fix for the next build, can you check whether there is a problem with an actual captured file and if so please attach the stimulus and captured files used.

 

[Bowkey]

Thanks, great to hear you've put in a fix for the next build. At the moment I'm using it to analyze a software crossover (using loopback recording in a DAW), so that causes the same troublesome leading zero samples, but of course this use-case is not very typical..

Btw; the only way to get the time reference signal to show up in my measured low-passed recording was by copy-pasting the original time reference in front of it (which works fine in the release build). To avoid the need to copy-paste the time reference when using external (low passed) loopback recording, it would be really useful if the 'Response' file in 'Import sweep recordings' could be a stereo file that has the timing reference (re-recorded through loopback) on one channel and the measured/processed signal on the other.

 

[Bowkey]

Hi John,

I've got a related issue importing sweep recordings. Below are two imports from the same measurement, but with a slightly different leading noise length. These are acoustic recordings, so with leading noise, not zeroes/silence. Noise (or silence for that matter) longer than approx 6.5 seconds before the timing reference starts seems to produce the unexpected yellow result. Would be great if you could fix this, so the length of the imported file is not critical anymore. I've also attached the .mdat file.

Many thanks,

Bowkey

 

[John Mulcahy]

Can you attach the import files and the sweep file?

 

[Bowkey]

Here you go, thanks for looking into this.

 

[John Mulcahy]

The file with a longer delay causes REW to miss the timing reference as it occurs past the point it stops looking - there is a message that the timing ref was not found. I can make it more tolerant of delays before the sweep is captured, but there is a risk that in captures without a long delay the timing ref search could process the whole sweep and prefer the timing signal at the end of the sweep to the one at the start and so miss the sweep altogether. It is best to avoid lengthy delays between starting recording and playing the stimulus.

 

[Bowkey]

Ah, I see. I'm running into this problem because I'm creating a DAW template where I'd like to export all measurements in one go, and I can't avoid some being later than others. But I'll just make sure there is a little less initial delay, so all the first timing references occur within 6.5s. Thanks!