Clock adjustment too large error 5.20.14 & measurement issue

[SimpleDude]

Hi All,

Have applied room treatments via acoustic panels and needed to do a new calibration with Dirac Live and post cal checks/corrections with REW, I had this issue last time (on 5.20.14) but never found a fix for it, running on Windows 10 connected via a DisplayPort > HDMI adapter going into the AVR for some reason I randomly get the "clock adjustment is too large" error, it happens most of the time and sometimes/rarely it doesn't, if it happens and I simply do the measurement again there is a chance it might not happen. This did not used to happen before on 5.20.13 and I think before a bunch of Windows updates, I did find a thread where John advised the user to download a newer "csjsound_amd64.dll" which I tried although I think that version of the file is already included in the newer 5.20.14 build ?, it didn't have any affect on the error, I would attach my MDAT file but am unable to post links.

Am I correct in believing when that errors happen it affects the impulse response so I can't check/verify the time alignment using the Impulse Response tab ?

Relevant info from my setup(I think) is:

• Windows 10 22H2 19045.3448
• RTX 3080
• Denon X3800H
• UMIK-1 connected via USB (have tried the cable it came with on both front and rear USB ports)

Also a problem I was facing unexpectedly which did not happen last time (nothing has changed hardware wise), was getting random large nulls in the frequency response which made it difficult if not impossible to apply corrections, picture below of Front Height Right speaker, Purple line is the first measurement, then I simply did the measurement again straight after and got the Blue line, the way I am measuring them is to use the File option the measurement screen, select a WAV file and then play a sweep using Windows Media Player and the corresponding .mp4 file for the speaker.

All in all I've been learning this for the last year and a half or so, each time I come to do a calibration again there is a different/new issue I face be it with the room correction software of choice or REW which is very much putting me off doing this anymore.

 

[sam_adams]

If you remove the smoothing from that plot there is a very good chance that what is displayed there is a dropout in the response. Whether it is caused by the mic or the method that you have chosen to generate the stimulus will require further testing on your part.

 

[SimpleDude]

Thanks for the response, it’s definitely a drop in the response, it did the same thing at different frequencies on different atmos speakers.

I think at this point in the future I will resort to using a laptop with a specific/older Windows 22H2 with HDMI direct from laptop to AVR and REW 5.20.13 which is what I used about a year back and was doing hundreds of measurements across weeks with no clock adjustment errors.