psychopy.org | Reference | Downloads | Github

Microphone-component-produced recordings stopped short for random trials

Dear all,

In my Trial routine, I have a Microphone component recording sound, which is followed by a Keyboard component that ends the recording (and thus the routine) whenever the subject presses a key. The Microphone component’s “duration” is thus irrelevant, and I’ve set it to a random high value (60s; most subjects take 10-15s long for a recording).

Most of the audio WAV data thus obtained is fine, however for some (random?) trials, these recordings are cut-off, and much shorter than they should be (as if the subject pressed a key to end the recording earlier than when they actually did). For these trials, the .LOG file suggests that nothing went wrong, for instance here below is the line corresponding to a recording that should have been 14s long but is instead only 2s long.

DATA AdvAudioCapture record: Record stopped early, new duration 14.066s

Am using 1.83.04 (due to problems with 1.85.01 that I report in a different thread). Any idea why this problem occurs? Many thanks!
Tudor

OS (e.g. Win10): Win 7
PsychoPy version (e.g. 1.84.x): 1.83.04

Please let me know if I should give more detail for this issue - which I still haven’t been able to fix myself - in order for it to perhaps become more understandable. Thanks again for any help with this.

Just reporting back to say that on a different PC (running a different version of PsychoPy), this bug does not appear, so it must be specific to either the initial PC’s configuration or to the specific version of PsPy used, 1.83.04

Sorry to say that the bug does, in fact, appear on other configurations as well, therefore some help with this would be very gratefully appreciated.

I found that this problem is in fact older and has been experienced by others as well (e.g. here).

And just to clarify that the mic component’s duration field probably itself misbehaves, since if it is left blank, all WAV outputs are void.

Dear tudor,

I am now having the same problem with v 1.90.2. Did you ever find a solution or workaround to this?

Hi Rob,
I’m afraid I never found a workaround for this, nor was a valid solution ever suggested by Psychopy’s support team. I ended up recording material outside of psychopy, which was considerably more cumbersome overal. Sorry to hear that even in the latest version this is still a problem!
Tudor

Thanks for your reply, Tudor!

Hello Tudor and Rob!!!

I’m having the same problem… did you ever find out how to fix the problem?

thank u!

Nope. In fact, other threads in which I had asked for help with the various unexpected behaviours in Psychopy have sadly stayed without a reply so I’ve other found some rough&ready workaround myself or just used other software where I could… Too bad, as psychoPy would still be my preferred medium to use for experiments, but I think there are by now too many requests from users and its devs are prbably overwhelmed

Hello tudor,
I’m still having this problem, every time i run the task, 2 or 3 random .wav files (out of 30) are empty or cut short and i cant figure out. its been a few months now and I can’t delay testing anymore… can you tell me what work around you used? is there another program that will allow recording one response per audio file so one doesnt have to seperate one long audio file into individual item recordings?
thanks

hi Miguel,
I\m afraid this was a while ago and don’t really remember… I think you’d be better off looking for a workaround outside psychoPy, else trying to open a new thread with this question and hoping you’ll be luckier getting anyone to answer than I was… good luck! Tudor