Psychtoolbox-3 has stopped running: what to do?

Bugs & Features
,
#1

I have recently encountered issues with getting Psychtoolbox-3 run using Matlab64-Bit. The operating system is OSX 10.10. A month earlier, I had no problems with running Psychtoolbox on the same device. Another group that uses this laptop has not experienced any issues – they use the same Matlab version and Psychtoolbox-3 for Apple OS X for their fMRI stimuli.

Here are the warnings and errors I have been receiving:

**
WARNING: Couldn’t compute a reliable estimate of monitor refresh interval! Trouble with VBL syncing?!? ! PTB - ERROR: SYNCHRONIZATION FAILURE ! ----

This is Psychtoolbox-3 for Apple OS X, under Matlab64-Bit (Version 30.0.12 - Build data: Oct 26 2015). Support status on this opeating system release: OSX version 10.10 is not supported or tested anymore.

Error using Screen.

**

If I update the operating system, will macOS Catalina 10.15.6 work? I do not want the other group using the laptop to experience any issues. I read on this forum that Catalina was not recommended for Pscyhtoobox.

I have found a different solution on the github website:

Switch resolutions within System Preferences --> Display Panel (Scaled). Select a different resolution then reselect the required/native resolution. Repeat this hack with each system restart.

What would the best solution be to get Psychtoolbox to run again?

I will provide further information if necessary. Please note I am new to forums like this one. I had nowhere else to go.

Thank you.

Monika

#2

Did you make up those output message from memory? If it really printed things like “Version 30.0.12” or “Build data:” then your disc drive is dying and corrupting files and there’s your explanation.

Sometimes it is transient failures due to some overload or malfunction of the machine, so a reboot doesn’t hurt.

Then Apples operating systems have and had various awful bugs that caused failure, and sometimes associated workarounds like the one resolution switching workaround your described helped. If that doesn’t do the trick and nothing else changed on your machine, well, we don’t support 10.10 anymore since a long time, or PTB 3.0.12, so with these you are on your own.

You could upgrade to PTB 3.0.16 which had hundreds of hours of work done to implement various new workarounds and hacks implemented to deal with many of the Apple bugs in 10.11 - 10.15. Ofc. that would also mean upgrading macOS to > 10.10. If you choose that route, i think the general rule of thumb is that the older the macOS, the less frelled up it is, but at the same time it means the OS is no longer supported with security updates and a hazard for that reason.

Btw. if the PsychtoolboxKernelDriver is correctly installed and used by PTB on a suitable graphics chip, it should provide more diagnostic info about causes of failure (full output needed) or more confirmation and peace of mind if things work correctly. If that driver isn’t used then it is also possible that there was silent failure before, even if nobody noticed.

If your Mac is old enough, you could install Linux on it for a much more reliable experience. Otherwise maybe try 10.11 or 10.12 or 10.13 if you don’t care about computer security updates. Or 10.14.6 Mojave for something ok’ish which still has 1 year of security update support.

I would avoid 10.15 Catalina at all cost, it is a disaster of an OS, with more relevant bugs for data collection than any other one before, and go at most up to 10.14.6.

1 Like
#3

Hi Mario,

Thank you so much for all the options you have provided. I will try them out.

To answer your question - yes, "Version 30.0.12” and “Build data:” were printed in Matlab. I did not change any wording in the text that starts and ends with “**” in my original post.

This being the case, would you still recommend testing the various options you suggested, or should I refrain from any?

Thank you,

Monika

#4

There was never a version “30.0.12” only “3.0.12” and “date” was never misspelled as “data”, neither was “operating system” misspelled as “opeating system”, so if this printout is real then your files are severely corrupted due to some serious malfunction or defect of that computer, or the download of PTB was somehow corrupted from the beginning. I would recommend any appropriate measure to fix that like running a full filesystem check and hardware diagnostic on that machine, redownloading PTB if those are the only defective files found. If it is a corrupted filesystem or dying disc drive then wiping and reinstalling the whole machine, replacing hardware etc. would be in order. Also checking what other experiment data from that machine was just corrupted data trash.

What hardware model / graphics card etc. is this?