| Reference | Downloads | Github

Bug-fix release 1.85.3

PsychoPy 1.85.3 has just been uploaded to the github releases page:

This release fixes several bugs but it does also add 2 new dependencies. Usually we avoid adding anything new within a release series (1.85) but on this occasion it is important in order that your monitor calibrations can be saved in a way that Python3 can read them. You should therefore download the full installer rather than just the patch file:

  • json_tricks and future are both required for this version
  • these are paving the way for updating to Python3 (which will be enabled in 1.86.x series of releases)


  • Problem with text stimuli on Windows not rendering if large (was a bug in pyglet, fixed in 1.3.0b1)
  • Whitespace/indentation guides were not toggling on/off correctly in Coder commit:4ebbdb9c5
  • Added status property to labjack U3 in Builder scripts commit:b4d16d3d6
  • Fix to iohub.launchHubServer() under Linux commit:a9392f62e
  • Fix to config file for iViewX on ioHub commit:590fb20de

I’ve actually noticed an issue here, which is that on macs with retina displays the screen is appearing on ly in the lower left corner. It appears to be a change in pyglet 1.3 forcing us to use retina but then not getting it quite right without help!

I’ve worked out how to support retina displays for stimulus rendering from here onwards (with pyglet 1.3) and that will be released in version 1.86 but for now I’ll upload a new Mac release called 1.85.3b that has the older pyglet 1.2 reinstalled.


Also note, 1.85.3 is now available from the useVersions selector. Happy Psychopy-ing!

1 Like

I have been receiving the following error only in version 1.85.3 when trying to play videos:


I was interested to see if there was a fix for this issue. I would use an earlier version, but displaying large text is needed for the experiment, which seems to be an issue with earlier version.


Oh. Possibly that means the latest versions of pyglet (it was the previous pyglet version that had the text-rendering bug) have now removed the sound-playing that we relied on. [sigh]

We do have an alternative movie player though that avoids this. If you’re in Builder ask the movie to play using the ‘moviepy’ backend which doesn’t have this issue Moviepy has superior performance in most cases too, and just has the one problem that we haven’t built in support for streaming audio (so very long movies with sound crash the experiment because the entire sound stream loads to memory in one go).