Initial PR for the pure-tone presentation task #17
Conversation
|
Thanks for getting this started! so one first element of feedback. wav files should not live in a git repo. It depends a bit on the size but just the sheer number of them, it would make more sense to create a datalad dataset, and expose that dataset as a submodule in |
…ad dataset in its location
|
Just removed the wav files' folder and replaced it with a datalad dataset. |
There was a problem hiding this comment.
After our discussion I tried using the value/volume values of the Sound class in psychopy
https://www.psychopy.org/api/sound/playback.html#sounddevice-sound
and it seems to work well (we use SoundDevice for the other tasks, but that can be changed if needed). I think it would be cleaner/faster than using files to be loaded and then shared, unless you need some fade-in/fade-out.
See other comments about how to fit your task in the rest of the software.
Also, the task should output a bids-like events files with the conditions/onset/duration/.
Let me know if you need more information to make the changes.
src/tasks/pure_tones.py
Outdated
| return filename | ||
|
|
||
|
|
||
| def stimulus_presentation(): |
There was a problem hiding this comment.
The task should be implemented as a subclass of the base Task class, and implement a _run() function where task design appears.
src/tasks/pure_tones.py
Outdated
| playsound(to_play) | ||
| i = i + 1 | ||
| print(f"Waiting for {ISI} second(s)") | ||
| time.sleep(ISI) |
There was a problem hiding this comment.
using time.sleep give chances for slippage due to concurrency for CPU access. A non-slip design is preferrable. See how ThingsMemory task is implemented for an idea on how to proceed.
src/tasks/pure_tones.py
Outdated
| time.sleep(ISI) | ||
|
|
||
|
|
||
| stimulus_presentation() |
There was a problem hiding this comment.
This file should only describe the task and is not meant to be ran as a standalone script.
…f stimuli presentation
…rameter difinitions in tasks (ses-pure_tones.py)
… to extract the subject number in the ID (ses-pure_tones.py)
|
Replacing playsound with psychopy seems to have created a distortion problem when reading the wav files but I can't find its origin. |
|
I pushed some changes to an audiometry branch on that repo (see commit 6b032a6 ). the 250Hz does not play well either, will look into it, maybe a limitation of sound card/hardware.... |
Still in development, this is a first iteration. This PR also contains the stimuli's .wav files