August 5, 2021 at 7:58 pm #7149agnesgaoParticipant
Hope you are well!
I’m running some issues with my audio recording experiment. It was working perfectly before, and we were saving the recordings to amazon server. However, when i tried to run more subjects recently, it would get stuck on the
“Please wait while the archive of your recordings is being uploaded to the server…” page forever, and nothing would be saved.
I knew you sent out an email regarding the audio limit in chrome, I was wondering if that applies to recordings as well? Could that be the main issue here?
AgnesAugust 6, 2021 at 9:48 am #7150JeremyKeymaster
I don’t think the audio limit on Chrome should have an impact on audio recording. Do you know if the issue only arises for participants with Chrome 92?
The message suggests that there is a problem with uploading the recordings, not necessarily with creating them. Did you change anything on the amazon side?
If possible, please share your project’s demonstration link with me, here or at firstname.lastname@example.org
JeremyAugust 6, 2021 at 1:14 pm #7151agnesgaoParticipant
I didn’t change anything Amazon’s end. Not sure about participants’ browsers, but nobody’s recordings were being saved.
Here’s the demo link with debugger on: https://farm.pcibex.net/r/TDFsfh/
Thank you so much for your help!
AgnesAugust 6, 2021 at 1:46 pm #7152JeremyKeymaster
I get a CORS error when I take your experiment. This means your server does not return the necessary CORS headers. When I try to access your website (agnesygao.com) I get a security warning, informing me that your certificate likely expired. This would explain why the CORS request fails, as your website is not considered secure. If your SSL certificate is expired, you should renew it, and see whether uploads succeed again then
Three additional notes:
– the URL you pass to
InitiateRecorderpoints to a PHP file, so in this case it doesn’t really matter that it is an “amazon server”. I thought you were using an API to process the request, as described in this guide
– I strongly encourage you to insert asynchronous
"noblock", ) at regular intervals in your experiment in addition to your final (synchronous)
UploadRecoringstrial: you have over 700 trials, with one recording per trial, which can rapidly make for a large archive of recordings to upload all at once at the very end of your experiment. Even after fixing the CORS issue, you might still have to wait for several (tens of) minutes on the “Please wait” message before the upload completes. You could include an asynchronous
UploadRecordingsafter each block, for example, so that when you reach the end of your experiment, there is only one block of recordings left to upload
– As far as I can tell, you are unnecessarily duplicating a
Templatepiece of code 12 times, with the only differences being which CSV file you use as a table, and which label you assign to the output trials. My advice is you concatenate all 12 CSV files into a single file, to which you add a “Label” column which will read “block1”, “block2”, …, “block12” for the corresponding rows. Then you can keep a single
Templatecommand that will start like this:
Template( "L1_all.csv" , row => newTrial( row.Label ,
- You must be logged in to reply to this topic.