Open imiraoui opened 1 year ago
What is the size of the .m4a file? The Whisper API has a limit of 25MB per file.
Is there an error message in the console (some instructions are on the readme for finding this)
Hi all, I just installed this plugin and got the same problem as imiraoui. I used voice memo in iOS to do recording, then share to Logseq to get a *.m4a file
I tried to follow the instruction to check console for hints of the problem, and it said "file not found"
Logseq referred to this link: ![New%2520Recording](../assets/New%2520Recording.m4a
, and it works. I can play the audio in Logseq
I tried to figure out the file path in the error message, and it is:
file:///Users/jake/Library/Mobile%20Documents/iCloud~com~logseq~logseq/Documents/daily_trial/assets/New%2520Recording.m4a
I can't open this link with Firefox, it said file not found.
So I tried to open a link with assets
directory in Firefox:
file:///Users/jake/Library/Mobile%20Documents/iCloud~com~logseq~logseq/Documents/daily_trial/assets
It listed a file named New%2520Recording.m4a
. But when I click into it, Firefox can play the file, but the path automatically changed to
file:///Users/jake/Library/Mobile Documents/iCloud~com~logseq~logseq/Documents/daily_trial/assets/New%252520Recording.m4a
You can see the file name has changed.
I tried to locate the file in Finder, and the path is /Users/jake/Library/Mobile Documents/iCloud~com~logseq~logseq/Documents/daily_trial/assets/New%2520Recording.m4a
. The file name is New%2520Recording.m4a
.
I don't sure why in "file:///" the file name is not the same as what I see in Finder..but I think the bug is here and I hope it will help you to fix the problem. @briansunter
Best, Jake
Screenshots:
Try installing the latest version; it appears that the previous version has a duplicated '/' in the URL. Or just change the URL from http://api.openai.com/v1/
to http://api.openai.com/v1
in plugin settings.
Thanks for this amazing plugin. I have shared a few voice notes to the logseq app in .m4a format. When I run the command '/whisper' on that block nothing happens and no text gets transcribed. Any idea how come?
If I run this on an empty block, I do get an error message telling me there is no audio file on that block so I'm not sure what I'm doing wrong...