Selur's Little Message Board
[BUG] Using Vidstab fails with error "<temp-file> is too small! (byteSize: 0byte)" - Printable Version

+- Selur's Little Message Board (https://forum.selur.net)
+-- Forum: Hybrid - Support (https://forum.selur.net/forum-1.html)
+--- Forum: Problems & Questions (https://forum.selur.net/forum-3.html)
+--- Thread: [BUG] Using Vidstab fails with error "<temp-file> is too small! (byteSize: 0byte)" (/thread-2104.html)



Using Vidstab fails with error "<temp-file> is too small! (byteSize: 0byte)" - damian101 - 05.10.2021

Version: build 20210718-1 from the AUR, on Arch Linux

Hybrid works fine, but when I try to use the Vidstab filter it quickly fails with error message "ERROR: "<temp-file> is too small! (byteSize: 0byte)".


Sidenote: Why is "linear" the default interpolation method in Vistab, instead of "bilinear"? Seems weird to me.


RE: Using Vidstab fails with error "<temp-file> is too small! (byteSize: 0byte)" - Selur - 06.10.2021

Might be a quote vs. tick issue assuming your ffmpeg version is compiled with vidstab.
Would need a debug output to know.

Quote:Why is "linear" the default interpolation method in Vistab, instead of "bilinear"?
In the transform plugin? Proably because 'linear' was first implemented. Smile

Cu Selur


RE: Using Vidstab fails with error "<temp-file> is too small! (byteSize: 0byte)" - damian101 - 07.10.2021

(06.10.2021, 05:48)Selur Wrote: Might be a quote vs. tick issue assuming your ffmpeg version is compiled with vidstab.
Would need a debug output to know.

Quote:Why is "linear" the default interpolation method in Vistab, instead of "bilinear"?
In the transform plugin? Proably because 'linear' was first implemented. Smile

Cu Selur

Update: Transcoding with Vistab filter actually works, but not together with the FFV1 encoder, if I select x264 with default settings it somehow works just fine.

I now got a debug output.


RE: Using Vidstab fails with error "<temp-file> is too small! (byteSize: 0byte)" - Selur - 07.10.2021

Seems to be a problem with %TRFFILE% no getting replaced during the job creation.
-> looking into it

Cu Selur


RE: Using Vidstab fails with error "<temp-file> is too small! (byteSize: 0byte)" - Selur - 07.10.2021

Okay, I know where the problem is.
When vidstab is a 2pass process, when x264&co are used Hybrid has an decoder and an encoder call and then simply copies the decoder call, modifies it and inserts it as an additional pass.
Problem is this doesn't work when there are no separate decoder&encoder calls.
-> I know what the problem is, but it will take some time to fix it. Smile

Cu Selur


RE: Using Vidstab fails with error "<temp-file> is too small! (byteSize: 0byte)" - Selur - 07.10.2021

Send you a link to a dev version via pm which hopefully fixes the problem. (jobs have to be recreated)

Cu Selur


RE: Using Vidstab fails with error "<temp-file> is too small! (byteSize: 0byte)" - damian101 - 08.10.2021

(07.10.2021, 17:21)Selur Wrote: Send you a link to a dev version via pm which hopefully fixes the problem. (jobs have to be recreated)

Cu Selur

Yes, that version works perfectly fine for my case, thanks a lot for the quick fix!