Is Transcoding to .MKV The Only Way To Keep Best Quality and To Be Able to Apply Time Remap Filter?

This is my first post here, after using Shotcut for the first time yesterday.

I am excited to be here, and I hope that I am posting in the correct sub-forum.

When attempting to use the Time Remap filter, I was prompted to “Convert to Edit-friendly…”, because the .MP4 video clip I had opened for my Playlist “…has B-frames, which are not supported by Time Remap”.

Accordingly, I first, I chose to convert to the “good” quality, medium sized file (Lossy: I-frame-only H.264 / AC-3 MP4) format, but that resulted in a clip of poor resolution (almost slightly ‘blurry’), so I then tried the “best” quality, biggest sized file (Lossless: Ut Video / PCM MKV), which matched the video quality (resolution) of the original clip.

I was surprised to see that the best quality .MKV file was 2.78 GB for a 1 minute and 22 seconds video clip.

  1. Is the size of the resulting file usual for the duration of the video clip (1.22)?
  2. If so, is there any other way I can go about converting / transcoding a video clip which will result in a smaller file, and will transcode quicker?
  3. Why does an .MP4 (i.e., not unusual format) need to be transcoded to apply the Time Remap filter?
  4. Do all .MP4 video clips always need to be transcoded in order to apply video (and perhaps other) filters?

Thanks for any replies.

MSI 15 Crosshair R6E
i7-12700H 2.70 GHz
16.0 GB RAM 1TB
3070 GPU

  1. That is normal.
  2. Choose the “better” option, which is sometimes called a visually lossless if not mathematically.
  3. It is not only MP4 files. The Time Remap lets you go in reverse, and when the video is compressed with B type frames the performance and reliability is too poor.
  4. No, only ones with B frames. You do not need to fully understand video compression and B frames to understand and accept this condition, but you can search the web to learn more about compression in general.

See also the docs

I was about t reply when I saw that Dan (Shotcut Leader) had done so. The only clarification I can make is for point 4).

No, in general you do not need to convert your video to apply video filters (other than the “Time Remap” filter). However, you may find it necessary if your video has a variable frame rate (VFR), rather than a constant one.

Thanks very much, Dan.

I appreciate your thorough and swift reply, and the link to the ‘Time Remap Filter’ doc.

I will try the “Better” option.

Thanks again.

Thanks very much, Elusien.

I must check the settings on my DSLR (on which I recorded the video) to see if the setting is VFR - I had thought it would be a constant frame rate, but I might be wrong.

Thanks again.

The overwhelming majority of DSLRs make constant frame rate videos. Time Remap prompted for conversion because the file had B-frames in it, as opposed to VFR. For purposes other than Time Remap, conversion is probably not necessary.

Thanks very much, Austin.

That was my supposition about DSLRs, so it is good for me to hear you confirm that most of them use constant frame rate.

It is a shame that the Time Remap filter needs to convert .MP4s, because it is going to be a feature that I consistently use in at least one series of videos.

Thanks again.

There are other ways of doing (simple) Time Manipulation without using the Time Remap filter. See the tutorial here:

Thanks very much, Elusien.

I came across that technique yesterday in a 2020 video by Ben Espanto, but he kept getting extra frames in the video, which he had to remove (not that great a chore, but annoying).

There were some speculations in the video’s comments about why these extra frames happened, and I would be grateful for your take on this, so that I can make this technique work.

Here is a link to the video

This topic was automatically closed after 90 days. New replies are no longer allowed.