Lossless video compression in Avidemux

[macduke]

Hi all.
Firts I would highlight that I'm not very good in understanding all video coder features.
Reading posts in this forum I understand that:
- avidemux 2.6.x includes 2 lossless video coders: HuffYUV and FF HuffYUV. Both don't have any configuration to set.
- avidemux 2.5.x includes 3 lossless video coders: the two above plus FFV1. All without any setting available.

Now I'm asking for experts help.

Are really the coders above lossless? So, really no degradation of video contents? With the hardcoded sets?

Moreover the 2 coders in 2.6.x are fully back compatible in 2.5.x? Can I code in 2.6.x and read in 2.5.x without (theorethical) problems?

Thanks to everyone that will answer my doubts.

           Mac.

[macduke]

No one have a clear answer?

[Grobe]

Hi.

If you're interrested to investigate this yourself, you can always take two screen records from a media player. First from a frame in the original video file, then from the same time in the resulting video file. Save it then as a loseless image format (PNG preferably).
Use Gimp or any advanced photo editing software and put the two images on top of each other. Then make the upper image (layer) appear as "difference". That will reveal if the images (the two frames of the original video and the output video) have any differences.
The result should be complete black. You could also try to adjust gamma to reveal slightly differences that under normal conditions cannot be seen with the eyes.

[mr_blonde778]

Are really the coders above lossless? So, really no degradation of video contents? With the hardcoded sets?

Short answer:yes, each of those makes a pixel for pixel copy of the source file the nuances between them is trivial and will not affect picture reproduction, though they file sizes may vary slightly you are still looking at huge file sizes in the tens of gigabytes. Just be sure and mux them in an Avi container. Also its important to consider interlacing, when they copy interlaced source material it reproduces a black scan line exactly where it would be in its given field, which looks great until you try and encode it down to x264 or other codec as this will give you interlacing artifacts most likely; however if the source is progressive then that is not an issue.

Moreover the 2 coders in 2.6.x are fully back compatible in 2.5.x? Can I code in 2.6.x and read in 2.5.x without (theorethical) problems?

Should not be an issue unless for some reason you were using FF HuffYUV and did not have the FF libraries installed, but it may still work

[AQUAR]

Just to add to mr_blonde778's reply.

HuffYUV is a very fast lossless compression algorithm.
Great for avoiding 'dropped frames' when capturing analog video (vhs, 8mm etc) whilst preserving all captured detail.
It is a compressor so raw video will be reduced in size (a fair bit). But its fully reversible on decompression meaning it will resolve to the original raw video (ie pixel to pixel is preserved).
Kind of a RAR file for video!