Applications to trim/join videos without losing any quality?

[immortal192]

I am looking for an application (or a set of applications) that allows me to trim and/or join files without losing any quality whatsoever (100% quality, not 99%). Besides this absolute requirement, ideally it would be free, easy to use, and portable (these are not necessary requirements).

Files I'm mostly dealing with are MKV, AVI, WMV, and FLV.

I did some brief googling and if I'm not mistaken, it seems that such a kind of application I'm looking for is usually not able to trim to the exact frame--is there one that does? If exact frame was a requirement and I cannot do it without losing a bit of quality, what kinds of applications would minimize that lost in quality?

Thank you so much.

[Baldrick]

Most free can't cut exactly like avidemux, xmedia recode, video to video converter, etc.

If you want to cut exactly then try Solveigmm Video Splitter, Videoredo h264 suite, tmpgenc mpeg editor. They will try to do a smart rendering and just reconvert at the cut.

[pandy]

Depends on context (video/audio codec and container).
I would give a chance to ffmpeg...

Bellow not tested script - i'm not the author (sorry for not crediting source either, but it should be not so difficult to 'google/bing/duckduck2go/whatever you like' it).

:: Edit the line below to match your path to the ffmpeg executable.
set path2exe=C:\example\ffmpeg.exe
(for %%i in (*.mp4) do @echo file '%%i') > list.txt
%path2exe% -f concat -i list.txt -c copy output.mp4
del list.txt

[KarMa]

....... without losing any quality whatsoever (100% quality, not 99%). Besides this absolute requirement.....

I cannot do it without losing a bit of quality, what kinds of applications would minimize that lost in quality?
Thank you so much.

You don't seem to know what you want. If you want 100% quality then you have no choice but to go lossless which will make the videos many times bigger. For example an hour of SD YUY2 lossless can be 20-30GB. HD will be many times bigger than that.

If you can deal with a bit of loss at the cut points you can use Virtualdub (free, for the Xvid AVI), and Videoredo for H.264/MPEG2

Files I'm mostly dealing with are MKV, AVI, WMV, and FLV.

Ok but those are containers. I'm going to guess you are wanting to cut up H.264, MPEG4 Pt2 (Xvid), and one of the WMV standards.

[PuzZLeR]

Yes, AVI and MKV, for example, are containers. They could contain a number of different video formats which could each need a different strategy.

If what's in the containers is indeed H.264, DivX, Xvid, MPEG-2, then keep in mind that these are highly compressed formats with key frames apart from each other (look into GOPs). You can cut and join without loss, but it would not be frame accurate. You can cut and join with frame accuracy, but it would not be without loss if it's not on a key frame as you would need to re-encode the few frames at the seams to ensure and maintain GOPs.

In other words:
It's 100% no loss with no guarantee of frame accuracy.
or
It's ~99.9+% no loss, more or less, with frame accuracy.

I would choose the latter personally. It's a small sacrifice.

There is no way around such formats. And that would also include WMV, which would also be the most difficult to edit due to Microsoft intentionally discouraging editing (something to do with protecting any DRM).

And each would likely need an editing tool of their own more dedicated to the format.

As per DV and lossless, this would be easy - 100% quality and frame accuracy - and any editor worth its salt should handle these with minimal effort if you know what you're doing (ex: smart render, staying on the same color space, etc.)

[immortal192]

Yes, AVI and MKV, for example, are containers. They could contain a number of different video formats which could each need a different strategy.

If what's in the containers is indeed H.264, DivX, Xvid, MPEG-2, then keep in mind that these are highly compressed formats with key frames apart from each other (look into GOPs). You can cut and join without loss, but it would not be frame accurate. You can cut and join with frame accuracy, but it would not be without loss if it's not on a key frame as you would need to re-encode the few frames at the seams to ensure and maintain GOPs.

In other words:
It's 100% no loss with no guarantee of frame accuracy.
or
It's ~99.9+% no loss, more or less, with frame accuracy.

I would choose the latter personally. It's a small sacrifice.

There is no way around such formats. And that would also include WMV, which would also be the most difficult to edit due to Microsoft intentionally discouraging editing (something to do with protecting any DRM).

And each would likely need an editing tool of their own more dedicated to the format.

As per DV and lossless, this would be easy - 100% quality and frame accuracy - and any editor worth its salt should handle these with minimal effort if you know what you're doing (ex: smart render, staying on the same color space, etc.)

I appreciate the the detailed explanation. Essentially what you're saying is that if I want 100% frame rate accuracy, the video quality would be 100% for the containers you've mentioned except for a few frames at the start and at the beginning of the video, which would be ~99%. Baldrick mentioned smart rendering and said that it reconverts the cut--does that solve this problem and allow me to have 100% quality for the entire video (including the start and end frames)? If not, how is smart rendering different from what you've mentioned? Now that I know that as high as ~99.9+% quality is kept and affects only a few frames at the start/end of the video (while it is 100% for the rest of the video), it is adequate for my situation but now I am just asking about smart rendering out of curiosity.

For my purposes it seems best to use the tools recommended by Baldrick (Solveigmm Video Splitter, Videoredo h264 suite, tmpgenc mpeg editor). What I need to do is use an application to determine what containers the videos I want to edit are and then find which of the 3 programs support lossless trim/joining?

Thanks.

[smrpix]

how is smart rendering different from what you've mentioned?

Smart rendering is what he mentioned. Since there's no numeric measurement of quality, 99% is a meaningless figure. Suffice it to say it's unnoticeable during normal playback. As for the number of frames re-encoded, that depends on how far your edit is until the next keyframe.

[PuzZLeR]

Hi again, I hope I didn't mislead with the "99.9%". What I meant is that 99.9% of the frames can be left untouched - they will be at the original 1:1 quality. The 0.01%, or whatnot, is how many frames will be affected. These were just hypothetical figures.

In other words, to cut/join such highly compressed formats, if done right, and you want to assure frame accuracy, is that only the few frames at the cut/join points will be affected, and that will depend on GOP size, etc., and quality drop will depend on encoding method, etc. But it's only a few frames, and frames that you won't really care for, or notice, in re-encoding. The rest of the video will be at 100% quality of the original.

Hopefully it's clear now.

And the method of doing this is typically called smart rendering. You only work on the parts that need it. Other editors may have different names for it, but it's generally the same concept.

Essentially what you're saying is that if I want 100% frame rate accuracy, the video quality would be 100% for the containers you've mentioned except for a few frames at the start and at the beginning of the video, which would be ~99%.

Careful on the terms here.

I meant the video formats, not the containers they're in. I meant what's inside the container.

Also, I meant "frame accuracy", which is different from "frame rate accuracy". What I meant by "frame accuracy" is that you can cut to the exact frame you want, but would need to re-encode a few frames near the cut/join if they're not key frames. If you don't do this, and you cut/join on a non-key frame, then you won't have frame accuracy. The cut will take place a few frames away from where you wanted it cut. You may see a bit of the previous scene, etc.

A straight cut would need to be on a key frame, otherwise your cut will be placed on an adjacent key frame, unless you decide to re-encode that gap. This is how highly compressed formats work in the first place. Their file size is compressed because of their key frames.

However, formats like DV and lossless are easier to cut and join.

For my purposes it seems best to use the tools recommended by Baldrick (Solveigmm Video Splitter, Videoredo h264 suite, tmpgenc mpeg editor). What I need to do is use an application to determine what containers the videos I want to edit are and then find which of the 3 programs support lossless trim/joining?

Certainly. Baldrick knows the tools likely better than anybody here. However, he, and the Forum, would give even better info if you tell us exactly which format is in those containers.

A good tool is MediaInfo (free). You can post, in text mode, what video format is in the MKVs, AVIs, etc, and we can recommend tools for you with absolute certainty this way.

[immortal192]

Ok, I used MediaInfo and loaded my entire library--the most popular formats are MPEG-4, Matroska, Windows Media, AVI, and Flash Video.

On my own time I'll need to read into the relationship between containers, formats, and codecs as well as transcoding, encoding, and transmuxing.

[deton24]

MKVMerge can trim and join videos losslessly without re-encoding, instantly (just HDD limits you).
It just copies segments into one mkv container.

You can even send mkv files to YT.
It's hell a lot faster than e.g. MP4 joiner.
And it's also free.