Looking for video compression software (Windows)

[Terrum]

Hi, I'm new to the forum and hoping someone can help!

I'm looking for a free/paid video compression software (free preferred, paid can be any price) for Windows that will compress a video 'automatically', or with little human interaction, that will compress a video's file size but not lose any noticeable quality in video. I have a lot of videos that have a bitrate higher than the actual visible output (such as VHS tape re-encodes) that are like 2GB or more when they come back as at at a maximum of 300mb when I upload and re-download off YouTube with hardly any visible video loss. I say 'automated' because I know there probably isn't a software that does this without having to configure it greatly, but perhaps there's a software that shows the video side-by-side so you can visibly adjust the quality (the original video on the left, and the quality adjusted one by the software on the right, for example)

Codecs the software supports are not that important, as long as it plays in VLC and, preferrably but not required, WMP. I recently found out about MiniCoder which does a great job from test videos (making high quality MKV videos with 30 minute long durations at ~480p at a crazy low file size of ~50MB each). But MiniCoder is far too advanced for me.

Hope someone can help or recommend me something. Many thanks for reading!

[poisondeathray]

eg. handbrake, staxrip, ripbot, megui , many others

But VHS is completely different source material - it's usually very noisy and difficult to compress. What I'm trying to say is different types of source material will compress differently. There is no 1 blanket solution for everything. If you have a very clean cartoon with lots of duplicate frames, it will compress many times more easily than a typical VHS source. So in those scenarios, you might filter (denoise, degrain, etc..) it to improve compressiblity

Nothing will show you an accurate side by side view - because you need to actually encode it to get accurate results. A preview might show you uncompressed results (pre-compression but post filtering) - not that actual end result

[Terrum]

eg. handbrake, staxrip, ripbot, megui , many others

But VHS is completely different source material - it's usually very noisy and difficult to compress. What I'm trying to say is different types of source material will compress differently. There is no 1 blanket solution for everything. If you have a very clean cartoon with lots of duplicate frames, it will compress many times more easily than a typical VHS source. So in those scenarios, you might filter (denoise, degrain, etc..) it to improve compressiblity

Nothing will show you an accurate side by side view - because you need to actually encode it to get accurate results. A preview might show you uncompressed results (pre-compression but post filtering) - not that actual end result

Thanks a lot for your answer! That's a really great and informative answer and helped quite a lot!

If this is the case, do you know if you could recommend me generic settings for video compression with any of the softwares you have provided, such as Handbrake? Such as YouTube's method, as it seems to compress any kind of video no matter how much it is moving or how still it is (such as cartoon backgrounds) with hardly any quality loss and a great file size.

Many thanks once more for your really helpful answer!

[poisondeathray]

If this is the case, do you know if you could recommend me generic settings for video compression with any of the softwares you have provided, such as Handbrake? Such as YouTube's method, as it seems to compress any kind of video no matter how much it is moving or how still it is (such as cartoon backgrounds) with hardly any quality loss and a great file size.

That's not true with youtube - if you have difficult to compress content (e.g. noise, or lots of motion like a video game or sports video), it will degrade substantially more than say, a static interview shot. The reason is youtube caps their bitrate - they allocate a maximum level which is quite low (bandwidth costs money)

Difficult to compress source material requires more bitrate to achieve a similar level of perceptual quality. Filesize = bitrate * running time. Thus larger filesizes required for higher complexity source material

With handbrake, you can try their "quality" based encoding mode and the filesize will be proportionally larger with difficult to compress material, smaller with easy to compress material. Most people use around 16-22 rate factor , lower values yield higher quality (larger filesizes)

[LigH.de]

This "constant rate factor" (CRF) method is only available for rather current codecs like x264; earlier codecs like DivX / Xvid did not yet provide this method. Many hardware players should already support the H.264 (MPEG-4 AVC) video format, but some may require that some constraints are respected. The HEVC format, created by e.g. the x265 encoder, is even more efficient but not yet widely supported, and encoding HEVC takes a lot more time than encoding AVC. So I would recommend x264 as the currently best compromise of a both fast and efficient video encoder with a compatible result (depending on the constraints you set up to respect player limits).

[Terrum]

If this is the case, do you know if you could recommend me generic settings for video compression with any of the softwares you have provided, such as Handbrake? Such as YouTube's method, as it seems to compress any kind of video no matter how much it is moving or how still it is (such as cartoon backgrounds) with hardly any quality loss and a great file size.

That's not true with youtube - if you have difficult to compress content (e.g. noise, or lots of motion like a video game or sports video), it will degrade substantially more than say, a static interview shot. The reason is youtube caps their bitrate - they allocate a maximum level which is quite low (bandwidth costs money)

Difficult to compress source material requires more bitrate to achieve a similar level of perceptual quality. Filesize = bitrate * running time. Thus larger filesizes required for higher complexity source material

With handbrake, you can try their "quality" based encoding mode and the filesize will be proportionally larger with difficult to compress material, smaller with easy to compress material. Most people use around 16-22 rate factor , lower values yield higher quality (larger filesizes)

Well, with all the videos I'm using I haven't seen a noticeable change in quality with the type of videos I am using. So would you know what such generic settings I should or could use with Handbrake to have such? Otherwise I'll try the "quality" based encoding mode you were talking about

[LigH.de]

The CRF method of the x264 encoder, provided by Handbrake as "constant quality" mode, tries to ensure that the quality loss during the reencoding stays below a certain threshold. If the quality loss is low enough, you won't notice any (the conversion is then "visually transparent", means, you can't tell apart the original from the copy). How small the CRF needs to be for you not to notice any loss, is your personal level of tolerance. Test it, try to convert a few movies with a few different CRF values, and discover how much loss you can accept before you rate it subjectively as annoying.