AVI Files Smaller Than WMV Files

[tutorpros]

I want to digest your replies before I answer.
Perhaps manono was right that I ask questions too fast before I read your replies well enough.

[jagabo]

Regardless of codec, frame rate, frame size, and bit depth, the equation I gave earlier is universal:

size = birtate * running time

It's the definition of bitrate.

Note that there are two main methods of encoding with x264: quaility and bitrate. With quality based encoding you specify the quality you want and the encoder uses whatever bitrate is necessary to deliver that quality. With bitrate based encoding you specify a bitrate and the encoder delivers whatever quality it can for that bitrate. when the bitrate of the two methods matches the quality is essentially the same.

[tutorpros]

Okay

[tutorpros]

I apologize for my late reply.
I've come to realize that this stuff is just beyond me.
It's not anyone's fault but my own, and I sincerely apologize for those of you who took the time to try and help me.
Perhaps, in time, I will get a firmer grasp on the terms that I don't understand, but for now, it just frustrates me that it's taking so long to learn.

[pandy]

Hi,
at first look I noticed that wmv bithdepth is 32 or 96??? is it ?possible? meanwhile x264 is just 24. That is first look. Also bitrate of x264 isn't half but almost 1/4. Don't worry about bithdepth, I think 24 is usual and hardly noticed, and most monitors is using it. If you can set ref frames to 8 it should be even better and also me method to umh, will improve quality in terms of size and quality IMHO. But encoding will last bit longer. That is first fast overlook.

Mediainfo may produce garbage, it is not 100% reliable info tool. Reported 32 bit may be related to format at the wmv decoder (AFAIR Windows decoder is used to decode wmv and this decoder may report 0RGB or 0YCbCr i.e. 32 bits). 32 vs 24 bit is common due memory and overall system architecture - 32 bits i.e. 4 bytes i.e. one longword, 24 bits i.e. 3 bytes i.e. 3/4 longword - to improve speed/performance longword or even higher factor of memory aligned transfers may be preferred (and they usually are preferred) - 8 bits is just ignored.
I would recommend to follow BD limitation - increasing ref may not provide significant quality gain and for sure will extend encoding time. In most cases compatible ref=3 is sufficient.

I apologize for my late reply.
I've come to realize that this stuff is just beyond me.
It's not anyone's fault but my own, and I sincerely apologize for those of you who took the time to try and help me.
Perhaps, in time, I will get a firmer grasp on the terms that I don't understand, but for now, it just frustrates me that it's taking so long to learn.

It is nobody fault - your confusion is justified. Just use decent modern video codec (H.264) and decent modern container (MP4, MKV) - WMV is dead - this is Microsoft decision - they patented WMV, they forbid to use this technology freely outside Windows and at some point stopped developing and promoting WMV - your experience is just confirmation of this. Don't search problems where they not exist.

[blud7]

1. The windows media codec is VC-1 for video and wma for audio, the container which 'holds' the video and audio is 'wmv'.

2. The h264/AVC codec for video and mp3 codec for audio is stored by your program in an 'avi' container.

So the container 'holds' or stores the video and audio in one file.

3. Storing h264/AVC video with mp3 audio in an avi container is non-standard and a force-fit.

4. h264/AVC is way better than VC-1 and many other codecs because it produces higher quality output when compared to a similar file size (file size is determined by bitrate)

5. It is advisable to store your h264/AVC video with aac audio in a superior container like mp4 or mkv.

6. mkv is more robust than mp4, but mp4 is more widely 'compatible' than mkv, although mkv is widely popular

7. I would suggest you tinker around with tools like Handbrake or VidCoder (recommended) which are free to download and use and will give good results for beginners.

[redwudz]

Not Latest Video News. Moving to Newbie/General Forum.

Moderator redwudz

[pandy]

1. The windows media codec is VC-1 for video and wma for audio, the container which 'holds' the video and audio is 'wmv'.

It may be not VC-1 (VC-1 is fork of WMV9 - variable frame rate is not allowed in VC-1 and many other changes like interlaced support etc etc) - IMHO Microsoft use ASF container (however extenssion of file is WMV)

[blud7]

1. The windows media codec is VC-1 for video and wma for audio, the container which 'holds' the video and audio is 'wmv'.

It may be not VC-1 (VC-1 is fork of WMV9 - variable frame rate is not allowed in VC-1 and many other changes like interlaced support etc etc) - IMHO Microsoft use ASF container (however extenssion of file is WMV)

Was just using his sample to illustrate the point but correct.

[tutorpros]

Well, I now understand what containers and codecs are.
And, a decent package is h264/AVC with aac audio in a MKV container.
Some of my frustration is gone.
Thank you.

I'll tinker around with various bitrates, frame rates, etc., to get good quality, but with a smaller file size.

[tutorpros]

Since my mind seems sharper today, I'm going to press my luck with another question...

I create my videos with snippets from mainly YouTube videos. They are for privately use only, so there's no copyright infringement.

In other words, I put the downloaded YouTube snippets into a movie maker program, and then create a new video.

I can download the YouTube videos in many different formats, but I've been using an MP4 container (I'm not sure about the codecs) and with MP3 audio.

If I intend to put my finished project video in h264/AVC, AAC audio, in an MKV container, does it matter what format I download the YouTube snippets in?

If so, then what format would you suggest I download them in?

[abolibibelot]

The source videos (which have already been compressed several times before they end up on YouTube...) will be re-compressed anyway when you export your work, so the source format doesn't really matter, as long as it's properly recognized by the movie maker program. But choose the one with the highest bitrate / size, to mitigate the generation loss as much as possible... And if at all possible, by all means, get a given footage from better source than YouTube, if you care about quality.

(Some videos on YouTube must have been re-compressed 20 times or more if people like you are making their own videos by shuffling bits of other videos downloaded from YouTube... at some point it's akin to someone trying to get nutrients and energy by eating feces produced by five fellow humans in a row, human centipede style ! :^D But, when you think about it, a large part of today's “culture” works just that way... people fed with garbage grow up to produce even more horrendous garbage...)

[tutorpros]

That is a great response.
I was leaning toward that point, but you confirmed it.
I guess there's only so much I can do.

I ended up downloading the YouTube videos into mp4, and then used the mp4 snippets to create a video in AVI.
The difference in quality between WMV and AVI in the final product was bearly noticeable (to me), but the file size was cut in half.

I learned from previous responses that the video/audio that I use in an AVI container is a forced fit, but all my videos play without incident.

What can happen with this forced fit format?
Will my videos breakdown after time or somesthing?

[KarMa]

Last I remember, AVI was never designed to deal with B-frames and so requires none linear frame orders when encoding and decoding (that's my basic understanding). Encoders like Xvid and x264 commonly include B frames but need these hacked frame orders and the decoders need to expect these frame orders when decoding AVI containers. In spite of requiring this hack, it seems like Xvid (and general MPEG4 pt2 encoders) is generally used in AVI containers with minimal problems but Xvids usage of B-frames is simple. With x264 (a encoder for the H.264 standard), the heavier reliance on B-frames and their higher reference frames and massive GOPs could cause many more problems in AVI. Personally I use x264 in AVI when capturing VHS tapes but I just use I and P frames with B-frames disabled. B-frames are usually the most bitrate efficient frame and so that's why they exist but they increase overall complexity.

It is nobody fault - your confusion is justified. Just use decent modern video codec (H.264) and decent modern container (MP4, MKV) - WMV is dead - this is Microsoft decision - they patented WMV, they forbid to use this technology freely outside Windows and at some point stopped developing and promoting WMV - your experience is just confirmation of this. Don't search problems where they not exist.

VC-1 is an open standard, free to implement but MPEG claims some patents on it. VC-1 files can also be put in MKV, AVI, M2TS, along with WMV. WMV9 is a encoder for VC-1.

[tutorpros]

Now you kind of lost me...

I put h264 and aac into an AVI container. I only intend to view it on my MX media player, and it works fine.
Is it plausible that I wake up one morning, and my file won't be compatible with my player anymore?

[Marsia Mariner]

To whom this may interest......

"AVI certainly wasnt designed to support b frames but on the other hand theres nothing which would make b frames in AVI illegal, the biggest problem isnt AVI but the various applications and APIs like vfw which are designed with zero delay codecs in mind"

==> https://guru.multimedia.cx/avi-and-b-frames/

"Unluckily the DirectShow technology lacks the ability to store PTS/DTS information for media samples too and various techniques are being used to work around this limitation."

==> http://blog.monogram.sk/janos/2008/06/08/b-frames-in-directshow/

Also: https://forum.doom9.org/showthread.php?t=167112

[jagabo]

I put h264 and aac into an AVI container. I only intend to view it on my MX media player, and it works fine.
Is it plausible that I wake up one morning, and my file won't be compatible with my player anymore?

Your current player should continue to work fine. But your next player may not be able to play the files properly. And if you need to edit those videos in the future you may have problems.

As was pointed out, the problem isn't the AVI container itself but rather programs and libraries that read it, most notably, Microsoft's VFW library (used by many editors, and which has not been updated in many years). In the old days, compressed frames were stored in the same order they were viewed. So a program could read the compressed data from frame 1, decompress it, and display that frame. Then read the compressed data for frame 2, decompress it, and display it. Etc. So video frames 1,2,3,4 are stored in the file in that same order, 1,2,3,4. But more modern codecs store the data out-of-order. They may store those same four frames as 1,4,2,3, confusing some programs. Some codecs have kludges to work around this problem (Xvid, Divx, X264 VFW) but not every program/player knows how to deal with them. Or, if you're writing your own program/library you can avoid using VFW and accommodate the out of order frames yourself.

So it's safest to store your h.264 video in a more modern container like MKV or MP4, containers that were originally designed for out-of-order codecs, or at least the programs and libraries that access them have been updated to handle it.

[tutorpros]

I'm starting to understand...and it makes sense because there have been times when I could not edit certain videos.

I've been downloading the YouTube videos into x264 AAC MP4 because it's the lightest file, but the movie maker I use doesn't let me export into MKV or MP4.

I have a converter that will convert my x264 AAC AVI into different codecs and containers including MP4 and possibly MKV.
But I think that I lost more video quality when I tried converting x264 AAC from AVI to MP4.
I haven't tried to switch the codecs because there are too many to choose from.

I also have duplicate videos in WMV containers, but I read further up in the thread that those might be in the same boat as the AVI's, regarding future compatibility.

I can probably download the YouTube videos into AVI but with different codecs, and perhaps export them the same way from movie maker.
If I don't lose quality, would that be a way to go?
And if so, which codecs should I use with AVI?

I don't know what b frames are.

[jagabo]

You always lose quality when reencoding with a lossy codec, even if the new file is 10 times larger. The issue is how much quality you lose. More modern codecs can compress to smaller sizes without losing much quality. A partial list of codecs in increasing ability to compress:

MPEG 1
MPEG 2
Divx, Xvid
h.264, AVC, VP8, VC1
h.265, HEVC, VP9,AV1

Of course, the settings used can make a big difference in the quality and encoding speed. So a bad encoding with h.265 could be larger and/or lower quality than a good encoding with h.264.

[pandy]

VC-1 is an open standard, free to implement but MPEG claims some patents on it. VC-1 files can also be put in MKV, AVI, M2TS, along with WMV. WMV9 is a encoder for VC-1.

WMV9 is video codec existing only on Windows (PC) ecosystem where VC-1 is video codec tailored for broadcast TV - they share some common concepts but VC-1 is standardized and WMV9 not.

[abolibibelot]

I can probably download the YouTube videos into AVI but with different codecs, and perhaps export them the same way from movie maker.
If I don't lose quality, would that be a way to go? And if so, which codecs should I use with AVI?

YouTube only uses MP4/H.264 and WEBM/VP8 (MP4/WEBM are the containers, H.264/VP8 are the codecs) for the highest quality streams (both being roughly equivalent in quality at a similar bitrate, but MP4 is more widely supported) ; so if a program or website lets you download a video from YouTube in AVI (most likely Xvid – as AVI can be a lot of things, for instance uncompressed RGB AVI can fill an entire HDD in a matter of minutes) it will have been transcoded with parameters over which you have no control, parameters which, like those used by YouTube, are “good enough” for playback and for people who don't see and/or don't care about the difference, but which will significantly reduce the quality you start with for your own creations. So you're better off downloading the “native” MP4 (which, again, is not native at all as it has been savagely transcoded by YouTube, before that by the editing software if it's not a straight-from-camera footage, by the camera itself, or by the DVD/BR authoring program... and so on...), and then, if it's not properly recognized, first try to convert it to another format without transcoding (MKV, TS), and only if it fails, transcode it to another format with a high enough bitrate that it won't degrade the visual quality too significantly. Ideally, that kind of intermediate conversion should be done with a lossless format like Lagarith / Huffyuv / MagicYUV / UTVideo, but it might be a tad overkill to work with videos downloaded from YT – or one could argue that, precisely because you start with a rather low quality, that's all the more reason to preserve it as much as possible, if the quality of the end result matters at all.


@KarMa :

Personally I use x264 in AVI when capturing VHS tapes but I just use I and P frames with B-frames disabled. B-frames are usually the most bitrate efficient frame and so that's why they exist but they increase overall complexity.

Isn't it supposed to be a bad choice as a capture format ? Maybe you're using x264 in lossless mode ? Is it really lossless, and how does it compare with specifically lossless codecs like the ones I mentioned above, in terms of compression ratio and speed / CPU usage, and is it editing friendly with those custom parameters you use ? Could you provide your usual command line for that purpose ?

[jagabo]

Lossless x264 is truly lossless (assuming you stay with the same chroma subsampling). It compresses about the same as the other lossless codecs.