virtualdub with wmv suport

[DJ_Izumi]

Zoom on that pretty picture.
Does it look familiar? Like, uh, JPEG?

Hey,

Ummmm, you are retarded. All forms of video and image compression are based on the same basic concept. Abusing the failures of the human mind to comprehend certian aspects. For instance, in a Mpeg1 and 2 the resolution of the chromonince (sp?) is actually scaled down to one quarter of the actual encoding resolution as the mind dosn't really notice. (Okay, you can change this value, but most decoders and encoders can't understand it) Also, area's are scanned for what's the same, what isn't the same, what moves and what dosn't move. This is why a solid white image saved to a Jpeg looks PERFECT and is really small, all it is is one big block of white. The more detailed the image is, the more data is needed to keep it look the same. In this process, the image is pretty much cut up into a little grid (Hence why you see 'squares' in compressed images and video). I congradulate you on realising th at Mpeg and Jpeg utalize a similar method of compression... You'd almost think that the Motion Picture Experts Group (Mpeg) and Joint Photographic Experts Group (Jpeg) were related in some... strange way... But don't let anyone in on the fact that you know... they might get you!

[danielwang]

Some additional info I might add.

Zoom on that pretty picture.
Does it look familiar? Like, uh, JPEG?

Hey,

Ummmm, you are retarded.

Dammit, Phade! I think it might become a trend to say that now...

All forms of video and image compression are based on the same basic concept. Abusing the failures of the human mind to comprehend certian aspects. For instance, in a Mpeg1 and 2 the resolution of the chromonince (sp?) is actually scaled down to one quarter of the actual encoding resolution as the mind dosn't really notice.

There are other methods that work mathematically, though. You're right, almost any lossless codec will use what you said, "perceptual compression" in order to trim the image down.

(Okay, you can change this value, but most decoders and encoders can't understand it) Also, area's are scanned for what's the same, what isn't the same, what moves and what dosn't move. This is why a solid white image saved to a Jpeg looks PERFECT and is really small, all it is is one big block of white.

The perfect white block can be stored as a simple value (a solid), and so can perfect gradients. As for motion, it started with deltraframing, which tracked only the changes. Windows Media scans the image (frame) for similar fields, used that information to track and comrpess object, and then subtract the image later.

The more detailed the image is, the more data is needed to keep it look the same. In this process, the image is pretty much cut up into a little grid (Hence why you see 'squares' in compressed images and video). I congradulate you on realising th at Mpeg and Jpeg utalize a similar method of compression... You'd almost think that the Motion Picture Experts Group (Mpeg) and Joint Photographic Experts Group (Jpeg) were related in some... strange way... But don't let anyone in on the fact that you know... they might get you!

Thanks, DJ_Izumi.

There's always Motion JPEG (MJPEG), which sounds confusing like a MPEG (or joint) standard...

[DJ_Izumi]

There's always Motion JPEG (MJPEG), which sounds confusing like a MPEG (or joint) standard...

Hey...! You got Jpeg on my Mpeg!
No! You got your Mpeg on my Jpeg!
Hey... this is pretty good.

Quick, name what comercial that's a rip off.

[Tab.]

Not to be confused with Mpeg-J

[Arigatomina]

Or use DirectShowSource in AVIsynth like Tab told you...

Is this in the guides? Is it a script that would make wmvs readable in Vdub? And if it can do that, but isn't in the guides, are there any other scripts you have to add in order to open a wmv file using that command?

[/ignorant newb]

[Tab.]

There are two standard ways of processing video info through windows, the older video for windows (vfw), and the newer directshow, both of which basically do the same thing. Usually you use vfw for editing and processing just because that's traditionally what it's use has been, and this is what avisynth and virtualdub are based on. To read stuff in through vfw you use the avisource() command in avisynth. In the case that the video isn't standard video for windows (avi is the only container that is), you need to use directshow to read it, which can parse and decode these streams. The same goes for avi files with directshow-only decoders such as DV type 1. With directshowsource(), though, you have to supply the framerate of the video being fed.

Once you have the clip in avisynth it's all the same, no matter what source.
The only thing is avisynth can't handle variable framerate files, which some rm and wmvs may be.

[danielwang]

VirtualDumb, yeah it's a great tool, but it sucks in one way:

When you try encoding "Div-X " 3.11 alpha, Virtual Dumb pops up a message box saying it's an illegal ripoff. Ok, it's an old version, and the newer ones doesn't do that with DivsuX 5, but stilllll

Why does it say it's illegal ripoff! OMGWTF they think they're doing

[danielwang]

There are two standard ways of processing video info through windows, the older video for windows (vfw), and the newer directshow, both of which basically do the same thing. Usually you use vfw for editing and processing just because that's traditionally what it's use has been, and this is what avisynth and virtualdub are based on. To read stuff in through vfw you use the avisource() command in avisynth. In the case that the video isn't standard video for windows (avi is the only container that is), you need to use directshow to read it, which can parse and decode these streams. The same goes for avi files with directshow-only decoders such as DV type 1. With directshowsource(), though, you have to supply the framerate of the video being fed.

Once you have the clip in avisynth it's all the same, no matter what source.
The only thing is avisynth can't handle variable framerate files, which some rm and wmvs may be.

Great explanation.

Variable framerate; Never thought that might the problem with my vids

Thanks.

[Tab.]

This is just far-fetched guesswork, but I'm thinking it might be because DivX 3 is an illegal ripoff.

[Scintilla]

All forms of video and image compression are based on the same basic concept. Abusing the failures of the human mind to comprehend certian aspects. For instance, in a Mpeg1 and 2 the resolution of the chromonince (sp?) is actually scaled down to one quarter of the actual encoding resolution as the mind dosn't really notice.

There are other methods that work mathematically, though. You're right, almost any <b>lossless</b> codec will use what you said, "perceptual compression" in order to trim the image down.

Um... if that's what it does, then it's not lossless, now is it?

Or use DirectShowSource in AVIsynth like Tab told you...

Is this in the guides? Is it a script that would make wmvs readable in Vdub? And if it can do that, but isn't in the guides, are there any other scripts you have to add in order to open a wmv file using that command?

Don't believe this is in the guides. It should work like this:

DirectShowSource("C:\mypath\myvideo.wmv")

I think you might also have to indicate the frame rate, but I don't know for sure, since for some reason, I can't get this to work at all for WMVs on my computer. So if there's some conversion I really really have to do (which there isn't), I'll use EO Video.