WMV to DVD, a peculiar case

[buttfacepoop]

Hello,

I have several gigabytes of 350Kbs 320X240 resolution WMV9 encoded video (read, so-so quality) that I'm trying to author a DVD from. The total time is between 28 and 35 hours, 7 separate files.

Since the resolution of the WMV is about half the resolution of a DVD file, I'd rather encode it as an SVCD file (this should create smaller size files right). However, since I have several gigabytes of video, I would like to put it on a DVD disc to minimize disc swapping.

My end goal is simply to have a couple (i'm hoping around 7, but up to 10) authored discs that I can stick in any DVD player to play.

1) Is my best option to simply burn several SVCD discs? (The maximum these can hold is 74 minutes right? Which means a lot of cds....)

2) Is it possible to re-encode the video into SVCD format and burn it to a DVD while still maintaining compatibility with standalone DVD players?

3) Should I just encode it to DVD format, though there is a 2 hour max? If so, to reencode is the best way to use AviSynth w/ QuEnc?

4) Any other ideas? :confused:

Thanks for the help!

[CWR03]

The 74-minute limitation is only for an audio CD. For SVCD, a file would be approximately 1 gig per hour. As far as a DVD, a standard single layer will hold two hours of video burned in a standard format to be compatible with a standalone player, regardless of the original format or file size.

If you're new to DVD burning and want something drag-and-drop, I recommend Roxio DVD Builder. It should be compatible with WMV (I've never tried it, but virtually everything else I've dropped into it has worked), and it has default menus that you can title easily to match the individual files.

[gircobain]

In case you have a DVD burner, IMHO you are better off encoding at half D1 (352x240) and authoring them as DVD

[buttfacepoop]

CWR03:

So the maximum amount of video while keeping compatibility with DVD players is 2 hours? Using an extremely low average bitrate of 1100kbs, I have been able to fit about 3.5 hrs (since the source is such low quality in the first place, the difference between the source and the output isn't too much). A DVD created at this bitrate will not be compatible with standalone DVD players?

gircobain:

I want to keep an aspect ratio of 4:3. Will encoding at half D1 (352x480) maintain this aspect ratio (do I just have to author them as 4:3 DVD?). I'll give this a shot tonite. With this half resolution I should be able to up the bitrate a little bit to something more acceptable.

Thanks for the help guys. :thanks:

p.s. So far, I've been able to use graphedit to extract the wma audio to wmv, and been able to use AviSynth & QuEnc to reencode the video @ 1100kbs. Thankfully, the source and the m2v look like they have the same length, so I think remuxing should work fine.

[Kika]

By the way: 352x240 isn't Half D1 which has the resolution of 352x480. ;)

[CWR03]

So the maximum amount of video while keeping compatibility with DVD players is 2 hours? Using an extremely low average bitrate of 1100kbs, I have been able to fit about 3.5 hrs (since the source is such low quality in the first place, the difference between the source and the output isn't too much). A DVD created at this bitrate will not be compatible with standalone DVD players?


I specifically said standard DVD is max two hours - even the DVD Rebuilder has the option for high compression, and I've been able to squeeze as much as three hours of low-quality files onto a disc. I've never tried one on a standalone player, though.

[JoeShrubbery]

Um... I routinely cram about 5 hours of video at Half D1 resolution with about a bitrate of ~1700kbps.

Do the math, 1700kbps + 192kbps for the audio is 1892kbps collectively. That's just about 236.5 kiloBYTES per second. 14190KB per minute. 851400KB per hour. That's ~831 megs per hour of video. You get about 4400megs of raw data space available on a standard single layer DVD, very roughly speaking.

4400megs / 831megs per hour = ~5.3 hours

That's theoretically about 5.3 hours of video on the one disc at that bitrate, although max runtime is decreased a bit because you lose some space to overhead (menus, etc.).


The original poster said all the clips that he'd be converting were at 320x240 as is. Quarter D1 (352x240) is allowed by the DVD specs, so he could encode at that and use an even lower bitrate to attain an even longer runtime than what I've just described.

In terms of flexibility, your suggested Avisynth & QuEnc option is definitely a strong contender for getting everything encoded the way you'd like, although when dealing with a large number of clips I strongly suggest looking into some sort of batch automation.

[buttfacepoop]

Thanks for the help Joe.

Am I right in thinking that encoding at Half D1 resolution just reduces encoding time rather than saves disc space? After all, the average bitrate is what controls the final file size.

So does Half D1 encoding, in addition to saving encoding time, somehow increase the quality of the final product (I can't imaging how - maybe increasing the vertical quality compared to the horizontal quality???)? Does Half D1 encoding work better than Full D1 encoding because of the reduced bitrate or something?

Since my original post, I realized that I was allocating way too much space for audio, so for this 3:45 WMV file I'm doing now w/ QuEnc I chose a bitrate of 2200kbs.

Thanks for your help.

[JoeShrubbery]

Half D1 is, surprise surprise, about half the resolution of Full D1 (and so has less fine detail than Full D1, provided that detailed was there in the first place). Yes, it takes less time to encode, but that's more of a side effect of having significantly fewer pixels to have to work with.

Fewer pixels means less data to be compressed, hence why you can use lower bitrates when encoding at that resolution without overcompressing the image and getting those unwanted nasty video artifacts. An even lower resolution (Quarter D1 at 352x240) has even fewer pixels to work with and so can handle being compressed at even lower bitrates without adverse effect, theoretically. You start to get diminishing returns fairly quickly at such low bitrates I find. As the resolution goes down, any artifacts that do crop up end up being much more visible as the artifact takes up a higher percentage of the total image.

Now given that your source files are all at 320x240, encoding at Full D1 (720x480) is just out and out overkill. You can't create detail that's not there so upsizing it that much is pointless and wasteful in both encoding time and needed bitrate. Encoding at Quarter D1 (352x240) would match your clip's original resolution the best, but as I mentioned any artifacts that do show up (and there probably will be a few here and there) will stand out more as the resolution goes down. If you upsize your clips to Half D1 (352x480) then you do have to use a higher bitrate and you're not gaining any detail, but any video artifacts from the mpeg2 encoding that show up will be smaller in the encoded image, and thus less noticable when watching. Of course any artifacts already present in your source files will still get passed on regardless of any of this.

Hope this help explain things a bit more, good luck!

[buttfacepoop]

Half D1 is, surprise surprise, about half the resolution of Full D1 (and so has less fine detail than Full D1, provided that detailed was there in the first place). Yes, it takes less time to encode, but that's more of a side effect of having significantly fewer pixels to have to work with.

Fewer pixels means less data to be compressed, hence why you can use lower bitrates when encoding at that resolution without overcompressing the image and getting those unwanted nasty video artifacts. An even lower resolution (Quarter D1 at 352x240) has even fewer pixels to work with and so can handle being compressed at even lower bitrates without adverse effect, theoretically. You start to get diminishing returns fairly quickly at such low bitrates I find. As the resolution goes down, any artifacts that do crop up end up being much more visible as the artifact takes up a higher percentage of the total image.

Now given that your source files are all at 320x240, encoding at Full D1 (720x480) is just out and out overkill. You can't create detail that's not there so upsizing it that much is pointless and wasteful in both encoding time and needed bitrate. Encoding at Quarter D1 (352x240) would match your clip's original resolution the best, but as I mentioned any artifacts that do show up (and there probably will be a few here and there) will stand out more as the resolution goes down. If you upsize your clips to Half D1 (352x480) then you do have to use a higher bitrate and you're not gaining any detail, but any video artifacts from the mpeg2 encoding that show up will be smaller in the encoded image, and thus less noticable when watching. Of course any artifacts already present in your source files will still get passed on regardless of any of this.

Hope this help explain things a bit more, good luck!

That was great! :thanks:

I've settled on Half D1, 2200kbs bitrate.