Question aboot virtualdub, Xvid, compression, frameserving..

[muttonbuster]

I'm a newbie so please bare with . What I'm trying to accomplish is to hard encode subtitles onto an avi, and then convert the avi into VOB with Nero so I can play it on a stand alone DVD player.

The source video is in Xvid (are Xvid and DivX just different types of compression for an avi?).

Here is the method I read off of a guide:

1) open the source avi with virtualdub
2) add theTextSub 2.13 filter
3) choose full processing mode
4) select video compression

From here I get lost. I realize a lot of this is personal preference, but for just hard encoding the subtitles onto the avi, which would be better: XviD MPEG-4 or DivX 5.2.1? Also how does MPEG-4 differ from MPEG-2?

Secondly (and I know that this is personal preference again) but I'm trying to keep the size of the newly created avi as small as possible. I tried using the sample that I got with subtitler and the smallest I seem to be getting the resulting file is roughly 7 times the original.

I'm not sure if frameserving is another viable option. I don't really understand what the difference is between having virtualdub compress the video and frameserving to another application (if indeed that's what frameserving is).

I was also wondering if it would just be easier to go straight to Nero and just add the subtitle option to the menu. My only concern is to get it to play in a standalone DVD player.

If someone can answer a few of these questions for me, I'd greatly appreciate it. Thanks

[Baldrick]

If you are going to convert to dvd do not recompress it again. You will lose lots of video quality by doing that. Use the frameserver or convert to a less compressed format like huffyuv instead.

And why keep the avi small?? the avi size does not matter at all when you convert to dvd. It is the runtime that effects the dvd output size.

[junkmalle]

which would be better: XviD MPEG-4 or DivX 5.2.1? Also how does MPEG-4 differ from MPEG-2?

They're both MPEG 4 codecs. Xvid is open source, Divx is a commercial product. Quality is roughly the same. I use Xvid because it's open source.

Secondly (and I know that this is personal preference again) but I'm trying to keep the size of the newly created avi as small as possible. I tried using the sample that I got with subtitler and the smallest I seem to be getting the resulting file is roughly 7 times the original.

The size of the output file is determined by the bitrate and the runtime. For example: 1000 kilobits per second * 1000 seconds = 1,000,000 kilobits. Divide by 8 to get kilobytes: ~125,000 kilobytes, or 125 megabytes. Remember the audio will take some room too.

If your intention is to make a Xvid AVI file (for viewing on the computer or a settop Divx/Xvid DVD player) I recommend using Xvid's Quantization mode (contstant quality). Pick the quality you want and encode in a single pass. The file turns out to be whatever size is required to maintain that quality. At a Q value of 2 the output file is almost indistinguishable from the input file. At 3 there will be some degradation but it will be hard to see without looking at still frames. At 4 you will start to see some degredation at normal playback speeds. You can use any value from 1 to 31 where 1 is the best, 31 the worst. You can use decimal values like 2.5 too.

I'm not sure if frameserving is another viable option. I don't really understand what the difference is between having virtualdub compress the video and frameserving to another application (if indeed that's what frameserving is).

If you're making a movie DVD you can frameserve to your MPEG encoder. Rather than saving the intermediate file to you hard drive then opening it with your MPEG encoder, you just have VirtualDub send the frames 1 by 1 to the encoder. No intermediate file, no intermiate compression, no loss of quality from that compression.

[aguillon]

I concur with junkmalle. Frameserving is a much better way of working with avi files in your case.

I ve done this many times. I add an extra step though (step number 2):

1- Open in Virtualdub
2- apply the smart resize filter
3- apply the textsub filter
4- Frameserve
5- Encode

Resizing before adding the subtitles is a step I use for coding in 4:3. It enables me to add a letterbox to my video, then add the subtitles into the black bars of the letterbox, which makes them more readable.

From there, select 'direct stream copy' (if needed, extract audio first and set it aside if you know your encoder does not deal with it very wel), frameserve (this creates a little vdr file which is a wrapper which contains one frame at a time coming from Virtualdub, passing it to the encoder) and encode (I use TMPGEnc). U can use VBR 2-pass for better results.

Lately, however, I ve been using DVD Lab Pro to master my DVDs, which means I can add selectable subtitles directly to the DVD. Hard coding the subtitles is no more needed. It s way cooler to design a DVD where the subtitles can be turned on/off.

[Abond]

Better use AVI2DVD it will make a DVD with selectable subtitles. You don't need to hardcode them.