Please help - Capture Resolution/Size question

[goldtop33]

I'm converting a bunch of 120 minute concerts on VHS to DVD. I'm using a 5900 series JVC S-VHS VCR w/TBC and going into my PC with an ATI 9800 All-In-Wonder Radeon card...

My question is this: I know that to keep the concert at its highest quality - I'd need to split the 120 minute show onto 2 DVDs....

My question is this:

Whats the best way to put a 120 minute concert onto 1 DVD?
From what I've read, I don't know if I should lower my capture resolution from 720x480 to 352x480 - as well as lowerering my bit rate conversion? Or use my max quality capture and shirnk it with DVDshrink or somthing similiar?

I appreciate any help!
Thank you!

[loughary]

My question is this: I know that to keep the concert at its highest quality - I'd need to split the 120 minute show onto 2 DVDs...

I don’t know if that’s necessary true with your source being VHS. I’ve encoded VHS at half D1 with a higher bit rate and the results were great. A lot of people in this forum swear that VHS source isn’t more then 352x480 anyway. If that’s true then I don’t know if you gain anything by capturing at a higher resolution.

[goldtop33]

thanks! - anyone else????

[The_Doman]

I depends a bit what kind of concert footage it is ofcourse, but with 2 pass VBR you should get easily 2 hours on a DVD in full resolution. I personally don't like half D1 at all because I loose too much detail with it. Even with my sources from VHS. But why don't you make 2 versions? One with Half D1 and other with full D1. Just to try what looks best for you.

[goldtop33]

What do you mean by "D1"?

[daamon]

Hi goldtop33,

loughary is leading you in the right direction. The resolution of VHS is 352 x 240, so capturing at 1/2 D1 - 352 x 480 / 576 (NTSC / PAL) is plenty enough.

Any more, means that you have more bits and so will need a higher bitrate to maintain quality. This, in turn, means that you will end up with a bigger file and so can fit less on one disk. You may wonder "why not capture at 352 x 240?". Well, because that's not a valid resolution for DVD (though 352 x 288 is).

Capturing at 1/2 D1, and hence not requiring high bitrates, means that you should easily get 120 mins of VHS onto DVD without (noticeable) loss of quality.

In fact, the lower bitrates have allowed people to get around 4 hours of VHS on to one DVD (bitrate = approx. 2,300kbps). But the you're the judge of the quality.

Hope that helps. Good luck...

[goldtop33]

thats helps tons - thank you so much!

So at a resolution of 352 x 480 - what should my max and target bit-rates be?

Also - should my GOP be open or closed?

[fmctm1sw]

thats helps tons - thank you so much!

So at a resolution of 352 x 480 - what should my max and target bit-rates be?

Also - should my GOP be open or closed?

What you will need is a bitrate calculator. I estimate that will work out to around 3500. You would not want to encode at 720x480 at this bitrate. That is why folks are correctly telling you to use 352x480. Also, 352x240 is a valid (NTSC) DVD resolution. You could lower the bitrate even more and get more time on one DVD. It is up to you though to determine what kind of quality you can live with. Do a search on GOP. I can't remember the advantages/disadvantages of them but I know it's been discussed. Maybe llok up a guide for TMPG, it may be explained in there.

[FulciLives]

Well the exact bitrate depends on the running time as well as the audio format.

For 120 minutes with PCM WAV audio you want to use a BITRATE of 3500kbps for the video. Either do CBR of if you do VBR then 3500kbps would be your AVG. Since you will be using Half D1 resolution (352x480) I say set your MAX to 6000kbps and your MIN to 1500kbps

For 120 minutes with AC-3 or MP2 again it will depend on the bitrate you use. Here you have an option because whereas PCM WAV audio has a fixed bitrate of 1536kbps you can use a range of values for AC-3/MP2 such as 192kbps, 224kbps, 256kbps, 320kbps or 384kbps (and AC-3 can go higher though that would be a waste).

If you do AC-3 then I would use 256kbps for the audio. In that case your video bitrate CBR/AVG would be about 4800kbps

If you do MP2 then I would use 384kbps for the audio. In that case your video bitrate CBR/AVG would be about 4680kbps

Please note that anything over 4000kbps is usually fine at CBR encoding for Half D1 resolution.

- John "FulciLives" Coleman

P.S.
I always do GOP closed.

[vcdforme]

352/240 is a legit DVD resolution.
If you use Tmpgenc, use the wizard and it will immediately
open a window offering you different settings and their
characteristics.
If I were you, being that these are VHS tapes, I would try
the Low Resolution 352/240 mode, 1850 bitrate. It's actually
not bad at all. Try it, if you don't like the 4 hours it offers, go up to
the next offering. Easy.

cheers

[FulciLives]

352/240 is a legit DVD resolution.
If you use Tmpgenc, use the wizard and it will immediately
open a window offering you different settings and their
characteristics.
If I were you, being that these are VHS tapes, I would try
the Low Resolution 352/240 mode, 1850 bitrate. It's actually
not bad at all. Try it, if you don't like the 4 hours it offers, go up to
the next offering. Easy.

cheers

352x240 is not enough resolution for NTSC VHS ... you need 352x480 ... that way you retain your interlaced video ... otherwise using a height of 240 throws out a field which makes a progressive video but also produces "jaggies" in the video.

- John "FulciLives" Coleman

[SatStorm]

Basicly, VHS is ~ 352 x 240/288 in a 720 x 576 canvas (according JVC, the inventor of VHS)

So, we can't capture direct to 352 x 240/288 without loses.
So, what we do? We capture higher so to preserve the info and then when we encode, we also convert (and resize) to a more VHS like framesize.

Inshort terms in theory, we capture at 352 x 576/480, we filter (a neccessary step for VHS - if we wish top results), and we encode to 352 x 576 / 480.
PAL users can also use the lower 352 x 288 framesize, but this is an advance subject with one limitation: You need a standalone to handle this framesize correct, which is not so usual, unfortunatelly.

In praxis, there also the variable of the capture card: The Capture cards don't capture all at the same way. Most cards capture at a specific native framesize and then resize - not always in the best way - in what we ask the to output.

So what we do? We capture in the most "compatible" way for all the type of the capture cards. And this compatible way is 704 x 576/480 or higher. Let's say that 704 x 576/480 is the safe mode about this. For the 60% of the cards it is an overkill (that doesn't harm), for the 25% of the cards is just right and for the rest 15% is a slighty (and unoticable for VHS/SVHS sources) picture loss.

Then we filter and resize to 352 x 576/480 which - we all agree - it is the best and safer (for compatibility with the standalone DVD players) framesize for VHS/SVHS sources converted to mpeg 2 interlace.

SO, in really short terms:
- Capture your tapes @ 704 x 576 (PAL) or 480 (NTSC)
- Filter your source
- Resize it to 352 x 576/480
- Encode it to mpeg 2 @ 352 x 576/480

Then Author and Burn it on whatever you wish....

[goldtop33]

Well the exact bitrate depends on the running time as well as the audio format.

For 120 minutes with PCM WAV audio you want to use a BITRATE of 3500kbps for the video. Either do CBR of if you do VBR then 3500kbps would be your AVG. Since you will be using Half D1 resolution (352x480) I say set your MAX to 6000kbps and your MIN to 1500kbps

For 120 minutes with AC-3 or MP2 again it will depend on the bitrate you use. Here you have an option because whereas PCM WAV audio has a fixed bitrate of 1536kbps you can use a range of values for AC-3/MP2 such as 192kbps, 224kbps, 256kbps, 320kbps or 384kbps (and AC-3 can go higher though that would be a waste).

If you do AC-3 then I would use 256kbps for the audio. In that case your video bitrate CBR/AVG would be about 4800kbps

If you do MP2 then I would use 384kbps for the audio. In that case your video bitrate CBR/AVG would be about 4680kbps

Please note that anything over 4000kbps is usually fine at CBR encoding for Half D1 resolution.

- John "FulciLives" Coleman

P.S.
I always do GOP closed.

For my capture card I use ant ATI AIW 9800 pro. It does not give me a minimum video bit rate - it just gives maximum, average, somthing called moition estimation quality? Aslo - I'm doing the VBR NOT CBR....

Also - I'm capturing straight to MPEG2, VS capturing to AVI - then encoding to MPEG-2 - thus, my audio is MP2. This is what lordsmurf suggests (although I know some people swear by AVI to MPEG2)

[goldtop33]

SO, in really short terms:
- Capture your tapes @ 704 x 576 (PAL) or 480 (NTSC)
- Filter your source
- Resize it to 352 x 576/480
- Encode it to mpeg 2 @ 352 x 576/480

Then Author and Burn it on whatever you wish....

I read that if I'm not going to edit the video - its best to capture directly to MPEG-2????

[dphirschler]

You could try 704x480. Your visual quality will still be quite sharp. In fact, if DVD would allow 640x480, then you could go with that and not really lose any sharpness. You can easily fit an entire 120 min concert on a single DVDR, and it look perfect. The problem for you is since it is a music concert, you probably want to keep the audio uncompressed PCM audio which takes considerable bitrate. So if you want the entire concert, you will have to go with AC3 audio, which is great, but not as good as PCM.


Darryl

[goldtop33]

why not just convert to audio to mp2?

[SatStorm]

I show you one alternative that gives IMHO the best picture quality results.
Also, I tried to explain why I suggest that method.
Beyond that, you can follow whatever suggestion you like!

I'm not a fan of realtime mpeg 2 capture. I use it only when I don't care for the quality and the filesize.

[goldtop33]

I understand and appreciate your opinion....

Correct me if I'm wrong - but the main concern between the AVI ot MPEG2 VS capturing directly to MPEG2 is that capturing to MPEG2 is too much on your system? (you are capturing AND encoding at the same time - resulting in a loss of quality)

[FulciLives]

I understand and appreciate your opinion....

Correct me if I'm wrong - but the main concern between the AVI ot MPEG2 VS capturing directly to MPEG2 is that capturing to MPEG2 is too much on your system? (you are capturing AND encoding at the same time - resulting in a loss of quality)

You should only capture direct to MPEG-2 for DVD use if you intend to leave it that way.

In short you CAN edit but it CAN be tricky but when I say "leave it that way" I mean no re-encoding. If you capture MPEG-2 and then have to re-encode it ... well ... you get double compression which is no good.

So if you capture direct to MPEG-2 try to use a bitrate that will fit your project on the DVD (use a bitrate calculator) so that you don't have to re-encode it.

Otherwise capture to AVI

I don't have an ATI capture card. My MIN, AVG, MAX settings based on how a 2-pass VBR works when encoding from a captured file. I have no experience with captures direct to MPEG-2 as I do all my captures as AVI then encode later. Same principals apply though to stuff like BITRATE determination etc.

If you can only capture as MP2 audio I would use 384kbps then convert that to 256kbps AC-3 audio for the DVD. Not all NTSC DVD players can handle MP2 audio so that is why it is best to use AC-3 audio instead.

- John "FulciLives" Coleman

[qlizard]

SatStorm, I basically follow this method, but I filter and resize only when I encode (I use procoder). Should I be filtering and resizing before encoding to achieve better quality? If so, using what, virtualdub?

SO, in really short terms:
- Capture your tapes @ 704 x 576 (PAL) or 480 (NTSC)
- Filter your source
- Resize it to 352 x 576/480
- Encode it to mpeg 2 @ 352 x 576/480

Then Author and Burn it on whatever you wish....

[SatStorm]

I use virtualdub to filter and resize. An alternative is avisynth. For VHS tapes as a source, avisynth / virtualdub are the same in practice (IMHO always)
Also I always framesave when I encode. Why? Once in a while an issue pop up, which always can be solve by frameserving. After so many years I prefer to do one step more - which doesn't harm and slow down the proccess only a bit, almost unoticable for my PC - and just be sure for the results, than try to deal with the "issues" when they manage to appear.
I hate the "issues" of this hobby :P

Procoder has some built in filters, and also TMPGenc. There are fine. Not excellent, like the filters of avisynth / virtualdub, but fine. More than enough for the needs of most users. Those applications filter before the actual encoding, so it doesn't really change something in the chain I suggest.
But I prefer to use encoders to encode and only to encode.
Got me?

[qlizard]

Yeah, I got you, thanks for the help. Makes a lot of sense. Sorry to the original poster for hijacking, but maybe it will help you too.

One more, Satstorm, which I've been wondering about. For the sake of speed through the process, I will probably keep doing like I have been (I may get burned some time like it sounds like you have, so then maybe I'll learn...). On Procoder, when you set up your source and target files, it gives you the option of filtering in both places (the source and the target). Does it matter which you choose? If you choose both, does it filter 2 times?

thanks

[SatStorm]

I never tested that fuction myself, so I don't know the answer on this.
As I said, I prefer to use the encoders for encoding, not filtering...