How to render file size to fit DVD?
 

Hello, I'm back with another quasi-newcomer question for Vegas 8.0 users; I've completed an NTSC DV widescreen format video that runs 1 hour, 43 minutes. I need to render the file to fit on a single DVD. When I render it using Mainconcept MPEG-2 for DVD Architect, it creates a 5.1GB file - too big to fit on a 4.7GB DVD. How do I adjust the compression settings to make a file that will fit? Thanks!

Sam

Your best bet is to use a bitrate calculator.
Here's a link (to a zipped file) to the one I use.
When you first start it up, click "Settings" (bottom left).
In the window that comes up, click the "1 kilobit = 1024 bits" option and set the "Audio Encoding type and bitrate" to 192 Kb/S (Vegas default for AC-3).
I set the Safety Margin to 5% whci, while a bit on the conservative side, guarantees that I'll never go over the limit.

To save you time, here's the settings I get.
CBR: 5,368,000
VBR: 8,000,000 / 5,368,000 / 3,216,000

BTW, recommended practice is to not exceed 8,000,000 as a maximum.
It has the potential of causing playback errors with low grade media and some DVD players.

just to add to what Mike said, a 4.7gig DVD is NOT 4.7gigs. It is actually 4.34 due to formatting, so keep that in the back of your head.
By using the bitrate calculator and AC3 audio (smaller file size which allows for higher bitrate) make sure you name the files the same; my movie.mpg and my movie.ac3-that way when you bring one into DVDA (if thats what you're using) both files will come in.
Also don't be surprised if the file size in the indicator is exaggerated. Again that's if you're using DVDA.

Don

Remember to set your Project Properies in DVD Architect to NTSC Widescreen too.

Sorry for the delay and thank you so much for the responses! The bitrate calculator did the trick. Thanks again!

Sam

Will the quality detoriate in this process?

590 divided by the minutes = bitrate.

Paul.

It depends.
To fit this video onto a DVD, you're compressing the source video by a factor of roughly 6:1.
Something has to give when you're throwing out that much data and picture quality is one of them.
Having said that, the quality of your source material has a lot to do with it.
The old rule of GIGO (Garbage In, Garbage Out) still holds true.
A 2 hr. DVD that was recorded on VHS will not look as good as a 2 hr. video shot on a good 3-chip camera.
I'm fortunate enough to have access to a good 3-chip camera.
I take care during shooting & editing to maximize quality and, consequently, have no qualms about doing a 2 hr. DVD as the quality does hold up.

"Maximizing Quality"
 

"I'm fortunate enough to have access to a good 3-chip camera.
I take care during shooting & editing to maximize quality and, consequently, have no qualms about doing a 2 hr. DVD as the quality does hold up."

Mike, I'm using FCP on a G5 Mac, and not Vegas on a PC... but your statement about "maximizing quality during shooting and editing" caught my attention. I shoot dance recitals and such with two or three "good" 3-chip cameras, and I'm curious how you set up your shoots so that you end up with "no qualms" about losing quality in compression... something that I ALWAYS seem to have at least some qualms about.

Sorry if this is off topic for this thread, but I'd appreciate any advice you feel willing to offer. If a reply post is inappropriate, feel free to email me.

Thanks a lot.

Skip Hall
Suffolk, Virginia

Some compression theory
 

Makes no difference when we're talking shooting for compression and best practices.
The DVD standard includes bitrates of MPEG2 video and AC3 or PCM or MPEG audio streams that every hardware "set-top" player is expected to be able to play back. Include a safety margin, which experience has shown is needed for wide distribution, and you end up with a video bitrate of around 8Mbps, or maybe 8.5 (megabits per second).

The DV tape standard runs about 25Mbps, so, right away, we're throwing away at least 2/3rds of the information in compressing for DVD. Since we can only fit about 4.3GB on a standard single-layer recordable DVD, once we get over about 70 minutes of material we have to throw away even more data. The practical maximum is (very) arguably somewhere between 1.5 and 2+ hours... depending.

And this is where it gets interesting - "depending" on how it's shot, how much motion is in the subject, how it's lit and exposed, how it's processed in post, and finally, how it's compressed.

Some subjects and shooting styles demand more bits to maintain quality. For example, if you are shooting a dance recital you usually have no control over the lighting, and you may be dealing with areas that are brightly lit and areas that are poorly lit in the same shot. What to do? A camera left on automatic will probably gain-up to get some decent exposure in the deep shadows. Usually this is going to lead to lots of noise in the blacks. From a compression perspective this is very bad, because every pixel is going to change color in every frame. Think about it - if we have a nice solid color in an area of the shot, it's not going to change so much, and compression is easy. Noisy blacks are just the opposite.

So, in an extreme situation (probably not real-world), every pixel changing every frame, the worst possible case, even 8 or 8.5Mbps isn't going to be enough to show without macroblocking (where pixels in 16x16 or 32x32 groups get assigned all to one color, because the compressor is constrained from giving enough bits to address the pixels individually).

Let's call this picture complexity.

OK, so now we know to keep the camera on manual, not let gain introduce noise into the blacks, and consequently we'll lose some shadow detail.

What else? Handheld camera will lead to high picture complexity. On a tripod with rapid moves - well, it will be pretty complex during a move, but, when using variable bit rate encoding (VBR) the encoder can give more bits to the short section of the move. And our eye tends to be more forgiving of compression errors during a short transition/move.

A moving subject. Water. Wind disturbing grass or the leaves of a tree.

The eye tends to be less forgiving of errors in what should be a smooth color transition, for example the sky.

Dissolves are way more complex than wipes or cuts. Better if they're quick.

What to do with the footage we have? Noisy, shakey, etc? You can tame some black noise using a "black restore" filter - kind of a secondary color corrector approach. The idea here is to make all those noisy dark greys black.

There are noise reducing filters - for Vegas there's one called Neat Video, that is being ported over to OS X, not yet for FCP, but I'm not up on what other filters might be available for FCP.

A little blur can sometimes help.

If the picture complexity has short peaks, VBR compression can help a lot.

A good green-screen shoot with a graphic background matted in can be very, very good for compression - because the picture complexity of the background is simple - it doesn't change.

BTW, these principles apply equally to compression for web distribution - WM, QT, Flash Video, etc.

Skip, Seth brought up a lot of excellent points for you to ponder so I'll just give a rundown on my approach.
Most of my videos are done for my work (community college) where I use a pair of JVC-550U cameras (3 X 1/2" CCD so low-light capability is great).
To give you an example of how good the low light is, I shot a fashion show once where the runway lighting was a long string of clear Christmas bulbs .
To my surprise, it turned out OK.
I always attend at least one rehearsal for all plays or performances that I shoot.
I make note of the brightest scene and set my iris (ALWAYS run it on manual) to that level for the entire show.
This is generally 4.0.
The only times I'll deviate from this is if it's a particularly dark scene in which case I'll open it up to 2.8 and sometimes 2.0 but, after this scene, I bring it back down again.
My (personal) feeling here is that, if the live viewing audience is seeing a dark scene, the TV viewing audience should see the same thing.
I've also shot various performances at my kids' grade school and, once again, I'm lucky enough to attend at least one rehearsal.
I also bring along a 9" field monitor that has been set up for color bars and use this as a reference.
I've done DVDs of shows that have run as long as 2 hr.
Using the bitrate calculator I linked to earlier in conjunction with a 2-pass VBR encode, no one has complained about the quality yet :-)
Hope this help somewhat.