Best bitrate? CBR or VBR?

[manono]

Hi-

Sometimes you have to use PCM. Remember that AC3 is compression, and sometimes that compression can alter the audio, especially on very damaged audio quality.

I guess I can buy that. I work with DVDs, and don't see WAV audio that often, and because there's always an AC3 audio track along with it, I can dump the PCM as it takes up way too much space. Although I've done a lot of reencoding of DTS and WAV audio to AC3 with no problems, the source quality has always been good. I think I can understand that if the WAV source is substandard, with, maybe, wind noises, crackling and other noise, distortion, and other kinds of problems, the reencode might amplify these things. Point taken. Thanks.

[videopoo]

Hi Scott

The Meastro scenerio was only to point out that when that line in the manual was written it was done by a inexperienced company with a sub standard authoring tool (at the time)

I'm not questioning your ability to create a spec compliant disc with a high bitrate - A spec compliant disc is just that. But, this thread is about PRACTICES. If you want to bump up the bitrate, check it with a bitrate analyser, give it a ECLIPSE and MEI report then test it in your player banks thats great. I never said bumping up the bitrate would create an out of spec disc but the likelyhood is far greater- especially for people who are not professional authors.

Ok..here comes the stupid (yet useful) analogy:

If you buy a truck with a certain towing capacity would it be a good idea to max out the weight in the bed just because you know you can and even though you woudn't have to for a specific job? I hope not...why would you want to come even close to ruining your suspension?


The quality difference from a 7000 encode to a 9000 would be fairly slight (certainly not dramatic) using a decent encoder. I'd really have to look close and you certainly wouldn't see a difference on a good hardware set-up. Aunt Sally watching her niece's wedding produced by Waldo1967 definitely wouldn't .

My point is, not everyone is a professional compressionist/author. Waldo1967 just wants his production on disc so he can get paid. He's not experienced or intrested in such things - besides he has more important things to do with his time. Going around on public forums where people are learning and saying to "bump up that bitrate - it will be just fine" is not good for business and certainly not good practices. Are you saying that player compatability is no way related to bitrate?

I'd like to hear your conclusions but with all the work you'll put into it coupled with your level of expertise I seriously doubt we'll have a problem But, I must say, you are the only individual I've ever met with an advanced knowledge of the DVD spec who has taken this stand - truely.

Eric

[videopoo]

Hi manono

Wow...thanks for that. I guess I stand corrected. But I don't see any video bitrates in the 9000 range there. Take out the audio and your looking at a video bitrate in the 6-7 range. I never said these rates don't go over 5500 - I was talking about avgs - but you proved me wrong (good for you!) In my previous post I never indicated setting your bitrate high will give you an out of spec disc but telling a gentleman who questions the skips and stutters on his wedding DVD its OK to bump up that bitrate is kinda wrong - don't you think? (don't worry, I'm not expecting an answer). BTW...I still don't agree that most stuttering is caused by bad media or burns...sorry to you as well...for you guys ,yes, since you know your shit, but for most production folks who think making DVDs is the same as copying VHS tapes...NO.

Also, what other crap have I been spewing in this thread? Just curious? I trying to help the guy out by telling him his bitrate might be set too high. Jeez..you practically berated the guy for God sake's - no help what so ever...

We can sit here all day and go back and forth but just don't sit there and say I've been giving out bad info and I have no idea what I'm talking about. I'm not here to ruffle feathers and play semantics

Good luck everyone.

[daamon]

There's no excuse, in my opion, for ever using PCM WAV audio for homemade stuff, where the extra bits freed up from converting the audio to AC3 can be used to improve the video quality.

I know what you're getting at and, in principle, I agree with the basis for your reasoning. lordsmurf has pointed out one such circumstance for using PCM audio, another would be if the homemade video runs at an hour or less. This would allow for the max video bitrate as well as enough room for the audio as PCM.

Why bother spending extra time encoding to AC3 or MP2 (where allowed), reducing quality (albeit negligibly) when it's not needed?

I'm not disagreeing with you per se, just pointing this out for anyone who's not aware of this fact who may be reading this thread for info or interest.

[Cornucopia]

videopoo--

Point taken. And some good points, too. Though your analogy doesn't quite hold for all the nuances of what I thought we were talking about. More appropriate might have been where one uses that same truck, but needs to carry a certain load a certain distance with the least amount of trips-thus the need to maximize/optimize.
Anyway, I think I'm going to try doing the compression test no matter what. I'm just curious now, is all. And if I learn that I shouldn't be pushing it, well, I'll have learned something valuable.
Now--bear with me--supposedly, the MPEG2 D1 rez "sweet spot" should be between ~6000-8000kbps, but that depends alot upon the source material. If you've been using Digibeta and better masters of 24p material, I wouldn't be surprised that you could get quite decent encodes in that or even lower rates. However, masters that, because of their technique and ancestry, may have high motion and detail needs--these masters NEED higher bitrate to keep them from incurring encoding artifacts. That, or you're gonna need to do alot of smart filtering/processing. Yes, players MAY baulk at a high bitrate. They SHOULDN'T, and I actually believe that most of the anecdotal problems referred to with players are just holdovers from the time when they weren't built as well. Chipsets have improved dramatically since '97-'98.

back OT...

Waldo1967,
Unless your Wedding vid music was extra special, it probably makes sense to encode to AC3 (try 224kbps for stereo). Use a GOOD video encoder and do multipass VBR--WITH preprocessing to smooth out any anomalies. Even if you DON'T want to go extra high, it'll surely be better than what you got right now. I'm assuming since you've shot video for 15 years, that most of the footage is tripod/steadicam, with some occasional handheld stuff, so you at least shouldn't have very much camera jiggle worries.

Scott

[lordsmurf]

I suggest 256k for home AC3, no lower. Go 384k if you have the space and it's important.

[Waldo1967]

Thanks a lot guys. Looks like I brought out some "good" discussion.

Just an update, the video that was skipping before on a cheap disc at 8000b/s that I reproduced on a quality disc at 6000b/s is working fine now. Could have been the disc quality, could have been the bitrate. Anyway, I've learned a lot here and I'll be back for more.

BTW, thanks for the support Videopoo. We all had to learn at some point by asking questions, and sometimes by asking stupid questions.

[lordsmurf]

Waldo, you should have just moved the data to a new disc. Dropping the bitrate was not a good choice. Process of elimination would have been to try ONE fix at a time.

[Ron B]

I learned a lot from this thread because almost everybody seems to know a little more about DVD encoding than I do. My philosophy on encoding is to get the material I need to put on a disc, shot and edited as cleanly as possible, use the highest bitrate for the video(don't usually go above 9000) that I can fit on the disc, then use AC3 audio unless I've got the space, then I usually just leave the PCM.
I have a crappy $40 DVD player that I got at Wal-Mart, I figure if the disc will play on it, it should play on most standalone players.
Every project is a little different and I'm using a couple different encoders, trying to use the best tool and technique for each job.