This is one time when I will differ with musicvid's opinion.
I've authored and burned in excess of 1,000 DVDs over the years at a CBR of 8,000,000 and have never had one returned because it wouldn't play.
It may help that I only use Taiyo-Yuden or Verbatim which are both good name brand media.

Mike

My experience is identical to Mike's: unless length is a concern, 8,000,000 CBR for my DVD's has never failed. Ever. I use Taiyo-Yuden watershield media.

That's quite OK Mike and I might agree, except you've never made a batch for a school that had all donated CM players? Those junk players can sometimes even stutter at the end of a 6Mbps VBR program (Verbatim discs and commercial movies). I had the misfortune of working in performing arts at just such a school for three years. CM (Curtis-Mathes) players also tend to show up a lot in grandparents' homes too, because they were sold so cheap at the time, and as you know, grandparents never throw away anything. Maybe some do work, but gotta target the lowest common denominator.

Thanks for all input!
I always use TY discs.
OK so no problem with 9416 average.
I´m just to render an opera (Cosi fan tutte) for a school and need an advice.
The first part is 1h 10 min and the second part 38 min.
I had first thought to make 2 discs but is it possible to
make one disc 1 hour 48 min with enough quality?
Mark says 5272 average so lets say 1h 53 min for headroom = 5024 average
is that good quality enough?
No meny in DVDA (singelmovie).

Why not author it onto a Dual Layer disc? If you use a good quality disc (I always use Verbatim+DL) there's nothing to fear. I've been using them for a long time now.

I have never done that before - how about playability for Dual Layer?
Are they not more sensibel for fingermarks and dirt?

Arthur, there is "something to fear." Burned dual layer discs will not play on up to 40% of home players in use. Two out of the five good players I own will either choke or not make it past the break point. If someone else will be getting the disc, dual layer is not a solution.

Aje, 5Mbps is plenty good quality for your opera. No one will be disappointed with the quality on a single disc. Use a 2,000,000 minimum bitrate (important), 5,000,000 average, and 9,500,000 maximum. Be sure to check 2-Pass.

Apparently our local OTA tv SD stations have poor DVD players as they hang the program with the station still locked in digitally and you get a black screen for a few seconds. This happens a lot with Gunsmoke and Mash. They get away with it because everybody thinks it is a digital dropout/fade.
JJK

Gotta pipe in here.

I've done lots of DVDs, always use the trivial formula:

Total bitrate = 600 / minutes. That's it. Then just subtract your audio bitrate (.2Mb/s for typical AC3), and you have your average video bitrate number.

For the opera, you have 108 minutes, which gives a video bitrate of about 5.35Mb/sec.
I'd definitely use VBR (5Mb/sec is really low unless you are using an exotic encoder such as CCE or have a gorgeous noise-free source video) with 9 for the max, 5.35 for the average, and 1 for the minimum. If you're using MainConcept set it for dual-pass and hope for the best. The first pass gives MC a look at how the bits need to be allocated for consistent quality, and the second pass does the encoding while making sure to meet your average bitrate target.

I think the official DVD spec (which all players are required to meet, at least when new), calls out for a maximum overall bitrate of just over 10Mb/sec. In my own experience I've seen some players have a harder time with such a high continuous rate. Some pundits say it's because reading a burned disk is harder than a pressed one, so the player makes mistakes and then has to back up and pick up the data again, making it hard to keep up with a high bitrate.

I did a double-layer release a few years back; 110 copies out, no issues or returns. I'd suggest the "trick" is to use Verbatim DVD+R DL (great reputation) and make sure the disk is bitset as "DVD-ROM" which makes a significant difference in older players.

EDIT: I've always burned disks with Nero. Unfortunately, Nero a few years back did a lousy job with double-layer, so I used CopyToDvd, which managed the layer break perfectly. I have no idea how contemporary burning software manages with double-layer.

Also, when I did the project, DVD-R DL had just come out but was a disaster. DVD+R DL, specifically Verbatim DVD+R DL, was the way to go. Don't know if the (-) folks have gotten their act together yet. The reason bitsetting was important is because older players had no idea what a +R disk was, so the idea was to fool them into thinking the disk was a plain old DVD-ROM, which they then happily played.

My experience was a little different.
The Verbatim DVD+R DL discs I burned on my one and only project for distribution eventually made a really nice mobile hanging in a classroom. The replacements I comped were 2-disc sets, and nobody complained.

EVERY time you burn a DVD or a Blu-ray or even a CD optical disc you create hundreds to thousands of errors per disc. YOU cannot really control the error rate other than select good media and burn at a calibrated slow-speed on a burner in optimal working condition. But absolutely the disc will have errors.

So the secret ? This works generally because playback of the optical media has SOME capability to error-correct as it plays. This works to some degree and is usually tolerable but certainly is not perfect.

THIS is why you avoid building discs at max data rates and I would suggest building dual-layer discs for distribution is for the foolhardy.

+ 1 to all those who question the reliability of dual layer disks. my (limited) experience is one of failure to play / pass layer on 60+% of machines it / they played on.

unless you've got a long program i use c/b of 8k. never had a problem in many years.

+1 on the 600/minutes formula. I've used that for hundreds of DVDs and it's never failed me.

5.3Mbps really isn't too low for an opera. High bitrates are needed when there is a lot of change from one frame to the next. Most operas tend to have little motion, confined to small areas of the stage, and usually rather slow moving. Most of the image remains pretty much static pretty much of the time. VBR encoding should handle the occasional faster or larger movements as long as there aren't too many of them in a row. (Of course, your particular opera may not follow this pattern.)

I recorded a "talking heads" play a few years back, about 4.25 hours of mostly people standing still or sitting on the set talking. Most all movement happened between scenes with the lights off. I encoded it at 2.1Mbps to fit on a single DVD5 and there were no discernable artifacts. I also tried another version at 6Mbps and split the three acts on individual DVDs. Visually they looked identical to the 2.1Mbps version, so we didn't bother making the 3-DVD set available.

I don't go above 8 for video (no more then 9 total). I don't use VBR, I've never had a case where it would be a great advantage over CBR except for space, but if I'm using a disc, after the initial render I'd be talking seconds in duplication times, not hours or even minutes.

I've seen some cheaper players not like big changes in VBR, never had anyone say they couldn't play a CBR disk, so I've stuck with what worked.

I've used the videohelp calculator for years. I like to do motion menus & such so I make sure I add in the time for those videos.

I have a bunch of render templates setup based on time: 1.5hr, 2hr, 2.5hr & 3hr, all in WS & 4:3. Then I just look @ my segment to render length & render to the proper template.

I've never had a problem with 8 Mbps CBR plus 192 kbps audio on Verbatim +R but I'm currently preparing 2 DVDs for Createspace and they want 6.5 Mbps max for video+audio, even for VBR peaks.

Their explanation includes: "DVD-R media is subtly different from replicated media. By sticking to an even lower maximum bit rate of 6.5 mbps, you're giving the player the chance to re-read some sectors if needed without interrupting the video. This provides similar benefits for scratched or dirty discs, making them a bit more likely to continue playing correctly."

So I'm somewhat reluctantly doing my video stream at 6.3 Mbps CBR to leave room for the audio.

I like to take clues from what is working in the marketplace. Worth noting that all my movies are pressed on DVD9 (big difference from burned), with the vast majority of titles between 5.75-7.0 Mbps VBR, no higher. I figure these studio folks already know the bottom line for the bulk of consumer players; witness there is often tons of filler and empty space on the commercial media. Logic suggests, if they could sell 2+ hr., 8Mbps discs without generating more returns, they would.

That's not to detract from those who are burning and playing 8Mbps CBR successfully. That's not an axiom, however. I actually watched one of those older CM players smoke when I brought one such disc as a raw cut of the final show to the cast party. No exaggeration.

Some players can't handle it.

",2Mb/s for typical AC3", is that 192?
I´m doing mostly musicfilms, concerts and operas recording audio separately in 24/48 and want good AC3 audio.
Isn´t 192 too low?
I always use at least 320 and mostly 384.
Could 384 combined with high video bitrate be to much for some players?
In other words - with audio 384 what will the critical bitrate border be for most players.
Or is 384 meaningless high for good audio?
Thanks for all input!
Aje

Typical video bitrate: 5000000-8500000
Typical audio bitrate: 192000-320000

I don't think you need to worry about the audio bitrate adding significant overhead. As long as the combined total doesn't exceed whatever you determine is a useful maximum you're ok.

The biggest issue I've had when testing DVD's on my own players (not handing out) is filing the DVD. It seems that when I get close to the 4.5gb limit I've had more issues then bitrate related. Keeping the disk under 4.2 gb's seems to make more reliable disks.

But that's just testing on MY players (a BD, computers & a couple DVD's, a cheap one & a nicer one). Because I've had an issues once or twice I don't hand out disks that are 100% full.

I've had occasional problems with the last minute or two not playing. Sometimes the player skips back to the menu; sometimes it just hangs. My solution is to include a couple extra minutes of black/silent at the end of the video. With VBR this takes up almost no space at all. I set an out point to make playback finish at the end of the content. Having this extra buffer at the end seems to help a lot.

OK no more than 4,2 GB and dark/silent end.
that sounds like reasonable and good advices.
i will follow them to - thanks a lot!
Aje

Re using Dual layer. I can only offer advice on my own experience. I'm not a 'mass producer' but I burn around 200 to 250 discs per year for my wedding clients. This past year, not a single return. Year before that, 1 I think. That's with a 4 year old burner - you would/should expect newer burners to be even better. I experimented with quite a few different DL's before letting them out 'into the wild'. - DL was very poor. Cheaper DL's of either - or + were also very poor. As mentioned above burning at too fast a speed also = poor reliability. I used an ancient Toshiba player as my main test player, and also included a cheapo 'supermarket' player and my BD player to test. EVERY single disc either SL or DL is tested in this player before going out. I appreciate that mass producers couldn't do this.
I've settled at Verbatim +DL burned at X 4. Variable bitrate of 8,500 max, average 6, min 2. Reason I don't go lower than 2 is because I've seen some quite 'flakey' fades to black at lower bitrates. (with SL and DL).

Bottom line is we all rely on our own experiences in the end. Two of mine - which will probably bring gasps of disbelief are; I don't use Tayo Yuden - I once had a batch of 30 discs with 10 failures! I don't use Sony HDV tapes - again after a major failure - this time with drop outs. Once bitten twice shy!

Oh, nearly forgot; they are no more sensitive to dirt/finger marks than a SL disc.