>>Can you render a file higher than 8,000,000 for better quality?

Personally, I stay about 6,000,000 and have had no compatibility issues with other people's set top DVD players. I have had clips around the 8 mark and have seen set top DVD players choke. Re-render the clip to 6, and the exact same stuff on which the DVD player was choking now played perfectly.

I decided some time ago to stick with the 6 figure so as to have good compatibility. You may have a 7.5 to 8 super high quality DVD...that will choke in some players. I don't want that. 6 looks terrific and plays well. (The default settings on the Main Concept stand-alone MPG2 encoder are also at 6.)

>>...Wouldn't it be better to render it from Veggas at a lower bitrate...

Sure, it would be better to have only *one (1)* AVI to MPG2 transcode. The reason you would not want to go so low is that it may not look so good. I personally have not run that low a bit rate, so I cannot comment. I think I may have had some footage that was about 4.5 variable bit rate, and it didn't look too bad.

The less generations you have between transcodes the better.

I have never had any problem with high or low bitrates. Here's the link to the document at MPEG.ORG that describes the bitrates that DVD players are required to handle:

mpeg.org

Remember that bitrates are for EVERYTHING the DVD player is handling: video, audio, subtitles, navigation, etc. I have often suspected that some problems that have been reported at higher bitrates are when people use PCM audio (which takes a lot of bits) and don't remember to leave room for the audio. The document above is quite explicit that DVD players HAVE to be able to decode 9,800,000 bps (total), and that virtually all of them use the same ASIC circuitry which is actually capable of handling decoding at even higher rates.

Now, I'm not saying that people haven't had problems with discs that have been encoded at high bitrates. That may be true. What I AM saying is that most likely there were other errors in the authoring of the disc, or there was a media incompatiblity, or something else caused the problem. My dad had an old Sony DVD player that refused to reliably play some of the discs I sent to him. Some of these used MPEG audio, and I suspected that was part of the problem. It might have been media. It definitely was NOT bitrate.

I wrote a bitrate calculator using Excel that gives you detailed insight into all the components of your DVD's bitrate, and makes sure you don't encode at higher than legal video bitrate. Nothing special about it, and if you have one that works, don't bother downloading it, but here's the link:

MPEG Bitrate Calculator

BTW, since Yahoo no longer provides "public" access to free Briefcase files, I can't give you one-click access. When you click on the above link, find the "MpegCalculator_JHM," right-click on it, and select "Save Target As." Feel free to download some of the short WMV or RM movie clips if you're in the mood to see some of my work.

bitrate calculator for you all http://www.videohelp.com/BitrateCalc.zip

also -- PLEASE keep in mind that the max bitrate values as found on mpeg.org are USELESS for the type of dvd's you are making -- namely BURNED dvd's vs. pressed dvd's ....

THE MAX. BITRATES FOR BURNED DVD'S ARE LOWER THAN FOR PRESSED DVD'S -- END OF STORY - THE FACTS ....

and different encoders have different max values that should be used - as well as bitrate spreads ..

B_JM:

I have no doubt that a burned disk is somewhat harder to read than a pressed disk (having to do with the contrast between pits and lands) and that therefore it might be harder for a player to keep up with a high-bit-rate burned disk. But I've never seen any documents that actually show this in a quantitative fashion. Do you know of any sources?

I've personally done all my stuff in VBR in CinemaCraft. For short videos, I'd set the average to 8 with a max of 9 and a min of 0. Never any problems with Ritek G04 or G05 blanks (-R) on a Pioneer -05 or -07 burner burning at max rated speed.

At the other extreme, the project I just finished had about 2:24 hours of video (2 hour documentary, the rest bonus stuff) and I also had to deal with 4-channel surround sound and an extra 2-channel commentary track. Final bitrate was 3.7 average, max 9, min 0. The encoder made about 4 passes, so it did about the very best job is could. The results look great, though I can start to see some minor artifacts in a few places if I look carefully. But 2 1/2 hours is about as far as I think I'll push a 4.38GB medium.

the max. bitrates for burned dvd's are lower than for pressed dvd's

Why is this?

I've seen all sorts of posts about this at doom9.org, dvdrhelp.com, burnworld, etc., but never any links to test or standards documents to back up the theory. I'm not in any way arguing with you or saying its wrong. However, without a link to a technical explanation, it just doesn't make sense to me. I just spent ten minutes searching using Google and Teoma and was not able to find a single document that supports this. What I can find are dozens of posts at the above-named sites where people say that recordable DVDs can't be recorded at as high a bitrate as those which are pressed. The only justification these people offer is that the error rate is higher with recordable media and that somehow this is going to make playback impossible at higher bitrates. This idea has at least one well-known proponent: Ralph LaBarge, whose work I respect. He wrote about this here:

DVD Compatibility Test

where he said:

"Use low to moderate bitrate encoding for video if the final product will be delivered on a recordable DVD. Although most DVD players and DVD-ROM drives can read recordable DVD discs, these discs often have a higher error rate during playback than replicated discs do, causing the player to reread some data sectors. Using a low to moderate (less than 7Mbps) data rate for video encoding will make the recorded disc a little easier to reread without a visible pause in the video playback, or audio dropout."

I added the emphasis in the above quote.

The problem I have with this statement is that I don't think the DVD player, when it encounters an error, stops, then re-reads the data. Instead, I'm pretty darn sure that, as with everything in computers, the error correction information is embedded in the data stream, and when errors happen, the data is reconstructed on the fly. There is no re-reading of the data, and the error correction is usually handled virtually instantly in the hardware (I actually designed error detecting and correcting hardware 30 years ago, and in jelly-bean logic days then AND/OR gates actually looked at all the data, including the error detection/correction data, and presented the correct result instantly -- there was zero difference in time between a packet that contained an error and one that didn't). If the error can't be corrected, then bogus data is passed to the decoder and you get a glitch. But -- and here's my point -- if you have enough errors to cause a problem, then that says to me that you've got a media problem, not a bitrate problem.

Put another way, if half the bits are bad, then half the bits are bad, whether the bitrate is low or high.

Now some people think that the error rate somehow gets higher when the player has to playback a disk encoded at a high bitrate. I think there is the assumption that the disk spins more quickly when reading a fast bitrate. I just did yet another search and tried to find some document that supports this. I couldn't find any such document. Now, I'm not saying that it ain't so. In fact, as I'm sure everyone knows, both DVDs and CDs do spin at different rotational speeds (unlike old vinyl records) depending on whether the information is being read near the center or near the edge. Also, the writing speed is adjustable, so I'm sure the reading speed is adjustable as well. However, However, I can't find any document that says that the disk rotates more slowly when the bitrate is low. Just as a quick thought experiment, if it WERE true that the disk spins more slowly at lower bitrates, then if you encoded at some of these really low bitrates, then the disk would barely be moving. Has anyone ever experienced that? I sure haven't.

Some have suggested that at a higher bitrate the DVD drive has to read more data, but that certainly isn't true: the disk holds 4.7GBytes and that is what must be read, regardless of the bitrate.

There is also the idea that something in the drive or the computer is going to have a hard time "keeping up" with something, but the data rate coming from a DVD is nowhere near extraordinary by modern standards.

Now, I am not trying to start an argument. If there is really a problem here, I'd definitely want to know about it, because I sure don't want to start getting returns from people that can't play my disks. However, I am in love with quality: I want the best sound, and the best photos, and the best video. I can easily see the difference between an encode done at 6,000,000 bps and one done at 7,500,000 bps, especially for the sports video I do. Therefore, I want to encode at the highest possible data rate.

I have never had a single disk returned or refuse to play (other than my dad's old Sony). From everything I've read, by FAR the biggest problem is media, and the second biggest problem is cockpit error (which is why videohelp/dvdrhelp.com, while a wonderful resource, is also a resource that needs to be used carefully -- lots of suspect andecdotes there).

If anyone can point me to actual scientific testing from a reliable source that shows a correlation between bitrate and playback reliability, where bitrate was the only variable in the test, I'd sure love to see it.

The source is me for one thing -- i've written on this extensively for years and years, published papers, worked in a consultant (and tester) capacity to several software encoding and authoring companies .. and trained many people in dvd and encoding techniques .. as well as even help develop both some codecs and software in these and similar fields ... I also hold several patents on both encoding and playback software and hardware .. etc., etc. ....

So with some small knowledge - i can sometimes say I know what I talk about (sometimes, still learn something new everyday).
I test dvd's in multiple machines from multiple media made in multiple ways ...

I've prob. burned over 20,000 dvd's and authored over a 1000 ....

Anyway - I am on videohelp.com / dvdrhelp.com and DMN (as a moderator on those sites) and doom9 (and several other sites), extensive testing and feedback from our users there all bear out the same results .. as i previously stated .. which is a pretty good indication from 'real world results' ... I mention that because some so called dvd tests i've seen were total, or nearly so, crap ...


It is really easy to see this a simplified demo of this in action -- use DVDinfoPro or NeroTools and check the read errors and read rate on burned vs. pressed disks ..

I don't doubt your expertise--you have many many comments here on this Board that indicates you've "been around the block" more than a few times.

What I'd like to see is actual proof that the burned disks do a worse job. For example, I seem to recall that a given drive must read a burned DVD-ROM disk at a somewhat lower speed overall than a pressed DVD-ROM disk. In other words, you put in a pressed disk, push "go", and x minutes later all the data is on your hard drive. Do the same thing to a burned DVD-R, and it takes a bit longer. I'm assuming this is the case--it would be an easy test to run.

But in either case the read rate would be something like 4x+. Since 1x by definition represents I think about 11Mb/sec, then that implies to me that even a burned disk has plenty of excess speed to burn (sorry for the pun), even if the margin is not as large as for a pressed disk.

BJ_M,

Given your knowledge and background, you must then know the answer: Why would DVD-R disks fail at high bitrates, but not at low bitrates, when the same doesn't happen with pressed disks? Is there a technical paper you can point me too, or perhaps have you written one? The player doesn't really move the laser back and try to re-read a bad section, as LaBarge implied in the quote I provided in my last post, does it?

Thus, if I author the identical program material, and make no changes except bitrate, what you and many other people are saying is that this is what I can expect:

              Low Bitrate       High Bitrate

            —————————————————|——————————————-

DVD-R      |     Pass        |      Fail    |

Pressed DVD|     Pass        |      Pass    |

            —————————————————|——————————————-

I'd sure like to know, in technical terms, what would cause that one cell in the above matrix to have "Fail" instead of "Pass" in it.

I guess I am genetically wired to reject explanations along the lines of "well, that's just the way it is," and instead always want to understand the actual mechanism behind the results. (I aggrevated a lot of my science teachers in high school.)

John, I do know from seeing various tests done of DVD burners that when the laser has a hard time reading the data on a disk, the vendor can have a number of things happen. Usually the disk is slowed down and the data is re-read. You can see this in the graphs produced by certain software products that are used to test a particular burner/medium combination--ideally the graph smoothly rises as the laser reads further and further out from the hub. But if there's a bad spot, the smooth graph line abruptly plummets, indicating that the drive has slowed down and is trying to re-read that bad area.

Read errors are very common in DVD media. The real issue is how many there are, since the error correction formulas can completely correct all but the most egregious cases.

Usually the disk is slowed down and the data is re-read.

Interesting. But, is the disk slowed down when the bitrate on the disc is lower? Remember, if recording at a lower bitrate is really going to produce a "more playable" DVD, there has to be some underlying mechanism that makes this so. If the disk is read in the same way, at the same rotational speed, regardless of the bitrate, and all that happens is that, with slower bitrates the reading is stopped for longer periods, then I can't see how a lower bitrate is going to change anything. Remember that the bitrate doesn't change anything about how the information is stored on the disk. Unless I'm missing something, a full low-bitrate disk and a full high bitrate disk have exactly the same number of bits on them, and they are exactly the same distance apart, etc. The drive rotational speed may change depending on bitrate -- that's one thing I'm trying to find out -- although I suspect it doesn't.

I'm still in the dark on this one.

now there is one thing we just can't duplicate or get a firm answer on, on recent models of dvd writers - to a degree ..

to wit:
"Does writing at lower speeds produce more compatable disks?"

The answer is complex and is dependent on media, burner, and burning software and burning speed and no hard and fast rules .. We have seen sometime MORE problems burning at some lower speeds even and sometimes the results make no sense (i.e 4x gives bad disks , 2x gives good disks).

I always burned at 4x for along time (1x at first - then 2x) , but now have been doing 8x and 4x depending which machine i'm using .. I dont ever burn at 16x ...


I said recent models of burners - because there were a few models that did have issues .. I also prefer the Benq 162x and pioneer 106 burners (and that is it)

There is good firmware hacks for the benq -- to taylor it for the media you use, etc .. with extensive support on this model ...

John,

Actually, Bitrate is the most important factor in compression. The lower the average bitrate, the more compressed the file will be and the less amount of space it will take up on a disk. Bitrate has everything to do with file size and that's why it needs to be adjusted for longer video's (over 1-1/2 hours) in order to fit on a 4.7GB DVD±R. Now, if you have a two our movie burned at 5000 Kb/s and the same movie burned at 6000 Kb/s the one burned at 6000 will take up more space on the DVD. They are both the same length in time, but one takes up more space on the DVD which requires the drive to spin faster increasing the risk of skipping data. However, probably neither of these bitrates will have an issue.

The BitCalculator I use has a pretty good explaination of this under the results section toward the bottom of the page.

one takes up more space on the DVD which requires the drive to spin faster increasing the risk of skipping data

I certainly understand that the player has to read fewer bits per unit time while playing back a movie encoded with a low bitrate (i.e., it is compressed a lot more). However, does the DVD drive actually spin slower when reading lower bitrates? If it does, then I can begin to understand how a lower bitrate might produce fewer errors, although compared to what happens in a DVD burner, even the fastest rotation in during DVD playback is pretty slow.

Back to the spinning. Are you saying that if I encoded a movie at a really, really low bitrate, like 500,000 bps, that I would see the disc spinning at 1/10 its "normal" speed that I see when I am playing a movie that was encoded at a more "normal" 5,000,000 bps average? It would be an easy thing to test using a DVD-RW. I think I have a drive around here that lets me look through a window and watch the disc spin. Has anyone ever seen a disc spin really slowly?

I bet this is not what happens though. If it did, then when I use VBR and have long stretches of still text credits (zero movement), I ought to see the disc almost stop, and then speed up when it hits the fast movement encoding. I don't think it does this.

But I don't know for sure. I'm just looking for a document that describes all this stuff. Again, my motivation is to make sure I produce DVDs that have the absolute best quality possible, but that will still play on everyone's DVD players. If all this talk about needing to keep bitrates down at 6,000,000 bps or below in order to ensure compatiblity is based on faulty information, then I don't want to sacrifice the quality of my work, especially if the REAL problem turns out to simply be lousy media. This, I'm sure we can all agree, is at the top of the list of what can cause DVDs to fail to play in certain players.

'does the DVD drive actually spin slower when reading lower bitrates"


No, the dvd player spins the same speed regardless of the bitrate (notwithstanding it spins at a different speed on the disk depending on position)