TMPGEnc says "clip total bitrate is too high for a DVD-Video" but it's not?

[sasuweh]

I hope somebody here can help me out. I'm using TMPGEnc DVD Author 3. I've got an m2v file which I am trying to turn into a DVD. I am also trying to include 2 audio tracks - 1 lpcm, 1 AC3. Here are the bitrates they are encoded in (according to mediainfo):

m2v - 7243 Kbps (9800 Kbps)
lpcm - 1536 Kbps, 48.0 KHz
AC3 - 448 Kbps, 48.0 KHz

I am making a DVD-9.

Here's the problem: With these 3 files loaded into DVD Author 3, here is the warning I get when trying to output the DVD:

"The clip's total bitrate is too high for a DVD-Video.
The Clip current total bitrate is 11784 Kb/s. However, in a DVD-Video the clip total bitrate should be lower than 9848 Kb/s.
If you ignore the alerts the DVD_Video you build will not conform to the DVD_Video standard. Do you wish to continue anyway?"

This makes no sense to me. I have used 2 different bitrate calculators to come up with the recommended bitrate for the video file, taking into consideration the total bitrate of both audio files. One recommended 7550 Kbps maximum bitrate for the video and the other recommended 7768 Kbps max, so I went with 7600 max bit rate when encoding. If you add the video and audio bitrates I listed above together, you will get 9227 Kbps.

So why is DVD Author 3 telling me the total is 11784 Kbps? That's what you get if you add the 9800 bitrate displayed in mediainfo in parentheses with both audio tracks. But it shouldn't be doing that should it? Gspot lists the m2v video bitrate as 7552 Kbps. So add 7552(m2v) + 1536(lpcm) + 448(AC3), and you get 9576 Kbps, still under the 9848 Kbps max that DVD Author 3 mentions.

So should I just completely ignore the warning and still make the DVD anyways? Why does mediainfo list 9800 Kbps in parentheses if it says the bitrate is 7243 Kpbs? It seems the 9800 that is reported in parentheses is what is throwing DVD Author 3 off. Any advice would be much appreciated.

[CogoSWSDS]

I've not used DVD Author 3, but have a question. Is it giving you this error before or during the muxing of the video & audio? If it's during or after the mux, then your encoder may not have honored your maximum bitrate. IIRC, Gspot only reports back what is in the headers, not what is actually in the file. Mediainfo may be actually scanning the file.

[Vidd]

So should I just completely ignore the warning and still make the DVD anyways?

Yes.

Why does mediainfo list 9800 Kbps in parentheses if it says the bitrate is 7243 Kpbs?

9800 Kbps is what is reported in the MPEG-2 stream header. Many consumer encoders will list 9800 if you encode with a DVD profile (even if you select a lesser max bitrate). CCE Basic also does this when the DVD encode profile is checked. But the video has an actual max of 7243 Kpbs.

[jagabo]

If that's a variable bitrate file the average may be 7243 but the peak bitrate will be higher -- as high as 9800 kbps. You will have playback problems during those peaks. Use Bitrate Viewer to examine the bitrate in detail.

[sasuweh]

Wow, thanks for all the quick replies! It is a variable bitrate file but I specified the maximum bitrate as 7600 Kbps. I did use CCE and selected the DVD profile. The message appears if I try the simulation or if I try to output the file.

So should I not be concerened then? It sounds like the header is reporting 9800 even though the bitrate is much lower. Will the header throw off the DVD specs though and affect playback, or is it the actual bitrate that matters?

Thanks for the mention of Bitrate Viewer. I wasn't aware of that program. I ran the m2v file in it, and it reported:

7541 average bitrate
8249 peak

Are those peaks going to throw everything off? What else can I do if I specify the max bitrate and it peaks anyways? Do I re-encode and specify a constant bit-rate in CCE? I thought the whole benefit of CCE is multi-pass variable bitrate encoding.

[filmboss80]

Do an experiment: try authoring without the LPCM audio track and see if you still get the same warning message. I have a feeling that the muxed .vob files, which include that LPCM track and a high-bitrate ac3 track, add up to something a bit overwhelming.

[sasuweh]

Okay, I just tried with only the AC3 file, and it still said the same thing, but now it says the total bitrate is 10248.

[jagabo]

7541 average bitrate
8249 peak

8249 + 1536 + 448 = 10233. That's slightly over the limit. But there's some question about what exacty "peak bitrate" means -- over what period of time you measure the peak. Bitrate Viewer has three modes -- second, GOP, and GOP enhanced (click on the graph icon at the left of the title bar for options). In any case, you are working very close to limit. I'd back off a bit on the video bitrate (average and peak).

[sasuweh]

Yeah, it does seem really close. I guess the best option then would be to re-encode my m2v over again but this time back off the bitrate by say 400 Kbps? Would that do it?

[manono]

Your authoring program doesn't know the true max bitrate until actually authoring. It can't tell you if you're going to get buffer underflows in advance. As Vidd said, that 9800 is just a figure in a header and probably means nothing. If you set the max bitrate as 7600 then you're good to go, assuming your encoder honors the bitrates you set. This is an old problem with TDA, taking that 9800 and adding to it the bitrates for the audio to then tell you you're over the limit, when in all probability you aren't. Go ahead and author it.

So why is DVD Author 3 telling me the total is 11784 Kbps?

It's adding together 9800+1536+448=11784. You can also safely ignore what Bitrate Viewer says as it handles NTSC bitrates incorrectly. There's still a chance you'll get buffer underflows, depending on the encoder you used, but you'll never know until you try.

[sasuweh]

I just ran the m2v through Bitrate Viewer again just to see what the different modes said, and here's what came up under the different modes:

GOP
Average Bitrate - 7541
Peak - 7613

GOP Enhanced
Average Bitrate - 7541
Peak - 7613

Second
Average Bitrate - 7541
Peak - 8249



EDIT: just read the new post after typing this, I guess the Bitrate Viewer is not to trusted then since I am using NTSC?

[manono]

Oh, sorry. I just clicked on the link. I was thinking you were using this Bitrate Viewer. My mistake. The one you're using is accurate.

[sasuweh]

Oh, good. Thanks again to everybody who tried to help me figure this out. I get the feeling I am probably all right as it is but to be on the safe side I guess I will remake the m2v with just a bit less bitrate.

[jagabo]

There's also the issue that problems are most likely to occur at the outer edge of the DVD. Using a few percent less than the entire DVD helps avoid that.

[Alex_ander]

You can do an experiment by remuxing it with Muxman. It simulates DVD buffer behaviour while muxing and if it shows no error, most likely it would be safe to leave it as is.

[sasuweh]

Jagabo, that's true. I hate it when it burns so close to the outer edge because I know there can be problems. Alex_ander, I didn't know that about Muxman, that's definitely useful information.

[sasuweh]

Okay, I tried an experiment and I'm totally frustrated here. I decided that since the large LPCM track is causing so many problems with maxing out the bitrate, I would instead encode the video at a higher bitrate and just use the 448 kbps AC3 audio file. So I used CCE to encode in CBR 9000 kbps, and I also checked "equalize each GOP's bit length" because I read that would help equalize the bitrate in CBR mode. So I scanned the new m2v in Bitrate Viewer, and it still has high spikes which will put it out of dvd specs. If I scan in GOP based mode or GOP Enhanced mode in Bitrate Viewer, it reports exactly 9000 kbps for both average and peak bitrate. However, in Second Based mode, it reports 9000 Kbps average bitrate but 9753 kbps peak, and moving the cursor around shows the cursor bitrate frequently over 9200 or 9300 kbps. Why am I having such a problem with CCE sticking to a specified bitrate? 9753 is quite a spike for a 9000 kbps CBR. With the audio it's out of spec again. Am I really going to have to back down to like 8200 kbps just because CCE can't seem to stick to the specified bitrate? At that point, I might as well lower it a bit more and fit it on a DVD-5. Bitrate calculator is saying 6800 kbps average VBR for a DVD-5. But I wanted to give it the highest bitrate I could and use a DVD-9.

Does anybody know what the difference is between the different modes in Bitrate Viewer? It defaults to Second based, so I'm assuming that is what I should be relying on, right?

[jagabo]

Does anybody know what the difference is between the different modes in Bitrate Viewer? It defaults to Second based, so I'm assuming that is what I should be relying on, right?

They are just different periods of time over which the peaks are calculated. Per second is obvious. GOP mode shows you the average bitrate for each GOP. The help file says this about enhanced GOP mode:

Enhanced GOP based: This is a similar method as the GOP based setting. The difference is a weighted algorithm that counts the frames each GOP has and shifts a GOP that is less than half the standard GOP size into the next cycle of calculation. This gives a more harmonized result than the raw GOP based method.

For example, a GOP that consists of a single I frame would have a very bitrate because I frames are much larger than P and B frames. But a 1 frame peak over 10000 kbps would not cause a problem for a DVD player. It's sustained bitrate peaks that are a problem. I don't know exactly how long a sustained peak you can have before there's a problem. But blending that short GOP with the next is a more accurate indication of the peak value that's important to a DVD player. I usually consider the per second mode as most "accurate". Ie, if I look at commercial DVD rips the 1-second report looks most like what you would expect.

Your results are a little odd though. I usually find per second mode reports lower peak bitrates, not higher.

[sasuweh]

Is it usual for CCE to create such high spikes? This is becoming much more complicated than it should be. I know the bitrates I want, but I can't rely on CCE to actually provide those bitrates. I can't even trust the recommendations by the Bitrate Calculators, because it gives numbers like 9050 kbps max for VBR settings, but since CCE is going way over that, I have to drop the max way lower, which then I'm afraid will throw off the project size and lower the quality.

[manono]

Have you actually tried to author yet? If not why not, instead of beating your head against a wall? I don't know what you encoded with originally (the one with a 7600 max bitrate), but if it was CCE using multiple pass VBR encoding, then I can pretty much guarantee it'll author properly. I don't use CBR encoding myself and don't know whether or not it abides by that 9000 you set, but I can see it going over during complex scenes because no CBR encoding is strictly CBR. Using VBR encoding, and using say, four or five passes, I can guarantee it won't go over the limit you set. How do I know this? Because I've done thousands of CCE encodes using a very high max bitrate (9500, usually). And they always author properly using Muxman. Alex_ander's suggestion was a good one. I don't use TDA for authoring and don't know if it notes buffer underflows or not, but Muxman won't complete a mux unless everything is in compliance.

[TimA-C]

I always had this problem with CCE and TMPG DVD Author. If you're happy that your files won't cause your player to glitch then you can either ignore the TMPGenc warning or use DVD patcher to set the bitrate value in the file's header before loading it into TMPGenc. I found that if I set CCE to a max bitrate of 8500kb/s then I generally had no more problems loading into TMGenc DVD Author, but then my audio was only a single track at 256kb/s.

[sasuweh]

Okay, here's where I'm at. I finally managed to get CCE to successfully encode my first file without nasty spikes, using 7 pass vbr. I set average bitrate of 6800 and max of 8800, and the biggest spike ended up being 8833, which is okay if I just use ac3 audio. So I'm happy with that. As a test, I also made a higher bitrate dvd using 8000 average bitrate, 6000 minimum, and 8800 max. Again, no problem - highest spike was 9096 so that's still okay. But that's just a bit larger than a DVD-5 can hold, so I'm not sure if it's worth it to use a DVD-9 for just a small increase in bitrate. I may just stick to the 6800 average bitrate one, since it looks great.

However, I went on to try it with a second file, and it's making nasty spikes again. The video is just under an hour, so I used the same higher bitrate settings as above - 8000 average bitrate, 6000 minimum, and 8800 max. This time, it gave me a spike of 10031 in Bitrate Viewer using Second mode. This was using 7-pass (8 including the first pass). Other than that it was great, but there is that one nasty spike - with audio it will bring it to over 12000-13000 kbps. Why won't CCE adhere to my max bitrate? I tried it twice and it does the same thing. The second time I added an extra pass hoping it would iron it out, but there was no difference. I know I can just burn it anyways, but I don't want the dvd to be out of compliance and maybe cause problems later on. Can I just ignore that one single spike or is it too high?

I ran it through Muxman as a test, and it completed with no problems. In the log, it lists the bitrate as:
Bitrate - avg: 8631935, min: 6416111 (lba 0), max: 9296815 (lba 13880).
So am I good to go with this?

By the way, DVD patcher looks pretty cool. If I use it to patch my header's bitrate, should I put in the maximum peak bitrate or the average bitrate under custom?

[manono]

As for DVD Patcher, it's useless for what you're using it for. Many, if not most, retail DVDs have the max bitrate listed as 9800 (check out a few) when, in fact, it's much lower than that, usually 7500-9000. As I said before, TDA warning you about that is a bug in the program and can be safely ignored. There's no way for it to know the real max bitrate until authoring.

If it authors using Muxman without buffer underflows, everything is OK. It looks to me like Bitrate Viewer in Second Mode doesn't apply when trying to decide whether or not it can be authored safely. One of the other two modes should be more applicable. Did the other 2 modes show it keeping to the max you set? And were the other above max figures you gave taken from that Second Mode also? Sometimes a video doesn't ever reach the max you set.

[sasuweh]

Well, I scanned it in Bitrate Viewer again, this time using GOP based, and the peak is only 8795 (average 8101). Same thing with GOP enhanced. That sounds much better. It did process fine through Muxman, so like you said, I guess it is okay then?

I really appreciate all the help.

[manono]

I suppose you're using TDA to make menus as well as for authoring the DVD. But, if all you want is a good muxer then Muxman is the better authoring program. To answer your question though, yes, I think it's OK. I said the same three days ago when first stumbling across this thread. And at the time I was just echoing Vidd who said the same a day before that.