CCE and Maestro Trouble! Chapter points OFF!!!!

[F-ING HOSTILE]

Hi- Recently I switched from TMPGEnc to CCE. I am amazed at the quality and speed of CCE, but have run into a slight problem. After encoding with CCE and running my mpv through "pulldown.exe", I
import the resulting m2v, along with the audio track, into DVD Maestro and begin authoring my movie. The problem comes when I import my "Chapter Xtractor"chapter points before compiling. All the
chapter points seem to be exactly one second early.I used the same settings as usual in chapter xtractor to create my chapter points (viewing and saving with the DVD Maestro presets).
I'm guessing the problem lies in the m2v CCE created, as I had never experienced this with TMPGEnc.

I am using Cinema Craft Encoder SP v 2.50.01.00.
My settings are based on both the "Robshot" (aka Getting The Best Out Of CCE) and "Roba" methods, as well as the "CCE FAQ"(Doom 9) and the original "User's Guide" that came with the program.
My template is as follows:

Video File: ES, MPEG 2, Multipass VBR (3 pass)
Video Settings: The following are checked: Add sequence end code, Progressive frames, Linear quantizer scale, Zig-zag scanning order, Luminance level=16 to 235, Intra DC
precision=Auto,Timecode=00: 00: 00: 00
GOP Settings: M=3, N/M=4, GOP header=1, SEQ header=1
Quality Settings: Image quality priority=5, Anti-noise filter disabled

If anything, maybe changing the Timecode to all zeros is where I went wrong, or I may need to do something differently in the creation of my chapter points.
Any advice would be greatly appreciated!!
THANKS!

[adam]

Well are you inputting your chapters into CCE? A chapter can only be placed on an I-Frame, so if you don't manually insert an I-Frame at each chapter point, via your encoder than your chpaters will always be about 1 second off.

The easiest way to do this is to make all of your settings in CCE and then save a project file. Open the file in a txt editor and scroll down to the very bottom. This is where you input the frame #'s for each chapter. You can get these from ChapterXtractor but its easier to use Scenchap. Just drag your ifo onto it and then hit file/save chapter files. Check CCE Chapter point (film) for ntscfilm encoded material and then hit save. It will create a new txt file with the chapter points, so just copy and paste them at the end of your CCE project file and save. Your chapter points should be completely accurate now. Make sure you set Maestro to drop frame though, as this could also make your chapters off.

A couple of suggestions regarding your settings. I know what that guide says that you are following, but it is wrong. Luminence should always be set to 0-255 when encoding a DVD source.

As far as image priority setting, 5 is WAY too low for DVD output. At the 2mbit range this would be good for a very action packed movie. At the 4-5mbit range you should be using something like 25 or so.

As far as linear vs non-linear quantization. Some say its best to use what the source DVD used, according to what a bitrate viewer reports, but bitrate viewer is very inaccurate about these things sometimes. According to the CCE manual, there is no advantage to ever using linear quantization and that non-linear is superior, so I recommend always using non-linear.

[Doggiedaddy]

This maestro guide only works for single PGC titles (most DVD's ARE). With this in mind, this guide should still work for most of the titles out there.

trilight's work
http://www.doom9.org/index.html?/mpg/maestro2.htm

trilight also has a VERY good scenarist guide if interested.This one uses a lot of the hard work by eyesonly and others with DoItFast4u and DoCCE4u. GREAT QUALITY full BU.
http://www.doom9.org/index.html?/mpg/doitfast-guide.htm

Dd

[bacardi/avt]

I know I should not ask a new question in this thread but since the reason why I wonder is here I do anyway....

@Adam, can you explain in more detail why the setting should be as you describe?

Luminence should always be set to 0-255 when encoding a DVD source.

And what does this setting really do?

As far as image priority setting, 5 is WAY too low for DVD output. At the 2mbit range this would be good for a very action packed movie. At the 4-5mbit range you should be using something like 25 or so.

I also have been using the guides on www.doom9.org and I am very happy with the results, but if I can get even better results with a different setting I would of course use that instead...

[BJ_M]

I know I should not ask a new question in this thread but since the reason why I wonder is here I do anyway....

@Adam, can you explain in more detail why the setting should be as you describe?

Luminence should always be set to 0-255 when encoding a DVD source.

And what does this setting really do?

As far as image priority setting, 5 is WAY too low for DVD output. At the 2mbit range this would be good for a very action packed movie. At the 4-5mbit range you should be using something like 25 or so.

I also have been using the guides on www.doom9.org and I am very happy with the results, but if I can get even better results with a different setting I would of course use that instead...

Luminence should always be set to 0-255 when encoding a DVD source becuase the source is already 7.5 IRE and setting it to other would then compress the Luminence level further ..

cases where you WOULD set 16-235 are using your own sources (except DV which is already 16-235 and also some other video sources), like for example animation rendering -- though properly, you would adjust levels before it gets to the encoder stage. Also some japan source videos (japan uses NTSC but full 0-255 unlike north america - at least in broadcast - DVDs may or may not be.)

[F-ING HOSTILE]

Well are you inputting your chapters into CCE? A chapter can only be placed on an I-Frame, so if you don't manually insert an I-Frame at each chapter point, via your encoder than your chpaters will always be about 1 second off.

I basically used the same method I learned on. This always worked fine with TMPGEnc!

I think I understand what you're saying about adding the chapter points in CCE before encoding. That "Robshot" method discusses it. According to that guide you can save the CCE project (after setting your template), and then open the ecl in notepad to view the text. Then I would use Chapter Xtractor to open the IFOw the CCE template (as described in the guide) and add the chapter text to the end of the ecl text, and save it as a new project file.
Does that sound correct?
I'm not 100% clear on the concept. I would be incorperating the chapters into my video BEFORE I encode it?
Would I then skip the "Import Chapters" step in Maestro?
Would the chapter points automatically be set when I import my video into Maestro, leaving me to just drag my AC3 and Video and compile?

Thanks for the help! I also appreciate all the advice on my template settings! It seems you can read several guides, and get several opinions on what a specific setting should be, many of which contradict the actual owners manual!

[BJ_M]

i could see where sometimes the time code between cce and maestro could get a little messed up --

if you start your time code in cce at the 1 hour mark , then you would set maestro to 0 based instead of asset based time code ...

same goes for importing chapter points (I frame insertion) into cce , either offset your chapter points or make cce start at 00:00:00:00

[F-ING HOSTILE]

As far as linear vs non-linear quantization. Some say its best to use what the source DVD used, according to what a bitrate viewer reports, but bitrate viewer is very inaccurate about these things sometimes. According to the CCE manual, there is no advantage to ever using linear quantization and that non-linear is superior, so I recommend always using non-linear.

Again I was kind of basing that setting on the "Robshot"and"Roba" methods. One said set it at 5, and the other said 15. The template description says that "the further left (the lower the number) the more bits that will be allocated to complex parts". That is why I chose 5.
Would having such a low number result in the rest of the video being allocated an insufficient bitrate?

Thanks again!!

[F-ING HOSTILE]

i could see where sometimes the time code between cce and maestro could get a little messed up --

if you start your time code in cce at the 1 hour mark , then you would set maestro to 0 based instead of asset based time code ...

same goes for importing chapter points (I frame insertion) into cce , either offset your chapter points or make cce start at 00:00:00:00

If I could go both ways, then I would prefer to just use Chapter Xtractor, and import the chapter points in before I compile (Just as I did with TMPGEnc). I have CCE set to 00: 00: 00: 00, so what would I need to change (in Extractor, I guess)to get them to line up?
Thanks!!

[BJ_M]

i could see where sometimes the time code between cce and maestro could get a little messed up --

if you start your time code in cce at the 1 hour mark , then you would set maestro to 0 based instead of asset based time code ...

same goes for importing chapter points (I frame insertion) into cce , either offset your chapter points or make cce start at 00:00:00:00

If I could go both ways, then I would prefer to just use Chapter Xtractor, and import the chapter points in before I compile (Just as I did with TMPGEnc). I have CCE set to 00: 00: 00: 00, so what would I need to change (in Extractor, I guess)to get them to line up?
Thanks!!

nothing as long as your movie also started at 0

[adam]

The chapters themselves are not added until authoring, in Maestro in your case. But chapters can only be placed on an I-Frame. So if you tell Maestro to put a chapter at 1:25:20, which is where it is on the DVD, but your nearest I-Frame is at 1:25:19, than that's where the chapter ends up. So if you want frame accurate chapters than you must manually insert an I-Frame anywhere you plan on later adding chapter points.

Your above process sounds fine, I just prefer to use Ifoupdate because its needed in the process anyway.

As far as image quality priority. Yes lowering the # does add more bitrate to high motion/complex scenes, but the more given to these scenes, the more that must be denied to flatter scenes. As a result, contour noise can appear. You always want to get the best compromise between the two. When dealing with relatively high bitrates, which you have even on a DVD5, you can get away with a higher image quality priority. With an avg of 4-5mbits you are going to have no problems achieving max bitrates in the high 9mbit range. So while its still important to allocate enough bitrate to these high motion/comples scenes, there is no reason to use such an extreme setting. You are goign to have plenty of bitrate for high motion scenes regardless. The CCE default setting (I'm using 2.5) is 25 and it recommends that the image quality be raised in proportion to your bitrate. At 25 I never see any mosquito noise when encoding DVDs, and I can't say that I've ever noticed any contour noise. When I lower this setting all it does is incrase contour noise on some sources.

Now when I encode at lower bitrates ie: ~2mbits I often go as low as 5 on very complex footage. But for high bitrates this is inefficient.

[BJ_M]

what i believe adam is saying goes for cce 2.5 , 2.66 and up use different settings

[adam]

Right sorry. 2.66 and up name this setting something different, and the scale is adjusted. Same basic principles apply, but you'll have to experiment to find what settings work best for you. The CCE manual for these versions gives good advice and suggested settings for these versions as well.

[disturbed1]

Luminence should always be set to 0-255 when encoding a DVD source becuase the source is already 7.5 IRE and setting it to other would then compress the Luminence level further ..

cases where you WOULD set 16-235 are using your own sources (except DV which is already 16-235 and also some other video sources), like for example animation rendering -- though properly, you would adjust levels before it gets to the encoder stage. Also some japan source videos (japan uses NTSC but full 0-255 unlike north america - at least in broadcast - DVDs may or may not be.)

Adam and I were discussing this in another thread, but I'll post here since it deals on the same subject.

The CCE color conversion is only valid for RGB input. It converts RGB color space to YCbCr output.

If avisynth is your input, it's already in YCbCr if YV12 or YUY2, so the setting is ignored.

If you use VFAPI converter to convert the d2v file to an avi, and you chose RGB in DVD2AVI then this option is used. It's also used for all RGB codecs (most DV).

[adam]

I've read this elsewhere Disturbed, that if the source and output are YUV then the luminence range set in CCE makes no difference. But I just can't ignore the fact that when I do two encodes back to back, each with different luminence levels set, I get noticably different picture on the tv. Assuming this setting does still affect luminence ranges even when the material is YUV, and assuming it is already stored in the 16-235 range ie: DVD, DV, then 0-255 would be the proper setting to use. So the way I look at it, if I'm wrong than it doesn't matter, and if I'm right than I'm safe. In any case, using a setting of 16-235 on a 16-235 source would never be correct under any circumstances, so I recommend that everyone simply use 0-255 in CCE for 16-235 sources.

[VinnySem]

I also have this issue with the chapter points being off and I load the CCE chapter points into DoCCE4U. The time shown in the Maestro chapter bar is the same as what is listed in the chapter file, but the picture is off. I don't know enough about what takes place during a reencode to determine what the culprit is for this drift. What I've resorted to is one by one going over each chapter point in maestro, comparing it to the original DVD, and moving the chapter until it is at the correct GOP. Although tedious, it is not time consuming, around 20 seconds per chapter point..

[borghe]

question. can I force TMPEG to manually insert I-frames? The first major project I encoded with TMPEG it automatically detected scene changes and started new GOPs. This worked beautifully. Unfortunately, I needed to compress the ogirinal again. This time I unchecked Close all GOPs as it seemed like I could and supposedly makes the file size smaller. Well, low and behold my GOPs were now all exactly 15 frames and the best I could do on chapter stops was within 1/2 second. Is it because I told TMPEG to leave GOP frames open? I am reencoding again now with the Close all GOPs option checked, and hopefully that will solve it. Although I know precisely where I want the chapter stops to go (thanks to VDub) so any way I could manually insert those into TMPEG would be great.

[BJ_M]

yes you can with TMPGenc . under GOP structure check off "force picture type setting" and hit "setting" go through and enter your I frame insertion ...

[F-ING HOSTILE]

yes you can with TMPGenc . under GOP structure check off "force picture type setting" and hit "setting" go through and enter your I frame insertion ...

Just out of curiosity, why is it that I can go back and encode the same movie with TMPGEnc, open Maestro, import my audio/video and Xtractor chapter points, and all is well?
What is different with CCE that I must now be concearned with the I-frame placement (where all the same things compile fine with TMPGEnc)?

Thanks a million for all the great feedback!

[F-ING HOSTILE]

The easiest way to do this is to make all of your settings in CCE and then save a project file. Open the file in a txt editor and scroll down to the very bottom. This is where you input the frame #'s for each chapter. You can get these from ChapterXtractor but its easier to use Scenchap. Just drag your ifo onto it and then hit file/save chapter files. Check CCE Chapter point (film) for ntscfilm encoded material and then hit save. It will create a new txt file with the chapter points, so just copy and paste them at the end of your CCE project file and save. Your chapter points should be completely accurate now. Make sure you set Maestro to drop frame though, as this could also make your chapters off.

Thanks for the help. This is essentially what I have been attempting to do, but perhaps the settings I have been trying to use with Chapter Xtractor are not right. I open Chapter Xtractor, import my IFO, choose the CCE preset (suggested by the "Robshot" guide) which is "Format: chapter=%f:" (which has no Header), and copy the new list of points to the bottom of my ecl list.

Is this correct?

Then after the encoding is complete, I import my audio/video into Maestro, drag both into the "Movie 1" screen as usual, use chapter xtractor to create my usual chapter points (with Maestro presets), and compile? If this correct then I am missing some little detail along the way
I might just be entirely off

Thanks A Ton!!

[F-ING HOSTILE]

Hey everyone, THANKS ALOT!!!

I think I finally figured out what the problem was. I don't think I was getting my I-Frame points set correctly. The "Robshot" method I was following gave a Chapter Xtractor format template(chapter=%f: ), but after using this to paste my chapter/I-Frame points to the bottom of my ecl project before encoding, and importing chapter points in Maestro, things still didn't line up!

Your posts got me to thinking about alternative IFO programs (since there seem to be no CCE presets in Chapter Xtractor), and I reviewed the "Roba" method (Doom 9) in which he recommends IFOUpdate. IFOUpdate had presets for CCE (to add I-Frames to my encoded video), and it also had DVD Maestro presets (for my authoring chapter points)!
I think using the same program for both did the trick. Every chapter point now seems to be in the correct place! Hopefully I've got it figured out!

Thanks again, I really appreciate the help

[adam]

The problem you had with ChapterXtractor may have been because you had it set to 29.97fps instead of 23.976fps. This will affect the chapter times. I think IfoUpdate is easier to use for this step, and I guess in your case its better as well.

Just for your info, you can test that your chapters are placed correctly in CCE before encoding. On the encode settings tab click the settings button which is just to the right of the line titled Input files. On the new screen that pops up double click on the file there and the chapter screen will pop up. You can scroll through your movie and manually add chapters, or you can select already input ones. So go through the chapters and make sure they where they should be, but now that you use IfoUpdate everything should work fine.

[BJ_M]

The problem you had with ChapterXtractor may have been because you had it set to 29.97fps instead of 23.976fps. This will affect the chapter times. I think IfoUpdate is easier to use for this step, and I guess in your case its better as well.

Just for your info, you can test that your chapters are placed correctly in CCE before encoding. On the encode settings tab click the settings button which is just to the right of the line titled Input files. On the new screen that pops up double click on the file there and the chapter screen will pop up. You can scroll through your movie and manually add chapters, or you can select already input ones. So go through the chapters and make sure they where they should be, but now that you use IfoUpdate everything should work fine.

bringing up the settings box brings up another good point which is not often mentioned .. it is in this box (which looks just like the main screen) where you add files if you want them joined together in the encode ..

not all types can be joined (though you will get no warning) ..

tmpgenc and mainconcept also can do this ..

[Faustus]

Strangely a friend of mine recently had a problem like this using DVD2SVCD to encode for him. He would use the chapter file that DVDdecrypter pulled down and it would put all the chapters at the end. I told him to still use ChapterXtractor and he's been fine since then, at least he hasn't had any further problems.

[F-ING HOSTILE]

Thanks again guys!!

Now that I think about it, I remember a few movies I encoded with TMPGEnc having sporadic chapter points (about every tenth film). I always wrote this off to carelessness, but it may have been the result of poor, or no I-Frame reference!

[dpcpro]

For me, the chapter points in Maestro are never accurate even when the movie doesn't require re-encoding. All I do is demux the video + audio, and then import the chapters. I have used the .chp that DVDDecryptor generates and also the one from ChapterExtractor with the same inaccurate results.

Anyone else have this experience and know how to fix it?

dpcpro

[Faustus]

If you insert the chapter point timecode into the CCE script the points "Should" not be off.

[F-ING HOSTILE]

For those still following this thread, I just figured out something very interesting, that may often be overlooked! Even after my last post (where I eagerly stated "all is well") I tested a few more DVD's and found that the chapter points were still off. I would encode and then assemble my entire project in Maestro, and view it side by side with the original in my media player to check the chapter points. The chapter points seemed to be consistently off (about one second later than the original). Even though I was inserting I frames in my ecl (with IFOupdate)before encoding, and using the same IFO program (IFOupdate) to create my Maestro chapter points, the chapters were still not exactly on cue.

I finally figured out what was causing the imperfection! With my project ready to compile, I right clicked on the gray chapter line of Maestro, and selected properties. Under the "Snap Chapter to GOP" setting, "Next GOP" was selected. When I switched the setting to "Nearest GOP" and imported the chapter points again, I was able to watch the points "snap" back to their precise locations. I'm guessing that choosing "Nearest GOP" allowed my chapter points to accurately "find" the I-Frames I had inserted earlier. Whatever the explanation, things are now PERFECT

Thanks for all the help!!

[Mr Grumpy]

.. it is in this box (which looks just like the main screen) where you add files if you want them joined together in the encode ..

tmpgenc and mainconcept also can do this ..

How do you do this? It's something I liked in CCE that I couldn't find a way of doing in TMPG.

[BJ_M]

tmpgenc = This function can be enabled in the Environmental settings window, General tab, "Open sequence files as a movie".

then all avi (only) must be in same folder and named in order (avi001.avi , avi002.avi, avi003.avi -- and so forth)


then output will be created as if these were all one file ..





this is different than an image seq.