Multiangle Back-ups W/ DVD-RB

[Jomile]

Are there any special settings in DVD-RB or in CCE that must be adjusted to process a multiangle dvd?

Such as in close GOP's and restrict auto-i-frame insertion?


Can the angles be removed?
How are they dealt with?

Thanks

Jomile

[the-warriners]

Multi angles are not currently supported in DVD-RB
If you want to process a M/A disk you need to remove the other angles first using a seperate tool.

[htc10825]

U can use CloneDVD or IC8(better quality) for multi-angels.

[Jomile]

htc:

I am currently using BIG3 method. (CCE). Are you saying the quality is better using BIG3 vs DVD-RB?

[Joergen]

Well if you consider not-backuping-at-all a level of "quality", then you could say BIG3 is better.

Due to all the bug hunting, jdobbs has not been able to start implementing multiangle. By the time you're finished fighting with BIG3, jdobbs mighthave already included it :p

[Trahald]

Originally posted by Joergen
Well if you consider not-backuping-at-all a level of "quality", then you could say BIG3 is better.

Due to all the bug hunting, jdobbs has not been able to start implementing multiangle. By the time you're finished fighting with BIG3, jdobbs mighthave already included it :p

ouch! ;)

[lab-one]

U can use CloneDVD or IC8(better quality) for multi-angels.

I think the statement reads, IC8 produces better quality than CloneDVD. I didn't take it as a "BIG3 vs. DVD-RB".

[htc10825]

So it was!

Using BIG3, you can get the best quality possable - by tweaking the encoding parameters, i.e. change GOP to 15 or/and use 1-pass vbr/constant quality etc...

Another important thing: BIG3 distribute the available bits over PGC/TT, but DVD-RB does it within cells - the "dynamic" option is depend on the bitrate distribution of original DVD. We know, the encoder which the big studio used is not very effecient. Thus the bitrate curve of the original are not optimal. The difference in quality is just like that between DVD2one(reduced at a constant factor) and IC8(redistributed the bitrate curve by analysing the complexity/compressability along the PGC/TT)

Conclusion:
If we say the quality of BIG3 by manual tweaking ist 100% then the RB-Method can reach 80%(for main movie) and 90%(for extras, mostly one cell per title)

But all above are refered to long movie the many extras. For the movies the avg. bitrate over 3.5M you will not able to tell the diff.

[jdobbs]

I completely disagree and believe your logic to be flawed. I don't think you understand how DVD-RB works. In dynamic allocation DVD-RB uses the originally assigned allocation as a baseline for assigning storage so it isn't limited to cell size. I think you are mistaken in assuming the studios don't know how to encode. Also the sample-size to bitrate distribution efficiency if far from a linear relationship and has a fairly early leveling point...

On the other hand it isn't a contest so it really doesn't matter. Use Big-3 if you think it's better.

[Joergen]

You're right jdobbs, and yes it isnt a competition.

And in any case, no hollywood movie is worth my time doing BIG3 now that DVD-RB is available, so I'd much rather keep those movies on the shelf until DVD-RB is ready to handle them, or strip the angles if they're not important.

[htc10825]

I think you misunderstood me. I'm highly appreciate your works sofar. I'd like use your program now and in the future. I only want to point out that the lack of user intervention of encoding parameters make a optimal encoding not possable. But it is for most movie not neccesary - due to higher bitrate. dynamic allocation DVD-RB uses the originally assigned allocation as a baseline for assigning storageI meant so!so it isn't limited to cell size I've never meant so!I think you are mistaken in assuming the studios don't know how to encode.the studio use hardware realtime encoder to reach very good quality at higher bitrate(usally >=4M), but never concerd during construction to be used at a lower bitrate (2.5M, 2.0M and lower). So use the originally assigned allocation as a baseline would not be optimal for encoding at lower bitrate. So it is better to leave the CCE encoder to make decision.Also the sample-size to bitrate distribution efficiency if far from a linear relationship...What did you mean?

By the way, I use RB to backup allmost all my DVDs, especially the one with BOV - in BIG3 you have to find out the button commands in vobs and add it manually in scenarist thus not simple. The last example is Charlie'e Angels 2, all extras/specials encoded at full D1 - the subtitles under half D1 still get mashed... I think it must be different to rebuild the vobs if the video resolution is hald D1... I hope you'll soon get rid of it...
I'll donate this program if the MA also works...

[Joergen]

Studios dont use hardware real time encoders for the final production masters, and they dont master in MPEG2.. lossy compression is not used until the final master stage.

Believe me, when in college a couple years back we used a studio dvd mastering lab with SGi computers that did multiple passes (several hours) on the final encode, and the authoring was done with scenarist. Funny thing though, these days a fast PC with CCE Basic and a good authoring package delievers the same quality with a fraction of the cost (the SGi setup cost about 80 000 USD).

In fact it would be interesting to know how many studio discs are encoded with our favourite encoder, CCE.

[jdobbs]

I can guarantee you that if a reputable studio contracted out a movie to be authored on DVD and it came back with a half-ass encode -- the authoring firm would be out of business very quickly. This level of encode and authoring job is generally done by highly skilled professionals (some of whom may be reading this post, as DOOM9 has a pretty significant following). They ain't home-hobbiests.

;)

[PleXuS]

i think if you look to qwality and time.
DVD Rebuilder is the best!

And when i check for qwality differents between DVD Rebuilder and BIG3 its pretty the same!

it's allmost a 100% copy! and DVD Rebuilder is damn good!

BiG3 Sucks becoz of slow tools like Scenarist.
With allot of subtitles i takes hours to get it done with scenarist!
Dvd Rebuilder takes only around 2hours and its done fully


So great job jdobbs!!

Maybe next time you can integrate a tweaked encoder in the tool :-)
But yeah modular software is the best :-)

GreetZ
PleXuS

[lamster]

Originally posted by Jomile

Can the angles be removed?
How are they dealt with?


Yes, they can be removed.

If you use DVD Decrypter to get the DVD onto your hard drive, then you can click Tools / Settings, and under File mode check the "Multi Angle Processing" checkbox.

For stuff already on your hard drive, you can use IfoEdit to remove the angles. Click on the "VOB Extras" button, and select "Remove Angles".

Here's a detailed example of how I did it for Shrek, using IfoEdit. (I have no idea if there's a better way; this is what worked for me.)

First I loaded VIDEO_TS.IFO and looked at VMG Overview to see which titles have multiple angles. Next, loaded the VTS_nn_0.IFO for that title set. Looked at the VTS Overview (initial screen) to see the program chain information. It lists which cells have multiple angles. Double-clicked on each to see what's in it - scrolled through it, and saw that the only place in the whole movie where separate angles were actually used was in the title where it says "Dreamworks Pictures presents" (for angle 1). Angle 2 had it in French, and angle 3 in Spanish. Since I didn't really care about having the text there match the language being used, I figured I was good to go.

I then clicked on the VOB Extras button, picked a destination directory, checked "Remove Angles", left "Keep Angle" set to Angle 1, clicked OK, which brought up the Stream List dialog, clicked "Strip It", and let it chug away. After it finished updating the VOB information, there's a dialog on which you have to click OK, then it completes the processing.

Finally, I copied the new directory back over the original. I then checked and saw that VIDEO_TS still said it had 3 angles, so I used IfoEdit to update the Video Manager Title Play Map table (*) in VIDEO_TS.IFO to specify 1 angle. After that, I was able to use Rebuilder to compress it down without any problems.

(*) Note that this is the actual "VMG_PTT_SRPT" table, which you get to by clicking the third line in the top IfoEdit panel. Don't confuse it with the VMG_PTT_SRPT summary which is shown in the VMG Overview (shown when top line in top IfoEdit panel is selected) - the latter is not editable.

[nimbles]

thank you so much lamster, I've been meaning back up my star wars I and II dvd's but thought i couldn't with rb,

So with decrypter- (as i'll be ripping the original disc) do i have to specify which angel gets kept or is there a default angle which decrypter can work out for itself?

[lamster]

Originally posted by nimbles
thank you so much lamster, I've been meaning back up my star wars I and II dvd's but thought i couldn't with rb,

So with decrypter- (as i'll be ripping the original disc) do i have to specify which angel gets kept or is there a default angle which decrypter can work out for itself?

There's no default angle per se; if you check the "Multi Angle Processing" selection, then "Retain Angle 1 in VOBs" gets enabled, but I think Angle 1 is picked just because it's the first angle.

To tell for sure, you can watch the movie using a program that displays which angle is selected, and see which one is being used for your setup. E.g., on Attack of the Clones, angle 1 was for English, and angles 2 and 3 were for French and Spanish. The only part that differed depending on angle was at the beginning, where the text scrolling off into the distance was correct for the selected language.

Note that I've never personally used DVD Decrypter to remove angles; I've only done it with IfoEdit.

As far as which angel gets kept, you'll have to ask Charlie. :)

[lrosado]

I cannot get DVDRB to process "You Got Served" multi angle movie but I removed the angles through dvddecrypter and removed all indications in the VIDEO_TS.ifo and the VTS_xx_x.ifo of angles and it still states in DVDRB that it cant handle multiangle dvd's. I even used IFOEDIT to strip the angle out and still the same result. I have read success using the above methods but I have had no luck. Can some one give me detailed instructions on how they do it to make sure I am doing it right? Thanks

Larry

[jdobbs]

Originally posted by lrosado
I cannot get DVDRB to process "You Got Served" multi angle movie but I removed the angles through dvddecrypter and removed all indications in the VIDEO_TS.ifo and the VTS_xx_x.ifo of angles and it still states in DVDRB that it cant handle multiangle dvd's. I even used IFOEDIT to strip the angle out and still the same result. I have read success using the above methods but I have had no luck. Can some one give me detailed instructions on how they do it to make sure I am doing it right? Thanks

Larry DVD-RB determines multiangled input by the flags in the IFO files. As I recall (I'm away from my code) I was looking at the first entry in the cell playback information table of the PGCs to see if a cell was listed as "angled". I'm not sure whether any of the preprocessing packages change the IFOs or not.

[lamster]

Originally posted by lrosado
I cannot get DVDRB to process "You Got Served" multi angle movie but I removed the angles through dvddecrypter and removed all indications in the VIDEO_TS.ifo and the VTS_xx_x.ifo of angles and it still states in DVDRB that it cant handle multiangle dvd's.

As I recall it, DVD-RB gave the same message for multiple angles and for ILVU. Is it possible that one or more VOBs have the latter?

If you open the VIDEO_TS.IFO, you'll see # of angles right there on the first screen. ILVU is less obvious; you have to open each VTS_nn_0.IFO and you'll see "IlVu" displayed in the program chains (if the DVD uses ILVU) in the same place that angles are listed.

[lrosado]

so how would I go about getting DVDRB to read and process these files?

[lamster]

Originally posted by lrosado
so how would I go about getting DVDRB to read and process these files?

One way would be to run IfoEdit against the VTS_nn_0.IFO that contains the ILVU, press the "VOB extras" button, and tell it to remove the seamless branch.

Be sure to test the results before writing to non-RW media.