K-DVD'S with DVD-Rebuilder & Quenc.

In the past I experimented with CCE & DVD-Rebuilder but that did'nt work out too well for making K-DVD,
With "cce" DVD-Rebuilder does NOT support the KVCD NOTCH MATRIX
(thanks phil for pointing that out)

But with Quenc DVD-Rebuilder DOES support the KVCD NOTCH MATRIX.

DVD-Rebuilder & quenc DON'T work really well for K-DVD
(Quenc actually did an amazing job, BETTER THAN I EXPECTED)
but the quality still sux !!

I did "Armageddon" (2hrs 20 mins long movie)
all menu's,all languages,all subs& all directors commentary & all music video onto 1 DVD+R 4.7 GB with some quality loss in 4 1/2 hours
(2pass high Quality)

A word to the wise tho ... Don't hate DVD-Rebuilder just because its a "Doom9" tool, It would be excellent
(if some1 invents a kvcd notch matrix patch for it)
Those who have tryed it ... I'd like ta hear some feedback

Make sure u add these filter lines to DVD-REBUILDER

Undot()
Deen()
LetterBox(16, 16, 16, 16)

...oh my GOD.....
dangerous comments in your post, read my feedback:

jdobbs give magnific suport for dvdrb..my personal opinion is that:
i never see suport like jdobbs(doom9) and vmesquita(kvcd) do.
see some members from doom9 that have my personal frienship:
Sh0dan,Wilbert,r6d2,gerty,neuron2,hakko,Tom Barry,oddyseus,ammck,dave82,Eye of Horus, girgobain,DSPGuru,Doom9,Kpex,2cool,jsoto,DDogg,Nic ,mrBass,dvd2svcd,Mug Funky,chickenman,wmansir,daphy,Bach(is gone from d9).....and lots more that i can't remember his usernames now but they know who they are!
see that in my posts in the doom9 forum,one of them always help me when i need....take a look at the end of the third post from hakko in dvd2avi faqs and see if you can found my username there...look at Quenc thread,and see how many posts i did there ...in audio forum...in dvd2svcd forum.....in avisynth forum...QuENC forum...ifoedit/vobedit forum...please,check. better, use the search with my username!
like i wrote in other thread here, i download all versions of dvdrb and the guides but still don't test!
nothing against doom9 forum, his members and tools...and believe me,i'm waiting for dvd2svcd with Quenc(Nic) a long time,you will find my post there showing it and my personal adimiration for jdobbs suport and i still don't use dvdrb!(this was posted there a long time)
a few minutes i did one question there and now i have someone helping me (and faster),take a look: http://forum.doom9.org/showthread.ph...threadid=79291
then...i don't "hate" anything from doom9 forum(tools and members) and i'm there everyday like here.
and i trust that this is the general opinion of the whole kvcd forum too.

i'm not a mod but your comments are dangerous in my opinion!

what's so dangerous about it,
yesterday some comments were made by some forum members
(i wont say no names)
about DVD-Rebuilder & doom9
personally I love doom9 & kvcd.net
those are the only 2 keepin it real out here !!!

Does it work cell by cell with Qenc like it does with CCE ?

If yes, trash the tool ! You CAN'T achieve a correct quality in 2pass if your work like that. That is so easy to undersand by anyone except the tool designer unfortunally.

The comments are made by ME, Philippe, Aka Dialhot and I signed them !
So much version of this tool that works in 2pass CELL BY CELL is nothing but the proof than people "there" should do other things than video editing.

There is nothing personnal againt noone, just this constant way of thinking they know what they do better than anyone on earth but don't see the slighest elementary problem in their speeches.
(and I don't even speak about using standard matrix and being astonished by the result even if this result is crappy compared to KVCD-Notch )

Now, if with Qenc the tool works differently, that's an other story...

Actually phil the quality sucks if it's not in cell by cell mode
its only decent if u use cell by cell but all in all
I like dvd rebuilder but que enc quality ain't that of cce
now if somone here can develop a patch that will patch dvd-rebuilder to use kvcd notch matrix then i will use it with cce but nothin else
(u can uncheck the cell by cell feature but i did tests &
my opion is dvd-rebuilder is awesome but not for kvcd as it doesn't let cce use KVCD NOTCH MATRIX

As I told to jorel I have an idea to use DVD-RB for step 1 (prepare) and step 3 (rebuild the DVD) BUT to do the step 2 (encoding) ouside the tool. This way we will have the hand on the encoding and will be abble to use any matrix we want

Some tests to do...

please keep me informed

A good idea, I was thinking along similar lines earlier today.

One would need a new Avisynth script (simply copy one of the generated ones and remove the Trim line), and then encode the entire VTS at once with whatever encoder. It would then be necessary to use a frame-accurate splitter to chop the resulting elementary stream into pieces precisely the same size (in terms of frame length) as the original cells. The new files, if named according to DVD-RB's naming scheme, would then be remuxed into the VOBs in the Rebuild phase just as if one had used the three-step process. Pretty simple, except for the splitting.

That said, I'm not sure what one would gain from doing it. CCE does a pretty amazing job even at fairly low bitrates, though arguably the custom KVCD parameters (matrix, GOP length, etc.) could be used if an ultra-low bitrate were needed.

My idea is different :
- merge the script (same as you)
- do the prediction on the whole movie for a 1-pass Qbased encoding (like we always do - at least Ialways do)
- encode each cell one by one, using the original script of DVD-RB, but with the Q computed in the previous step.
It SO much easier

Not needed
In fact I didn't do it yet just because I did my test in perl on my own PC and I really don't have the time to see a way to distribute that to people on the forum

again NOT in cell-by-cell 2pass encoding. And NOT if you don't use the KVCD-Notch matrix (2 points we saw in other threads so don't need to double post here).

A good idea. Shall we test it? If you want to give it a try, make a new thread for the results .

I've been coding in Perl on an almost daily basis for about five years, so I won't have any problem understanding it . Once we've got a nice Perl script going, and our testing is successful (of course it will be...), I can maybe whip up a Win32 native app to do the same thing.

(Note: We don't have to limit ourselves to CCE. It would almost certainly be possible to automate TMPGEnc to do the same thing, though CCE is much faster, obviously.)

Well, I've achieved exceptional results from cell-by-cell 2-pass and the standard MPEG-2 matrix with video up to 3 hours in length. But that's another topic for another thread .

Try the "Notch" matrix, and that should be able to push your 3 hour footage per DVD to 4 hours per DVD, with the same visual quality (I do 5 hours with slight loss, but still excelent quality on my HDTV)
Give it a try. It works fine with CCE

-kwag

I'll definitely try it on my next low-bitrate encode. What GOP length do you recommend for NTSC and the Notch? 15 or 18?

Hum... my script was only something to merge the avs generated by dvd-rb. Very simple !

But in fact I realize now that you don't even need to do that, you can do a script by hand as that is you that tell to DVD-RB the filters you want to put in the avs script.

Or perhaps I forgot some element of these avs that are important.

@Dialhot, and all:

I made a new thread for kicking around ideas for a custom encode phase for DVD-RB. Please come contribute!