Newbie's questions

[Godai85]

i have to apologise in advance, 'cause i don't speak english very well.
i know that most of the answers can be found on web, but i looked for without success, so i hope you could help me :P
here's my questions about dvd-re pro 1.26.5 w/ cce sp:
1)i read that yuy2 conversion is necessary for some reasons, but also if i don't use avs filters ? and if i use hc instead of cce ?
2)i work only with pal dvds that i think are always interlaced, so Deinterlace with DECOMB and interlaced=true are reccomended options or not ?
3)for a newbie, is it important to use (and understand what are :P) other matrices ?
4)i know that there isnt't the best setting for cce, but what are the most recommended for a wide range of materials (2d animation, cg animation, slow movies, action movies, dark movies, ...) ? i read (bias 20 or 25 or 30) (pass 3 or 5) (qual 16 or 32 or 20) and many more... (i don't mind about time, it is ok also if the pc work for 24h :P)
5)where is Dynamically Assign Cell Bitrates ? is enable bitrate redistribution ? if it is, is it stable or a "beta" option ?
6)months ago (when i don't use dvd-re) i read that cce sp2 isn't recommended instead of cce sp 2.7. is it always true ?
7)how can i modify % of each "part" of a dvd ? i read that i need rb-opt with dvd-re free, but the pro version have an internal tool that can do anything (video segment viewer/editor)?
8)if i use hc encoder, how can i modify advanced settings (if exists) ?
9)how can i set target size of 4485MB ? with CCETargetSectors=2296320 ?
10)do i have to add max_bitrate=9000 ? is it recommended ?

stop for the moment :P
thanx for all (future) answers ;)

[jdobbs]

1. YUY is the colorspace used internal to CCE, so YUY is fed to CCE by DVD-RB to avoid another conversion. If, however, you select HC as your encoder -- YV12 is fed to the encoder.

2. I'd recommend you just let DVD-RB do its job. It looks at the flags and responds appropriately. The exception may be in cases where something is inappropriately flagged as interlaced when it is indeed progressive. In those cases just select "Disable Interlaced" under "OPTIONS/AVS EXPERT OPTIONS"

3. No. In most cases you are best served by using the encoder's standard matrix. I almost always do.

4. The defaults are set the way they are to cover the widest spectrum of sources. If you leave them set, you are good for most sources.

5. "Dynamically Assign Bitrates" is no longer an option -- because there is ever a time when it shouldn't be used. Enable Bitrate Redistribution is still being tested... but it is stable.

6. They are virtually identical. The differences are outside of the engine itself.

7. The Segment Editor is the tool for doing what you want.

8. The "Settings" menu gives you the options you should be using.

9. That value would give you a target of 4.455GB. The formula is TargetSectors * 2048 * 1.015 = Size_In_Bytes

10. No. Leave the setting at the default and let DVD-RB automatically set it for you.

[djskribbles]

DVD+R is 4482mb and a DVD-R is 4489MB. i'd recommend you keep the target below 4482MB. i use CCETargetSectors=2254000 as the sector value you enter here is not completely accurate. if you want to be more accurate, i recommend you use the slider in the preview/edit feature and slide it to 4473mb, setting it higher will result in a size too large for a DVD+R.

[Godai85]

ok but i still have some doubts...

1)i modified cce settings and i can't remember the defaults, but i believe that passes was 2 so... what are the differences between basic and sp ? (isn't 4 passes better? it takes me 3-4hours)
2)what settings do you use ? also if you use a setting for every film genre.
3)do you recommend to use sp2 or sp 2.70 ?
4)is "Enable Bitrate Redistribution" recommended ? (but... bias is unuseful without it?)
4.1)and "apply redistribution to all vts..."?
4.2)and "always use hc..."?
5)is menu encoding recommended ?
6)keep closed captioning applies to pal dvds also? it usually takes much MB ? i prefer to preserve every subtitle

[djskribbles]

ok but i still have some doubts...

1)i modified cce settings and i can't remember the defaults, but i believe that passes was 2 so... what are the differences between basic and sp ? (isn't 4 passes better? it takes me 3-4hours)
basic is limited to two passes. more passes is better, but anything more than 3 is excessive.
2)what settings do you use ? also if you use a setting for every film genre.
personally, i use the default for precision and vbr bias, but i use 3 passes for all movies. i use the default matrix for 3700-4300 bitrate, anything below, ill use AUTOQ1 (AVAMAT6). anything higher and if there isn't too much compression needed, ill use FOX matrix. i also use xvid for the decoder as its fast and precise.
3)do you recommend to use sp2 or sp 2.70 ?
i've compared the two and as jdobbs said, they are exactly the same as far as PQ. speed may differ slightly but personally, i would just use SP 2.70 simply because the encoder actually kind of runs minimized whereas SP2 keeps popping up and its kind of annoying if you use your computer while encoding. otherwise, it doesn't really matter imo.
4)is "Enable Bitrate Redistribution" recommended ? (but... bias is unuseful without it?)
4.1)and "apply redistribution to all vts..."?
4.2)and "always use hc..."?
i'm by no means an expert on this, but i think this is only worth it if the film requires a lot of compression or has a lower bitrate.
i think the apply to all vts is only required for episodic discs or multiple angle films correct me if i'm wrong?
and the "always use hc" fixes a bug that you may encounter with some movies if you use CCE.
5)is menu encoding recommended ?
afaik, menu encoding should give you a little better quality for everything else as it re-encodes the menu as well, otherwise its untouched but i'm not totally sure on this one either.
6)keep closed captioning applies to pal dvds also? it usually takes much MB ? i prefer to preserve every subtitle
dunno, will have to wait for someone else on this one.

[robot1]

6)keep closed captioning applies to pal dvds also? it usually takes much MB ? i prefer to preserve every subtitle
As far as I know, only NTSC video have closed captioning.
Anyway, I've never seen a Pal DVD with closed captions.
But I'm not expert.

[Godai85]

thanx to all, but... :P

1)so 4 passes could do worse job than 3 ? or can be a little bit (and, for someone, insignificant) better ?
2)can you remember me the default settings please?
3)is it important to use various matrices ? or is mostly the same ?
4)what do you mean by "use xvid for the decoder" ?
5)sp2 can't do a better (even a little) job than sp ?
6)"i'm by no means an expert on this, but i think this is only worth it if the film requires a lot of compression or has a lower bitrate" sorry but i can't understand the sentence :°(
7)""always use hc" fixes a bug" how can i recognise the bug ?

[blutach]

As far as I know, only NTSC video have closed captioning.
Anyway, I've never seen a Pal DVD with closed captions.
But I'm not expert.Correct.

http://www.dvddemystified.com/dvdfaq.html#1.45

1)so 4 passes could do worse job than 3 ? or can be a little bit (and, for someone, insignificant) better ?Of course it can't do a worse job. But, the effect is only very marginally better.

3)is it important to use various matrices ? or is mostly the same ?jdobbs (and djskribbles) have already answered this question. But it is a matter of personal preference.

7)""always use hc" fixes a bug" how can i recognise the bug ?If the prepare stops and gives you an error while using CCE for redistribution. Search the forum to find out about this. And try it yourself. You can't break anything.

Regards

[Godai85]

[color=blue]jdobbs (and djskribbles) have already answered this question. But it is a matter of personal preference.

ok, but i would know if a different matrix can do a very different ('n' better) job.

remains these questions:
2)can you remember me the cce default settings please?
4)what do you mean by "use xvid for the decoder" ? (instead of dscaler, cyberlink, purevideo, etc. ?)
5)sp2 can't do a better (even a little) job than sp ?
6)"i'm by no means an expert on this, but i think this is only worth it if the film requires a lot of compression or has a lower bitrate" sorry but i can't understand the sentence (please use other words :P)

[Godai85]

i'll write here to avoid a new 3d:
when i set a % for the feature vts, it applies only on a few "v0x/c0x" (i don't know how to call them). why ?
if i set 80% i will that all vts is compressed by 20%. instead of this, dvd-rb compress some "chapter" and the others are "no reencoded". it's boring to click on every "chapter" -_-'

ps: i can't view video (black screen) in the segmet/viewer editor. is it normal ?
pps: why dvd-rb increase audio size ? for example, if the original ac3 is 200MB the result is 210... not only when it add some black frames (and why this ?)
ppps: sorry for my english XD

[jdobbs]

It will apply to all that are within the area you are changing. The "not reencoded" is an indicator that they are already small enough to meet the criteria set. It is also possible that it is a ILVU segment -- which can't be changed without carefully matching it up with alternate paths.

If the segment editor isn't working, you have installed other software that caused the internal Windows routines to fail. Do a search and you'll find which ones are the usual suspects.

The audio is a bit-for-bit exact duplicate of the original DVD-RB doesn't change it.

[Godai85]

so is it normal that part of the vts are re-encoded and many others are not ? i mustn't change it ?

last dvd that i made has an ac3 o 223 instead of 228... also subtitles size has changed... 1 (of 4) is 1MB smaller and 1 is 1MB bigger...
another "problem" is that sometimes the colours are "modificated". for example, in an animation film, the gold colour becames more yellow. is it normal cause of the colour space conversion ?

[jdobbs]

last dvd that i made has an ac3 o 223 instead of 228... also subtitles size has changed... 1 (of 4) is 1MB smaller and 1 is 1MB bigger... As I said, the audio doesn't change... so whatever you are using to do the comparison is incorrect.

[Godai85]

i opened the original and the compressed one with dvd shrink...
same length but different audio sizes.. is it a liar ? :D

for the other questions ?

[jdobbs]

If it says the DVD-RB audio is a different length than the original... yes -- it's a liar. Of course "inaccurate" may be a more appropriate term.

[blutach]

DVD Shrink's initial analysis does not go thru every pack and can be quite inaccurate.

Regards

[Godai85]

ok, next time i'll do deep analysis ;)

however, i still have some doubts on "It will apply to all that are within the area you are changing. The "not reencoded" is an indicator that they are already small enough to meet the criteria set. It is also possible that it is a ILVU segment -- which can't be changed without carefully matching it up with alternate paths."

if i click on action->reset to defaults in the editor it unchecks every "no reencode". i did a new compression with all "chapters" reencoded and it works... so why by default (but not "reset to defaults") dvd-rb checks some "no reencode" ?

[Godai85]

no answer ?

[blutach]

If segments are too small or ar ILVUs, it won't re-encode them. Also, after bitrate distribution, some segments are marked as "no re-encode" - i.e. the redist'n says "these segments should be as they were".

Regards

[Godai85]

ok... but why, if i click on action->reset to defaults in the editor and do a new compression with all "chapters" reencoded, does it work ? (i saw the resulting dvd to be sure)

ps: in this case the sizes are similar to other "chapters" and there aren't ilvus (if i undestrand it correctly. no alternative video or multiple angles). so seems that, in this case, is wrong to set no reencode...

[blutach]

Why wouldn't it work? You have just changed the bitrates for the various segments and lengthened the time it takes to encode (the no re-encode segments would have been extracted, now they are encoded at the bitrate in the segment viewer/ECL file).

It is not "wrong" to set no re-encode, nor is it "right". If segments are tiny, it's irrelevant and not worthwhile. If the prog marks a segment as no re-encode and you want it to be compressed (or vice-versa), mark it so in the segment editor.

Regards

[Godai85]

ok, but i said that the sizes are very similar to other segments. so the algorithm that decides if a segment need or not to be compressed is not optimal (imho), cause i think is better to compress the entire feature title at 90% instead of half title at 80%. am i the only person that think this ? :P

it is a bit different if the algorithm decide that in case a particular segment, if compressed by 10%, becomes visibly worse than an other compressed by 20%.
so the question remains: why are some segments not reencoded ? based on whitch algorithm ?

ps: vbr bias applies to every single segment or to the whole titleset (so the bitrate can vary in the range of the entire title) ?
if a segment is compressed in a way, and the other in a different way the result isn't homogeneous o_O'

[Godai85]

up :P
probably only jdobbs can explain that :/

[jdobbs]

I don't want to appear unresponsive -- but I think the answers get thinner because you continue to ask again even though they have been answered (look back through the thread).

DVD-RB doesn't make decisions to reduce at different levels in different segments. BUT... if a value that you provide would bring a segment to a bitrate that is higher than the original -- it is left unchanged instead. You can't make the copy better than the original, so going higher would only waste bitrate while not accomplishing anything. Even if it was bypassed -- the result would be a larger difference between segments (less homogeneous) since it would move the bitrate in an UPWARD direction.

All segments are not created equal and segment bitrate allocations are based upon the complexity of that segment. It is either a result of the original allocation's distribution or the REDISTRIBUTION function. The percentage shows the reduction based upon the preselected bitrate.

As for bias.... I don't have a clue what you are trying to say. The bias is applied to all segments based upon your choice. So what does that have to do with any of this discussion? I'm confused. The bias just decides whether the encode will have higher bitrate swings or move more toward CBR. It has no impact (nor should it) as to whether a segment should or should not be reencoded.

I guess you are entitled to your opinion as to whether it is "optimal" -- but IMHO you don't seem to understand how it works, so it might be difficult to make calls as to how to make it more "optimal".

[Godai85]

yes, it's a bit difficult to me 'cause i don't speak english well...
but:
-you're telling that even if i set 90% in a segmet it could produce an higher bitrate ? i'm confused... i thought that 90% applies even to the average original bitrate (what is the preselected bitrate ? i can't understand it... but i can understand preselected percentage...). i'll do a test: i set a chapter that was preset to noreencode by dvd-rb to 90% and i see if, after the "forced" encode, it has an higher bitrate (but i'm pretty sure that it is impossible...)
-however, is it the only explanation to the "no reencode preset" ? (excluding ilvu and size)
-what i would to know is if the bias applies to the whole titleset (so it "determines" a minimum ad a maximum bitrate for all the vts) or to the every single segment (so exists many minimum and many maximum cause bitrate can vary from minimum to maximum of each segmet). i think that the first is the right.
but this is less important than the other questions (cause in this case i can't do anything. but i would know if is it better to set 80% with "no reencode" preset or to set all segments to 90%. this example assuming that 50% of segments are "no reencoded" so 20%*50%=10%*100%).

[blutach]

-you're telling that even if i set 90% in a segmet it could produce an higher bitrate ?No. 90% means 90% of the original bitrate of the segment.

Regards

[jdobbs]

[Edit] I'm a little frustrated and am respectfully bowing out of this thread.

[Godai85]

ok, don't worry...
but i still think that "if a value that you provide would bring a segment to a bitrate that is higher than the original" is not possible and blutach confirm that...

[jdobbs]

You are, of course, incorrect -- and have underlined the reason I have decided to walk away from this conversation. Since you obviously know more than I do about the application I wrote, I figure you could answer it better than me. If REDISTRIBUTION would have brought the value to something higher than the original -- it would limit it to "NO REENCODE" -- exactly as I said.

Maybe my response wasn't clear... I don't know. But, your lack of understanding doesn't equate to the statement being wrong.

Sigh...

[Godai85]

i didn't want to seem assumptive.
i've said that i don't know english very well, in fact i didn't understand that, previously, you mean "If REDISTRIBUTION would have brought the value to something higher than the original -- it would limit it to "NO REENCODE".
this sentence is simpler for me :P i haven't asked many times "why are some segments not reencoded ?" if this was the first answer (instead of repeating ilvu and already small size) ;)
i thought that redistribution couldn't set a bitrate higher than the original (and i thought that the blutach's answer confirm that).

now i can't, but i will try if, setting for example 99%, size becomes bigger than "no reencode" (in this case bitrate should become higher than the original and so the size, right ?)
if so, obviously, "no reencode" is better than every percentage that makes the segment bigger.

thanx & sorry ;)

[jdobbs]

You might want to post a capture of the segment editor -- because I'm thoroughly confused as to what the issue is... I thought you wondered why:

1. The percentages were different between segments (some were 80% and some were 90%) -- and the answer would be REDISTRIBUTION.

2. Some segments were set to "no reencode" -- and the answer would be either ILVU or REDISTRIBUTION (limiting the up side).

[Godai85]

now i don't have the dvd that has this problem (over the edge). so i can't post images.
however

1: i know and i don't mind cause when i increase the feature percentage i set automatically all segments to that percentage (i can't think any alternative)
2: this is the problem. some segments were set to "no reencode". i want to know if it is better to set the same percentage (bigger then the default cause i decrease extra) to all segments (so all are encoded) or only to the segments that are not set as "no reencode". now i think that the 2nd option is better, cause force encoding to the "no reencode" segments even if at 90% is worse. right ?

edit:
with a different dvd i obtain this example:
http://img168.imageshack.us/img168/3937/exlx7.jpg
in this case only this segment is set to "no reencode" (probably cause the size is very small). however in the dvd that i have the "problem" (over the edge) many segments are set to "no reencode" and all have sizes similar to other segments.
unfortunatly if i force encode and i set 99% the size becomes smaller, like the bitrate... i still don't understand :°(
is this case different from "over the edge" ? if it is not, i cant's understand "If REDISTRIBUTION would have brought the value to something higher than the original -- it would limit it to "NO REENCODE" cause i can set a smaller bitrate/size -_-'

[blutach]

You don't have the DVD? You surely are aware of Rule 6.

Regards

[Godai85]

my parents (divorced) have 2 flats. i can't buy 2 copy of the same dvd ;)

edit: i verified it. also with over the edge (many segments which size is also >200MB are set to no reencode) if i force encode and set 99% the bitrate and the size become smaller...
http://img265.imageshack.us/img265/3673/exfr0.jpg
i can't imagine that is better to compress at 80% some segments and no reencode others instead of encode all segments at 90%...

[jdobbs]

i can't imagine that is better to compress at 80% some segments and no reencode others instead of encode all segments at 90%...As I've said (several times now), that is the result of REDISTRIBUTION. It means the bitrate distribution profile has changed somewhat based upon the new bitrate and in order to keep a constant quality level one segment has to be compressed more (or less) than another.

If you want the entire stream to be compressed at a fixed reduction percentage -- turn REDISTRIBUTION off. But continually arguing, making the same comment, or asking the same question is accomplishing nothing.

i can't imagine that is better to compress at 80% some segments and no reencode others instead of encode all segments at 90%...Try to imagine it. It is a often a fact and is the only reason REDISTRIBUTION was added as an option in DVD-RB. It is particularly useful on poorly encoded VBR or pure CBR sources.

This time I'm truly outta' here.

[Godai85]

ok, this time i take it (cause you talk about quality level and not bitrate [an 80% segment could have same quality level as a no reencoded segment, but not necessarily the same bitrate]. in same way it confirm my supposition "it is a bit different if the algorithm decide that in case a particular segment, if compressed by 10%, becomes visibly worse than an other compressed by 20%" that was ignored).

but... in this case, every time that i use the editor i can make a worse job :(
redistribution is useful ONLY if i don't change the preset percentages, right ?
is there a method to increase percentage maintaining the redistribution ?
even if i set "steal space from extras" i don't have what i want (i should remove some extras and increase feature percentage).


come on, don't abandon me and sorry me again :P

EDIT: yahoo! i think i found the answer myself (don't kill me XD) i have to reduce percentage of extras and only then click on "allocate saved space to feature". now all by doubts have a solution :D
however... this means that i have done bad backups until today d'oh! XD