Watch out with DVD-RB-PRO 1.02 !

[apfraats]

Having used DVD-RB-PRO 1.02 with CCE 2.67 and I came across a problem.

This problem , at least when I used 1.02, was causing DVD to be out of specs because of not DVD-compliant bitrates that could occure anywhere in the stream.

The total must be equal or less then 10,080 kbps including video, all present audio tracks and even subtitling as this is all multiplexed in one main stream and that stream must be =< 10,080 kbps.

Because of a probably present bug in 1.02 (I contacted JDOBBS about it, but he has NOT confirmed this !, so I'm at risk here) the bitrate could well go over this max of 10,080 kbps.

This is especially the case when you did a run on a DVD and preserving the DTS track with a 5.1 track.

CCE was given a bitrate of 0.5 to 9000 kbps in VBR for video, despite of any lowering that should occure regarding the release notes.

There should be a 'Dynamic Peak Analysis'to prevent CCE going over the original bitrate anywere and the soundtracks preserved should be substracted.

But for some kind of reason this did not happen with CCE (and maybe other encoders, didn't test it)

It was 'just' set to 0.500 kbps to 9000 Kbps always, even if it was raining...

What happens if you have a DVD with DTS and one 5.1 soundtrack you would both like to preserve ?

Yep: 9000+768+448 could happen or worse when more audio comes in.

9000+768+448= 10216 or even more when more audio comes in.

As you can see 10.216 is a bit more than 10.080 , so 2 of my players had playback problems with a disc were video_bitrate was high (at 9000 even !)

Stutterinmg occured and DTS audio disruption.

Every player will behave differently because this is out of specs and UNDEFINED !!!!!!

So if you have done DVD's with 1.02 PRO and CCE , make sure especially with DTS and 5.1 that they are in specs, or redo them with the released 1.03.


Immediately after installing 1.03 I saw adjusted max_bitrates for CCE, not the 9000 again.

(watch it in the item.elc or the window right under the CCE progress window display).


Again, despite contacting him, he has NOT confirmed this bug.

He hasn't put any release notes too with 1.03 in my case at least.

Just to warn people who used 1.02 and now MAY have out of spec backups from their DVD's.......

If you don't believe me, try 1.02 with CCE and look at the CCE's reported bitrate range, or open item.ecl if you can find it.

After installing 1.03 the problem seems to go away and max_bitrate should never be exceeded even not using DTS and 5.1 audio track.

I used a VBR_BIAS of 0, and a verry complicated source (that had already high bitrates and distortions causing even higher bitrates to come out of CCE !!)

So be warned and in case of a hit concerning this , redo the backup with 1.03.

It's also a reason to immediate install 1.03 and not wait !

[blutach]

The source (if it was compliant) could not have been out of spec, so I wonder how the re-encoded version could be.

Let's not start bashing jdobbs here please - you have 3 days to serve of your ban at doom9; let's do it in peace here please.

Regards

[ElBoricua433]

just to let you know, a new version came out of DVD Rebuilder

[cakebox50]

The source (if it was compliant) could not have been out of spec, so I wonder how the re-encoded version could be.

Let's not start bashing jdobbs here please - you have 3 days to serve of your ban at doom9; let's do it in peace here please.

Regards

output bitrate > input bitrate can happen.
Read JDOBBS release notes on 1.02
He introduced a mechanism called 'Dynamic Peak Analysis' to prevent this.
But what apfraats has seen, i have seen too.Using 1.02 and CCE gives
standard range 500-9000 for each cell/segment.So reaching 9000 is possible,
and adding DTS and 5.1 to this, apfraats has right.I'm glad I do not have to
review my backups, becuase I only use 1* 5.1 and a 2 chnnel stream, so not
breaking any compliancy.I always remove the DTS track for optimum video
results.

[blutach]

I have made backups with both DTS and DD 5.1 (Passion of the Christ is one I can recall).

I have 2 of the most finicky players in the world as far as DVD specs is concerned - SONY and NAD - and neither baulked on this title.

But who's to know? You could be right. You say 1.03 does not exhibit this behaviour, so all is well now in any event.

Regards

[cakebox50]

It CAN happen with 1.02 , not it WILL happen.

You have luck, not pushing up CCE the the specified limits, that's all.

And indeed with 1.03 all is correct again, but dvd's done with 1.02 will not recover automatically by this event

[blutach]

Well, I don't use CCE as I consider it greatly inferior to Procoder 2 - maybe I am lucky

Regards

[apfraats]

Well CCE is my personal favorite so far, but i'm always seeking the truth, and open for ANY improvement.....

So it would be nice if you could give me some why's on this.

Because DVD-RB-PRO is my tool to use and JDOBBS still hasn't left us,

(referring to the sunny days of DVDSHRINK when the author decided to spend his time in a more money making way.... good luck to him for the work done of course but sorry for the tool, ehich stopped, same with DVD-DECRYPTER, also stopped of threatening legal action that we in the Netherlands even laugh at, as long as it lasts, we may freely use whatever software we like and even develop what we want so far, because freedom of acting here is still valuable, so even DVD's may be legally copied for strict private use only by the person himself and even not having to own it, lucky we are not already like the US.)


I like to hear some more arguments.....

Nope I'm not a CCE freak, but a while ago it was considered best quality for most money , so if procoder does the job better i would greatly be happy.


Especially if I can get rid off the annoying blockiness that (maybe it's me) CCE tends to have at lower bitrates IMO.

How manny passes can you do with it ?

Is it better preventing the result from blocking at same (lower) bitrates ?

Does it have less GIBBS artificacts ( noise around sharp contrasted areas such as bright white letters on a darker gray background), using lower bitrates as VBR is implying on such scenes (steady bright white text on a darker background) ??

What are the options for best quality ?
Is it going to cost me $495,- or something like that, even for trying ?

Please be a bit more specific about the WHY not the THAT



oh yeah, and how is it speeds relative to CCE 3 pass (1 VAF 2 M2V) ?

Thanx.

[UncasMS]

- procoder is WAY better at lower bitrates than cce and thus artefacts will be less

- procoder always does 2 passes

- there is a procoder express version available as trial in the *let's edit* trial iirc

- its speed is slower than cce or hc but i consider it worth the time

[ElBoricua433]

well, if you don't have the money also, you can also try HC Encoder, it also does a good job with those lower bitrate scenes

[apfraats]

Ok so you say it's better, well I'm a person that's used to have an attitude like: First let's see it, then believe it.


I already used HC, but this is still in beta and I experienced a completely strange problem with it doing Lord of the rings.

Forward step, fast and normal playback went fine on my very strict Sony.

But when I played resversed the LOTR backup sample step by step, it suddenly gives disorted frames and even switches the player completely off.

I contacted Sony and they said this was impossible., well I asked them to come an have a cup of coffee

But whatever I do, using other discs everytime at exactly the same spot, it just switches off, only when playing reverse step by step.

So I just don't trust too much the HC encoder at this moment, even is nobody seemed to have this problem.

It's weird because other movies went fine, and I did not have a problem playing them in step reverse..........

So it's very very strange.

Then HANK315 released a new version and explaned the Motion Vectors were sometimes pointing to 1/2 a dot outside the picture frame...... Maybe that was the problem. But what will be next ??? Still a Beta.

Assuming this should be solved now, but HC didn't perform much better then CCE when using adapted encoder matrices. I used high motion, high VBR scenes that even had blocking at some frames in the original !!! (don't worry, it happens more often on originals then one might expect, often only visible at slow/still playback.

So my trial is done by forward and backward skipping at slow speed through te scene, and the Sony makes very little steps and shows almost near each other frames..... Then when using this method every little blocking that even doesn't seem to be there at normal playback speed shows up.

I used Lord of the Rings: The Two Towers, where there is a high motion high action scene in the beginning with the dragon-like monster and the wizard.

Even here again the scenes on the original showed some blocking already when this particular scene starts....

But when the wizard lowers its staff and there is a lot of bright light suddenly comming of it, there is a part that will be great to test encoders and different matrices.

The original can just keep things straight here, and when compressing to <70% the CCE encoder with the right matrix can do it almost perfect too.

For this I use a testclip and a low CCE_TRAGETSECTORS and HC_TARGETSECTORS setting so I can do quick comparissions.

Didn't see much difference there concerning HC and CCE, and HC is still beta and CCE is 'proven technology', so the choice is easily made.

CCE with default matrix will fail however....

but HC with the same matrix does too.

It's very difficult to differ between them.

But for a freeware encoder HC is very promising for what Ive seen of it.

So now I can try procoder. But I do'n't know if the test-version gives the right results.

UncasMS said:

- procoder is WAY better at lower bitrates than cce and thus artefacts will be less

- procoder always does 2 passes

- there is a procoder express version available as trial in the *let's edit* trial iirc

- its speed is slower than cce or hc but i consider it worth the time

So it defenitely worth a try.

I however think the procoder express version is NOT having the same quality as procoder 2. And again I have to do my own testing and comparing.

So i go and see if I can get my hands on it otherwise in a legal matter or letting somebody do the testing which has this encoder and will run the testsample for me, as I can send it using www.yousendit.com with the XXX_TARGETSECTORS setting and have me the result mailed back.

So if anyone has time for it, feel free to contact me using PM.

I'll send the testclip and the result can be returned they same way.

I would grealy appreciate it. The testclip is handled pretty fast as it's really not so big. I used DVDREMAKE-PRO to hide the most blocks (=CELLS) on the movie, so it's running in reasonable fast time.

The speed of Procoder doen't really is of any importance as long as it's not unreasonable 10+ times longer then CCE 3 pass.

I have more not superfast but fast enough computers that I keep working all day/night and I use DVD-RB-PRO with low priority settings so it runs in LOW-PRIORITY, keeping the machines available for doing preparation work and such.

(I think DVD-RB-PRO will be responsible for my CPU-wearing.... )

the movies goes in batch on each system at 100% CPU for hours and hours on and on. Even with filters used and not used to see what's better.

(My electricity bill will kill me next year )

So if anyone has PROCODER 2 BEST VERSION , and would try my testclip and send the result back, I would be very pleased.

Now I have to force CCE using abnormal alla 'apfraats' matrices with almost all high values in all cells except for them around the 0,0 coordinate. This seems to filter out precision, but macroblock function will have more bitrate thus preventing too early blocking.

So anyone offering me testclip run with procoder ?????

please PM me.

[apfraats]

- procoder is WAY better at lower bitrates than cce and thus artefacts will be less

- procoder always does 2 passes

- there is a procoder express version available as trial in the *let's edit* trial iirc

- its speed is slower than cce or hc but i consider it worth the time

Huuuuuh as I don't know my way around there, can you provide a link ?

thanx

[UncasMS]

havin read your rather lengthy reply two questions come to mind:

1. you say you dont give much on other people's finding/preferences
...fine with me

you say you want to test and see for yourself

...even better

BUT why do you ask in the first place and why give a reply THAT long


2. you doubt that a procoder EXPRESS version will provide you with useful results compared to procoder2

why is this?
how come you know beforehand?

and most importantly: would you say that the engine for cce BASIC works less good than the one used in the SP version?


+++++++++

and btw: NO i dont have a link to the *let's edit* trial

maybe this can help:

http://register.canopus.com/US/support/tech_support/techsupport_demosoft.asp

[apfraats]

You turn my remarks the other way around, as happens more often.

I didn't say I didn't care about other persons opinions, but I have seen so many discussion about what should be the best, that for sure you have different camps.

I always ask why a particular encoder is better and if it's shared by more people it's just common sense to TRY and SEE for yourself.

If anyone would say DVDSHRINK is the best in quality, I even won't bother at all.

Always try it and test it yourself, it's partly subjective findings from people to decide what they think is best.

So if I hear about HC and more people are calling it 'the best', I'll just try and see.

If people do this for Procoder I want to know why.

If the answer has any ascepcts that seems important to me, such as less blockiness compared to X, I just gonnan try and see.

If I would have to believe everything that everybody is calling the best, I probably will be in doubt untill 2050 or so.....

So first I ask for more specific motivation about WHY something is regarded better. And you gave intersting response, and that's why I wanna test it.

So I DO care what other people say, but it's simply a contradiction to asssume anybody is right because all encoders would be the 'best'.

So I ask, evaluate, and going to test and see for myself.

I think this is a procedure everybody is using.

See nothing wrong with that. So after the Procoder test I will be able to agree or disagree with you. now I can't say anything about it at all.

That's why people test and see for themselves as there also is a subjective aspect included here.

Id you stated it will be generating lees artifacts, inclusing blockiness, it's sure a reason for me to try. Testing is knowing. That's all.

Yep, CCE BASIC is worse then CCE SP, not because of the engine that probably is much the same one, but because SP can do multiple pass VBR and CCE basic is resticted to 1 VAF and one VBR M2V pass.

So thanks to you I noticed Procoder as worth giving it a try.

I cannot state anything on it, before i've done evaluation myself.

Looks just basic logic to me.

Thanks for the link by the way, I will jump into it.

Also I suspect PROCODER EXPRESS to be giving less Quality as PROCODER 2 , or what should else be the reason to have these two versions in market at the same time, but I could be wrong.

Probably I would compare Procoder Express with CCE basic and Procoder 2 with CCE SP.

As I wanted BEST quality I start with Procoder 2, cause Procoder Express being better then Procoder 2 would be less likely....

so to prevent loss of time I start with the best.

Maybe finally going to use EXPRESS becasue of another aspect: Money