[quote="Dialhot"] I know we don't use this in mencoder but it is possible to encode quality based.

and Prodater's Idea is oke If you encode really slow and wan't to see what your movie's like it could be usefull.
The bat file want's to open mencoder so many times that my computer is giving me alot of errors you said to use as many samples as minutes in the movie well I shut it down after 50 :P . And I had 50 little m2v files 0kb large was your batfile going to join them?


Edit: the quote went wrong

You quoted all the thing except the part you really need to read : a previewer based on a sample is USELESS for an encoder that is not quality based. The result of memcoder is average based

To taste a soup with a spoon is relevent as all the soup have the same taste so you can just take a sample and extrapolate the taste of all the soup.

the preparation of a soup is "quality based"

To taste a cherry and extrapolate on the taste of all cherries carried by the tree is irrelevant because fruits didn't received the same amount of sun whole other the branches.

the growth of cherries is "average based"

Mencoder is not a soup cooker, it's a cherries tree.

THAT is why your approach in making this script is useless. The idea was interresting (and the realisation clever), but asking some advice before to work on it would have prevent you for what I called "lost of time".

@rds
If you like to do things with no goal, do as you want...

@koekies
I think there is a post from Inc just above that explains why the part of the manual of quoted isn't usable.
(if it was, we would have did it since a long time insteed of loosing time in 2pass encoding ).
That is typically the kind of sentence inspired by guides posted with fair knowledge of the problem I wanted to avoid.
NO the tool of prodater CAN'T give you any idea on how will be your movie. And I explain all above.

Cherries and soup...

Edit :
Of course I didn't mentioned this line but I won't forget it...

Lol well I am a newb after all, I dont fully understand Mencoder fully im afraid.

__________________
Regards.

Michael.

NEvertheless my answer that you quoted was out of the subject so you can read the remainder of the thread to increase a little your knowledge about mencoder

In first pass you will obtain many log files maybe of 0kb because is so little the info for these log files. In second pass you will obtain many m2v chunks. When Mencoder finished, batch file will join all these chunks in one file and will delete all chunks. (Still has a bug, chunks are note joined in order, I am working over that)


--------------------------
Visit: Intermediate guide: MencodeMe/Win32 - Avisynth - MakeAvis by Prodater64.
Visit: KVCD - MencodeMe - Auxiliar Task - KVCD Docking Gate by Prodater64.
Visit: Mencoder scripting with AVSEdit and Guide for Multiple Files by Prodater64.
Visit: Mini-guide quick and easy - DVD to (S)KVCD with MencodeMe by Maurus.
Visit: Mencode-me: a newbie oriented GUI - 0.23 is out! by VMesquita.
--------------------------

Thanks I am learning a lot about Mencoder all the time and this thread has some very interesting points of view and does indeed increase my knowledge of this encoder.

So thanks guys.

__________________
Regards.

Michael.

But the samples won't really give what it would look like. The bitrate might be the same.
But in the second pass mencoder spreads it really different than mencoder would do with small samples. Wouldn't it? If you encode a small part of the movie with the same bitrate as you would encode your full movie bitrate viewer would have really different bitrates at those parts.

-Koekies

You perfectly understood what I wanted to say.

Talking now "neutral" and with --no offense-- at all ... :

The problem at KVCD in here actually is that some developing users are releasing guides on a tool which is not actually enough known in its habbit related to some output formats.
A guide is MEGA usefull for people which want to "get into the materia mencoder" .... and thats cool! And we all do apreciate that peoples try to bring such interesting (even if not in final state) Tools more near to other people who want testing it.

Theres only one "little" thing which results beside that.
"Guides" generally at KVCD stand for a way to get successful encodings (noobs/new visitors do see that this way ).
And so many new users take guides as "a way to get that phenomenal compilant KVCD "thing" working" ... so they dont see its "real" purpose even if guide athours do mention it thats just a "workout" to see what that encoder is able to do.
So visitors try and ... of course in such a not "final" state like in the case of mencoder the chance of 70/30% is given that they fail!
This will produce only one opinion finally: "KVCD doesnt work and isnt compilant" ... means a simple failure on one encoder which is known "by us" will give a bad image to KVCD seen in a whole. Because those visitors or mega newbies do judge very simply and will give that opinion to other places.

Well what I did say now is "just" a way how it "could" end up .... as many visitors/Noobs do directly enter the "peak" of our developings/testings.

So these words are my little quiet "thinking" and are not an offense to authors of mencoder guides!


I personally think whe in here have many straight discussions and till now every diskussion got "even" a constructive final end. Thats the way in here!

But such personal adressed "senseless" attacks don't enhance friendly atmospheres in here, neither they are constructive ---- just a hint!

And as you are a mod in here I personally would apreciate if you could switch back to "moderate" replies again as the past showed us that this is your normal way.




Spoken generally:

I do quote Jorel:
"Friendship is all"!
Maybe I do quote it to often as a final conclusion for me, ... but everything is contained in these few words which is needed.

Best greets
Inc.

But if you allready had the divx2pass.log and would use that for the samples and al the samples would be
made with vpass=2 would the bitrate be spread more like it would in the real movie.
But that doesn't really matter because then it would take really long to make a sample
and that wouldn't be like the whole idea of quickly making a sample (which was a great idea prodater ) originally was.

Each pass is outputted in a MovinameNumber.log file, so info of each chunk is present for second pass.

--------------------------
Visit: Intermediate guide: MencodeMe/Win32 - Avisynth - MakeAvis by Prodater64.
Visit: KVCD - MencodeMe - Auxiliar Task - KVCD Docking Gate by Prodater64.
Visit: Mencoder scripting with AVSEdit and Guide for Multiple Files by Prodater64.
Visit: Mini-guide quick and easy - DVD to (S)KVCD with MencodeMe by Maurus.
Visit: Mencode-me: a newbie oriented GUI - 0.23 is out! by VMesquita.
--------------------------

yes but that's a totally different log than you get from the first pass of the entire movie.

you could test it by making a log by scanning the entire movie. And joining the logs in the right sequence from all your samples in the right order and compare.
But that's alot of work and I'm very shure they'll be different

Well, I don't knew that before. I need to see results for myself (and I wiil do) but, in any case it is a nice toy. I like to see that slices.


--------------------------
Visit: Intermediate guide: MencodeMe/Win32 - Avisynth - MakeAvis by Prodater64.
Visit: KVCD - MencodeMe - Auxiliar Task - KVCD Docking Gate by Prodater64.
Visit: Mencoder scripting with AVSEdit and Guide for Multiple Files by Prodater64.
Visit: Mini-guide quick and easy - DVD to (S)KVCD with MencodeMe by Maurus.
Visit: Mencode-me: a newbie oriented GUI - 0.23 is out! by VMesquita.
--------------------------

you could test it with quality based encoding and mencoder has anyone ever tried that before? those results should be about the same.
Like I said it was a good idea. But with my mencoder build I get amazing speeds 80fps it doesn't take so long to encode the movie for me. But I can only encode in 720x480 :P

-Koekies

This is a 50 fps limiting area here. Didn't you see the signs ?

Updated 07/JUN/2004: Fixed bad order of chunks when rebuilding m2v final stream.

Look at first post.


--------------------------
Visit: Intermediate guide: MencodeMe/Win32 - Avisynth - MakeAvis by Prodater64.
Visit: KVCD - MencodeMe - Auxiliar Task - KVCD Docking Gate by Prodater64.
Visit: Mencoder scripting with AVSEdit and Guide for Multiple Files by Prodater64.
Visit: Mini-guide quick and easy - DVD to (S)KVCD with MencodeMe by Maurus.
Visit: Mencode-me: a newbie oriented GUI - 0.23 is out! by VMesquita.
--------------------------

Hi everyone,

Quality based encodes are not possible with mencoder by just varying vqscale. This can produce terrible bitrate peaks, plus the difference between vqscale=2 and vqscale=3 is enormous and vqscale=4 already looks bad. It's not impossible to do quality based: one can tweak rate control equation and vratetool so mencoder doesn't care about average, and then vary scplx_mask and lmin (and maybe other parameters) to get different "quality". But I don't think it makes sense to go ahead with it. Mencoder gives good quality with 2 pass VBR and it's very fast. But if anyone has interest in this, check the old mencoder thread, the one with more than 20 pages.

On the other hand, I can see where this tool can be really useful. One of the reasons I like to use OnePass Quality based with CCE (besides speed), is that I can somehow garantee that the quality will be good. If I manage to fit my video matherial in CCE using Q30, quality will be good for me. Q30 may mean 1000 kbps or 4000 kbps, depending on the source matherial, because compressibility varies a lot. This was not possible with mencoder, till this Prodater64 idea. Now I can encode a sample at vqscale=3, calculate the average and see how far it is from the average I'll use, so I can know the quality of the final product. So let's say that to encode some movie and fit in on CD I have to use 1200 kbps. I encode a sample at vqscale=3 using Prodater sample, and it comes out as 3000 kbps. So I know quality will be bad, and I'll go for 2 CDs or lower resolution. Or maybe it comes out as 900 kbps. So I know I can use a higher resolution.This can save precious time!

And what do you do if you sample don't go other 1500 but you have birate peaks above 10.000 in the whole movie ? Hugh ... ?

Note: which tool do you use to calculate the average ? A simple calc or something else ?

What I meant is that the samples are small parts of the movie and your tool aply's the average bitrate in the whole movie
to a small part of the movie but in the fully encoded 2 pass movie the bitrate of your small part isn't the same as the
average bitrate of that part in the fully encoded 2 pass movie.
So there fore you wouldn't really see how the bitrate would be spread over the movie after a 2 pass encode and what
the quality would be. Do you get my point now?