StatsReader 2.1

[malkion]

I noticed the 1.0 tree of XviD still comes bundled with StatsReader 2.1.
Would there be any advantage of using the stats generated from StatsReader in place of the original .pass file generated by the 1st pass for your 2nd pass encode?

What other purposes could the StatsReader serve for me besides inserting manual key frames in my 2nd pass?

Asking, since 1.0 no longer uses external compression scaling (forgive my ignorace, but am I correct in assuming this?)

[Koepi]

StatsReader is just included for i.e. checking your first pass file size. It can't export xvid-1.0-style stats - and due a harddrive crash i lost my sources for it so i can't implement that.

Regards
Koepi

[Clock]

Well, didn't need this for a long time, but is there any software available out there, that really analyses my ".pass" files?

The Statsreader tells me stuff about single frames, but I'd also like to have global informations, like those shown during encoding...

I like to keep my PC running when I'm not home, and sometimes give him various videos to encode. But, of course, only the last XVid-Debug-Window will stay alive.

So, any hints for me?

Clock

[Aktan]

You can use the java stats viewer aval here:

http://forum.doom9.org/showthread.php?s=&threadid=67639&highlight=java

[Sharro]

Koepi!! :-)

Family and work stole my IRC sessions....

I still encode though :-) (Did a few encodes with 1.1 and run a few SSIM's VHQ for B-FRames looks very promising and damn I still agree with you on curve compression :-))

Still this is not the purpose of my post.

I'm trying to find a decent way of averaging the quant distribution for 3 movies that I want to fit on a dvd...

I was thinking of using merge stats for this but it's a small haslle to encode this way...

Before I summed first pass sizes + audio size for the 3 movies and calculated % of each on total size, then used this % to distribute target filesize.

Doesn't work .... some movies are getting a much better quant dist. than others... Is there a way to get a decent average quant distribution on the 3 movies for a dvd size using your stats reader and merge function ?

All the best,

Sharro

[Koepi]

Heya Sharro,

nice to see you online again! :-)

I fear this isn't possible with stats reader. What you can do though is adopting your avs:
a=mpeg2source("a.d2v")
b=mpeg2source("b.d2v")
c=mpeg2source("c.d2v")
d=a+b+c

d.crop(...)
d.resize(...)
return d

or something like that. Only issue here is that you need to cut the final XviD :-)

That should result in a real averaged quant distribution at least.

Best regards
Koepi

[Sharro]

Thanks for your reply Koepi.

eheh I had already thought of that still I hope this is something that someone looks at in the future it would also be usefull for averaging quant dist on episodic dvd's.

Is there a way to get a coef that relates first pass size + audio size + total frames ?

Cause in my % calculatios nr of frames is left out...

Off topic sorry...

All the best and send my best regards to Syskin.

Sharro