Quantcast FFmpeg vs FFvfw vs Mencoder ? - Page 5 - digitalFAQ.com Forums [Archives]
  #81  
02-23-2004, 02:31 PM
Amenophis Amenophis is offline
Free Member
 
Join Date: Feb 2004
Posts: 14
Thanks: 0
Thanked 0 Times in 0 Posts
Quote:
Originally Posted by digitall.doc

Well, I already used all this settings, with lots of macroblocks during the explosion, that didn't improve raising bitrate from 5000 to 8000. I didn't try adding noise, but I still don't understand why the encoder doesn't make use of higher bitrate avaliable to improve the image and avoid this macroblocks...
Maybe you've to lower the min.quanziter (=1). some movies can be compressed completery at quant2 with about 4000 kbit/s, so you need to lower the quant if you want to use more bitrate
Reply With Quote
Someday, 12:01 PM
admin's Avatar
Site Staff / Ad Manager
 
Join Date: Dec 2002
Posts: 42
Thanks: ∞
Thanked 42 Times in 42 Posts
  #82  
02-23-2004, 02:55 PM
bilu bilu is offline
Free Member
 
Join Date: Jan 2004
Posts: 341
Thanks: 0
Thanked 0 Times in 0 Posts
@Amenophis

quantizer=1 should be avoided. See this post:

http://www.kvcd.net/forum/viewtopic.php?p=63511#63511


Bilu
Reply With Quote
  #83  
02-23-2004, 02:55 PM
digitall.doc digitall.doc is offline
Free Member
 
Join Date: Jul 2003
Location: Valencia (España)
Posts: 741
Thanks: 0
Thanked 0 Times in 0 Posts
Hi amenophis:
I will try min quant=1, but I think I read in man-page he didn't advise min quan=1, but don't remember why.
About your statement:
Quote:
some movies can be compressed completery at quant2 with about 4000 kbit/s
when I encoded it again with vbitrate=8000 and got the same filesize, I thought that the encoder already had compressed completely the film and couldn't improve it even raising bitrate. But, how can I take profit of higher bitrate, if any, to avoid the macroblocks in fast scenes? I'll try ypur quant=1 suggestion. Maybe adding noise?. Any suggestion anout adding noise in mencoder?
Where did you take the parameter value for vi_ vb_ and so? Are the applyable to any film?
EDIT: phew, bilu, I was slower for just a second, friend
Reply With Quote
  #84  
02-23-2004, 03:02 PM
bilu bilu is offline
Free Member
 
Join Date: Jan 2004
Posts: 341
Thanks: 0
Thanked 0 Times in 0 Posts
Quote:
Originally Posted by digitall.doc
Well, I already used all this settings, with lots of macroblocks during the explosion, that didn't improve raising bitrate from 5000 to 8000. I didn't try adding noise, but I still don't understand why the encoder doesn't make use of higher bitrate avaliable to improve the image and avoid this macroblocks...
I didn't try SATD instead of default (?) SAD to see if this improves quality (avoids these macroblocks), but I didn't want to make encoding even slower...
Did you look at the source? Maybe the blocks are already there...
Besides noise or sharpening I have very few ideas left

Quote:
From a theoretical point of view... what values, thinking in quality terms, would you advise for b_qfactor, vi_qfactor, vb_qoffset, and vi_qoffset
Until I try mencoder myself, and based on my Xvid experience, I think the defaults are fine. Others may test and present a better parameters than me - experience DOES count here

Quote:
And what about a command line for 2pass encoding
Same goes here
Have a look my previous posts about the average bitrate for the 2nd pass.

EDIT: can you post your command-line? Are you using B-frames?

Bilu
Reply With Quote
  #85  
02-23-2004, 03:09 PM
Amenophis Amenophis is offline
Free Member
 
Join Date: Feb 2004
Posts: 14
Thanks: 0
Thanked 0 Times in 0 Posts
additionally you could use a matrix with low values, so you get more bitrate/quality without changing the quantizer
Reply With Quote
  #86  
02-23-2004, 03:14 PM
bilu bilu is offline
Free Member
 
Join Date: Jan 2004
Posts: 341
Thanks: 0
Thanked 0 Times in 0 Posts
Quote:
Originally Posted by Amenophis
additionally you could use a matrix with low values, so you get more bitrate/quality without changing the quantizer
Others can help you here, I'm curious too

Bilu
Reply With Quote
  #87  
02-23-2004, 05:14 PM
digitall.doc digitall.doc is offline
Free Member
 
Join Date: Jul 2003
Location: Valencia (España)
Posts: 741
Thanks: 0
Thanked 0 Times in 0 Posts
Here's the last mencoder settings I've tried
Code:
mencoder -of mpeg -ovc lavc -nosound -lavcopts vcodec=mpeg2video:intra_matrix=8,9,12,22,26,27,29,34,9,10,14,26,27,29,34,37,12,14,18,27,29,34,37,38,22,26,27,31,36,37,38,
40,26,27,29,36,39,38,40,48,27,29,34,37,38,40,48,58,29,34,37,38,40,48,58,69,34,37,38,40,48,58,69,79:inter_matrix=16,18,20,22,
24,26,28,30,18,20,22,24,26,28,30,32,20,22,24,26,28,30,32,34,22,24,26,30,32,32,34,36,24,26,28,32,34,34,36,38,26,28,30,32,34,36,
38,40,28,30,32,34,36,38,42,42,30,32,34,36,38,40,42,44:
mbd=2:vrc_minrate=300:vrc_maxrate=8000:vbitrate=8000:vrc_buf_size=1835:vmax_b_frames=2:naq:trell:cbp:scplx_mask=0.5:
mv0:vqsquish=0:vqblur=0:vqcomp=1:vrc_eq=tex:vqmax=4:keyint=15:aspect=16/9 D:\Temp\1.avi -o D:\Temp\encoded.mpg
I still have to get rid off vqcomp, vqmax, naq and scplx_mask as you advised. As you can see I'm making use of notch matrix.
And no, macroblocks aren't in the source (Star Wars Episode II DVD).
About 2-pass, I don't really ask for a working command, we'll have to test. I ask, for instance, how do I state which functions just working in first pass and which in second, and so on.
Quote:
Until I try mencoder myself,...
Go and try it, man.
Reply With Quote
  #88  
02-23-2004, 06:35 PM
rds_correia rds_correia is offline
Free Member
 
Join Date: Apr 2003
Location: Chinese Democracy starts now!
Posts: 2,563
Thanks: 1
Thanked 0 Times in 0 Posts
Quote:
Originally Posted by bilu
Until I try mencoder myself,...
Quote:
Originally Posted by digitall.doc
Go and try it, man.
Now I think digitall.doc has it all: you really have to try this baby, bilu
__________________
Rui
Reply With Quote
  #89  
02-24-2004, 05:16 AM
digitall.doc digitall.doc is offline
Free Member
 
Join Date: Jul 2003
Location: Valencia (España)
Posts: 741
Thanks: 0
Thanked 0 Times in 0 Posts
Hi Rui:
have you played with 2pass encoding?
I'm curious about how to set mencoder to do that.
I'd try it and see if is so slow to forget it.
Well, we played a lot with ffvfw, and now playing with mencoder. And it seems that with mencoder we managed
to avoid main ffvfw drawback: too high bitrates. Don't you think so?.
It seemes to me that we limited bitrates in mencoder.
BUT, maybe still far from getting the best of it, because in my fast action scene test I still get too much blocks (seen on PC monitor).
Did you see the last settings I posted in this thread, following bilu advise?. What do you think about them?. What can we improve in
order to get a better output?.
He, he, enjoying a lot with this...
Reply With Quote
  #90  
02-24-2004, 05:52 AM
bilu bilu is offline
Free Member
 
Join Date: Jan 2004
Posts: 341
Thanks: 0
Thanked 0 Times in 0 Posts
Code:
mencoder -of mpeg -ovc lavc -nosound -lavcopts vcodec=mpeg2video:intra_matrix=8,9,12,22,26,27,29,34,9,10,14,26,27,29,34,37,12,14,18,27,29,34,37,38,22,26,27,31,36,37,38,
40,26,27,29,36,39,38,40,48,27,29,34,37,38,40,48,58,29,34,37,38,40,48,58,69,34,37,38,40,48,58,69,79:inter_matrix=16,18,20,22,
24,26,28,30,18,20,22,24,26,28,30,32,20,22,24,26,28,30,32,34,22,24,26,30,32,32,34,36,24,26,28,32,34,34,36,38,26,28,30,32,34,36,
38,40,28,30,32,34,36,38,42,42,30,32,34,36,38,40,42,44:
mbd=2:vrc_minrate=300:vrc_maxrate=8000:vbitrate=8000:vrc_buf_size=1835:vmax_b_frames=2:naq:trell:cbp:scplx_mask=0.5:
mv0:vqsquish=0:vqblur=0:vqcomp=1:vrc_eq=tex:vqmax=4:keyint=15:aspect=16/9 D:\Temp\1.avi -o D:\Temp\encoded.mpg
Remove vmax_b_frames=2, default will be B-frames disabled.
B-frames are not that good on high action, or else there wouldn't be a switch like this for the 2nd pass:
Quote:
vb_strategy=<0-1>
strategy to choose between I/P/B frames (pass 2):
0 always use the maximum number of B frames (default)
1 avoid B frames in high motion scenes (will cause bitrate misprediction)
Remove vqquish=0, default is vqsquish=1 and seems better.
Remove vqcomp=1, not needed when using vrc_eq=tex.
Remove vqmax=4, this may limit mencoder's capability to keep in the 8000 Kbps range.

If not enough, then try:

Remove scplx_mask=0.5 and naq, spatial deblocking without temporal influence may end up looking bad on high action scenes too, who knows?
Use default matrix instead of the Notch matrix.


Bilu
Reply With Quote
  #91  
02-24-2004, 06:00 AM
bilu bilu is offline
Free Member
 
Join Date: Jan 2004
Posts: 341
Thanks: 0
Thanked 0 Times in 0 Posts
Quote:
Originally Posted by rds_correia
Quote:
Originally Posted by bilu
Until I try mencoder myself,...
Quote:
Originally Posted by digitall.doc
Go and try it, man.
Now I think digitall.doc has it all: you really have to try this baby, bilu
OK, let me tell you a story

I have two machines at home, one is a Pentium 200 w/ 64 Mb RAM and 6 GB, the other *was* a PIII-850 with 128 MB RAM and 20 GB.

I said *was* because the board is dead: it lived one extra month with a borrowed stronger power source, but it's dead again. Some condensers/resistors/whatever are leaking.

I already checked out yesterday on the FIC boards site that the FIC PT-2007 from my P200 will handle the 20 GB disk from the other machine.

But that disk was from a machine running XP and NTFS 5.1 and has documents, I don't wanna risk mounting it on my P200 VectorLinux 4.0 .

I don't have finantial conditions for an upgrade at this precise moment, or else I would have already backed up that disk and shoved it into my P200 - this will be my DVD backup machine !

Until then my 6GB disk isn't big enough for that


Bilu
Reply With Quote
  #92  
02-24-2004, 07:12 AM
rds_correia rds_correia is offline
Free Member
 
Join Date: Apr 2003
Location: Chinese Democracy starts now!
Posts: 2,563
Thanks: 1
Thanked 0 Times in 0 Posts
Quote:
Originally Posted by digitall.doc
Hi Rui:
have you played with 2pass encoding?
I'm curious about how to set mencoder to do that.
I'd try it and see if is so slow to forget it.
Hi Digitall.doc,
No I haven't tried 2-pass because I never liked the idea of having to wait almost twice the time. Don't get me wrong, I aim to find the best quality but I'm not willing to wait one week for a movie encode.
Quote:
Originally Posted by digitall.doc
Well, we played a lot with ffvfw, and now playing with mencoder. And it seems that with mencoder we managed
to avoid main ffvfw drawback: too high bitrates. Don't you think so?.
It seemes to me that we limited bitrates in mencoder.
By now it does seam so but I'm not quite sure. What about those "buffer underflows" and those "PTS SCR"? Maybe they appear when we try to constrain the bitrate, and I think I could tell you that "buffer underflows" are bad when you get a lot of them on very dark low action movie parts.
Quote:
Originally Posted by digitall.doc
BUT, maybe still far from getting the best of it, because in my fast action scene test I still get too much blocks (seen on PC monitor).
That's what I'm saying
Quote:
Originally Posted by digitall.doc
Did you see the last settings I posted in this thread, following bilu advise?. What do you think about them?. What can we improve in
order to get a better output?.
Maybe I'm not cut off for understanding mencoder explanations.
So I'm maybe not the right person to give that kind of advise.
I will test your command line though.
It seems that Bilu can explain it perfectly even without having tested the encoder and still I don't get it.
It's as if you were all reading more/different documentation than the one I have, which I'm sure you're not.
I'm completely missing something.
But this is my 1st time with you guys trying to understand an encoder.
Maybe next time I don't pose so many questions.
Quote:
Originally Posted by digitall.doc
He, he, enjoying a lot with this...
I've been my friend but not so much in the last days since sometimes I don't understand a word that you're all saying...
But I can't wait for some guy to unveil the best command line for mencoder, yet.
Cheers
__________________
Rui
Reply With Quote
  #93  
02-24-2004, 08:08 AM
digitall.doc digitall.doc is offline
Free Member
 
Join Date: Jul 2003
Location: Valencia (España)
Posts: 741
Thanks: 0
Thanked 0 Times in 0 Posts
Hi all,
bilu, yesterday night I tried all your suggestions, that patiently you had already suggested earlier . Know what?,
to my susprise it didn't change anything (filesize neither visual quality) .
I'll explain: I changed to vqsquish=1 and removed vqcomp=1. Changed to vrc_eq=tex. Removed vqmax=4
(and let default). And believe me, nothing changed.
Tried removing scplx_mask=0.5 and naq, and again to my susprise filesize raised (well, not really surprising
since you stated that scplx_mask filters image, but it was supposed to decrease image quality,... and bigger
filesize??)
Then I remembered that nocht matrix is supposed to cut frequencies and lower bitrate, so I tried using default
matrix instead of the Notch matrix. File size grew, but no image improvement.
My interest in higher bitrate and bigger filesize is getting better quality (thinking on KDVD). Don't mind lower
bitrate, but I want to get the better quality possible with this encoder, and if filesize is too big, I'll filter, apply Notch
matrix, or whatever needed to fit the film in media. But I'm not getting to this point. Image quality is quite good, but
still didn't get rid off these blocks. I tried adding noise (-vf noise=lumaa50) but still (less) blocks.
Any advise?
Reply With Quote
  #94  
02-24-2004, 08:57 AM
bilu bilu is offline
Free Member
 
Join Date: Jan 2004
Posts: 341
Thanks: 0
Thanked 0 Times in 0 Posts
scplx_mask=0.5 works like a deblocker, so it's normal that by disabling it the movie filesize grows.

I think we need to focus more on that scene
Please give some feedback about what Bitrate Viewer reports.
Maybe you need to rise bitrate up to the 9800 limit.

Or change to SATD (slower)

Bilu
Reply With Quote
  #95  
02-24-2004, 09:00 AM
bilu bilu is offline
Free Member
 
Join Date: Jan 2004
Posts: 341
Thanks: 0
Thanked 0 Times in 0 Posts
@rds_correia

I'd like to know specifically what you're not understanding.
If you understand you can help me explain to others

Please post your questions, I'm sure someone else will have the same doubts.

About FFVFW, maybe some of the misconceptions made before with mencoder (like the abusive use of vqscale which doesn't respect bitrates) are the same being done in FFVFW.

Could someone post what's being used in FFVFW now?


Bilu
Reply With Quote
  #96  
02-24-2004, 10:39 AM
digitall.doc digitall.doc is offline
Free Member
 
Join Date: Jul 2003
Location: Valencia (España)
Posts: 741
Thanks: 0
Thanked 0 Times in 0 Posts
@bilu,
hmmm, if I encode a film without b-frames, won't the file grow really a lot?. And, is it DVD compliant?.
I mean, GOP for DVD must be 15, isn't it?. And GOP structure for DVD compliancy must be
IBBPBBPBBPBBPBBI... isn't it?
I'm not just looking for a solution for my scene, but learning the way to do our encodings, to get the better
quality possible, but always compliant with DVD standard.
I may try to raise up 9800, but as I told you I raised from 5000 to 8000, without changing filesize, neither
quality. Maybe was due to Notch matrix.
I'll try SATD, and post here the "slowering" magnitude, and filesize and quality.
What about 2-pass?. A fast first pass, and second pass for quality?. Any ideas?. Will it really take double
encoding time?. I would like to test it, but I don't know how.
Reply With Quote
  #97  
02-24-2004, 11:28 AM
kwag kwag is offline
Free Member
 
Join Date: Apr 2002
Location: Puerto Rico, USA
Posts: 13,537
Thanks: 0
Thanked 0 Times in 0 Posts
Quote:
Originally Posted by bilu
Remove vmax_b_frames=2, default will be B-frames disabled.
B-frames are not that good on high action, or else there wouldn't be a switch like this for the 2nd pass:

Removing B frames will make worse quality, no matter what encoder
That is, if bitrate is kept at the same reference.
Removing P frames is even worse quality.
If you remove B frames, you have to compensate with a higher bitrate, and then quality is better.
Removing B and P frames is the BEST quality, but only at very high bitrates ( ~50,000Kbps or more ).
MPEG-2 is used that way in broadcast studios.

-kwag
Reply With Quote
  #98  
02-24-2004, 11:34 AM
rds_correia rds_correia is offline
Free Member
 
Join Date: Apr 2003
Location: Chinese Democracy starts now!
Posts: 2,563
Thanks: 1
Thanked 0 Times in 0 Posts
Hya Kwag,
Have you given mencoder a try yourself?
Would like to share your results with us?
Thanks pal.
Cheers
__________________
Rui
Reply With Quote
  #99  
02-24-2004, 11:38 AM
kwag kwag is offline
Free Member
 
Join Date: Apr 2002
Location: Puerto Rico, USA
Posts: 13,537
Thanks: 0
Thanked 0 Times in 0 Posts
Quote:
Originally Posted by rds_correia
Hya Kwag,
Have you given mencoder a try yourself?
No, I haven't had time
Quote:
Would like to share your results with us?
I will, when I have some time to fool around with it

-kwag
Reply With Quote
  #100  
02-24-2004, 11:56 AM
incredible incredible is offline
Free Member
 
Join Date: May 2003
Location: Germany
Posts: 3,189
Thanks: 0
Thanked 0 Times in 0 Posts
Send a message via ICQ to incredible
Quote:
Removing B frames will make worse quality, no matter what encoder
That is, if bitrate is kept at the same reference.
I underline that as P Frames ONLY work predictious means they refer to the Frames BEFORE where B Frames also do refer to FOLLOWING frames.

If you would only use P Frames you could receive a stream where an visible sudden quality "refresh" occurs when the playback reads an I Frame.
Cause in the Frame run within the I Frame Interval the Frames would get only worse and worse
Reply With Quote
Reply




Similar Threads
Thread Thread Starter Forum Replies Last Post
FFMPEG: Ffvfw VIDEO CODEC kwag Video Encoding and Conversion 364 08-12-2005 07:49 AM
FFMPEG: Curious about H.263 in ffvfw poerschr Video Encoding and Conversion 14 02-25-2004 07:54 PM
FFMPEG: Observation about ffvfw poerschr Video Encoding and Conversion 28 02-24-2004 05:50 PM
FFMPEG: Do ffvfw and mencoder/ffmpeg give the same results? Razorblade2000 Video Encoding and Conversion 4 02-06-2004 04:23 PM
FFMPEG: XMPEG 5.03 and ffvfw kwag Video Encoding and Conversion 2 02-05-2004 10:57 AM

Thread Tools



 
All times are GMT -5. The time now is 07:38 AM  —  vBulletin © Jelsoft Enterprises Ltd