Quantcast Mimicking Motion Adaptiveness Blur with Mencoder. - Page 6 - digitalFAQ.com Forums [Archives]
  #101  
06-11-2004, 02:25 PM
Dialhot Dialhot is offline
Free Member
 
Join Date: May 2003
Posts: 10,463
Thanks: 0
Thanked 0 Times in 0 Posts
Quote:
Originally Posted by Zyphon
Karl those screenshots look awesome
We definitely don't have the same eyes and I understand why I can't love this encodere.

FYI, Marty's anti-rad jacket on the second snap is supposed to be yellow. A plain, uniform, bright, yellow tint. Not this undefined grainy green-yellowish things you can see on mencoder output

Same (but less visible) for the color of the terrorists' pickup on third snaphot.

I'm currently DL the sample... stay tuned !

Edit: Okay, let's go for a quick comment !

Grainy, grainy, grainy.
Look at Marty's face when he closes the DeLorean's door (sec 6). Look at his jacket when he is driving and his arm just after (sec 16 to 18 ), look at the terrorist faces also.
In one words : ANYTHING is like if a sand strom passed in front of the camera while they were shooting.

Note: I only speak about the bad things. "forgetting" the good ones (for instance there is absolutly no mosquitoes despite the resolution and target size and only few pixelated part). But personally I can't really stand watching a whole movie like this. The only idea that comes in mind watching that is "this video need a GOOD denoiser"
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
  #102  
06-11-2004, 03:02 PM
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 Dialhot
Grainy, grainy, grainy.
Phil,

I think you haven't seen the original VOBs
They are worse
I'll post a screenshot, so you can compare, and then maybe you'll see the difference

-kwag
Reply With Quote
  #103  
06-11-2004, 03:26 PM
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
Here's your answer Phil

Original Anamorphic VOB:
Edit: Talk about a BAD quality DVD



4:3 Letterboxed KDVD:



Edit: I think mencoder filters are doing a pretty damn good job

Second comparison:

VOB:


KDVD:


-kwag
Reply With Quote
  #104  
06-11-2004, 03:41 PM
Dialhot Dialhot is offline
Free Member
 
Join Date: May 2003
Posts: 10,463
Thanks: 0
Thanked 0 Times in 0 Posts
Whoaow. You definitely need the THX-labelized PAL release then .
This source is AWFULL
Okay, mencoder did "what it could" on that, I have to admit.

Just to know : how many underflows error message for the whole movie ?
And time to encode ?
Reply With Quote
  #105  
06-11-2004, 03:51 PM
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 Dialhot
Whoaow. You definitely need the THX-labelized PAL release then .
This source is AWFULL
Yes it stinks
Quote:
Okay, mencoder did "what it could" on that, I have to admit.

Just to know : how many underflows error message for the whole movie ?
And time to encode ?
None. Zero. Nada

The next code was not my final mux. That's the mux I did with an MP2 audio I had already encoded at 44.1Khz. The sample you download has 44.1 audio,
My final mux was with AC3 audio.

Code:
  Scanning video stream for a sequence header and pulldown type ...
    2:3 pulldown detected ...
 
  Scanning video stream for pictures ...
  Found 166937 picture headers.
  Video stream information
    Stream length : 1566633911
    Total time (seconds) : 6962
    Sequence start : 9891
    Sequence end : 0
    No. Pictures : 166937
    No. Groups : 9891
    No. I Frames : 9891 avg. size 37234 bytes
    No. P Frames : 45756 avg. size 12214 bytes
    No. B Frames : 111290 avg. size 5746 bytes
    No. D Frames : 0 avg. size 0 bytes
    Horizontal size : 704
    Vertical size : 480
    Aspect ratio : 0.6735
    Picture rate : 29.970 frames/sec,  2:3 pulldown detected
    Bit rate : 1000000 bytes/sec (8000000 bits/sec)
    Computed avg rate : 282500 bytes/sec (2260000 bits/sec)
    Computed max rate : 994750 bytes/sec (7958000 bits/sec)
    Vbv buffer size : 229376 bytes
    CSPF : 0
 
  Scanning audio stream for access units information
  Found 266606 audio frame headers.
  MPEG audio stream information
    Stream length : 97501623
    Syncwords : 266606
    Frames : 76173 size 365 bytes
    Frames : 190433 size 366 bytes
    Layer : 2
    CRC checksums : no
    Bit rate : 14000 bytes/sec (112 kbit/sec)
    Frequency : 44.1 kHz
    Mode : 0 stereo
    Mode extension : 0
    Copyright bit : 0 no copyright
    Original/Copy : 0 copy
    Emphasis : 0 none
 
  Multiplexing information
    Video stream data rate : 994750 bytes/sec (7958000 bits/sec)
    Audio stream 1 data rate : 14000 bytes/sec (112000 bits/sec)
    Overhead data rate : 18050 bytes/sec (144400 bits/sec)
    Total data rate : 1026800 bytes/sec (8214400 bits/sec)
 
  Multiplexing file k:\bttf.mpg
  Finished multiplexing k:\bttf.mpg


-kwag
Reply With Quote
  #106  
06-11-2004, 04:34 PM
audioslave audioslave is offline
Free Member
 
Join Date: Mar 2003
Location: Sweden
Posts: 725
Thanks: 0
Thanked 0 Times in 0 Posts
@kwag
Well, I don't know how you do it but I'm still getting underflows!
I'm using the setting you posted a few pages earlier...

BTW Where can I download the MEncoder version you're using?
__________________
AudioSlave
Reply With Quote
  #107  
06-11-2004, 04:42 PM
Dialhot Dialhot is offline
Free Member
 
Join Date: May 2003
Posts: 10,463
Thanks: 0
Thanked 0 Times in 0 Posts
Quote:
Originally Posted by kwag
Yes it stinks
Yeah and this points out one of the weakness of mencoder I already mentioned during my very first attempt with it : the internal filters of mencoder can't handle quality of sources in the same range we use to do with avisynth.

There is only ONE denoiser implemented in mencoder and it's easy to understand that you can't fix all situation with so few. It's like if avisynth script were limited to Deen or Convolution3D. Mencoder misses also a good temporal only filter.
Reply With Quote
  #108  
06-11-2004, 04:59 PM
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 Dialhot
Quote:
Originally Posted by kwag
Yes it stinks
Yeah and this points out one of the weakness of mencoder I already mentioned during my very first attempt with it : the internal filters of mencoder can't handle quality of sources in the same range we use to do with avisynth.
I meant the VOB stinks
The encoded sample looks MUCH better
Quote:

There is only ONE denoiser implemented in mencoder and it's easy to understand that you can't fix all situation with so few.
Sure, but it's not bad at all

-kwag
Reply With Quote
  #109  
06-11-2004, 05:22 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
Yes Kwag, the images you post are impressive.
They're from KDVD, aren't they?. I see you use maxrate=5000, what's your maximum bitrate for this KDVD?.

I ask because I tested your command-line, with a DVD source and for SKVCD output, and yes, the filter chain really clean nice the film, but with a maxrate=2500 I get peak of 4014 in BitrateViewer .

I posted here, in this same thread some questions I have about your mencoder parameters, but I'm afraid you missed that post.
What's going wrong?, is it the source?, is it that I encode at 704x576 (used noise=2th)?.
How, do you think, can I control those too high bitrate peaks?.
Thanx for your help.
Reply With Quote
  #110  
06-11-2004, 06:41 PM
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 digitall.doc
They're from KDVD, aren't they?.
DVD source (VOB) -> KDVD
Quote:
I see you use maxrate=5000,
No.
Quote:
what's your maximum bitrate for this KDVD?.
8,000Kbps. You can see that in Bitrate viewer.
Quote:

I ask because I tested your command-line, with a DVD source and for SKVCD output, and yes, the filter chain really clean nice the film, but with a maxrate=2500 I get peak of 4014 in BitrateViewer .
Quote:

I posted here, in this same thread some questions I have about your mencoder parameters, but I'm afraid you missed that post.
What's going wrong?, is it the source?, is it that I encode at 704x576 (used noise=2th)?.
SKVCD at 704x576
Didn't you just said you encoded for SKVCD output
Quote:
How, do you think, can I control those too high bitrate peaks?.
Thanx for your help.
I have no Idea. I think it's a PAL issue, because I don't have any problems on NTSC.

-kwag
Reply With Quote
  #111  
06-11-2004, 06:44 PM
Dialhot Dialhot is offline
Free Member
 
Join Date: May 2003
Posts: 10,463
Thanks: 0
Thanked 0 Times in 0 Posts
Quote:
Originally Posted by kwag
I meant the VOB stinks
I had understood

Quote:
The encoded sample looks MUCH better
But a correct avisynth script would have done this better. That's what I wanted to say.

Note : I find your script lost a LOT of details in the bald skrull of the guy on the left (don't remember his name) on the third snap. Marty's face is correctly denoised but this is done with too much sacrifice.

Quote:
Sure, but it's not bad at all
Fortunally ! But what can you do with this source that bumps this unique filter out of its limits ? Nothing.
But with avisynth...
Reply With Quote
  #112  
06-11-2004, 06:47 PM
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 Dialhot

Note : I find your script lost a LOT of details in the bald skrull of the guy on the left (don't remember his name) on the third snap. Marty's face is correctly denoised but this is done with too much sacrifice.
Remember, these are still test parameters.
I'm not done with the filter chain yet

-kwag
Reply With Quote
  #113  
06-11-2004, 06:50 PM
Dialhot Dialhot is offline
Free Member
 
Join Date: May 2003
Posts: 10,463
Thanks: 0
Thanked 0 Times in 0 Posts
Quote:
Originally Posted by kwag
Remember, these are still test parameters.
I'm not done with the filter chain yet
I know, but if everyone just say in front of your snapshots you will start to think that they are right
Reply With Quote
  #114  
06-11-2004, 07:13 PM
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 Dialhot
Quote:
Originally Posted by kwag
Remember, these are still test parameters.
I'm not done with the filter chain yet
I know, but if everyone just say in front of your snapshots you will start to think that they are right
Well, any volunteers to try it with AviSynth
I'm pretty sure the difference will be minimal, and the encoding time to achieve at least the same, will be much longer.
It's a fact that mencoder's internal filters are much faster than AviSynth. So for me, I'm very pleased with those results, just as they are
Can they be improved Of course
Is it worth it Of course
But there will be a price to pay, which usually turns out to be "time".

-kwag
Reply With Quote
  #115  
06-12-2004, 07:33 AM
Dialhot Dialhot is offline
Free Member
 
Join Date: May 2003
Posts: 10,463
Thanks: 0
Thanked 0 Times in 0 Posts
Quote:
Originally Posted by kwag
It's a fact that mencoder's internal filters are much faster than AviSynth. So for me, I'm very pleased with those results, just as they are
Don't make me laught
You admitted yourself that the new line dropped the speed by two ! This added to the 2pass encoding I'm sure you have now an encoding time longer than tmpgenc !
I already have this with a coding line issued from packshot and it doesn't use sab

Encoding time under mencoder : 6h per pass
Encoding time under tmpgenc : 10h

With all your filters you gave a long kiss goodbye to the speed benifit and you know it.

You can do the test on BTFF by yourself. Reinstalling tmpgenc does not take so long
Reply With Quote
  #116  
06-12-2004, 08:15 AM
Koekies Koekies is offline
Free Member
 
Join Date: May 2004
Location: Amsterdam
Posts: 124
Thanks: 0
Thanked 0 Times in 0 Posts
Is the MA avisynth script and mencoder really that slow?
Is the quality better than with tmpgenc?
If it's only a bit slower you could just use that because waiting 12 hours for an encode or 14 isn't that big a deal :P
Reply With Quote
  #117  
06-12-2004, 09:07 AM
Dialhot Dialhot is offline
Free Member
 
Join Date: May 2003
Posts: 10,463
Thanks: 0
Thanked 0 Times in 0 Posts
Quote:
Originally Posted by Koekies
Is the MA avisynth script and mencoder really that slow?
Speed is a relative notion, as you said just after.

Quote:
Is the quality better than with tmpgenc?
We are not talking about mencoder vs tmgenc. We are talking about mencoder internal filters vs avisynth ones. If I put on the side the biggest problem (for me) of mencoder - that is the birate peaks and the buffer underflow - and focus only on the filter problem, I'm not againt using mencoder here BUT with an avisynth script as input.
But in this case, the question to know if tmpgenc will be faster or not is not even needed to be asked

Quote:
If it's only a bit slower you could just use that because waiting 12 hours for an encode or 14 isn't that big a deal :P
Waiting 14h insteed of 12 to obtain the result you see in the snapshot above + not be sure to have a stream readable on your standalone (Kwag said he didn't have any underflow on this but I put that on the luck as this is clearly an issue that nobody control completly)... THEN this is a big deal
Reply With Quote
  #118  
06-12-2004, 09:40 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
Quote:
Originally Posted by kwag
Quote:
I posted here, in this same thread some questions I have about your mencoder parameters, but I'm afraid you missed that post.
What's going wrong?, is it the source?, is it that I encode at 704x576 (used noise=2th)?.
SKVCD at 704x576
Didn't you just said you encoded for SKVCD output
OK, if SKVCD just refers to 480x576 (PAL), I don't know how to call it.
What I mean is I'm encoding mpeg2, 25 fps, 704x576 res, with maxrate=2500 and avg=1230.
Quote:
Originally Posted by kwag
Quote:
How, do you think, can I control those too high bitrate peaks?.
Thanx for your help.
I have no Idea. I think it's a PAL issue, because I don't have any problems on NTSC.

-kwag

So, just nothing to do about it, isn't it?.
It's sad not having to do nothing to sort it out.
Even more, when all my previous tests (3 DVD sources, 2 AVI sources and a home made film: my older son) went OK without a problem. And suddenly, in my last 3 DVD sources, this .

I hope you find a way for PAL users...
Reply With Quote
  #119  
06-12-2004, 11:15 AM
Koekies Koekies is offline
Free Member
 
Join Date: May 2004
Location: Amsterdam
Posts: 124
Thanks: 0
Thanked 0 Times in 0 Posts
I asked if the quality is better because tmpgenc is faster and it would be a waste of time encoding it in mencoder then wouldn't it
unless the quality is better.
Reply With Quote
  #120  
06-12-2004, 12:41 PM
Dialhot Dialhot is offline
Free Member
 
Join Date: May 2003
Posts: 10,463
Thanks: 0
Thanked 0 Times in 0 Posts
Quote:
Originally Posted by Koekies
I asked if the quality is better because tmpgenc is faster and it would be a waste of time encoding it in mencoder then wouldn't it
unless the quality is better.
I definitely sure that with the "old" method (avisynth + tmpgenc) you will have a really better result on sources like this version of back to the future AND you will have a quicker result.

For the speed, my experiences have already proved that. For the quality, only a direct comparison can give the answer.
Reply With Quote
Reply




Similar Threads
Thread Thread Starter Forum Replies Last Post
blur between scene changes is really noticeable? bigggt Video Encoding and Conversion 5 02-08-2004 06:09 PM
Avisynth: More Blur On Action Scenes! Jellygoose Avisynth Scripting 3 10-25-2003 12:08 PM
Avisynth: I can see the blur on scene changes! audi2honda Avisynth Scripting 11 07-08-2003 09:46 PM
KDVD has Motion blur? grc Video Encoding and Conversion 8 06-05-2003 08:06 AM
TMPGEnc: Motion Search Precision - Highest Qual or Motion Est Search MoovyGuy Video Encoding and Conversion 1 08-28-2002 11:00 PM

Thread Tools



 
All times are GMT -5. The time now is 08:17 AM  —  vBulletin © Jelsoft Enterprises Ltd