06-04-2003, 11:00 AM
|
Free Member
|
|
Join Date: Sep 2002
Posts: 108
Thanks: 0
Thanked 0 Times in 0 Posts
|
|
@kwag and jellygoose: I'm using the same trigger value for 480x480 encodes and it works just fine.
|
Someday, 12:01 PM
|
|
Site Staff / Ad Manager
|
|
Join Date: Dec 2002
Posts: 42
Thanks: ∞
Thanked 42 Times in 42 Posts
|
|
|
06-04-2003, 11:06 AM
|
Free Member
|
|
Join Date: Jun 2002
Location: Germany
Posts: 1,288
Thanks: 0
Thanked 0 Times in 0 Posts
|
|
Quote:
Originally Posted by girv
@kwag and jellygoose: I'm using the same trigger value for 480x480 encodes and it works just fine.
|
good news girv thanks!
btw: Where did STMedian Filter go in the latest script
__________________
j3llyG0053
|
06-04-2003, 11:07 AM
|
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 Jellygoose
So at KVCDx3 resolution, the current script should also work with no blinks or flashes right?
|
Yes. The flashes and blinks were related to the on/off switching of the filters. At least that's the only logical conclusion I have, because now I can't duplicate the problem
-kwag
|
06-04-2003, 11:12 AM
|
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 Jellygoose
btw: Where did STMedian Filter go in the latest script
|
Geee thanks Jellygoose , I forgot to add it, and the samples I posted were done with the filter included. I just fixed the script
Too many cycles are wearing my brain out :P
-kwag
|
06-04-2003, 11:55 AM
|
Invalid Email / Banned / Spammer
|
|
Join Date: Aug 2002
Location: Brasil - MG - third stone from the sun
Posts: 5,570
Thanks: 0
Thanked 0 Times in 0 Posts
|
|
Quote:
Originally Posted by Jellygoose
Quote:
Originally Posted by girv
@kwag and jellygoose: I'm using the same trigger value for 480x480 encodes and it works just fine.
|
good news girv thanks!
btw: Where did STMedian Filter go in the latest script
|
i see without STMedianin in the last and before the last too but don't post.
i was afraid to bore with some that i don't understand.
thank you all.
doing the sample number 39738390,8
|
06-04-2003, 12:30 PM
|
Invalid Email / Banned / Spammer
|
|
Join Date: Aug 2002
Location: Brasil - MG - third stone from the sun
Posts: 5,570
Thanks: 0
Thanked 0 Times in 0 Posts
|
|
i did samples with and without the (forgoten)STF and can't see differences.
with the filter i got little less size, but in pc the quality is the same.
anyone did the same?
see differences
in the last news i can't see some strange artefacts in
the moviments on the background using 480x480.
is really better
|
06-04-2003, 01:11 PM
|
Free Member
|
|
Join Date: Jan 2003
Location: Puerto Rico
Posts: 423
Thanks: 0
Thanked 0 Times in 0 Posts
|
|
[quote] Quote:
Originally Posted by jorel
i did samples with and without the (forgoten)STF and can't see differences.
with the filter i got little less size, but in pc the quality is the same.
anyone did the same?
see differences
|
I Dont use STMedian I use Deen but that only a persopnal preference you can use Conv 3D, MipSmooth or any other to replace it. If you want to use STM use it with a low setting like Kwag has it I think it works better that way. Now TemporalSoften that one i wouldn't live out, is the best of the smoothers n the fastest.
|
06-04-2003, 01:14 PM
|
Free Member
|
|
Join Date: Jun 2002
Location: Germany
Posts: 1,288
Thanks: 0
Thanked 0 Times in 0 Posts
|
|
[quote="ovg64"] Quote:
Quote:
Originally Posted by jorel
i did samples with and without the (forgoten)STF and can't see differences.
with the filter i got little less size, but in pc the quality is the same.
anyone did the same?
see differences
|
I Dont use STMedian I use Deen but that only a persopnal preference you can use Conv 3D, MipSmooth or any other to replace it. If you want to use STM use it with a low setting like Kwag has it I think it works better that way. Now TemporalSoften that one i wouldn't live out, is the best of the smoothers n the fastest.
|
The only difference is that STMedian Filter does only spatial smoothing, the way we use it right now, while C3D or MipSmooth are combining Spatial and Temporal Filtering...
__________________
j3llyG0053
|
06-04-2003, 01:57 PM
|
Free Member
|
|
Join Date: Apr 2003
Location: Brockton, MA
Posts: 134
Thanks: 0
Thanked 0 Times in 0 Posts
|
|
Now that the dust has settled after evryone's hard work I thought I might share my experience for adding filters for not so clean material/captures.
UnDot and DNR2 do a good job of removing noise and are a balance of speed and quality. Actually, Kwag, you might want to consider adding UnDot to the current script for DVD's since it has virtually no effect on encoding time.
|
06-04-2003, 03:28 PM
|
Free Member
|
|
Join Date: Jun 2002
Location: Germany
Posts: 1,288
Thanks: 0
Thanked 0 Times in 0 Posts
|
|
Right now, all I can say is that the latest script still works best for me, but I'm currently testing aSharp Filter, which is an Adaptive sharpening Filter, and the results are promising...
Testing is such a long-taking procedure for me...
__________________
j3llyG0053
|
06-04-2003, 03:31 PM
|
Free Member
|
|
Join Date: Jan 2003
Posts: 275
Thanks: 0
Thanked 0 Times in 0 Posts
|
|
Which Sharpen Filter, JellyGoose
EDIT:
Now, I read that you're testing "aSharp"
|
06-04-2003, 03:52 PM
|
Invalid Email / Banned / Spammer
|
|
Join Date: Aug 2002
Location: Brasil - MG - third stone from the sun
Posts: 5,570
Thanks: 0
Thanked 0 Times in 0 Posts
|
|
Quote:
Originally Posted by Bchteam
Which Sharpen Filter, JellyGoose
EDIT:
Now, I read that you're testing "aSharp"
|
believe me friend Bchteam,
asharp is the best sharper that i know.
i'm doing samples for 2 days to see and compare
using in the Kwag's scripts(and i test all scripts after any change)
changing unfilter to asharp and compare any encode.
in all asharp is champion!
do a test and see
asharp(2,3)...(very gently) or asharp(2,4)...(more sharp)
|
06-04-2003, 04:06 PM
|
Free Member
|
|
Join Date: Jan 2003
Posts: 275
Thanks: 0
Thanked 0 Times in 0 Posts
|
|
Sounds very Promising, Jorel
Am I supposed to put "aSharp" Before(Like UnFilter) or After resizing
@Kwag
It's time for updating the script...
...again
|
06-04-2003, 04:23 PM
|
Invalid Email / Banned / Spammer
|
|
Join Date: Aug 2002
Location: Brasil - MG - third stone from the sun
Posts: 5,570
Thanks: 0
Thanked 0 Times in 0 Posts
|
|
Quote:
Originally Posted by Bchteam
Sounds very Promising, Jorel
Am I supposed to put "aSharp" Before(Like UnFilter) or After resizing
@Kwag
It's time for updating the script...
...again
|
Bchteam,
all "sharp" filter tested before resize give
more sharpness and less size...ever
|
06-04-2003, 04:32 PM
|
Free Member
|
|
Join Date: Jan 2003
Posts: 275
Thanks: 0
Thanked 0 Times in 0 Posts
|
|
Yes I know, Jorel
I just required your opinion as final confirmation
I just made my first sample and I'm very curious...
|
06-04-2003, 04:33 PM
|
Free Member
|
|
Join Date: Jan 2003
Location: Puerto Rico
Posts: 423
Thanks: 0
Thanked 0 Times in 0 Posts
|
|
Hi all if this is thru than its worth using asharp. I have tried it and so far so good, but keep in mind the filter is very strong and at a high setting will add file size.
What it does:
basically, this filter is a very common unsharp mask, simply because unsharp
mask is the most pleasant sharpenning technique for my eyes ^^
but i've added 3 simple ideas i had about sharpening :
- adaptive thresholding to avoid noise enhancement
- block adaptive sharpenning to avoid DCT block edges enhancement
- block based adaptive thresholding to avoid DCT block edges enhancement
|
06-04-2003, 04:43 PM
|
Free Member
|
|
Join Date: Mar 2003
Location: Sweden
Posts: 725
Thanks: 0
Thanked 0 Times in 0 Posts
|
|
Hmm... This ASharp filter sure sounds interesting! What settings do you guys recommend for a medium to high action DVD movie? I'm currently encoding "A Knight's Tale" and it would be nice to test the filter on that movie! Do you use Bicubic or Linear resize together with ASharp?
|
06-04-2003, 05:02 PM
|
Invalid Email / Banned / Spammer
|
|
Join Date: Aug 2002
Location: Brasil - MG - third stone from the sun
Posts: 5,570
Thanks: 0
Thanked 0 Times in 0 Posts
|
|
Quote:
Originally Posted by Bchteam
Yes I know, Jorel
I just required your opinion as final confirmation
I just made my first sample and I'm very curious...
|
Bchteam,
i got a little less CQ after prediction but big more quality
and of course,after prediction the final size will
be the same using any filter.
audioslave,
the recomendation from the developer is asharp(2,4)
but it will encrease the size and the Kwag's scripts,
don't need huge sharpness.
better is (2,2) or(2,3).
see how this filter encrease the contrast
ps:
where is our master?
(Kwag of course)
|
06-04-2003, 05:12 PM
|
Free Member
|
|
Join Date: Mar 2003
Location: Sweden
Posts: 725
Thanks: 0
Thanked 0 Times in 0 Posts
|
|
@jorel
What about the other two parameters? No need for them?
And again:
Do you use Bicubic or Linear resize together with the filter?
|
06-04-2003, 05:27 PM
|
Free Member
|
|
Join Date: Apr 2002
Location: Puerto Rico, USA
Posts: 13,537
Thanks: 0
Thanked 0 Times in 0 Posts
|
|
@dano,
Thanks for the tip on "undot". It's now in the script.
@jorel,
Thaks for your heavy testing with "asharp".
I've removed "unsharp" and used "asharp".
@all,
Thanks for your patience
I've re-written the script, eliminating ALL dynamic blur and mergeluma(blur) calls.
I've used negative values in asharp to produce exactly the same dynamic results, but without the DAMN bugs that the "blur" and family of functions.
I had to do this, because I still found sporadic flashes on several encodes I did today. So I tried asharp, but used it as a "blur" engine instead of as a sharpener. Only at the beginning of the script it's used to oversharpen, just like unfilter was used.
Also, the speed of the script is much faster by removing all the 'blur" functions.
So give it a whirl, and post your comments. I can now say that I didn't find ANY blink or flash anywhere on scenes that DID have flashes with the previous script. The functionality of the new script is identical to the previous one.
-kwag
|
All times are GMT -5. The time now is 10:42 PM — vBulletin © Jelsoft Enterprises Ltd
|