Welcome to Doom9's Forum, THE in-place to be for everyone interested in DVD conversion. Before you start posting please read the forum rules. By posting to this forum you agree to abide by the rules. |
26th February 2016, 16:42 | #961 | Link |
I'm Siri
Join Date: Oct 2012
Location: void
Posts: 2,633
|
nah nah nah nah nah, some kernels do that, and some kernels don't, generally non-linear kernel (median, bilateral, etc) based sharpeners never ever produce halos, but they do feature the risk of enhancing those already existed halos
even if the kernel itself gives halos, it's possible still, you can always just non-linearize the sharpening diff with sigmoidal like expressions to remove halos at the cost of some sharpness |
26th February 2016, 21:13 | #962 | Link |
Registered User
Join Date: Feb 2002
Location: California
Posts: 2,691
|
The sharpening functions used in VideoFred's scripts are the ones that do introduce halos, so the absence of halos indicates, to me, that sharpening was not used, or used very sparingly (i.e., he used really small -- or zero -- values for the sharpening parameters).
|
26th February 2016, 23:00 | #963 | Link |
Registered User
Join Date: Feb 2016
Posts: 2
|
@John :
Thank you for the compliment. In total it took me about three months to build the machine and convert about 11 hours of film. You are absolutely right about being conservative on the sharpening. In the examples shown I use USM_sharp_ness= 30 .For lower quality movies (I had a few that were almost transparent) I used even less sharpening. I don’t know about halos, but if the sharpening is too high you get noise that appears around fast moving objects, something I find very annoying. To maintain a good dynamic range, the program slowly adjusts the exposure time frame-to-frame to push the histogram as high as possible without saturating. For movies that show almost no degradation ( like the examples I shown) I use a gamma correction of 0.5 during capture ( 12bit raw -> 12bit gamma corrected -> 8bit final (drop 4 lsb bits) ). Pol |
27th February 2016, 14:39 | #964 | Link | ||
Registered User
Join Date: Oct 2014
Posts: 34
|
Quote:
Quote:
Watch this: As you can see there is a pretty good sharpening. But no halo artifacts. Last edited by EHarlen; 27th February 2016 at 16:15. |
||
27th February 2016, 18:41 | #965 | Link | |
Registered User
Join Date: Feb 2002
Location: California
Posts: 2,691
|
Quote:
From that video: Before: After That halo above the leaf was not there in the original. Last edited by johnmeyer; 9th October 2019 at 16:37. Reason: replace photobucket images |
|
29th February 2016, 19:13 | #966 | Link | |
Registered User
Join Date: Dec 2002
Location: Region 0
Posts: 1,436
|
Quote:
|
|
6th March 2016, 16:19 | #967 | Link |
Registered User
Join Date: Jul 2014
Posts: 5
|
script
Hi!
I have an industrial camera which can shoot 1600x1200 (or less) RGB32 @ 20, 15, 7.5, 3.75 fps. I have Eumig mark 610d with obturator fan cut-off I am trying to capture avi at 20 fps on camera, with 6 fps set on Eumig (8 mm film normal) i am using this script with getdups Avisource("E:\test\1.avi") loadplugin("GetDups.dll") fc=Framecount() AssumeTFF() ConvertToYV12() GetDups(mode=0) AssumeFps(16) ChangeFps(25) Trim(0, fc*2/3) Here is what i've got after the script: The speed is not normal, and it's twitchy The question is: maybe i can use modified (somehow) Fred's script (or any other) to get the normal speed ? (actually, i dont need any cleaning or stabilizing of the film due to time it takes, only speed normalization, if it is possible). Thanks! |
6th March 2016, 18:34 | #968 | Link |
Registered User
Join Date: Feb 2002
Location: California
Posts: 2,691
|
Well, whatever that script is doing, it sure is creating a mess. There are all sorts of missing frames (i.e., frames that are NOT dups that have been removed). You need to take a completely different approach to removing duplicates. I use TFM and TDecimate, using some very unusual settings, in order to recover good frames from a 24 fps sound film projector captured with a Sony FX1 NTSC camcorder with the shutter speed set to 1/1000 second.
Your situation is far, far easier, and you should be able to use fairly standard settings for TFM and TDecimate. All you need to do is get the Cycle and CycleR parameters set correctly. |
14th March 2016, 23:55 | #969 | Link |
AviSynth plugger
Join Date: Nov 2003
Location: Russia
Posts: 2,183
|
1. what do you get without ChangeFPS(25) ?
if resulted frame sequence is not smooth (dubbed or missed frames), try tune GetDups parameters (BTW, new version 0.94 is out with additional parameters) 2. then we can help you with framerate conversion (or without it) Please provide also input video (before getdups). johnmeyer, the existense of GetPups is due to situation is NOT far, far easier ( IMO). In particular not-stable projection speed.
__________________
My Avisynth plugins are now at http://avisynth.org.ru and mirror at http://avisynth.nl/users/fizick I usually do not provide a technical support in private messages. |
12th July 2016, 12:38 | #971 | Link |
Registered User
Join Date: Jun 2016
Posts: 15
|
Why video is cropped second time by using est_left est_top est_right and est_bottom? Can left, top right and bottom parts be harmful for depanning?
Code:
est_left=20 est_top=60 est_right=60 est_bottom=60 #crop values for special Estimate clip stab_reference= source1.crop(20,20,-20,-20).colorYUV(autogain=true).crop(est_left,est_top,-est_right,-est_bottom) Be default winx - number of columns (width) of fft window (default = 0 for maximum power of 2 within frame width). winy - number of rows (height) of fft window (default = 0 for maximum power of 2 within frame height). Could the FFT window be expanded to match stab_reference or is this avoidable? Would this cropping be ok? Code:
est_left=20 est_top=20 est_right=20 est_bottom=20 #crop values for special Estimate clip stab_reference= source1.crop(est_left,est_top,-est_right,-est_bottom).colorYUV(autogain=true) data=DePanEstimate(stab_reference,trust=trust_value,dxmax=maxstabH,dymax=maxstabV, winx=stab_reference.Width , winy=stab_reference.Height) Code:
result5= overlay(source1,greyscale(stab_reference),x=est_left,y=est_top).addborders(2,2,2,2,$FFFFFF).Lanczos4Resize(WS,HS) |
12th July 2016, 12:57 | #972 | Link | |
Registered User
Join Date: Dec 2004
Location: Terneuzen, Zeeland, the Netherlands, Europe, Earth, Milky Way,Universe
Posts: 689
|
Quote:
Fred.
__________________
About 8mm film: http://www.super-8.be Film Transfer Tutorial and example clips: https://www.youtube.com/watch?v=W4QBsWXKuV8 More Example clips: http://www.vimeo.com/user678523/videos/sort:newest |
|
15th September 2016, 23:48 | #973 | Link |
Registered User
Join Date: Feb 2002
Location: California
Posts: 2,691
|
I'm still using more or less the same script I posted a few years ago. This fall I hope to spend a little time improving it by using some of the things VideoFred and StainlessS created for improving color balance.
Here's a sample of some of the improvements I made on some 16mm film I did for a client last spring: 16mm Before & After Last edited by johnmeyer; 15th September 2016 at 23:49. Reason: fixed link |
2nd October 2016, 21:17 | #974 | Link |
Registered User
Join Date: May 2016
Posts: 235
|
I'm still a bit new at this. I get the script working but there is only comparing, I try to modify the avs-files in the download to get the result, but I just can't get it to work. Can someone post the scripts with only the result? I realy want to try it out.
|
5th February 2017, 03:59 | #976 | Link |
Registered User
Join Date: Jan 2017
Posts: 1
|
real time transfer
First of all, i want to say a big thanks for this script.
I am doing real time telecine at 25fps. I seen some warnings about this script is for frame by frame. Here is my comparison video I know how to remove the split screen by removing the "S" But i want a full screen result as here in the comparison with NO interpolation because its 25 fps recorded with camera. Anyone here getting some good results with real time? thank you |
5th February 2017, 14:09 | #977 | Link | ||
Registered User
Join Date: Dec 2004
Location: Terneuzen, Zeeland, the Netherlands, Europe, Earth, Milky Way,Universe
Posts: 689
|
Quote:
If your file contains duplicate frames, denoising will not work as it should be. If possible remove duplicate frames before using my script. Quote:
Fred.
__________________
About 8mm film: http://www.super-8.be Film Transfer Tutorial and example clips: https://www.youtube.com/watch?v=W4QBsWXKuV8 More Example clips: http://www.vimeo.com/user678523/videos/sort:newest |
||
21st February 2017, 16:52 | #979 | Link |
Registered User
Join Date: May 2014
Posts: 19
|
Wanted to show you all another one of our color restorations.. this one was horrid to say the least.. here is a split side by side comparison..
(I'm aware of the slow frame rate, was fixed in our final non-comparison release) I'm semi-satisfied w/the results..the saturation isn't where it should be, but because so much color was gone it was hard to pump it up anymore w/out it looking fake. This was prior to building a new PC and new workflow. We're going to use less avisynth and more Virtualdub plugins. The newest version of the plugin we use uses Nvida Cuda and we have a new video card (nvidia GTX 1070) enroute to speed that up. Couple it with the new i7 6800k and 2x SSDs in RAID 0 - we're cookin'! -t |
21st February 2017, 17:24 | #980 | Link |
Registered User
Join Date: Feb 2002
Location: California
Posts: 2,691
|
I couldn't get the video to play ("No video with supported format and MME type found"). I am able to see a single frame thumbnail, but that thumbnail looks very "muddy." However, perhaps it is "dimmed," as thumbnails often are.
|
Thread Tools | Search this Thread |
Display Modes | |
|
|