New Plugins and Utilities

[arakar]

A new version 2.4 of AviSynth Front is available. It adds proper look and feel under Vista and HD video support.

AviSynth Front is a frontend for AviSynth 2.5. It is not intended to be a full featured editor for AviSynth scripts, but rather focuses on ease of use and simplification of everyday tasks. It uses templates that you assemble into a script and then apply to a video or a list file.

Note that the link has changed:
http://arakar.ifastnet.com/andrej/software/

Andrej

[vcmohan]

FQFocusDefocus : This new plugin operates in frequency domain and either (re)focus a mildly defocussed image or defocus an image. 3 patterns of PSF are provided. Additional parameters minimize ringing.

[vcmohan]

Seamer (http://www.avisynth.org/vcmohan) :- Seamer resizes an image by identifying seams through different metrics and carves them out. It has two functions TRSeamer and WSSeamer. Both work in unsupervised mode only.

Seam carving is unlikely to be used in regular video processing, but due to enormous interest generated an attempt is made. These are very slow.

Works in RGB color space only.

Discussion here (http://forum.doom9.org/showthread.php?t=135735)

[vcmohan]

DeBarrel: This plugin works in RGB format only, and corrects for barrel type or pin cushion type of distortion. Such distortions are usually seen on images taken with extreme wide angle or telephoto. Mostly these distortions are noticeable on still images.

Discussion (http://forum.doom9.org/showthread.php?t=136083) thread.

[Wilbert]

Bilateral filter on GPU: http://forum.doom9.org/showthread.php?t=136370

[Gavino]

GRunT is a plugin which makes the Avisynth run-time environment (http://avisynth.org/mediawiki/Runtime_environment) much easier to use.

Features:
- Simple, natural and robust way to pass variables into a run-time script from 'outside'
- A run-time script can be evaluated in its own independent scope
- Run-time functions can be called from a user function
- Run-time functions can be applied to any frame (relative to the current one)
- Additional variant of ConditionalFilter with single boolean expression
- Fixes a fairly serious bug in the run-time system
- Lightweight plugin extending the standard run-time environment, minimal time and memory overhead
- 100% backwards compatible with existing scripts

Details and download here.

[vcmohan]

ReTarget:- ReTarget plugin is a newly developed plugin which through seam carving downsizes a video for the target display device. This has facility to use either forward or backward energy criteria and for marking upto 5 areas in the image for protection or erasing.

Discussion (http://forum.doom9.org/showthread.php?t=139583)Thread.

[vcmohan]

Spinner: This plugin rotates a frame or selected part of it through any desired angle. The rotation axis coordinates and angle of rotation can be in floating point. Option for varying parameters linearly along clip and specify a rectangular area to undergo rotation is available. Precision may be nearest neighbour, bilinear Lagrange and Lanczos interpolation.
RGB, YUY2 and YV12 formats are processed.
Discussion (http://forum.doom9.org/showthread.php?t=140019) thread

[vcmohan]

Perspective : A plugin to correct for distortion caused by camera not being square with image plane. The distortion transforms a front view to a perspective view. This can be corrected and image brought back to front view by this plugin. It can also be used to transform a front view image to a perspective view. Works in RGB format only.

Discussion (http://forum.doom9.org/showthread.php?t=140367)thread.

[vcmohan]

Unfurl : This plug in stitches frames into a panorama excluding the overlap. Specially developed for a case where camera is stationary filming a canvas which is rolled one end and unrolled from other end.
Discussion (http://forum.doom9.org/showthread.php?t=140854) thread.

[-=lDooml=-]

i am new to the forum
just start using it
please anybody can redirect me where is limited sharpen fatser and is it good to use with fft3d
i want plugin
please redirect or give here

[vcmohan]

FXpanse (http://avisynth.org/vcmohan/FXpanse/FXpanse.html) : This plugin creates a super scene out of a panned video. Panning can be in any direction.
Discussion thread here (http://forum.doom9.org/showthread.php?t=141545)

[vcmohan]

Peep is a plugin to display in an elliptical window image from a corresponding part of a different clip. Varying Blending, tapering off of blend are possible in addition to location and size of window.

Discussion at http://forum.doom9.org/showthread.php?t=143432

[Gavino]

GScript is a plugin that extends the Avisynth scripting language to provide additional control-flow constructs:
- multi-line conditionals (if-then-else blocks);
- 'while' loops;
- 'for' loops.

Rather than trying to simulate these constructs with functions, GScript effectively extends the language syntax, making it easy to use the new constructs in a natural way and in arbitrary combinations.

Details and download here.

[jj30]

added:

AviSynth Front ....

You can download it from: http://users.volja.net/arakar/andrej/software/avsfront23.zip

Anyone have a mirror? I am new enough to AviSynth to really want to take this for a spin ... thx.

[Wilbert]

http://rapidshare.com/files/265873478/avsfront23.zip.html

[tin3tin]

Lua and Cairo in Avisynth: Overlua (http://forums.animesuki.com/showthread.php?t=52757)

[StainlessS]

Hi Wilbert,

New version Filter ExBlend() v1.02b. Date 07/Oct/2010

New Thread Here: (http://forum.doom9.org/showthread.php?t=157337) or download via sig.

YUY2, YV12 Progressive source.

Avisynth v2.5+ filter for detecting pairs of blends and replacing them with
a pair of IDENTICAL un-blended frames. (Requires MultiDecimate decimation).

The filter is specific to a particular type of blending, where the source
was 24 FPS Progressive, and converted to 30 FPS (eg Film to Video CD) by
blending as in below:-

A B C D # 24 FPS Source (6 blocks of 4 frames = 24 FPS)

A B cb cd D # Blended to 30 FPS (6 blocks of 5 frames = 30 FPS)

The original frame C was thrown away, leaving a repeated sequence of 3
good frames and 2 blended frames. The framerate may then have been
slowed by 0.1% to 29.97 FPS for NTSC (American) color TV standards.

The filter will also deal with similarly blended 25 FPS to 30 FPS video
in which case the blending would be as below:-

A B C D E # 25 FPS Source (5 blocks of 5 frames = 25 FPS)

A B cb cd D E # Blended to 30 FPS (5 blocks of 6 frames = 30 FPS)

ExBlend() recovers the frame that was thrown away, from the pair of blends and
does not rely on the clip staying in sequence. It is quite easy to set up,
(works pretty gud out of the box).

StainlessS (Stainless Stephen)

[StainlessS]

ColorYUV2() v1.0 Chroma Gamma Fix by Stainless Stephen (StainlessS on Doom9)
Creates DLL as ColorYUV() but named ColorYUV2().

This plugin (ColorYUV2) is a slight modification of the great ColorYUV.

ColorYUV() does not actually implement chroma Gamma correction, and so the example
in Avisynth docs has never worked (Still does not as of v2.5.8). It is taken as a
parameter (gamma_u & gamma_v) and stored somewhere, but pretty much forgotten after that.
For chroma, a large +ve gamma produces an input/output graph shape a bit like an S
(the top half looks just like large +ve luma gamma), sort of like this:-
_
/
_/
It pushes chroma outwards from the mid chroma point (128), but in smaller steps the
further away from the mid point it orginally was.

and a large -ve gamma like this:-

__/
/
It pulls chroma inwards towards the chroma mid point, but in bigger steps the further
away from mid point it originally was.

Try out large (eg +100) for one or both gamma_u and gamma_v for some nice effects that
you have never got before from colorYUV().


ColorYUV2(), Uses alternative text rendering method to render text for Analyze=true,
if you dont like this then just use coloryuv() instead.

ColoYUV2() is not meant as a replacement for ColorYUV(), mearly a stop gap for the
chroma gamma correction until the fix is implemented directly in Avisynth.


ALL Args (and most code) identical as Avisynth ColorYUV()

LINK DELETED

Link to anything ColorYUV2 now here (http://forum.doom9.org/showthread.php?t=156774)

[Wilbert]

ColorYUV() does not actually implement chroma Gamma correction, and so the example
in Avisynth docs has never worked (Still does not as of v2.5.8). It is taken as a
parameter (gamma_u & gamma_v) and stored somewhere, but pretty much forgotten after that.
There is not a good way to compute Y'Cb'Cr', since gamma is a RGB concept. See: http://forum.doom9.org/showthread.php?t=110439

[StainlessS]

@Wilbert,

This post can be deleted. Thankyou.

[mantis2k]

With ColorYUV2, I was able to increase the saturation on my blood red washed out source, using Gamma_u, but instead of the red coming back there is lot's of blue "sparks" (same with TMPGEnc YUV Gamma correction slider). Therefore, those new functions do seem useful, but there is always a trade off with any video treatment.

[Gavino]

It seems that I have stepped in something nasty, the developers seem not to want such a function. Perhaps I should make ColorYUV2(), go away. :)
The function you have implemented seems useful - perhaps you could just rename the parameters to something other than 'gamma'. ;)

[StainlessS]

@Gavino,

Have been a little busy of late, but renamed gamma_u and gamma_v to spow_u and spow_v.
(well, moved functionality to, rather than renamed). Also added args of ramp_u and ramp_v
which implement the identical equivalent of gamma_y for chroma. I managed to resist the
stong urge to call them gamma_u & v again. These two new ramp_x functions are intended
to allow non linear fixes for bad chroma, I thought it might be a good idea based on something
a fellow said in the link provided by Wilbert. Anyway, the eg ramp_u and spow_u can be combined
and together with off, gain, cont, can produce a lot of weird shaped graphs. Have also added
spow_y functionality, thought that it could not do any harm. Small problem though with spow_y,
if in range 16-235, mid point is not 128 but about 126, I have not bothered about this, dont seem
worth any worry. Could maybe mess with this based on whether coring or PC->TV were called for.
Also, PC->TV is actually done before number mangling is applied, to convert to 16-235 before eg
gamma_y does not seem correct in order to me. Would seem better to have TV->PC before and
PC->TV after mangling. (lut creation) Anyways, maybe you have some thoughts on the 126/128
thing or anything else here said. If you wanna see the ramp_x (gamma_y) and spow_x being
combined with off_x gain_x and cont_x, I've knocked up a small (8kb rar) ColorYUV2 graphing
program so you can better visualize the output (would be pretty impossible without it). Intend
to include this with next version of ColorYUV2(), also usable with standard ColorYUV(). Also have
nearly completed a Levels() grapher with arguments conversion to ColorYUV() style, will also include
that with upcoming version. EDIT:- Will also include VB source for both with ColorYUV2().
See below for ColorYUV2_Graf.EXE, 8KB Rar :-

EDIT:- Link Removed: see Post #69


May not be back for a few days, but eager to read any thoughts you have. :)

[IanB]

... It seems that I have stepped in something nasty, the developers seem not to want such a function. Perhaps I should make ColorYUV2(), go away. :)
Yes, something really quite icky, as I said in the referenced thread, "none of these functions are Gamma", I don't have a problem with applying power functions to chroma, I just have a problem calling it gamma.I managed to resist the strong urge to call them gamma_u & v again.That's the spirit ;)

SPOW is not a bad term, not so sure about RAMP though :confused:

Also the functions are not commutative, i.e. order is significant. Question is which way around is more useful. :cool:

[StainlessS]

@ IanB

Thanx for the reply:

not so sure about RAMP though

hows about 'rpow_u' rather than ramp, or do you have a better alternative?
('r' as it sort of looks like lower case r when + maximum)
would be nice to find out before I put my foot in it again.

[pbristow]

Hi! Regarding this one (mentioned in the first post, under "Japanese plugins":

[GPU_001.zip (by thejam79)]: GPU versions of several AviSynth plugins (GPU_BilinearResize, GPU_LanczosResize, GPU_Convolution3d, GPU_TemporalSmoother, GPU_ColorYUY2).

I can't seem to find an active source for this anywhere. Can anyone help?

[Wilbert]

Hi! Regarding this one (mentioned in the first post, under "Japanese plugins":

[GPU_001.zip (by thejam79)]: GPU versions of several AviSynth plugins (GPU_BilinearResize, GPU_LanczosResize, GPU_Convolution3d, GPU_TemporalSmoother, GPU_ColorYUY2).

I can't seem to find an active source for this anywhere. Can anyone help?

http://www.avisynth.info/?%A5%A2%A1%BC%A5%AB%A5%A4%A5%D6#x44606bb
http://www.avisynth.info/?GPU%a5%d7%a5%e9%a5%b0%a5%a4%a5%f3 (more gpu plugins)

[pbristow]

Thanks Wilbert! Gonna try them out this evening, I think. =:o}

[Jenyok]

I am very sorry,
.
http://seraphy.fam.cx/~seraphy/progr...p_20080325.rar
403 Forbidden
Request : GET
Object : /~seraphy/program/WarpSharp/warpsharp_20080325.rar
.
Where can I download WarpSharp 2008 version ?
Please, put WarpSharp 2008 version to public file servers, such as "depositfiles", "rapidshare" and so on.
Thank you.

[jmac698]

Updated to 1.02 Dec 18/2010

Capture Glitch Analyzer Ver. 1.02 by jmac698 (glitchanalyzer.avs)
A video with pseudo-timecode is generated then analyzed to detect temporal capture problems such as duplicated, missing, or field-reversed frames

Why: Some capture cards drop or dupe frames with no other indication. This will interrupt pulldown sequences making IVTC impossible. It also affects various multicap techniques.

Update: I found the solution to the original problem; VirtualDub drops frames by default (and my last 5 years of captures are all ruined). Only the following should be checked:
correct timing
auto disable sync

Requirements: ColorMatrix (optional), only tested on AviSynth 2.58 (in some of my other scripts, 2.6 caused problems)


#Capture Glitch Analyzer Ver. 1.02 by jmac698 (glitchanalyzer.avs)
#A video with pseudo-timecode is generated then analyzed to detect temporal capture problems such as duplicated, missing, or field-reversed frames
#1.01: Bug Fixes; DarkBlue wouldn't join with Colourbars due to differing fps, undefined variable sep, more convenient video sources
#1.02: Proper fix for CreateSolidBackground fps (need 30000.0 to assume float)

#Why: Some capture cards drop or dupe frames with no other indication. This will interrupt pulldown sequences making IVTC impossible. It also affects various multicap techniques.

#Requirements: ColorMatrix (optional)

#Useful tools: AvsP, HCEncoder, VirtualDub

#Usage level: Diagnosis probably requires an intermediate level user familiar with encoding, capturing videos and script setup, and some data analysis skills.
# However, fixing diagnosed capture problems automatically may require advanced scripting knowledge

#Use:
#1. Modify this script to set mode to generate and type to "NTSC" or "PAL" then encode the resulting diagnostic video
# If encoding to mpeg2 use interlaced, top field first (e.g. in HCencoder)
#2. Capture the diagnostic video (e.g. in VirtualDub)
#3. Set mode to analyze, set the captured video filename and the timecode log output filename
#4. Run the analysis by playing back this script once (i.e., load the modified glitchanalyzer.avs into MediaPlayer and play)
#Load the resulting timecode.txt into a spreadsheet and
# note where the timecode skips or duplicates numbers (a perfect capture will count from 0 to 15 repeatedly). It is up to you to interpret this.
#The darkblue intro will read as constant 0, the lightblue outtro will read ideally as constant 15, but in reality will may be random, 0's or 15's. Adjust slicelevel in readtimecode or the value of lightblue or ignore.

#Problems:
#The timecode may not be decoded if the video is horizontally or vertically shifted beyond tolerance, and/or the video levels are beyond tolerance.
#It is up to you to shift, crop, and adjust levels as necessary.
#Glitches spanning more than about 4 frames should be double-checked with the labels contained on the video, because of ambiguity in the sequence (I haven't thought this out further).
#Bluescreens typically generated by a VCR due to loss of signal may lead to unreliable timecodes. Adjust slicelevel in readtimecode so that a constant timecode is decoded.

#Results
#Some capture cards/software/driver combinations will drop frames
# with no indication. One example: duplicate a frame (freezeframe), next frame, skip frame,
# then back to proper frame. This 'hiccup' repeats every 13-15 frames. The dupe is numerically exact.
#This sequence would look like: 0,1 0,1 2,3 6,7... (when it should be: 0,1 2,3 4,5 6,7)
#The next step is to find a way to deal with the missing frames.
#Restoring missing frames could be done by capturing twice, and merging the two videos together (which is easier said than done). Interpolation is another approach to restoring missing frames.
#It's also possible that you don't need the missing frames anyhow, because the video is anime or silent film which cotains many duplicates due to the low frame rate.

#Other uses: Changes in field parity can also be detected. The colourbars can verify levels. The 'darkblue' can help measure noise. The script is educational. Other codes like Closed Captioning could be decoded.

#How it works: First some backgrounds are generated, frame/field count is added, then a call to AddTimecode inserts a test signal
# into the last 6 lines of video. Although only two lines are needed, another problem with
# capture cards is de-centred capturing. An analog signal has 486 lines, but capture cards
# only record 480 of them. They should be centred. Mine is two lines too high, thus
# cropping the first two lnes. Therefore, 6 lines gives some leeway.
#The background 'darkblue' is meant to give time for "Play" (on your playback device) and other symbols to disappear
# before the timecode part appears. For VHS this also gives time for the transport/tape
# to stabilize. And it gives you time to start the capture!
#The analyze function decodes two lines of pseudo-timecode into a 4 bit number,
# which counts from 0 to 15 and repeats. This is enough to detect several combinations
# of duplicated, missing, parity reversed, or rearranged frames. An absolute timecode would be more difficult to program.
#With glitches affecting more than about 4 frames, you should look at the actual video
# to verify the absolute timecode, which is written in text.

#Functions:
#CreateColourbars(string type, string colourspace, float length) Creates colourbars with type as NTSC or PAL, colourspace as REC709 or REC601, length in seconds
#CreateSolidBackground(string type, string colourspace, float length, int c) Creates a solid background with type as NTSC or PAL, colourspace as REC709 or REC601, length in seconds, colour as YUV in c
#LabelVideoWithCounts(clip v) Labels a video with frame and field counts
#whiteline(clip template) Creates a white line using template for video properties
#blackline(clip template) Creates a black line using template for video properties
#createcode(clip template,int code) Creates a video representing a 4bit timecode using template for video properties
#createcodefields(clip template,int code) Creates a video representing a 4bit timecode using template for video properties. The video has two time codes on two lines, for use in interlaced video.
#AddTimecode(v) Overlays a timecode onto the supplied video, using createcodefields for generation.
#readtimecode(clip videoin) Extracts the video lines representing a timecode and decodes it, labeling the results on the returned video and also writing the results to the timecode file
#ShiftVideo(clip v, int hshift, int vshift) Shifts a video in any direction, where hshift<0 means left and vshift<0 means up. Use only multiples of 2.
#ShowTimeCode(clip v) Extracts the video lines representing a timecode.

videoindrive="C:\"#You may find this useful if you use external hard drives
videoindir="project001\multicap\"
videoinname="glitchdiagnostic.avi"
videoinfilename=videoindrive+videoindir+videoinname
#You may find it easier to just uncomment the line below:
#videoinfilename="C:\MyVideos\glitchdiagnostic.avi"
dataoutdrive="C:\"
dataoutdir="project001\multicap\"
dataoutname="timecode.txt"
dataoutfilename=dataoutdrive+dataoutdir+dataoutname
#videoinfilename="C:\project001\multicap\framedrops.txt"
#Various test videos
v1=AviSource("C:\project001\multicap\svideo-source-glitch.avi").trim(209+117,0)
v2=MPEG2Source("C:\project001\multicap\Pinnacle PCTV 800i.0.SVideoInput.d2v", cpu=0).trim(51,0)
vid=v1

generate=1#Used to set mode to create the diagnostic video
analyze=2#Analyze a supplied video
demo=3#Both generate and analyze a virtual video and generate a timecode file
mode=demo
type="NTSC"#Set to NTSC or PAL. You definitely need to set this.
#Important: if you don't have the optional ColorMatrix plugin installed, please set REC601 below!
colourspace="REC709"#Choose REC709 when encoding to mpeg2, REC601 for divx. It is only for creating correct colourbars.
DarkBlue=CreateSolidBackground(type,colourspace,15,$10B848)#Specify video system, colourspace, time in seconds, and colour in YUV (directly in the colourspace of your choosing, i.e., it's not converted).
LightBlue=CreateSolidBackground(type,colourspace,15,$7EB848)#Joke: Booktree 848 is a nice colour!
CreateColourbars(type,colourspace,60)
mode==analyze?vid:last
mode==analyze?nop:LabelVideoWithCounts#Label the last video with frame and field counts
mode==analyze?nop:AddTimecode#Add computer readable timecode to last video (this is not related to the timecode in professional VCRs)
mode==analyze?nop:darkblue.assumefps(last.framerate)+last+darkblue.assumefps(last.framerate)
sep=","
WriteFileStart(dataoutfilename, """ "Current_Frame"+sep+"Top_Timecode"+sep+"Bot_Timecode" """)
WriteFileEnd(dataoutfilename, """ "End of Timecode for the file "+videoinfilename """)
mode==generate?last:readtimecode#Decode timecode embedded in video and log timecodes
#ShiftVideo(0,0)#Used for development; can be used to align incoming diagnostic video
#ShowTimeCode#Show only the extracted timecode video lines for debugging

#--- Functions are defined below. You may want to split this out to another file and Import. ---
function CreateColourbars(string type, string colourspace, float length) {
fps=type=="NTSC"?30000/1001:250000/1001
colorbars(width=720,pixel_type="YV12").trim(0,length*fps/2-1).ConvertAudioTo16bit#weave'd down to half the time
colourspace=="REC709"?ColorMatrix(mode="Rec.601->Rec.709"):last#If you get an error message here, set colourspace to REC601
}

function CreateSolidBackground(string type, string colourspace, float length, int c) {
fps=(type=="NTSC")?30000.0/1001:25000.0/1001
blankclip(Round(5*fps),width=720,pixel_type="YV12",fps=fps,audio_rate=48000,stereo=true,color_yuv=c)
}

function LabelVideoWithCounts(clip v) {
#Add framecount and field counts to a video
v
assumetff
separatefields
top=last.selecteven
bot=last.selectodd
#all this is to remove color shadows; where the text overlays, color is desaturated, due to YV12
tu=top.utoy
tv=top.vtoy
bu=bot.utoy
bv=bot.vtoy
top1=top.ScriptClip("""subtitle("top field = " + string(current_frame*2),x=-1,y=100,size=20,text_color=$00FFFFFF,font_width=17)""")
bot1=bot.ScriptClip("""subtitle("bot field = " + string(current_frame*2+1),x=-1,y=120,size=20,text_color=$00FFFFFF,font_width=17)""")
top=ytouv(tu,tv,top1)
bot=ytouv(bu,bv,bot1)
interleave(top,bot)
weave
ScriptClip("""subtitle("frame = " + string(current_frame),x=-1,y=140,size=40,text_color=$00FFFFFF)""")
assumefps("ntsc_video")
assumetff
}

#creates a 2 pixel high white line, quarter width, with the same video properties as the template
function whiteline(clip template) {
template.crop(0,0,template.width/4,-(template.height-2)).tweak(sat=0,bright=255)
}

function blackline(clip template) {
template.crop(0,0,template.width/4,-(template.height-2)).tweak(sat=0,bright=-255)
}

function createcode(clip template,int code) {
#make 4 bit code from 4 quarter width horizontal lines
code=code % 16
b3=code/8
code=code-b3*8
b2=code/4
code=code-b2*4
b1=code/2
code=code-b1*2
b0=code
template
stackhorizontal(b3==1?whiteline:blackline,b2==1?whiteline:blackline,b1==1?whiteline:blackline,b0==1?whiteline:blackline)
}

function createcodefields(clip template,int code) {
code=code*2
interleave(createcode(template,code),createcode(template,code+1))
pointresize(template.width,4)
assumefieldbased.assumetff.weave
crop(0,0,0,-2)
}

function AddTimecode(v) {
#adds a 4bit timecode on the last 6 lines of video
#e.g. frame 0, top field reads 0000, bot field reads 0001
ScriptClip(v,"""
overlay(last,createcodefields(last,current_frame),y=last.height-6)
""")
}

function readtimecode(clip videoin) {
ScriptClip(videoin,"""
t=last
w=last.width
q=w/4#this is used to divide a video line into 4 horizontal parts to locate bits in the timecode
h=last.height#this is used to locate the timecode lines, and should compensate for NTSC or PAL
slicelevel=126#this luma video level is used to distinguish between 1 and 0 bits in the timecode ('data slicer' is a common term in electronics for this purpose)
sep=","#the separator character used for the output data
pointresize(last.width,last.height*2)
b3tv=crop(0,(h-4)*2,-3*q,-3*2)#the leftmost 1/4 of a horizontal video line contains the highest bit value of the timecode
b2tv=crop(q,(h-4)*2,-2*q,-3*2)
b1tv=crop(2*q,(h-4)*2,-q,-3*2)
b0tv=crop(3*q,(h-4)*2,0,-3*2)
b3bv=crop(0,(h-3)*2,-3*q,-2*2)
b2bv=crop(q,(h-3)*2,-2*q,-2*2)
b1bv=crop(2*q,(h-3)*2,-q,-2*2)
b0bv=crop(3*q,(h-3)*2,0,-2*2)
b3t=b3tv.averageluma>=slicelevel?1:0#a 'white' line (i.e., luma>=slicelevel) represents a 1 bit
b2t=b2tv.averageluma>=slicelevel?1:0
b1t=b1tv.averageluma>=slicelevel?1:0
b0t=b0tv.averageluma>=slicelevel?1:0
b3b=b3bv.averageluma>=slicelevel?1:0
b2b=b2bv.averageluma>=slicelevel?1:0
b1b=b1bv.averageluma>=slicelevel?1:0
b0b=b0bv.averageluma>=slicelevel?1:0
tc=b3t*8+b2t*4+b1t*2+b0t#top timecode
bc=b3b*8+b2b*4+b1b*2+b0b#bot timecode
t
subtitle("decoded top field timecode="+string(tc))
subtitle("decoded bot field timecode="+string(bc),y=16)
WriteFile(dataoutfilename, "string(current_frame)+sep+string(tc)+sep+string(bc)")
""")#tip: this could be potentially modified into a closed caption decoder as well
}

function ShiftVideo(clip v, int hshift, int vshift) {#Shift a video in any direction, hshift>0 shifts right, vshift>0 shifts down, use multiples of 2
v
l=hshift<0?-hshift:0
r=hshift<0?0:hshift
up=vshift<0?-vshift:0
dn=vshift<0?0:vshift
crop(l,up,0,0).addborders(0,0,l,up)
crop(0,0,-r,-dn).addborders(r,dn,0,0)
}

function ShowTimeCode(clip v) {#Show only the extracted timecode video lines for debugging
v
w=last.width
q=w/4#this is used to divide a video line into 4 horizontal parts to locate bits in the timecode
h=last.height#this is used to locate the timecode lines, and should compensate for NTSC or PAL
crop(0,h-4,0,2)
}

[jmac698]

Pan Scan Utilities Ver. 0.1 by jmac698
Utilties for pan, zoom, and aspect ratio conversion of video

Shift - will pan a video in any x,y direction
Zoom - will zoom into the center of a video without resizing it
Aspectconvert - can pillar box or letterbox

Can be used with Animate to do a Ken Burns effect (panning over stills to add action to a documentary video, zooming in on a person your talking about).


function shift(clip v, int sx, int sy) {
v#shift a video sx, sy pixels, sx<0 is left, sy<0 is down
pointresize(width*2,height*2)#to allow 1 pixel shifts in YV12
sx=sx*2
sy=sy*2
l=(sx<0)?-sx:0
r=(sx<0)?0:sx
t=(sy<0)?0:sy
b=(sy<0)?-sy:0
crop(l,t,-r,-b)
addborders(r,b,l,t)
bilinearresize(v.width,v.height)
}

function zoom(clip v, float zx, float zy) {
v#Zoom but cropping to original size
pointresize(width*2,height*2)#to allow 1 pixel increments in YV12
w=int(zx*width/4+.5)*4
h=int(zy*height/4+.5)*4
addleft=(zx<1)?(width-w)/2:0
addtop=(zy<1)?(height-h)/2:0
bilinearresize(w,h)#zoom
left=(zx>1)?int((w-v.width*2)/2):0
top=(zy>1)?int((h-v.height*2)/2):0
crop(left,top,-left,-top)#Keep only centered portion of zoomed video
addborders(addleft,addtop,addleft,addtop)
bilinearresize(width/2,height/2)#back to normal size
}

function aspectconvert(clip v, int w, int h) {
v#Extract a wxh video from original, e.g. a square cut out of a rectangle, centered
#If source is smaller than dest, center source in dest
#If source is less wide than dest, center crop source (i.e. pan scan)
#If source is taller than dest, center crop source
#a square (dest) on a widescreen (in) does a center cut. A widescreen on a square does a letterbox.
pointresize(width*2,height*2)
w=w*2
h=h*2
left=(w<width)?(width-w)/2:0
addleft=(w>width)?(w-width)/2:0
top=(h<height)?(height-h)/2:0
addtop=(h>height)?(h-height)/2:0
addborders(addleft,addtop,addleft,addtop)
crop(left,top,-left,-top)
pointresize(w/2,h/2)
}

[jmac698]

New Filter: Deep Color Tools Ver 0.3a by jmac698
Scripted functions to manipulate 48bit color video. Reads 10bit mov files. These are the first Avisynth functions to support 10 bit and above video. Ver 0.3, bugs fixed. Tested to work with v210 file.

#Deep Color Tools Ver 0.3 by jmac698
#Scripted functions to manipulate 48bit color video. Reads 10bit mov files. These are the first Avisynth functions to support 10 bit video.
#Bugs: Only 9 bits of 10bit files are used. Tweak only supports cont and bright and calculates like coloryuv, there's some small numerical errors (rounding in mt?)
#Ver 0.3a: renamed to Deep Color Tools (from Deep Color Utilities) since this was the only name I can remember :) Updated documentation.
#Ver 0.3: fixed bug in clamprange, deepgreyscale, code cleanup (thanks Gavino from Doom9!)
#Ver 0.2: Fixed bugs with deeplastlo in all functions, successfully tested with a v210 file, added compatibility with yuy2,
# order of operations in tweak, chroma bug in tweak
# new functions: deepramp, deepcrop, deepgreyscale
# cleaned up code
#Requires masktools 2.0a48 (http://manao4.free.fr/masktools-v2.0a48.zip), qtsource (http://www.tateu.net/software/dl.php?f=QTSource), Quicktime 6+ or QT Lite (http://www.free-codecs.com/download/QT_Lite.htm)
#Uses an internal format with two videos for the high and low bytes, audio in low is ignored, format tracks the pixel format
#Functions:
#deepread(string fn) - reads a v210 mov file and saves to deeplast
#deepgreyclip(int "luma",string "pixel_type") - sets deeplast to luma (a 16 bit number), defaults to YV12
#deeptweak(float "hue", float "sat", float "bright", float "cont", bool "coring") - adjusts brightness and contrast in 16bit, including coring
#deepcrop(l,t,a,b)- crops deeplast
#deepramp(string "pixel_type",int "bits") - creates a luma ramp of bits, defaults to 10 bit yv12. The range is unrestricted.
#deepgreyscale - converts deeplast to greyscale
#The high byte is always displayed as a preview. There is no dithering or rounding.
#To come: more downcolor options, color correction, colorbars, more functions, resizing
Global deeplasthi=blankclip(pixel_type="YV12").tweak(bright=-255,coring=false).tweak(bright=16,coring=false)
Global deeplastlo=blankclip(pixel_type="YV12").tweak(bright=-255,coring=false).tweak(bright=0,coring=false)
Global format=0#0=YV12, 1=YUY2, 2=RGB

fn="G:\project001a\deepcolor\v210.mov"
deepramp
#deepread(fn)
deeplastlo#this line shows if you have data in the lower bits

function add16k(float n) {
string(n)+" x y 256 * + +"
}

function mult16k(float n) {
string(n)+" x y 256 * + *"
}

function multadd16k(float m, float a) {#multiply by m then add a
string(a)+" "+string(m)+" x y 256 * + * +"
}

function clamphi(string s, int n) {
s+" "+string(n)+" > "+string(n)+" "+s+" ?"
}

function clamplo(string s, int n) {
s+" "+string(n)+" < "+string(n)+" "+s+" ?"
}

function clamprange(string s, int low, int high) {
clamplow(s,low).clamphi(high)
}

function b0(string s) {
s+" 255 &u"
}

function b1(string s) {
s+" 65280 &u 8 >>"
}

function b2(string s) {
s+" 16711680 &u 16 >>"
}

function b3(string s) {
s+" 4278190080 &u 24 >>"
}

function deeptweak(float "hue", float "sat", float "bright", float "cont", bool "coring") {
coring=default(coring,true)
bright=default(bright,0)
cont=default(cont,1)
clamplower=coring?16*256:0
clampupper=coring?235*256:65535
#Make sure deeplast is yv12 before using masktools (yuy2 is stored as yv12 double height to preserve color)
preprocess
#do Contrast then Brightness, then clamp. -bright can bring a cont value back in range, ie 162*2=328, -100=228
#Note: this is operating like coloryuv gain,off because tweak takes into account tv levels (ie (y-16)*cont+bright+16)
tweakloexpr=b0(clamprange(multadd16k(cont,bright),clamplower,clampupper))
tweakhiexpr=b1(clamprange(multadd16k(cont,bright),clamplower,clampupper))
newlo=mt_lutxy(deeplastlo,deeplasthi,yexpr=tweakloexpr,chroma="copy second")
newhi=mt_lutxy(deeplastlo,deeplasthi,yexpr=tweakhiexpr,chroma="copy second")
global deeplastlo=newlo
global deeplasthi=newhi
postprocess
deeplasthi
}

function deepgreyclip(int "luma",string "pixel_type") {
pixel_type=default(pixel_type,"YV12")
global deeplasthi=blankclip(pixel_type=pixel_type).tweak(bright=-255,coring=false).tweak(bright=luma/256,coring=false)
global deeplastlo=blankclip(pixel_type=pixel_type).tweak(bright=-255,coring=false).tweak(bright=luma%256,coring=false)
Global format=pixel_type=="YV12"?0:pixel_type=="YUY2"?1:2
deeplasthi
}

function deepread(string fn) {
global deeplasthi=qtinput(fn,color=2,audio=true,mode=1,raw="v210")
global format=IsYUY2(deeplasthi)?1:0
qtinput(fn,color=2,audio=true,mode=1,raw="v210",dither=1)#9th bit
#low bit to high bit
getlowbyte(last,deeplasthi)
global deeplastlo=last
deeplasthi
}

function deepcrop(l,t,a,b) {
global deeplasthi=crop(deeplasthi,l,t,a,b)
global deeplastlo=crop(deeplastlo,l,t,a,b)
deeplasthi
}

function preprocess {#work around limitations of masktools
global deeplasthi=format==1?deeplasthi.pointresize(deeplasthi.width,deeplasthi.height*2).converttoyv12:deeplasthi
global deeplastlo=format==1?deeplastlo.pointresize(deeplastlo.width,deeplastlo.height*2).converttoyv12:deeplastlo
}

function postprocess {#undo temporary format
global deeplasthi=format==1?deeplasthi.converttoyuy2.pointresize(deeplasthi.width,deeplasthi.height/2):deeplasthi
global deeplastlo=format==1?deeplastlo.converttoyuy2.pointresize(deeplastlo.width,deeplastlo.height/2):deeplastlo
}

function deepramp(string "pixel_type",int "bits") {#Create an increasing luma ramp in 10bit deepcolor, full range
#more useful as an internal diagnostic than a video test pattern
pixel_type=default(pixel_type,"YV12")
bits=default(bits,10)
global deeplasthi=blankclip(pixel_type=pixel_type)
global deeplastlo=blankclip(pixel_type=pixel_type)
Global format=pixel_type=="YV12"?0:pixel_type=="YUY2"?1:2
n=int(pow(2,bits))
lowbits=bits>8?bits-8:0
vlowbits=int(pow(2,lowbits))
v1=n-vlowbits
v2=vlowbits-1
v3=vlowbits<8?8-lowbits:0
ramphiexpr="x "+string(v1)+" &u "+string(lowbits)+" >>u"
ramploexpr="x "+string(v2)+" &u "+string(v3)+" <<u"
#How this works: for 10 bits, you count as: 00000000 00, 00000000 01, 00000000 10, 0000000 11, 00000001 00
#the low 2 bits we put into the low byte, high 2 bits, or shift up 6 times, or (8-2) times, this is value v3
#now we have 0,64,128,192... for the low byte
#In the high byte, you can see a pattern of repeating values, 0,0,0,0,1...
#so we could just /4 to get this but if the video width were 1024 we'd reach 256
#so we & with (1023-3), this is v1. Now we can divide by 4 or shift right 2. This is lowbits.
preprocess
global deeplasthi=mt_lutspa(deeplasthi,mode="absolute",yexpr=ramphiexpr)
global deeplastlo=mt_lutspa(deeplastlo,mode="absolute",yexpr=ramploexpr)
postprocess
deeplasthi
}

function deepgreyscale {
global deeplasthi=greyscale(deeplasthi)
global deeplastlo=greyscale(deeplastlo)
deeplasthi
}

function getlowbyte(clip vround, clip vtrunc) {#vround may have some pixels +1 compared to vtrunc; put the difference in the high bit
Overlay(vround,vtrunc, mode="Subtract", pc_range=true)#e.g. y:129-128=1, 128-128=0. u,v are 128 or 129
coloryuv(gain_y=127*256,off_u=-128,off_v=-128)
coloryuv(gain_u=127*256,gain_v=127*256)
}

function notdeeptweak(clip v,float "cont", float "bright") {#just for a quick demo
v
coloryuv(gain_y=(cont-1)*256,off_y=bright)
}

[frustum]

jmac698 --

Very nice! I just had an hour to kick the tires and haven't done much testing yet. One roadblock is I don't have quicktime on my PC and don't want it on my PC. So I set up a virtual machine, installed quicktime, your script, and its dependencies. Here is my simple script:

film = "z:\original-uncompressed\01.avi"

loadplugin(".\mt_masktools-25.dll")
loadplugin(".\QTSource.dll")
import(".\deepcolor.avs")

deepread(film)
#deeplastlo

Running this script under virtualdub, I do see the video. If I uncomment the last line to see the lsbs, I see black, confirming that my "10b video" is really 8b video in a 10b format.

What is weird is that the film has 5023 frames, but if I view deeplastlo, virtualdub says that I have 240 frames.

[jmac698]

New Filter: VCR Simulator ver. 0.2 by jmac698
Simulates an old recording by introducing line jitter and reduced frequency response. Besides being an interesting special effect, this modelling can be used to test various filters for cleaning up analog recordings.

#VCR line jitter simulator Ver. 0.2 by jmac698 - with special thanks to Gavino :)
#requires both gscript and grunt, and noise generator (see Warpsharp web page)
#Ver 0.2: add noise, more realistic jitter, minor fixes
colorbars(pixel_type="YUY2").trim(0,10)
pointresize(width,height*4)#get line resolution by working around min height 4 limitations
h=height
splitlines(4)#split to lines of height 4 - note the jitter function requires height>=4
jitter
#return last
MergeLines(h)#merge until height=h
noisegenerator(true,55)
blur(1).blur(1).blur(1)
pointresize(width,height/4)
temp=last
u=temp.utoy
v=temp.vtoy
u=u.bilinearresize(30,height).bilinearresize(320,height)#it's actually better than this, placement of transitions is finer
v=v.bilinearresize(30,height).bilinearresize(320,height)
ytouv(u,v,temp)

function SplitLines(clip c, int n) {
Assert(c.height%n == 0, "Clip height not a multiple of 'n'")
Assert(!(c.IsYV12() && n%2==1), "'n' must be even for YV12 clips")
nStrips = c.height/n
c = c.ChangeFPS(nStrips*Framerate(c)).AssumeFPS(c) # Repeat each frame 'nStrips' times
BlankClip(c, height=n) # template for ScriptClip result
GScriptClip("c.Crop(0, (current_frame%nStrips)*n, 0, n)", args="c, nStrips, n")
}

function MergeLines(clip c, int h) {
n=h/c.height #n is number of frames to merge, e.g. 480 split to 4line has n=120
c.SelectEvery(n)
GScript("""
for (i=1, n-1) {
stackvertical(c.SelectEvery(n, i))
}
""")
}

function shift(clip v, int sx, int sy) {
v#shift a video sx, sy pixels, sx<0 is left, sy<0 is down
pointresize(width*2,height*2)#to allow 1 pixel shifts in YV12
sx=sx*2
sy=sy*2
l=(sx<0)?-sx:0
r=(sx<0)?0:sx
t=(sy<0)?0:sy
b=(sy<0)?-sy:0
crop(l,t,-r,-b)
addborders(r,b,l,t)
bilinearresize(v.width,v.height)
}

function jitter(clip v) {#note: a good tape could typically be +-3 without a TBC, distribution is not realistic
ScriptClip(v,"""
shift(rand(2)+rand(2)+rand(2)+4,0)
""")
}


VHS simulator, VHS emulator, VHS effect

[pbristow]

Nice one, jmac698! (I haven't tested it, but it the idea is a good one and the implementation looks OK to me).

For a more realistic distribution of the jitter, try changing:
rand(6)
to:
rand(2)+rand(2)+rand(2)

That should give a roughly gaussian distribution over the the range 0 to 6 (average 3, standard devation around 1-ish).

:)

[jmac698]

Try it now - I can't believe how much it looks like a VCR, I could swear I just capped it - try a movie.

[jmac698]

New Filter: RGB<=>YCgCo Ver. 1 by jmac
Requirements: masktools v2a38+
Description: Converts between RGB and YCgCo. YCgCo is a low correlation colorspace used in H.264 for better compression efficiency.
Usage: Load your source, convert to RGB, call RGB2YCgCo, convert to YV12, compress. Should encode well.
To playback, use the Avisynth filter in FFDShow and enter converttorgb.YCgCo2RGB

#RGB<=>YCgCo demo by jmac, requires masktools v2a38+, tested in avisynth 2.58
#The demo just shows colorbars, but what is happening is a conversion back and forth between colorspaces
#Theoretically you would convert to YCgCo and compress, this will compress about 10% better than YUV
#However, it's unclear how to send this to an h.264 encoder. There's probably a way to arrange the pixels to make it work.
#To test compression, just feed as YUV to a compressor. The Cg/Co channels will be subsampled. Playback the result as YUV, the script will decode it.
#Ver 1.0 - Does correct, full range conversion both ways, in 8bit
#Testing...
RGB2YCgCo
#for realtime playback in ffdshow, from h.264 with YCgCo, erase all lines above
#except I don't know the format return from ffdshow
YCgCo2RGB

function RGB2YCgCo(clip v) {
#convert an assumed RGB clip into Co,Y,Cg
#~ Y = 0.25 * R + 0.5 * G + 0.25 * B
#~ Cg = -0.25 * R + 0.5 * G - 0.25 * B + 0.5
#~ Co = 0.5 * R - 0.5 * B + 0.5
v
#put planes into yuv for masktools (awkward)
#RGBintoYV12
r1=ShowRed.converttoyv12(matrix="PC.709")#R->Y
g1=ShowGreen.converttoyv12(matrix="PC.709")#G->Y
b1=ShowBlue.converttoyv12(matrix="PC.709")#B->y
r=YtoUV(r1,r1,r1.pointresize(r1.width*2,r1.height*2))#y->y, y->u, y->v
g=YtoUV(g1,g1,g1.pointresize(g1.width*2,g1.height*2))#y->y, y->u, y->v
b=YtoUV(b1,b1,b1.pointresize(b1.width*2,b1.height*2))#y->y, y->u, y->v
#Note: remember b=u=Cg=z, r=v=Co=x, g=y=y=y
mt_lutxyz(r,g,b,yexpr="x 4 / y 2 / + z 4 / +",uexpr="y 2 / x 4 / - z 4 / - 128 +",vexpr="x 2 / z 2 / - 128 +",chroma="process")
YV12intoRGB
}

function YCgCo2RGB(clip v) {
#convert an assumed Co,Y,Cg clip into RGB
#~ R = Y - Cg + Co
#~ G = Y + Cg
#~ B = Y - Cg - Co + 1
v
#put planes into yuv for masktools (awkward)
r1=ShowRed.converttoyv12(matrix="PC.709")#R->Y
g1=ShowGreen.converttoyv12(matrix="PC.709")#G->Y
b1=ShowBlue.converttoyv12(matrix="PC.709")#B->y
r=YtoUV(r1,r1,r1.pointresize(r1.width*2,r1.height*2))#y->y, y->u, y->v
g=YtoUV(g1,g1,g1.pointresize(g1.width*2,g1.height*2))#y->y, y->u, y->v
b=YtoUV(b1,b1,b1.pointresize(b1.width*2,b1.height*2))#y->y, y->u, y->v
#Note: remember b=u=Cg=z, r=v=Co=x, g=y=y=y
mt_lutxyz(r,g,b,yexpr="y z + 128 -",uexpr="y z - x - 256 +",vexpr="y z - x +",chroma="process")
YV12intoRGB
}

function YV12intoRGB(clip v) {
#place v,y,u data into r,g,b planes
#this result has no visual meaning (however, closest approximation) but is only for performing arithmetic on the data
v
r=vtoy.converttorgb(matrix="PC.709")
g=greyscale.converttorgb(matrix="PC.709")
g=g.pointresize(width/2,height/2)#Y was doublesized
b=utoy.converttorgb(matrix="PC.709")
MergeRGB(r,g,b)#u->b, v->r, y->g
}

[StainlessS]

ClipClop, New Plugin for range/frame editing of clips, see here:-

http://forum.doom9.org/showthread.php?t=162266

Planar, YUY2, RGB32, RGB24.

Does NOT affect Audio (as source clip).

Clipclop is a simple plugin to replace ranges in a source clip with the same range, from a replacement clip.
Supports up to 255 replacement clips, with unlimited number of replacements into output clip.

[StainlessS]

There is a great little utility that I thought deserved another mention here on-site, I did a search and
AMP WinOFF has not had mention since 2004. AMP WinOFF is a utility whose best attribute is
that it can shut down/hibernate etc, the computer under various conditions, mainly that where the CPU
usage falls below a user selected level of activity. Most similar type programs pretend they do this, but what
they measure is "User Activity" rather than CPU activity, not quite the same, and probably of limited use
to Doom9 members. I have been using WinOFF since about 2004, perhaps earlier and have always
been very happy with it. I noticed recently that a number of threads were concerned with adding shutdown
type capability into eg Encoder GUI's, with AMP WinOFF, this becomes completely un-necessary,
especially if doing several jobs simultaneouly, you would not want to shutdown just because one of those
jobs has completed, that type of utility function should be implemented at a machine level.

However, having multiple shutdown options are quite ofter handy, and the Hibernation options
are always a good idea (goes to sleep, wakes up and continues as if nothing had happened).

This is typical of how I used to use WinOFF using TMPGEnc (dont use TMPGEnc anymore).
Encode in TMPGEnc, Hibernate after 1 hour of processing, auto boot at a certain time in the
morning (Bios Setup), Auto Logon to system, continue encode in TMPGEnc, use TMPGEnc
implemented ShutDown On Completion. All automated.

Anyway, have just visited WinOFF website and found that a new version was released in Nov 2010,
(not updated mine for several years) and thought I should let you all in on the secret.

Below from AMP WinOFF Website:-

AMP WinOFF 5.0.1

Description

AMP WinOFF is an utility for scheduling the shutdown of Windows computers,
with several shutdown conditions and fully configurable. Some of its features are:

Several modes/conditions for shutdown planification:
At certain date/time (e.g. 12:00 AM).
After a period of time (e.g. 1 hour and 15 minutes).
When there isn't user activity (i.e. no mouse/keyboard input).
When there isn't CPU activity.
When there isn't network activity.

Option of set several conditions at the same time for the planification. Examples:
Shut down at 08:00 PM or when there is no user activity for 15 minutes.
Shut down when the CPU load is below than 1% and the network transfer is below than 1Kb/s.
Several types of shutdown: power off, restart, close session, lock session, administrative shutdown/restart, sleep and hibernate.
Immediate shutdown and session lock from the taskbar icon menu.
Several security options including anti-close protection and password protected access to the configuration.
Option of showing a display with the active shutdown planification.
Option of executing a program, capturing the desktop, and/or hang up the modem before the shutdown.
Command line support for batch processing.
Dual English/Spanish version.

http://www.ampsoft.net/utilities/WinOFF.php

[StainlessS]

From Previous post above:


This is typical of how I used to use WinOFF using TMPGEnc (dont use TMPGEnc anymore).
Encode in TMPGEnc, Hibernate after 1 hour of processing, auto boot at a certain time in the
morning (Bios Setup), Auto Logon to system, continue encode in TMPGEnc, use TMPGEnc
implemented ShutDown On Completion. All automated.


You can "Auto Logon" to the system using AutoLogon tool from
SysInternals (Microsoft). here: (EDIT: Updated Feb 2011)

http://technet.microsoft.com/en-gb/sysinternals/bb545027

[StainlessS]

Prune(), New plugin for Trim & Splice.

Prune is a simple plugin to trim() multiple source clips and splice the results into a new clip.
Supports up to 256 source clips, with unlimited number of trims/splices into output clip.

Video:- Planar, YUY2, RGB32, RGB24.

http://forum.doom9.org/showthread.php?t=162446

[vampiredom]

New Utility and Function: MpegAutoIndex and MpegAutoIndexSource()

http://forum.doom9.org/forumdisplay.php?f=69

All requirements are included in the distribution.

MpegAutoIndex provides easy script-controlled indexing for MPEG2, h264/MPEG4 and VC1 sources via Donald Graft's indexing tools. As of 2011/11/01, h264 .mkv files are also supported. [note: .mkv containers, VC1 formats and interlaced h264 require a license from neuron2.net (http://neuron2.net)]

The package includes the indexers - and expands h264 support to include .mp4 and .mov containers via the included mp4box and faad executables. Likewise, it now includes ASF2VC1 to enable accurate indexing of VC1 in .wmv containers.

MpegAutoIndexSource() uses NicAudio for sourcing a large variety of audio formats. It also includes SoundOut for easy exporting of AC3 and WAV formats, which can be particularly helpful for 5.1 channel encodings, etc. The function provides many options for audio input/processing/output.

When opening .VOB files, MpegAutoIndexSource() automatically detects and indexes the entire set.

# MPEGAutoIndexSource(filename [, options])

# Open an MPEG4 file
MpegAutoIndexSource("foo.mp4")

# Open an AVC stream using DGIndexNV and hardware double-rate deinterlacing
MpegAutoIndexSource("foo.mts", NV=true, args="deinterlacing=2")

# Open an AVC stream using DiAVC and hardware double-rate deinterlacing
MpegAutoIndexSource("foo.mts", DI=true, args="deinterlacing=2")

# Open an NTSC MPEG2 DVD stream using DGIndex, force it to 24P (film)
# Downmix the multichannel audio to stereo, Normalize the audio to 99%
MpegAutoIndexSource("VTS_01_1.VOB", FieldOperation=1, Downmix=true, PeakAudioLevel=0.99)

# Open an AVC stream from BluRay using DGIndexNV, forcing 24p
# Downmix the multichannel audio to stereo, normalize it to 100%
# Output the audio as a 256kbps AC3 file
MpegAutoIndexSource(
\ "12345.m2ts",
\ NV=true,
\ args="fieldop=1",
\ Downmix=true,
\ PeakAudioLevel=1.00,
\ AC3Out=true, AC3Bitrate=256
\ )

# Open an VC-1 .wmv file
MpegAutoIndexSource("vc1.wmv", NV=true)

# Open a .mkv file with h264 video
MpegAutoIndexSource("vc1.wmv", NV=true)

[StainlessS]

New utility, MeGUI AutoEncode Batcher

Utility to allow group selection of *.AVS files for batching to MeGUI AutoEncode.

Here:-
http://forum.doom9.org/showthread.php?t=163343

[StainlessS]

New utility, DGIndex Batcher v0.01 (Previous post for MeGUI AutoEncode Batcher is new also, v0.12).

Utility to allow group selection of files for batching to DGIndex.

EDIT: Can call above previously posted "MeGUI AutoEncode Batcher" to pass generated AVS files
for it to AutoEncode batch to MeGUI.

EDIT: Both utilites are now updated.

Here:-
http://forum.doom9.org/showthread.php?t=163369

[redfordxx]

New plugin, RedAverage (http://forum.doom9.org/showthread.php?p=1537311#post1537311).

Set of filters to sum, average and merge "unlimited" number of clips.
Input and output can be independently 8bit or 16bit (for higher precission).

[StainlessS]

ShowChannels() v0.1

Plugin for Avisynth 2.5+.

Planar, YUY2 and RGB.

Simple plugin whose only function is to display the average Y,U and V
values for a YUV frame or R,G, and B for an RGB frame.
Also shows accumulated average for all frames visited so far.

Typical output for a YUV clip:-

347 ] Frames Visited = 348

This Frame Accumulated
Y U V Y U V
Ave 77.00 125.52 125.42 70.91 125.87 125.78
Min 4 80 95 0 69 83
Max 255 177 165 255 190 174
~Min 14 95 101 12 92 101
~Max 234 167 148 235 167 151

See here:-
http://forum.doom9.org/showthread.php?t=163829

[DrPhill]

I have created a new plugin in C. It adds a 'watermark' to a video.

A 'watermark' deforms the image as if it were 'embossed' into it. Teh watermark is defined by the user as a black and white image (still or moving). The intensity of the effect is controllable through parameters.

This filter currently only works in RGB32, and has not been exhaustively tested, though it is known to work on at least three machines.

I would dearly like someone with real C skills to check the code for obvious bugs, portability issues and general silliness.......

The source and release are available from SourceForge (https://sourceforge.net/projects/watermarkplugin/).

Any and all comments, feedback and suggestions are welcome, but preferably to this thread (http://forum.doom9.org/showthread.php?t=163870) as I have email notification on it.

Example of output (http://www.youtube.com/watch?v=4Jxb3Y4aq2w)

Thanks

[panzerboy]

http://sourceforge.net/projects/avisynthgmplyr/files/?

This is an implementation of GIMPs layers as AVIsynth merge filters.
See the GIMP docs for the types of merges.
http://docs.gimp.org/2.6/en/gimp-concepts-layer-modes.htm

I've cribbed the code from the GIMP source (mostly the pixel arithmetic) and inserted it into the SDK example 1.7.
Amazingly, it all seems to work!

So there's 20 merge filters in AvisynthGimpLayer.dll (must think of a better name).

Normal(clip, clip, opacity=float)
Multiply(clip, clip, opacity=float)
Divide(clip, clip, opacity=float)
Screen(clip, clip, opacity=float)
GimpOverlay(clip, clip, opacity=float)
Burn(clip, clip, opacity=float)
Dodge(clip, clip, opacity=float)
Hardlight(clip, clip, opacity=float)
Softlight(clip, clip, opacity=float)
GrainExtract(clip, clip, opacity=float)
GrainMerge(clip, clip, opacity=float)
Difference(clip, clip, opacity=float)
Addition(clip, clip, opacity=float)
Subtract(clip, clip, opacity=float)
DarkenOnly(clip, clip, opacity=float)
LightenOnly(clip, clip, opacity=float)
Hue(clip, clip, opacity=float)
Saturation(clip, clip, opacity=float)
Colour(clip, clip, opacity=float)
Value(clip, clip, opacity=float)

They work on RGB or RGBA clips only.
The last four Hue, Saturation, Colour and Value have a RGB-HSV conversion and back.
So they will probably run extra slow.

So why?
I play with mandelbrots with the Fractal Extreme program.
I started off with static pictures now I tend to do zoom movies.
With the static pictures I often generate multiple copies of the mandelbrot with different colour palettes and mappings.
These get layered in GIMP.
Lately I've been doing similar with the movies using Merge, Mergechroma and MergeLuma.
I wanted to do a movie of a older static picture
http://www.flickr.com/photos/panzerboy/6233908023/
, to get an match I needed the same layering that GIMP provides.
http://youtu.be/Nk5aeJi22bg

[wonkey_monkey]

I had a need of this simple filter - there may be a better way to implement it, but the work's done now so I thought I may as well share it.

http://horman.net/simpleselect.zip

simpleselect(clip, f1 [,f2...,fn])

Return only frames f1, f2,...,fn of the source clip, e.g.

simpleselect(clip, 2, 3, 5, 7, 9, 13) # return a 6-frame clip comprising frames 2,3,5,7,9,13 of clip

David