AvsP v2.0.2 (10/27/07)

[AlanHK]

I was making some screencaps using "save image as" to use as logo masks.

Anyway, after some hairpulling when the logo filter failed, I eventually discovered that the screencaps were odd sizes.

If I use a fixed zoom ratio (25%, 200%, etc) the screencap is as expected, the frame size.

But I had been using zoom set at "fit inside window". That made the screencaps scale to the zoom level. Same happens with "fill window".

This seems more like a bug than a feature to me...

[chipzoller]

Request: Add a larger library of the type-as-you-script plugins and their functions. I like having all the functions come up once typing, for example, lanczosresize( and would like to see this appear for more plugins.

[azsd]

nice tools! I have to start three Virtualdubmod instance to load differnt scripts, then click between taskbar until I found this one.

Is there any option to make the priview windows floatable or dock verticality on the right of editor?
there were lots of blank on the right of scripts.

I've tried mainSplitter.SplitVertically but takes no change.

[gomguk]

AvsP 2.0.2 Korean language file.
Making(Credit) by "Hani"(name).

Download URL:
http://zizybe.bcpark.net/data/avsp202_korean_lang.zip

Please register avsP official website. ^^

[blausand]

Hi all,

as many of you know an open source solution is still missing for compressing video with alpha channel.
Now, i found two avi-CoDecs that process AlphaChannel:

Lagarith,
which is lossless and therefore too heavy for VJing and
Softlab MJPG with Alpha,
which requires WxH resolution multiples of 16,
installs fine in Version 5

However, this did not work for me:

mask() some video with another as RGB32 in Avisynth
compress via avs2avi,
choose Softlab MJPG+Alpha,
reopen that .avi in AVSP,
pass it through showalpha
-> it shows only black.


SO, is that RGB32 correctly transferred to the CoDec by avs2avi, qwerpoi?

Thanks for your time anyway! AVSP is highly appreciated over here!
Give it another release, qwerpoi!

PS.: I screen the transparency as known in Adobe etc. with the following little script. You can append it to every script you are working on, instead of Showalpha().
@qwerpoi: I think transparency should display like this as default!

src=AVIFileSource("source.avi", pixel_type="RGB32")
ImageReader("path.to\TransparencyCheckers.png",pixel_type="RGB32").trim(0,src.framecount).crop(0,0,src.width, src.height)
layer(last,src)
Result looks like this:
http://www.silvertree.de/images/TransparenzPlayer.png

[GasKid]

Hi,
Excuse me if it is a stupid question, but how can I import my filter customizations?
There's an export button in "AviSynth function definition...", but I can't find where to import.
I've looked at the help file, but it looks like it's written for old version of AvsP. It says there's an import command under File menu, but it's not there in v2.0.2
Could anyone please help me? Thanks a lot.

[gzarkadas]

Hi,
Excuse me if it is a stupid question, but how can I import my filter customizations?
There's an export button in "AviSynth function definition...", but I can't find where to import.
I've looked at the help file, but it looks like it's written for old version of AvsP. It says there's an import command under File menu, but it's not there in v2.0.2
Could anyone please help me? Thanks a lot.

Import does not seem to be implemented yet. Either you can wait a future release (but apparently you don't want to wait; so did I :D) or you can edit directly the file "filterdb.dat" inside the folder where avsp.exe is with a text editor and add your custom entries under the appropriate section.

Most probably you will need to do that only for sections [PLUGINS] and [USERFUNCTIONS]. And perhaps for section [COREFILTERS] if you want to put your own custom defaults to some arguments.

For those sections you must observe the following rules to succesfully enter/edit entries without messing things up:

Keep exactly one empty line between filter/function definitions.
All added entries in the above two sections are multiline entries, such as in the example below:

DirectShowSource_DirectShowSource(
string filename= (*.*) [, ...],
float "fps",
bool "seek"=true,
bool "audio"=true,
bool "video"=true,
bool "convertfps"=false,
bool "seekzero"=false,
int "timeout"=60000 (1000 to 300000 by 1000),
string "pixel_type"="AUTO" ("YV12"/ "YUY2"/ "RGB32"/ "RGB24"/ "YUV"/ "RGB"/ "AUTO"),
int "framecount",
string "logfile",
int "logmask"=35 (1 to 512)
)

The first line of entry is the name of filter/function followed by an opening parenthesis.
NB: For plugins, put in front of the filter/function name the plugin's DLL name and join the two names with an underscore. This will make available both the fully qualified and the normal names of the filter.
On each following line you place, one by one in the order accepted by the filter/function, the filter/function's arguments. Use a comma to end each argument line (except of course the last argument).
For each argument you can specify defaults, etc. (see the example above and the contents of filterdb.dat for details; sorry, I don't have time to explain further).
The last line of the entry is a closing parenthesis alone.


Some final notes:

If you want to add/edit entries in other sections such as [SCRIPTFUNCTIONS] or [CLIPPROPERTIES] you must then put the entire entry in one line.
You can have multiple sections with the same name in the file; AvsP will parse them all.
This provides the ability to put all your new custom entries in separate sections from the default ones shipping with AvsP, which is both better documentation for tracking your changes and less prone to make default entries dissapear from the AvsP GUI should you make a typing or entry-formatting error.
Make a backup copy of original "filterdb.dat" before starting to edit it. This will save you time, if errors are made, to restore in a stable version of the file.
It is better to put all of your customisations in a separate file end edit there, then copy and paste to "filterdb.dat". Apart from quicker editing, this will also create a backup of your customisations.


Happy customisation :).

[GasKid]

Thank you gzarkadas for this nice guide. It works fine.

[Red Wine]

Greeting to the great community!
As you probably understand I'm newbie here, however, I'm watching the forum constantly.

Since I discovered AvsP I was so impressed and I decided to build my own tiny tool in order to automate the procedure to encode video in combination with AvsP.

Later I thought why not share my tiny tool? After all it might be helpful. Quick AVI Creator (my tiny tool) works nicely and in combination with AvsP produces fine results.

However I would like to ask AvsP developer if it's possible a (special?) build with some fields disabled. Since I know nothing about python, I'm unable to do it by my self.

Specifically I want disabled the fields Input file and Output file in the encode video window and the field container in the bitrate calculator window.

I guess you probably want to take a look on my tiny tool.
Here's a download link for QAC (sorry for rapidshare link I don't have a server to upload it).

http://rapidshare.com/files/114015821/Quick_AVI_Creator.7z

EDIT: If you have downloaded from the above link please download the updated Quick Avi Creator.exe which eliminates a couple of reported bugs,
otherwise download the updated package. Sorry....

Updated exe only:
http://rapidshare.com/files/114237910/QAC.zip

the package:
http://rapidshare.com/files/114238337/Quick_AVI_Creator.7z

Thanks in advance

[Red Wine]

Never mind, forget the above post, I'll survive without these features.

Occasionally the silence is louder than 1000s words.

Apologies bosses if I have bothered you...

Cheers

[tin3tin]

Well I just came across the Aldrin (http://trac.zeitherrschaft.org/aldrin/) python/wxpython module/API (http://doc.zeitherrschaft.org/aldrinapi/) which seems to offer a lot of shortcuts on building timeline/sequencer gui's.

I'm not that much into python/wxpython yet to dare to try to code a timeline/sequencer gui for AvsP, but it would be great if ex. tabs with different scripts could be converted into drag n' drop/scaleable boxes in a timeline as a typical videoediting gui.

The timeline view would mainly be used for editing clips/scripts together timing wise.

Would this be at all possible?

[unskinnyboy]

qwerpoi, I've stumbled across something, which I believe is a weird bug in AvsP. When attempting to fix field-blending and ghosting issues on a source, and after restoration of the source to progressive frames, when stepping through the frames in AvsP, random blends appear in random frames, and only at times, after extensive forward and backward navigation using the <-/-> keys. The same issue doesn't happen in VirtualDubMoD.

To demonstrate -

Blending observed in AvSP (look at the younger guy's head):
http://img338.imageshack.us/img338/623/avsp465blendingop5.th.jpg (http://img338.imageshack.us/img338/623/avsp465blendingop5.jpg)

The same frame, when you just jump to the frame in VirtualDubMoD:
http://img338.imageshack.us/img338/4077/vdm465noblendingwn9.th.jpg (http://img338.imageshack.us/img338/4077/vdm465noblendingwn9.jpg)

Source used is here (http://www.sendspace.com/file/iiq1cc), and the following was the script used:
Mpeg2source("...sample.d2v")
greyscale()
interp=nnedi(-2)
yadifmod(mode=1, edeint=interp)
Mrestore(dclip=last.crop( 6, 78, -8, -82).Spline36Resize(704,384))
crop( 6, 78, -8, -82).Spline36Resize(704,384)
mc_spuds(frames=3,strength=4)
I had another person verify the same behavior, just to make sure that it wasn't just me.

[DarkZell666]

Never mind, forget the above post, I'll survive without these features.

Occasionally the silence is louder than 1000s words.

Apologies bosses if I have bothered you...

Cheers

... be patient will you ? ;)

[talen9]

@unskinnyboy: perharps you had more than one script, open (and previewed) in AvsP, which uses MVTools? Perhaps two very similar versions of this very script?

If you use the same value for the "idx" parameter, this can give problems to the output, 'cause it makes the two different elaborations share several memory areas or such (i'm not so keen in MVTools internals); anyway, that's a behaviour which I already happened to see.

When it happens, try to visualize the preview; then, right click on the preview area and "Release all videos from memory" should fix it. Your next step should be to differentiate the "idx" values in different scripts.

[elguaxo]

@talen9: only this one script was open.

[unskinnyboy]

@unskinnyboy: perharps you had more than one script, open (and previewed) in AvsP, which uses MVTools? Perhaps two very similar versions of this very script?

If you use the same value for the "idx" parameter, this can give problems to the output, 'cause it makes the two different elaborations share several memory areas or such (i'm not so keen in MVTools internals); anyway, that's a behaviour which I already happened to see.

When it happens, try to visualize the preview; then, right click on the preview area and "Release all videos from memory" should fix it. Your next step should be to differentiate the "idx" values in different scripts.Like elguaxo said above, only this one script was open and I wasn't using MVTools either. And FYI, he was the one whom I asked to double-check this issue for me, just to make sure that I wasn't seeing things.

[talen9]

Hmm very strange then, 'cause to me it seems exactly the kind of artifact that a misuse of MVTools/idx can produce.

BTW, your script calls mc_spuds, which in turn heavily relies on MVTools ... OTOH, if it was some obscure bug with it, it should appear on VDub too :confused:

[foxyshadis]

What does the source frame look like at that point? It's likely avsp is resetting mrestore's context and vdub isn't, or vice versa, since conditional functions really aren't designed to skip around in.

[unskinnyboy]

BTW, your script calls mc_spuds, which in turn heavily relies on MVTools ... Indeed. I totally forgot about MC_Spuds using MVTools internally. But still, like you said it doesn't affect VirtualDubMoD.

What does the source frame look like at that point? It's likely avsp is resetting mrestore's context and vdub isn't, or vice versa, since conditional functions really aren't designed to skip around in.I believe this is it:

http://img176.imageshack.us/img176/9131/sourcesameframeax1.th.jpg (http://img176.imageshack.us/img176/9131/sourcesameframeax1.jpg)

Frame number scripted on the frame is a bit off, of course, since this hasn't been restored yet.

Was any of you able to replicate the issue?

[Rippraff]

@unskinnyboy

In post 162 you're comparing different frames.

Cu Rippraff

[unskinnyboy]

@unskinnyboy

In post 162 you're comparing different frames.

Cu RippraffNo, I'm not. It looks different because of the problem I mentioned. To avoid manual error is why I have current_frame subtitled on the frame.

[Vesi]

I'm getting this type of error:
Traceback (most recent call last):
File "AvsP.pyo", line 6291, in OnSliderReleased
File "AvsP.pyo", line 8925, in ShowVideoFrame
File "AvsP.pyo", line 9467, in PaintAVIFrame
File "pyavs.pyo", line 322, in DrawFrame
File "pyavs.pyo", line 301, in _GetFrame
File "avisynth.pyo", line 277, in GetFrame
WindowsError: exception: access violation reading 0x3F0E39DA

And also when I work with my script Avsp sadanly get close and also few other error i get during working. I'm facing these things for the first time and don't know what is the reason.

[Zarxrax]

Hmmm, I'm wondering if this project is dead or just on hiatus or something? I haven't seen any response from qwerpoi in quite a while now.

[bidmead]

It won't be dead in any meaningful sense, because everybody's got the source code. qwerpoi's done a lovely job setting this up, so I guess it's there for anyone who feels like picking it up and running with it. But obviously it would be courteous to get in touch with him first <qwerpoi.avsp@gmail.com>

--
Chris

[cantonesejim]

AvsP wont run any longer on my system heres what the log file says with a freshly downloaded and unpacked copy

Traceback (most recent call last):
File "run.py", line 5, in <module>
File "AvsP.pyo", line 73, in <module>
File "pyavs.pyo", line 31, in <module>
File "avisynth.pyo", line 4, in <module>
File "ctypes\__init__.pyo", line 415, in __getattr__
File "ctypes\__init__.pyo", line 340, in __init__
WindowsError: [Error 126] The specified module could not be found

[Adub]

What have you done recently on yours system? Have you done a fresh install? Or have you removed any software recently?

[Dreassica]

Woudl it be possible to write framenumbers or ranges to a designated txt file in combination with a python script?
I'm trying to filter single frame flashes that usually get missed by mvdegrain using conditional reader/writer.
That works fine, but it's kind of a hassle to have to switch to a text file to write down the framenumbers.
It would be handy if i could do it with the press of a button in avsP, via assigning a hotkey to a python script addon.

[Zarxrax]

You can get the current frame number by using avsp.GetFrameNumber()
I don't know enough about python to tell you how to output that to a file though. I'm sure it's possible, however.

You can use the following macro to output the framenumber to a window:
frame = avsp.GetFrameNumber()
avsp.WriteToScrap('\n%i' % frame)

Then when you are done, just select it and then copy paste to a file.

[MoUCLAs]

hello, i am trying to compare my source and the encoded mkv. I open both with avsp but when i choise on specific frame i get different frames to each of them.. Also whan i press next frame it goes 3-4 frames not 1. Why is happening this?

[nurbs]

What command do you use to load the videos? DirectShowSource isn't necessarily frame accurate.

[mikeytown2]

Try one of these filters to load your vid
http://avisynth.org/mediawiki/External_filters#Source_Filters

It's where I get alternatives to DirectShowSource() (http://avisynth.org/mediawiki/DirectShowSource)

[MoUCLAs]

for the mkv i use:
DirectShowSource("Encode_Video.mkv")

and for the source i use:
MPEG2Source("Source_video.d2v")
LoadCplugin("yadif.dll") Yadif()
Crop(2,16,-2,-16)
Spline36Resize(640,336)

What else can i use?

edit: I saw now the second post. I will look one for the mkv. thanks

edit2: ok it worked. I use:
FFmpegSource ("encoded_video.mkv")
It just needed to add in the avisynth plugin directory the FFMpegSource.dll , downloaded from here:
http://ivtc.somestuff.org/new/
Thanks boyz once more

[nurbs]

FFMpegSource is frame accurate on mkv.

[torpare]

I tried out using VirtualDub as external player for previewing, but decided I preferred the internal preview. There seemed no way to revert other than by retracing my steps (ie removing the path to VDub from the box labelled 'External player' in 'Program Settings -> General tab'). So that's what I did.

Ever since, every time, a blank preview window comes up. I can't see any way to restore the internal preview. No doubt I'm missing something obvious but it isn't obvious to me!:confused:

Can anybody please tell me what I should be doing that I'm not?

EDIT
Please ignore this (I have a lot to learn).

[flebber]

I cannot seem to start avsp 2.02. I receive an error when I attempt to run avsp.exe.

Traceback (most recent call last):
File "run.py", line 1, in <module>
ImportError: No module named os

I am on windows xp sp3. Anyone had this before ?

[Sagekilla]

Sounds like something went wrong with your installation. Only reason that (should) pop up is that, exactly as the error message says, a module is missing. Try downloading it again and copying over the lib folder.

[flebber]

I have done that with no luck. Should it matter that I have a current install of python 2.6.1 on my pc ?

[Sagekilla]

Your installation of python shouldn't matter at all, since (I'm not 100% sure on this yet) the working python code should be run out of /lib/ and no where else. I have Python 2.5.2 though, so I could be wrong.

[tekNerd]

Hey, guys.
Mine worked just fine. I installed python3 for something and now it doesn't work anymore. I mean, it starts as a process, but there is nothing else, plus this process has no threads.
After all, does the python installation has anything to do with it?
(Avs 1.3.4 does work)

Strike that. I reinstalled avisynth and works now. Bo!

[Leak]

I installed python3 for something and now it doesn't work anymore. I mean, it starts as a process, but there is nothing else, plus this process has no threads.
After all, does the python installation has anything to do with it?
To quote Wikipedia...

Version 3.0 (http://en.wikipedia.org/wiki/Python_(programming_language)#Version_3.0)

Python 3.0 (also called "Python 3000" or "Py3K") was designed to break backwards compatibility with the 2.x series in order to repair perceived flaws in the language. The guiding principle was to "reduce feature duplication by removing old ways of doing things".

[Sagekilla]

@Leak: That shouldn't matter though since if AvsP was compiled in a similar way to how I compile my programs, then the modules needed to run should be entirely self contained within the AvsP directory. Nothing outside of that should be loaded or required.

[Zarxrax]

I was wondering if anyone has kept a list of updated filter definitions beyond those that are included in avsp?

[spyOne]

Hi

I dont get image in avsp when loading a m2ts file, only thing i get is a black screen and no error's.

tried graph file (.grf)
tried DGAVCDecode (.dga)
and tried also DirectShowSource("C:\.....\00057.m2ts")

No error's just a black screen, also refreshed image and no luck.

Using avisynth 255.

[spyOne]

Hi

I dont get image in avsp when loading a m2ts file, only thing i get is a black screen and no error's.

tried graph file (.grf)
tried DGAVCDecode (.dga)
and tried also DirectShowSource("C:\.....\00057.m2ts")

No error's just a black screen, also refreshed image and no luck.

Using avisynth 255.
Solved, changed avisynth version from 255 to 258.:)

[MADAJ]

anybody knows why I get this error?
although just a few plugins are installed
it crushed when I added medianblur.dll
and I REALLY need this dll

Traceback (most recent call last):
File "AvsP.pyo", line 5804, in OnMenuVideoRefresh
File "AvsP.pyo", line 8822, in ShowVideoFrame
File "AvsP.pyo", line 9256, in UpdateScriptAVI
File "pyavs.pyo", line 186, in __init__
File "avisynth.pyo", line 117, in Invoke
WindowsError: exception: access violation reading 0x6D65746F

[fvisagie]

Hi Authors and All,

Last night I dipped my toes in AvsP and like everyone else I am very much impressed. Well done!

There are a couple of things I haven't figured out yet that I was hoping you could help me with, please.

Let's say the scenario is I'm working on a large amount of holiday footage, meaning I'll be discarding most material.

1. I work through the footage, using the Trim editor to add Trims to Keep. After a while my time for today is up and I need to stop and save my project. So I click Apply and save. Now the video preview shows the results of my editing. How do I change the video preview's context back to the full source to resume editing?

2. After a while I've collected many Trims and need to refresh my memory by previewing a particular one. How do I do that?

3. Especially with holiday and documentary movies much of the footage contains off-camera speech (narration usually). I found using an external player for audio and then transferring frame references back to AvsP to be very cumbersome and inaccurate. What easy way is there of reviewing audio in AvsP?

If what I'm hoping for isn't possible in AvsP, please be kind enough to suggest an alternative workflow that will work for this kind of scenario?

Many thanks,
Francois

[Gavino]

1. I work through the footage, using the Trim editor to add Trims to Keep. After a while my time for today is up and I need to stop and save my project. So I click Apply and save. Now the video preview shows the results of my editing. How do I change the video preview's context back to the full source to resume editing?
Just comment out the Trim line added by the editor.
2. After a while I've collected many Trims and need to refresh my memory by previewing a particular one. How do I do that?
Not sure exactly what you mean, but you can perhaps use the bookmarks feature for this.
Note that you can always temporarily alter the result of the script by commenting out parts or adding a 'return' at some intermediate point.
3. ... I found using an external player for audio and then transferring frame references back to AvsP to be very cumbersome and inaccurate. What easy way is there of reviewing audio in AvsP?
The AudioGraph (http://avisynth.org.ru/docs/english/externalfilters/audiograph.htm) plugin can help with correlating the sound track with frames.

[fvisagie]

Not sure exactly what you mean, but you can perhaps use the bookmarks feature for this.
Note that you can always temporarily alter the result of the script by commenting out parts or adding a 'return' at some intermediate point.

Say the constructed timeline looks like this:

a1 = OpenDMLSource(...)
a2 = OpenDMLSource(...)
...

clip1 = a1.Trim(...)
clip2 = a2.Trim(...)
clip3 = a1.Trim(...)
...

return clip1 + clip3 + clip2 ...


After a while there's a long list of Trims and I can no longer remember exactly which Trim in the return statement returns which footage.

What is the quickest way of sampling in random order the contents of individual Trims/clips? E.g. sample clip3's contents, sample the contents of clip1, sample the contents of clip2 and now you have a good idea of that part of the timeline?

The AudioGraph plugin can help with correlating the sound track with frames.

Although that helps somewhat, it doesn't provide the complete answer for my footage, unfortunately. I use Aegisub's spectrograph when transcribing speech and although that helps, background noise in most footage is sufficiently white noise-like to make it impossible to visually distinguish speech.

It seems the only solution really is to have audio playback - preferably with video - in AvsP so that I can insert Trim points at the various cursor positions. Can some AVS user function do this, and what would be its basics?

[fvisagie]

To recap, the footage I work with contains lots of off-camera speech which needs to be considered when cutting.

Using an external player for rendering audio and then manually transferring frame references back to AvsP turned out to be very cumbersome and inaccurate.

Also, background noise usually makes it impossible to visually distinguish speech with e.g. AudioGraph.

Therefore I'm left unable to work with audio in AvsP and it seems I won't be able to continue working with it, much to my regret.

However, I thought my experiences might be of some value to the developer(s), hence this feature suggestion for adding audio rendering, preferably with synchronised video playback.

Kind regards,
Francois

[AlanHK]

Therefore I'm left unable to work with audio in AvsP and it seems I won't be able to continue working with it, much to my regret.

AvsP can't play video, let alone audio -- it just displays the results frame by frame. To add that would be a considerable job, and seems unlikely. (Has development stalled, by the way?)

I might suggest VirtualDubMod, which incorporates a simple AVS editor (one of the mods to the straight VirtualDub). This can insert Trims in the AVS for selected ranges.

Or set VirtualDubMod to be AvsP's "external player" and use both.