Avisynth-remove grain problem and Canopus DV codec problem

[SerbianBoss]

I have two questions which is really annoying.

1. I capture with Canopus ADVC 110 in DV format to computer.

Then i make this script (by Fulcilives)

LoadPlugin("RemoveGrainSSE2.dll")
avisource("D:\CAPTURE\capture.avi")
SeparateFields()
odd=SelectOdd.RemoveGrain(mode=16)
evn=SelectEven.RemoveGrain(mode=16)
Interleave(evn,odd)
Weave()
ConvertToYuY2(interlaced=true)

and when i try to import to CCE than write this:




I have Athlon64 so i use SSE2 version, but i tryed with RemoveGrainS.dll and the situation is the same.
The situation is the same when i line ConvertToYuY2(interlaced=true) remowe at line 3(after avisource).

I copy file RemoveGrainSSE2.dll to system32, system and in Avisynth plugin folder.

Where i wrong?

2. I installed Canopus DV codec that i can load DV avi files in virtualdub mod. But when i try to load some DV avi file it said something about Decoder and i cant load file in virtual dub.

But when i installed Panasonic DV codec i can normally load DV avi files in Virtual dub.

So question is: Can i load DV avi files in virtual dub just with Canopus DV codec?

best regard

[guns1inger]

The error message is being generated by Avisynth, and is tell you you are in the wrong colour space for the RemoveGrain filter. Try adding a ConvertToYV12(interlace=True) after the AviSource call.

Panasonic's DV codec has always been required for Virtualdub. Unless something has changed in the very latest version, this is still the case.

[SerbianBoss]

For my first question i fix it how you say(after avisource) but i add

ConvertToYV12()

Problem for question two is still, any suggestions?

best regard

[FulciLives]

If you want to use the RemoveGrain filter in your script and plan on using the script with CCE then your script should look like this:

LoadPlugin("RemoveGrainSSE2.dll")
avisource("D:\CAPTURE\capture.avi")
ConvertToYV12(interlaced=true)
SeparateFields()
odd=SelectOdd.RemoveGrain(mode=16)
evn=SelectEven.RemoveGrain(mode=16)
Interleave(evn,odd)
Weave()
ConvertToYUY2(interlaced=true)

As for the Canopus/Pansonic DV codec "conflict" if that is what it is ... I have no input on that I am sorry to say.

BTW did you ever get the script to work with Convolution3D? With Convolution3D you only need the ending ConvertToYUY2(interlaced=true) since the source (DV AVI) is already YUY2 so it will work fine with the YUY2 version of Convolution3D (but beware there is an inferior YV12 version of Convolution3D).

From reading the RemoveGrain docs I get the sense that it is best used for non-interlaced or progressive video as opposed to interlaced video although your usage of it for interlaced video is correct but it just seems to be a filter that works "best" with non-interlaced or progressive video.

In short I think Convolution3D is the better choice for interlaced video.

- John "FulciLives" Coleman

[guns1inger]

What exactly was the error message ? "it said something about Decoder" is pretty vague and unhelpful.

I believe the reason you need to use the Panasonic codec is because it is a VFW filter, whereas the others are directshow and cannot be accessed by Virtualdub.

[SerbianBoss]

Regard FulciLives,

I succeed to work with Convolution 3d and the results are excellent.

But my(your) script only work when ConvertToYuY2(interlaced=true) is after avisource

LoadPlugin("Convolution3d.dll")
avisource("D:\CAPTURE\capture.avi")
ConvertToYuY2(interlaced=true)
SeparateFields()
odd=SelectOdd.Convolution3D (1, 32, 128, 16, 64, 10, 0)
evn=SelectEven.Convolution3D (1, 32, 128, 16, 64, 10, 0)
Interleave(evn,odd)
Weave()

When i use this script when is ConvertToYuY2(interlaced=true) is at the end in CCE i get this Error:



Why is that?

And whent the same script (where is ConvertToYuY2(interlaced=true) is at the end) i get this error:



I capture in Win DV type 2 and than make this scripts.

best regard[/b]

[guns1inger]

The only time I have seen the CCE error reported is when a pirated version has been used - you may have been ripped off by an unscrupulous seller.

The script above works because it converts the colour space of the video correctly for the filter that is about to be called. This is why it works when the colour space is converted after the avisource() command but before the Convolution3D call. When you shift the colour space conversion to a point after the filter is called, the filter returns an error because the colour space is incorrect.

Why are you moving the colour space conversion to the end of the script ?

The basic order of things is

Load filters
Load video
Set colourspace for filters
Call filters
Set colourspace for encoder

If possible, use filters that all work in the same colourspace to avoid too many conversions.

edit : a little while later . . .

Google found the following http://forums.afterdawn.com/thread_view.cfm/1/97052 regarding your error i the context of DVD Rebuilder. Q15 and Q16 of the second post might help you find an answer.

[FulciLives]

LoadPlugin("Convolution3d.dll")
avisource("D:\CAPTURE\capture.avi")
ConvertToYuY2(interlaced=true)
SeparateFields()
odd=SelectOdd.Convolution3D (1, 32, 128, 16, 64, 10, 0)
evn=SelectEven.Convolution3D (1, 32, 128, 16, 64, 10, 0)
Interleave(evn,odd)
Weave()

If this works then I would just leave it like this ... in other words don't worry about the fact that you need to put ConvertToYUY2(interlaced=true) so early in the script instead of at the end of the script.

Although I was under the impression that DV AVI is YUY2 but I never really did any real work with the DV AVI format so I could be wrong.

Please note that your Convolution3D settings are rather "strong" meaning you might be destroying "too much" actual image detail along with the "video noise".

I found that for most captures the following Convolution3D setting is good enough:
Convolution3D (0, 6, 10, 6, 8, 2.8, 0)

You can also change the first number from "0" to "1" ... some people think it works "better" if you use "1" instead of "0". So it would look like this:
Convolution3D (1, 6, 10, 6, 8, 2.8, 0)

Of course you can "play" around with the values and find what works "best" with your source as every source is different.

- John "FulciLives" Coleman

[SerbianBoss]

Thanks for answers

@FulciLives

It seems that DV avi is not have YUY2 colorspace.

Convolution filter is excellent and the resault is very nice.

I most using settings 1, 32, 128, 16, 64, 10, 0 because my DV avi have too much noise. And with this settings i clear almost all noise and the video detail is ok.

You said that is no difference when i put ConvertToYuY2(interlaced=true) after avisource. Because when i put at the and it sais about colorspace (either in CCE either in virtual dub)

And one more think,

Because my DV avi files is bottom field in CCE i turn off option Output top field first stream, and in Offset line i type 0. I think that is OK.



best regard

[FulciLives]

And one more think,

Because my DV avi files is bottom field in CCE i turn off option Output top field first stream, and in Offset line i type 0. I think that is OK.



best regard

I have a different version of CCE so my setting is different ... you will have to encode and burn to a DVD and play it back on a TV to see if it is OK or not.

- John "FulciLives" Coleman

[SerbianBoss]

thanks fulcilives for all help.

Users who have CCE SP can tell does my settings about field order is right or not?

I capture with ADVC 110 in DV avi which is bottom field. Than i load Avisynth script in CCE an for parametar for field i set like in this picture:



I disselect Output top field (because DV avi is bottom) and in offset line i write
0. I insert 0 when i read this from CCE manual:

Specify the field order of output stream. If it is selected, an output
stream is top field first. If not, an output is bottom field first. This
setting works on MPEG-2 output.
It is important to set correct field order, because incorrect field
order causes stuttered motion. To set field order correctly, you have
to care about Offset line setting.
• If the field order is same between the source and the output,
Offset line should be 0 or even number.
• If the field order is different between the source and the output,
Offset line should be 1 or odd number.

Is this ok?

thanks[/code]

[SerbianBoss]

any comment?

[guns1inger]

I use 2.70, and set the offset lne to 0 and it works fine.

Honestly though, what does it cost you to encode 5 minutes of footage as a test ?

[SerbianBoss]

I encoded first with 0 and is Ok

But when i encoded as 1 the movement was jerky.

I am in Europe and using Pal Dvd.

I asked because i see in tutorial on videohelp that said, if you are using NTSC you shold set to 0 and if you are using Pal you must set 1.

I think that Pal users must select 0 (bottom field)

[SerbianBoss]

I have just one more question about my avisynt script.

Where in this script is better to put crop() function?

LoadPlugin("Convolution3d.dll")
avisource("D:\CAPTURE\capture.avi")
ConvertToYuY2(interlaced=true)
SeparateFields()
odd=SelectOdd.Convolution3D (1, 32, 128, 16, 64, 10, 0)
evn=SelectEven.Convolution3D (1, 32, 128, 16, 64, 10, 0)
Interleave(evn,odd)
Weave()

Does is better to put before ConvertToYuY2(interlaced=true) or after?
And i think that add border function go at the end?

best regard

[FulciLives]

One school of thought is to crop and add borders after any filters.

This does make things go a bit slower though than if you crop first then filter then add borders.

So crop and add borders last or crop first then filter then add borders.

Here is an example:

LoadPlugin("Convolution3d.dll")
avisource("D:\CAPTURE\capture.avi")
CROP HERE
ConvertToYuY2(interlaced=true)
SeparateFields()
odd=SelectOdd.Convolution3D (1, 32, 128, 16, 64, 10, 0)
evn=SelectEven.Convolution3D (1, 32, 128, 16, 64, 10, 0)
Interleave(evn,odd)
Weave()
ADD BORDERS HERE

That is one way of doing it ... here is the other way:

LoadPlugin("Convolution3d.dll")
avisource("D:\CAPTURE\capture.avi")
ConvertToYuY2(interlaced=true)
SeparateFields()
odd=SelectOdd.Convolution3D (1, 32, 128, 16, 64, 10, 0)
evn=SelectEven.Convolution3D (1, 32, 128, 16, 64, 10, 0)
Interleave(evn,odd)
Weave()
CROP HERE
ADD BORDERS HERE

- John "FulciLives" Coleman

[SerbianBoss]

thanks fulcilives,

What you prefer: first example or second?


When i first crop and add borders at the end (your first example) CCE work little faster 0.50 speed.

But when i dont crop CCE work little slower about 0.47

And one more question.

For crop and for add borders i use this code:

crop(8,4,-8,-12)
AddBorders(8,8,8,8)

Is that OK?

best regard

[FulciLives]

thanks fulcilives,

What you prefer: first example or second?


When i first crop and add borders at the end (your first example) CCE work little faster 0.50 speed.

But when i dont crop CCE work little slower about 0.47

And one more question.

For crop and for add borders i use this code:

crop(8,4,-8,-12)
AddBorders(8,8,8,8)

Is that OK?

best regard

Crop after filter is slower but considered "better"

Here is how you use the Crop command:
Crop(left, top, -right, -bottom)

Here is how you use the AddBorders command:
AddBorders(left, top, right, bottom)

So in your example you are cutting 8 from the right, 4 from the top, 8 from the left and 12 from the bottom.

Then you are adding 8 all around.

Assuming you captured 720x480 and you have "noise" at the bottom you want to crop ... this is how I would do it. Generally cutting 12 from the bottom is enough to remove the video noise you get from videotape sources.

- John "FulciLives" Coleman

[SerbianBoss]

I capture 720x576 because i am in Europe

You thing that this script is better, OK

LoadPlugin("Convolution3d.dll")
avisource("D:\CAPTURE\capture.avi")
ConvertToYuY2(interlaced=true)
SeparateFields()
odd=SelectOdd.Convolution3D (1, 32, 128, 16, 64, 10, 0)
evn=SelectEven.Convolution3D (1, 32, 128, 16, 64, 10, 0)
Interleave(evn,odd)
Weave()
CROP HERE
ADD BORDERS HERE

I just one more question.

When i work frameservering from Virtual dub, where in this script to put
Trim(when i cutting in VD)?

best regard

[FulciLives]

LoadPlugin("Convolution3d.dll")
avisource("D:\CAPTURE\capture.avi")
TRIM HERE
ConvertToYuY2(interlaced=true)
SeparateFields()
odd=SelectOdd.Convolution3D (1, 32, 128, 16, 64, 10, 0)
evn=SelectEven.Convolution3D (1, 32, 128, 16, 64, 10, 0)
Interleave(evn,odd)
Weave()
CROP HERE
ADD BORDERS HERE

For trimming purposes you can use this:

avisource("D:\CAPTURE\capture.avi")

It will be much faster and easier to trim.

- John "FulciLives" Coleman

[SerbianBoss]

OK, this is my finally script

Thanks fulcilives, if i have more questions i asks

best regard

[SerbianBoss]

For trimming purposes you can use this:

Code:
avisource("D:\CAPTURE\capture.avi")

It will be much faster and easier to trim.

I dont understand what you want to say with this?

Does i need just to put TRIM after avisource?

best regard

[FulciLives]

This is your final script right?

LoadPlugin("Convolution3d.dll")
avisource("D:\CAPTURE\capture.avi")
TRIM HERE
ConvertToYuY2(interlaced=true)
SeparateFields()
odd=SelectOdd.Convolution3D (1, 32, 128, 16, 64, 10, 0)
evn=SelectEven.Convolution3D (1, 32, 128, 16, 64, 10, 0)
Interleave(evn,odd)
Weave()
CROP HERE
ADD BORDERS HERE

Of course you need to put in the "real" complete commands for TRIM, CROP and ADDBORDERS but that is script.

All I am saying is that if you want the editing in VirtualDubMod to go quicker you can just use:

avisource("D:\CAPTURE\capture.avi")

Or you can take the original script and "REM OUT" the lines. Anytime you put the "#" in front of a line it is ignored in the script.

So make your script look like this:

#LoadPlugin("Convolution3d.dll")
avisource("D:\CAPTURE\capture.avi")
#ConvertToYuY2(interlaced=true)
#SeparateFields()
#odd=SelectOdd.Convolution3D (1, 32, 128, 16, 64, 10, 0)
#evn=SelectEven.Convolution3D (1, 32, 128, 16, 64, 10, 0)
#Interleave(evn,odd)
#Weave()
#CROP HERE
#ADD BORDERS HERE

The only line the script will process is avisource("D:\CAPTURE\capture.avi") which is all you need to load it into VirtualDubMod and make your edits. After editing import the edits into your script so it is just after the AviSource line then delete all the "#" for the final script.

- John "FulciLives" Coleman

[SerbianBoss]

This is my final script, normally with crop and add borders values:

LoadPlugin("Convolution3d.dll")
avisource("D:\CAPTURE\capture.avi")
Trim
ConvertToYuY2(interlaced=true)
SeparateFields()
odd=SelectOdd.Convolution3D (1, 32, 128, 16, 64, 10, 0)
evn=SelectEven.Convolution3D (1, 32, 128, 16, 64, 10, 0)
Interleave(evn,odd)
Weave()
crop(8,4,-8,-12)
AddBorders(8,8,8,8)

I was doing like this:

This whole script insert in Virtual dub mod, i do some cutting, than in script i put TRIM and than frameservering in CCE.


So, you say that i insert this script in VD than i do cuttind, and when i finish cuttinf than delete all the "#" for the final script. And than do frameservering. Am i write?

#LoadPlugin("Convolution3d.dll")
avisource("D:\CAPTURE\capture.avi")
#ConvertToYuY2(interlaced=true)
#SeparateFields()
#odd=SelectOdd.Convolution3D (1, 32, 128, 16, 64, 10, 0)
#evn=SelectEven.Convolution3D (1, 32, 128, 16, 64, 10, 0)
#Interleave(evn,odd)
#Weave()
#CROP HERE
#ADD BORDERS HERE

[FulciLives]

This is my final script, normally with crop and add borders values:

LoadPlugin("Convolution3d.dll")
avisource("D:\CAPTURE\capture.avi")
Trim
ConvertToYuY2(interlaced=true)
SeparateFields()
odd=SelectOdd.Convolution3D (1, 32, 128, 16, 64, 10, 0)
evn=SelectEven.Convolution3D (1, 32, 128, 16, 64, 10, 0)
Interleave(evn,odd)
Weave()
crop(8,4,-8,-12)
AddBorders(8,8,8,8)

I was doing like this:

This whole script insert in Virtual dub mod, i do some cutting, than in script i put TRIM and than frameservering in CCE.


So, you say that i insert this script in VD than i do cuttind, and when i finish cuttinf than delete all the "#" for the final script. And than do frameservering. Am i write?

#LoadPlugin("Convolution3d.dll")
avisource("D:\CAPTURE\capture.avi")
#ConvertToYuY2(interlaced=true)
#SeparateFields()
#odd=SelectOdd.Convolution3D (1, 32, 128, 16, 64, 10, 0)
#evn=SelectEven.Convolution3D (1, 32, 128, 16, 64, 10, 0)
#Interleave(evn,odd)
#Weave()
#crop(8,4,-8,-12)
#AddBorders(8,8,8,8)

Exactly !!!

VirtualDubMod will be much more responsive without the filter in place or any of the other stuff.

When the script is done you load the script directly into CCE.

- John "FulciLives" Coleman

[SerbianBoss]

Ok

best regard

[SerbianBoss]

I do little different.

I load this scripr in VD.

LoadPlugin("Convolution3d.dll")
avisource("D:\CAPTURE\capture.avi")
ConvertToYuY2(interlaced=true)
SeparateFields()
odd=SelectOdd.Convolution3D (1, 32, 128, 16, 64, 10, 0)
evn=SelectEven.Convolution3D (1, 32, 128, 16, 64, 10, 0)
Interleave(evn,odd)
Weave()
crop(8,4,-8,-12)
AddBorders(8,8,8,8)

And i just do cutting. Then i go Tools-Script editor and click import frameset as trims. And then i copy all the code (now with trims).

Than i close VD, open notepad and paste. Now i have script with trims. And than i load that script in CCE and do encoding.

On that way, i do not use frameservering, just use VD to cut scenes and than i copy TRIMS.

What you think about that way?

[FulciLives]

If you go back and read my original guide that IS how I tell you to do it.

I thought you knew that part already

- John "FulciLives'" Coleman

[SerbianBoss]

Sorry, now i really understand you

I ask you one more question

I have 6 DV avi files(about 13 GB) and i want to join them together as one.

How do that in this script?

LoadPlugin("Convolution3d.dll")
avisource("D:\CAPTURE\capture.avi")
ConvertToYuY2(interlaced=true)
SeparateFields()
odd=SelectOdd.Convolution3D (1, 32, 128, 16, 64, 10, 0)
evn=SelectEven.Convolution3D (1, 32, 128, 16, 64, 10, 0)
Interleave(evn,odd)
Weave()
crop(8,4,-8,-12)
AddBorders(8,8,8,8)

thanks

[FulciLives]

Anytime I have tried to join AVI files in a script I get sound that is out-of-sync.

So I convert to MPEG-2 files with muxed/multiplexed audio then join using MPEG2VCR or VideoRedo

Of course you are working with DV AVI which is a much better format in terms of A/V sync.

If you visit the AviSynth homepage and follow the link called MANUAL you will see a list of commands and how to use them.

http://www.avisynth.org/

However this is the "best" way to do it as far as I know:

Say you have 3 files and they are called, cap1.avi, cap2.avi and cap3.avi
AlignedSplice(AviSource("cap1.avi"),AviSource("cap 2.avi"),AviSource("cap3.avi"))

Good Luck

- John "FulciLives" Coleman