RGB in AviSynth (and MeGUI)

[byme]

Is possible use the RGB in the MeGUI without creating artifacts?

The IV12 creates darkened and blur


YV12

http://www.plurkpix.com/pix/7sd.png


RGB

http://www.plurkpix.com/pix/7sf.png

[Guest]

I deleted your OT signature spam. Do not restore it.

[byme]

I deleted your OT signature spam. Do not restore it.

ok, sorry

[byme]

or as an good alternative to megui?

[Didée]

It doesn't matter. You can use RGB on the road to the encoder, but even then, at latest the encoder will use YV12 anyway.

[byme]

then also with Avidemux converts it to YV12?

[kemuri-_9]

afaik, the only program that megui has that can work with non YV12 colorspace is ffmpeg,
but most encoders (even within ffmpeg) use YV12 because of the amount of raw data it has compared to RGB makes YV12 more compressible.

if you're looking to compress it well, then convert it to YV12 within avisynth (may need to specify a matrix to keep the colors right).
if you're just looking to compress it compared to the raw format, then LAGS (lagarith) or FFV1 work for RGB.

[byme]

ffmpeg do not understand how it works :(

I understand few English and I have some difficulty

...

sorry if I go off topic.....

I would like to use this logo

http://www.plurkpix.com/pix/7tN.png

PSD http://www.psdnow.com/download/file/1tmGSosOEH/


the result is enough disastrous

http://www.plurkpix.com/pix/7tK.png


can someone help me?

[poisondeathray]

I don't understand what you are trying to do? Use that image and make it into video sequence using x264 and MeGUI?

What script are you using? ImageSource() ?

[byme]

I don't understand what you are trying to do? Use that image and make it into video sequence using x264 and MeGUI?

What script are you using? ImageSource() ?
I use

VD_Logo(filename="C:\edit\temp\logo",x=8,y=8,alpha=255,transparent=true,xr=0,xg=0,xb=255,tolerance=127)

[byme]

which GUI recommended to use for ffmpeg?

I tried AutoFF but does not work

I tried MEncoder but I will not meet

[poisondeathray]

One option: you can use avidemux's logo filter to superimpose a logo from a .png or .bmp image, and encode to x264

e.g. TDK apple movie trailer:

http://i38.tinypic.com/r8i9n6.png

[Guest]

VD_Logo(filename="C:\edit\temp\logo",x=8,y=8,alpha=255,transparent=true,xr=0,xg=0,xb=255,tolerance=127) Is your logo file a 24-bit BMP as required for this filter? If so, what is your complete script?

[byme]

Is your logo file a 24-bit BMP as required for this filter? If so, what is your complete script?

the script fo MeGUI is


Import("C:\edit\VD_Logo.avsi")
LoadPlugin("C:\edit\gradfun2db.dll")
Import("C:\edit\GradFunkMirror.avs")

####

AviSource("C:\edit\video.avi")
ConvertToRGB()
VD_Logo(filename="C:\edit\temp\logo.bmp",x=8,y=8,alpha=255,transparent=true,xr=0,xg=0,xb=255,tolerance=127)
ConvertToYV12()
GradFunkMirror()


why?


however I solved the problem of logo (apart the problem of the YV12)

This is the result

http://www.plurkpix.com/pix/7xO.png

[Guest]

why? Why what?

Post a link to the logo.bmp file and I'll figure out what the problem is.

[byme]

Why what?

Post a link to the logo.bmp file and I'll figure out what the problem is.

you are right :p


.. Now I thought the problem with ffmpeg

[Guest]

I asked you for two things and you didn't provide either of them.

And I don't know why you say I am right, because I did not make any assertions in my post.

I'm hearing the Twilight Zone music in my head.

[byme]

Excuse me, maybe I have not explained well

I do not speak English very well

[Guest]

Will you give me the thing I need to help you, that is, a link to the logo.bmp file?

Or have you solved your problem and don't need my help now?

[byme]

I solved the problem of the logo but not the problem of the YV12 and ffmpeg

[Guest]

You haven't told us about these new problems. What is your ffmpeg command line?

[byme]

I haven't a command line..... I did not understand how ffmpeg :o

[Guest]

OK. Then what is the YV12 problem that you still have?

[byme]

that I wrote on the first page

the IV12 creates darkened and blur

.. you can explain how ffmpeg?

kemuri-_9 writes:

the only program that megui has that can work with non YV12 colorspace is ffmpeg

[Guest]

You never told us how you made the result in the first post!

Tell us exactly what you did to make that happen.

This thread is beginning to resemble a Monty Python sketch.

[byme]

aaargh! as I never told us?



Import("C:\edit\VD_Logo.avsi")
LoadPlugin("C:\edit\gradfun2db.dll")
Import("C:\edit\GradFunkMirror.avs")

####

AviSource("C:\edit\Nuova cartella\_All Music_rgb.avi")
ConvertToRGB32()
VD_Logo(filename="C:\edit\temp\logo.bmp",x=8,y=8,alpha=255,transparent=true,xr=0,xg=0,xb=255,tolerance=127)
ConvertToYV12()
GradFunkMirror()

http://img183.imageshack.us/img183/5638/logolilladx8.png (http://imageshack.us)

http://www.weepic.com/out.php/t3202_Snap1.png (http://www.weepic.com/out.php/i3202_Snap1.png)

[Guest]

What happens if you leave off the GradFunkMirror() line?

One more time, can you give me the logo.bmp file? How many times do I have to ask?

[byme]

GradFunkMirror use this to remove the imperfections of color created by YV12

the logo is always that

http://img183.imageshack.us/img183/5638/logolilladx8.png (http://imageshack.us)

not there?

but this logo does not interest me most

the logos that now using are these

http://www.picpanda.com/images/62asoymsj2bphenjq0o.bmp (http://www.picpanda.com/)

http://www.picpanda.com/images/de1nl0ajolo05dog26zg.bmp (http://www.picpanda.com/)

[Guest]

I asked what happens when you do it without GradFunkMirror! You did not answer me. The point is that GradFunkMirror may be the cause of your problem.

[byme]

no, not GradFunkMirror to create the problem

the result is the same with or without GradFunkMirror

http://img183.imageshack.us/img183/5638/logolilladx8.png

http://www.picpanda.com/images/62asoymsj2bphenjq0o.bmp

http://www.picpanda.com/images/de1nl0ajolo05dog26zg.bmp

[Guest]

I don't see any problem in the images you just posted. Are you jerking my chain?

This works fine for me (don't import anything because I put them in the code below, use the script as is):

loadplugin("gradfun2db.dll")

global VD_plugins = "d:\Virtualdub 1.8.6\plugins"

function VD_Logo(clip clip, int "x", int "y",
\ int "alpha", bool "transparent",
\ int "xr", int "xg", int "xb",
\ int "tolerance", string "filename",
\ bool "animate", int "start", int "duration", int "loops",
\ int "fadeinlen", int "fadeoutend", int "fadeoutlen")
{
LoadVirtualdubPlugin(VD_plugins+"\logo.vdf", "_VD_Logo", 0)
return clip._VD_Logo(default(x,5), default(y,5),
\ default(alpha,128), default(transparent,true)?1:0,
\ default(xr,0), default(xg,0), default(xb,0),
\ default(tolerance,10),
\ default(filename,VD_plugins+"\logo.bmp"),
\ default(animate,false)?1:0,
\ default(start,0), default(duration,0), default(loops,0),
\ default(fadeinlen,20), default(fadeoutend,300),
\ default(fadeoutlen,20))
}

Function GradFunkMirror( clip c, float "strength" )
{
strength = default( strength, 1.2 )
w = c.width()
h = c.height()
vflip = c.FlipVertical()
hflip = c.FlipHorizontal()

stackhorizontal( hflip.crop( w-16, 0, 16, h ).addborders( 0, 16, 0, 16 ),
\ stackvertical( vflip.crop( 0, h-16, w, 16 ), c, vflip.crop( 0, 0, w, 16 ) ),
\ hflip.crop( 0, 0, 16, h ).addborders( 0, 16, 0, 16 ) )
gradfun2db( strength )
crop( 16, 16, -16, -16 )

Return last
}

BlankClip(color=$808080)
ConvertToRGB32()
VD_Logo(filename="e:\tmp\logo.bmp",x=8,y=8,alpha=255,transparent=true,xr=0,xg=0,xb=255,tolerance=127)
ConvertToYV12()
GradFunkMirror() Replace BlankClip() with your AVISource() line.

[foxyshadis]

He's stressing about the very faint chroma blurring that ConvertToYV12 causes; it washes out the bright white and black edges somewhat. Sorry, byme, there is no cure. Besides, any encoding will only blur them even more, so it isn't worth worrying.

Use translate.google.com if you have to, or people might not be able to help you.

[Guest]

He's stressing about the very faint chroma blurring that ConvertToYV12 causes; it washes out the bright white and black edges somewhat.
I see no such effect on the output from the script I posted above. Can you be specific about the specific problem he is now talking about?

[byme]

I don't see any problem in the images you just posted. Are you jerking my chain?

This works fine for me (don't import anything because I put them in the code below, use the script as is):

loadplugin("gradfun2db.dll")

global VD_plugins = "d:\Virtualdub 1.8.6\plugins"

function VD_Logo(clip clip, int "x", int "y",
\ int "alpha", bool "transparent",
\ int "xr", int "xg", int "xb",
\ int "tolerance", string "filename",
\ bool "animate", int "start", int "duration", int "loops",
\ int "fadeinlen", int "fadeoutend", int "fadeoutlen")
{
LoadVirtualdubPlugin(VD_plugins+"\logo.vdf", "_VD_Logo", 0)
return clip._VD_Logo(default(x,5), default(y,5),
\ default(alpha,128), default(transparent,true)?1:0,
\ default(xr,0), default(xg,0), default(xb,0),
\ default(tolerance,10),
\ default(filename,VD_plugins+"\logo.bmp"),
\ default(animate,false)?1:0,
\ default(start,0), default(duration,0), default(loops,0),
\ default(fadeinlen,20), default(fadeoutend,300),
\ default(fadeoutlen,20))
}

Function GradFunkMirror( clip c, float "strength" )
{
strength = default( strength, 1.2 )
w = c.width()
h = c.height()
vflip = c.FlipVertical()
hflip = c.FlipHorizontal()

stackhorizontal( hflip.crop( w-16, 0, 16, h ).addborders( 0, 16, 0, 16 ),
\ stackvertical( vflip.crop( 0, h-16, w, 16 ), c, vflip.crop( 0, 0, w, 16 ) ),
\ hflip.crop( 0, 0, 16, h ).addborders( 0, 16, 0, 16 ) )
gradfun2db( strength )
crop( 16, 16, -16, -16 )

Return last
}

BlankClip(color=$808080)
ConvertToRGB32()
VD_Logo(filename="e:\tmp\logo.bmp",x=8,y=8,alpha=255,transparent=true,xr=0,xg=0,xb=255,tolerance=127)
ConvertToYV12()
GradFunkMirror() Replace BlankClip() with your AVISource() line.

the result does not change

thanks anyway for the interest

[Guest]

What result are you talking about?

[EuropeanMan]

Foxyshadis is right...there is deviation in the form of blurring in chroma. That's the result byme is referring to. He/she wants to eliminate that in the conversion process to the colourspace.

Now he/she knows it is not possible. It would have been better IMHO to get to lag, then feed that to megui and it gets converted to yv12 automatically...instead of doing it all at once to megui...again, that's just MY opinion.