videoFred
20th May 2005, 10:10
I'm not an expert in colorspace, and need good advice.
Lets begin with my camera: output of this machine vision camera is YUV(4:2:2).
Progressive and uncompressed, as far as I can see.
With the fire wire connection, I capture direct to harddisc.
As a codec for these AVI files, I use Huffyuv.
But how do I configure Huffy to begin with?
Huffy converts to YUY2, right?
Then, I load the Huffyuv files into my Avisynth script.
Here is my filter chain:
1) MVTools MVDenoise: requires YV12
2) Depan: requires YV12
3) MSharpen: works with RGB32, YUY12, YV12
4) Cropping and resizing
5) HDRAGC: requires RGB
6) Save the file again with Huffyuv: back to YUY2 again, right?
So my file is converted 4 times! from the original YUV to YUY2-YV12-RGB then back again to YUY2.
My questions are:
1) My source is 1024 x768, progressive. But at the point I use the HDRAGC filter, there is already a resizing 720x576. Should I use ConvertToRGB32("rec709", false)?
2) At the end of the script, should I use ConvertBackToYUY2 or should I leave this to Huffyuv?
3) How do I configure Huffyuv for capturing and at the end of the script, in my case?
My end target is Mpeg2, for DVD use.
4) Any other suggestions to improve quality?
Fred.
Lets begin with my camera: output of this machine vision camera is YUV(4:2:2).
Progressive and uncompressed, as far as I can see.
With the fire wire connection, I capture direct to harddisc.
As a codec for these AVI files, I use Huffyuv.
But how do I configure Huffy to begin with?
Huffy converts to YUY2, right?
Then, I load the Huffyuv files into my Avisynth script.
Here is my filter chain:
1) MVTools MVDenoise: requires YV12
2) Depan: requires YV12
3) MSharpen: works with RGB32, YUY12, YV12
4) Cropping and resizing
5) HDRAGC: requires RGB
6) Save the file again with Huffyuv: back to YUY2 again, right?
So my file is converted 4 times! from the original YUV to YUY2-YV12-RGB then back again to YUY2.
My questions are:
1) My source is 1024 x768, progressive. But at the point I use the HDRAGC filter, there is already a resizing 720x576. Should I use ConvertToRGB32("rec709", false)?
2) At the end of the script, should I use ConvertBackToYUY2 or should I leave this to Huffyuv?
3) How do I configure Huffyuv for capturing and at the end of the script, in my case?
My end target is Mpeg2, for DVD use.
4) Any other suggestions to improve quality?
Fred.