On Sat, 3 Jun 2006, E.Chalaron wrote:
> Got some troubles here with probably y4mstabilizer
Oh, you've got a problem but I don't think it's with y4mstabilizer ;)
> However, things are getting tricky because when I use stream of a picture of
> 1244x934 or 1008x764 , guess what ... it works :-)
And that's the type of problem you've had in the past as I recall.
Certain sizes would cause y4mscaler to compute scaling factors which
overflowed and caused the program to abort. That would break the
pipeline and you'd have things "mysteriously" come to a grinding
halt.
> BTW frames are square pixels and prog.
> Any ideas ?? thanks a lot
This is for displaying on a computer from the looks of it - can't
be putting 1024x768 on a DVD ;)
> #!/bin/bash
> reel_name=$1
> lum=$2
> (find . -name \* -type f | sort | xargs cat) | yuyvtoy4m -w 1116 -h 832 -a 1:1
Ok - I'll ask, why "-name \*"?
find . -type f | sort | xargs cat
will work just fine.
> -i p -r 17:1 | y4mscaler -v 1 -O chromass=444 | y4mstabilizer -a 0.7 | yuvd
> enoise | y4munsharp | yuvcorrect -v 0 -Y LUMINANCE_"$lum"_16_235_16_235 |
> y4mscaler -v 0 -O size=1024x768 -O chromass=422 | y4mshift -b 8,8,1008,752 |
When having problems it's a GOOD idea to turn on as much logging
as possible. 'y4mscaler -v 0' is disabling the logging from the 2nd
y4mscaler. Not sure if the extra info will help but it can't hurt ;)
> INFO: [y4munsharp] Luma amount: 0.300000
> INFO: [y4munsharp] Luma threshold: 4
> INFO: [yuvcorrect] yuvcorrect 1.9.0 (16-11-2002) is a general image
> correction utility for yuv frames
> INFO: [yuvcorrect] (C) 2002 Xavier Biquard <[EMAIL PROTECTED]>, yuvcorrect
> -h for usage, or man yuvcorrect
> INFO: [y4mscaler] Frame number 1
> xargs: cat: terminated by signal 13
THERE is the explicit symptom of the problem. 'cat' is getting a
SIGPIPE because something later in the pipeline is exiting and
breaking the pipe.
What version (y4mscaler -V) of y4mscaler are you using? I thought
Matt had changed the scaling calculations to use either floating point
or 64bit arithmetic to get around the problem you'd encountered in the
past.
The other possibility is that with lengthy pipelines it's easy to
introduce a syntax error. It's easy to, especially when cutting&pasting
to accidentally create an incorrect option to a command, leave out a
| or other error. Can be hard to find - good indentation and so
on can help make those errors easier to spot/avoid. That's why I find
something like this:
(find . -type f | sort | xargs cat) | \
yuyvtoy4m -w 1116 -h 832 -a 1:1 -i p -r 17:1 | \
y4mscaler -v 1 -O chromass=444 | \
y4mstabilizer -a 0.7 | yuvdenoise | \
y4munsharp | \
yuvcorrect -v 0 -Y LUMINANCE_"$lum"_16_235_16_235 | \
y4mscaler -v 0 -O size=1024x768 -O chromass=422 | \
y4mshift -b 8,8,1008,752 | yuvflip -H | \
y4mtoqt -o "$reel_name"
preferable to:
(find . -type f | sort | xargs cat) |yuyvtoy4m -w 1116 -h 832 -a 1:1 -i p -r
17:1 | y4mscaler -v 1 -O chromass=444 | y4mstabilizer -a 0.7 | yuvdenoise |
y4munsharp | yuvcorrect -v 0 -Y LUMINANCE_"$lum"_16_235_16_235 | y4mscaler -v 0
-O size=1024x768 -O chromass=422|y4mshift -b 8,8,1008,752 |yuvflip -H | y4mtoqt
-o "$reel_name"
:)
Cheers,
Steven Schultz
_______________________________________________
Mjpeg-users mailing list
Mjpeg-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/mjpeg-users
