junglemike
13th June 2003, 06:16
Hi everyone.
I am trying to understand influence of first pass size to compressibility.Sorry for my basic questions - i'm pretty newbie here.
I am trying to compress captured material to Xvid and experiencing some quelity problems. Unlike DVD rips, my captured files are very hard to compress - even with bitrabes like 1100-1300 quality is not as good as i compress dvd with 800-900.
Here are my settings: I am using bt848 card and capturing video in 768x576 pal using mjpeg or huffyuv codec. The source is pretty noisy since i have regular cable(not digital)
Xvid(14052003) settings: motions search=6,vhq=4,qpel,chroma,bframes=3/150/100/0,chroma optimizer,all other=default.
So i'm testing some filters and their settings to find out hot to encrease compressiibitlity. And i would appreciate your opinion very much. I apply smart deinterlace filter, some slight denoise filters.
The dilemma is which resize filter to uze? I read some info about different resize methods here on this forum but i still want to try them all and see how they work with my particular source. My dest. resolution is 448x336. I'm doing this way: apply all the filters that i need,and make first pass(only first pass). So each first pass for each resize methods. The file sizes of these passes differ from each other. So if i understand it correctly smaller file size means more compressibility? And final second pass will look better?(with the same settings as first pass)? In general, Does the first pass mean that Xvid is trying to sompress with maximum quality? I will appreciate if anyone will explain these things to me.
Another question about speed: I personaly think that quality is much more important than speed, but , anyway, for some reason speed is really slow, i was wondering if there is some problem in my settings or something. I have Celeron 1700 with 133mhz sdram memory. 4 minute clip takes almost 2 hours to compress (only first pass with below settings) It is really slow. a 2-hour movie would take about 60hours! Isn't it strange?
I am trying to understand influence of first pass size to compressibility.Sorry for my basic questions - i'm pretty newbie here.
I am trying to compress captured material to Xvid and experiencing some quelity problems. Unlike DVD rips, my captured files are very hard to compress - even with bitrabes like 1100-1300 quality is not as good as i compress dvd with 800-900.
Here are my settings: I am using bt848 card and capturing video in 768x576 pal using mjpeg or huffyuv codec. The source is pretty noisy since i have regular cable(not digital)
Xvid(14052003) settings: motions search=6,vhq=4,qpel,chroma,bframes=3/150/100/0,chroma optimizer,all other=default.
So i'm testing some filters and their settings to find out hot to encrease compressiibitlity. And i would appreciate your opinion very much. I apply smart deinterlace filter, some slight denoise filters.
The dilemma is which resize filter to uze? I read some info about different resize methods here on this forum but i still want to try them all and see how they work with my particular source. My dest. resolution is 448x336. I'm doing this way: apply all the filters that i need,and make first pass(only first pass). So each first pass for each resize methods. The file sizes of these passes differ from each other. So if i understand it correctly smaller file size means more compressibility? And final second pass will look better?(with the same settings as first pass)? In general, Does the first pass mean that Xvid is trying to sompress with maximum quality? I will appreciate if anyone will explain these things to me.
Another question about speed: I personaly think that quality is much more important than speed, but , anyway, for some reason speed is really slow, i was wondering if there is some problem in my settings or something. I have Celeron 1700 with 133mhz sdram memory. 4 minute clip takes almost 2 hours to compress (only first pass with below settings) It is really slow. a 2-hour movie would take about 60hours! Isn't it strange?