Log in

View Full Version : 2 pass question


beautifuljesus
23rd December 2001, 11:52
Can anyone one tell me if this is a waste of time or not.

I always use 2 pass mode when ripping movies, but because of time I always have to run overnight. I set up Gordian Knot to do a 2 pass encode in Div 4. I set up my resolutions, cropping, bitrates etc then I just compress the video (I prefer to do the audio myself as I usually end up with sync problems when splitting a large movie if I let Gordian Knot handle the audio itself). When I open thee dvd2avi file and hit save and encode I always check the "both passes" option. From the docs that I have read there seems to be an awful lot of checking and tweaking to be done between rips. In the morning, GKnot tells me it took 6 hours and that both passes were completed normally. Now, my question is, is there any difference in quality between a one-pass encode and a two-pass encode when I don't touch any settings in between the passes. Does GKnot automatically calculate what it needs to complete a better quality 2nd pass? Or is it just doing a 2nd pass without really changing/improving anything?
The same question applies to EasyDivX. Yet i think it may be working because

1. in EasyDivx there is nothing else to do, just let it get on with it and finish.
2. The movies I encode on 1 or 2cds are usually very good quality.
3. when I split a movie over 2 cds in virtual dub I had to tinker about a bit as a direct split near the middle (between key frames) caused 2 uneven files (1st was about 520mb and the 2nd was 630mb) or something like this before i had added audio!

Is this to do with constant bitrate vs variable bitrate? I would presume so. But i still find this 2nd pass confusing as there are so many different docs regarding different programs and processing methods.

Some easy to understand explanations would be cool!

Thanks.

soujir0u
23rd December 2001, 12:00
I think Gordian Knot leaves your settings alone. Why don't you try checking the file to see if your original resolution matches the resolution of the final file.

I always encode the first pass and 2nd pass separately so that I can adjust the resolution, resizing type manually and run another first pass for better quality (if needed).