Single Pass or Two Pass?

I use Xvid to encode movie(.vob), using the Autogk. However, doing a two pass encode need 5-6 hours(movie is two hours length, encoding at 640*480), it is painful. I notice that there is an option "Target Quality" which is using single pass. Will the motion scene of the movie degrade if using single pass? Also, will the final file size terrible? (ie. Will the file size bigger than that of the two pass very much?) Hope to hear from those who has such experience. Thx..

No, the only setbacks of single-pass, constant-quality encodings are - You have no precise control over the output-size; your resulting file may be just a tad too big to burn onto a CD-R. On the other hand you can determin a feasible quantizer to use with a tool called 'Enc' quite precisely enough. - Whenever you aim for a fixed filesize / don't have space in abundance (definition of 'space in abundance': 1 DVD-R per movie ; i.e. you can encode at constant quantizer=2) a two-pass encoding, if done correctly, will yield a slightly higher perceived quality than single-pass. Because a) constant quant undershoots in most cases as you can't afford to overshoot; hence you don't reach the lowest quantization possible for the aimed-at filesize and b) if you use just a little bit of curve-compression (about 10% perhaps) in two-pass you have a good chance that some frames where you don't realize it will be compressed with quant=5 while others where you do realize the diference will be encoded at quant=3. That's theory though and might not really work out that way. As you see; the difference gets all the less important the more space/quality you want to invest into the encoding.

Since I always encode movie which is of intensive action/motion, will single pass result in the badly compressed motion screen? Previously with Divx 5.2.1 that was the experience. I feel ill when seeing those broken screen..

Quality takes time.

Well, for 120 minutes movie(DVD .vob format, most likely 4G size), if using Autogk to encode(Quality 75% which is the default setting, the tutorial said it is Quat 4), what seem to be the output file size? :confused: (Thx jsquare, but I am still struggling with Autogk, GK is so professional)

Yes, everyone who wants to encode at fixed quantizer but still cares about filesize must use Enc or other other means of filesize-prediction. There's no point in asking what quality-setting you should use, miniyoyo.