View Full Version : How do I use Gknot, when filesize is not important and quality is #1?
QuakerOatz
25th November 2003, 05:40
In the past I've used Gknot for DVD --> 1or2 CDs using Divx 5. Now that I have a DVD burner I'd like to up the quality of my Rips, by picking high quality bitrates and high resolutions (~640x480 for 4:3).
Things I've learned/read:
- I should be using "One pass quality based" on encoding. Q=2=100%?!?!!?
- I should be using VirtualDub.
A few questions arise:
1. Should I even be using Gknot, I've been trying using "Caculate Avi File Size" and specifying bitrate and res?
2. What steps should I go through using VirtualDeb regarding VOB files. Do i have to append all the VOBs? (slow!) and then Mux?
3. What is a nice high quality bitrate/resolution to encode to, if you didnt care too much about file size, but obviously want to compress it a bit from the massive VOBS?
jggimi
25th November 2003, 15:42
Gknot's calculator is designed for multipass encoding to a specified size or specified bitrate. But many people use Gknot as an AviSynth generator. All you need do is click "Save" rather than "Save & Encode" from the Save AVS window or the Edit AVS window.
If you want the video output in DivX format, then a quality based one-pass encoding will provide fixed quantizer encoding. 100% used to mean quantizer 2, in current releases of DivX 5, quantizer 2 = 98% and quantizer 1 is now equivalent to 100%. The DivX 5.1 user guide (http://www.divx.com/support/guides/DivXGuide51.pdf) describes how quantizers are used in the codec.
If it were me attempting this, I would avoid any filtering of the DVD video stream, other than any necessary Inverse Telecining / or Deinterlacing. That means no pre-processing in the DivX codec, and no spatial or temporal filtering in the AviSynth script. I might crop letterboxing, but only on 16 pixel boundaries (8 each on top/bottom or left/right), and I would also eliminate resizing from the .avs script, since I use AVI players that can adjust aspect ratio to 16:9 or 4:3.
Of course, if you really want maximum quality, you wouldn't be transcoding from one lossy MPEG codec to another. Instead, you'd backup 1 DVD-9 to 2 DVD-Rs (http://www.doom9.org/mpg/dvdbackup-guides.htm).
QuakerOatz
25th November 2003, 16:52
Thanks, I think I'm starting to understand the process here, please correct me if I'm wrong:
In a HQ process where file size isn't important,
1) DVD decrypt
2) DVD2AVI to demux audio and create D2V file.
3) Use Gknot to pick crop dimensions and create AVS script. Also encode audio?
4) Open Virtualdub(mod) to open avs file, pick quant.(2), resolution and compression options then compress.
5) Mux with Vdub...
Done?
Thanks!
jggimi
25th November 2003, 17:17
Yep.
vBulletin® v3.8.11, Copyright ©2000-2026, vBulletin Solutions Inc.