Hi guys,
I was just wondering about the KVCD specs that Kwag published recently and I'm a bit puzzled about some details:
1 - Kwag mentions "It's also recommended to set "Scene change detection", if the encoder permits." but I have seen on this forum Kwag saying that if u use CQMatic u will have to disable this feature from the encoder. Should it be disabled just for prediction and reenabled by the time u start encoding???
2 - Kwag says "Suggested bitrates are a minimum of 64Kbps and a maximum of 3,000Kbps. For MPEG-1 and MPEG-2 in this bitrate range, a DC component precision of 8 bits is recommended." but when I download the latest KVCDx3 MPEG-2 template it comes configured to 300-2500Kbps and with DC component of 10. Again should this be disregarded when encoding and just used for prediction with CQMatic???

I say this because more than once I've seen Kwag posting in this forum that the bitrate should be set to .57*AvgBit-2000. But I've also seen Kwag saying that if we shoot for a 2 CD encode then we should set the high bitrate to 2500......
3 - Now the most confusing: the KVCDx3 MPEG-2 template comes configured to High quality (slow) motion search precision. Again I've seen Kwag saying that we should use "motion estimate search (fast)". What to do???
Do some of these settings differ if one is using avisynth 2.08 or 2.52?
I'm using 2.08 because it seems very reliable and because I can use MStacker to build it easy. But if someone tells me to use 2.52 because it's faster (which I haven't been able to proove) than 2.08 when using MA because one can set "motion search precision" to "motion estimate search (fast)" and the quality will be the same as in 2.08 but faster then maybe I can start understanding why you guys are all very excited about 2.52.
I think that someone should start some kind of reference KVCD specs when using TMPG with CQM but that's only me thinking...