So, I'm testing my new computer's capabilities and I decided to do this by finishing the first half-hour of ESB v2.0 and trying to render it the same way I would an AVCHD version - this led me to two realizations:
A) My new computer rocks!
B) I tried rendering the first half-hour of ESB v2.0 with the same settings I used for the v2.1 AVCHD of SW and the quality of the AVCHD encode isn't very good compared to the results I was getting for the SW AVCHD, so this means I'll need to create a new set of settings for ESB's rendering. As I'm not very knowledgeable about such things, I could use some help with that - this is the script I used (written for me by Jan, who I'm not sure is still around here but I sure owe that guy for helping me immensely back when I was rendering SW v2.0):
"K:\\FINAL ENCODE\x264.exe" --pass 1 --bitrate 6500 --bluray-compat --level 4.1 --preset slow --keyint 48 --sar 1:1 --slices 4 --vbv-bufsize 15000 --vbv-maxrate 15000 --open-gop --weightp 0 --colorprim bt709 --transfer bt709 --colormatrix bt709 --me umh --merange 64 --deblock -1:-1 --qcomp 0.8 --stats ".stats" --output NUL "K:\FINAL ENCODE\ESB WP 1.avs
"K:\\FINAL ENCODE\x264.exe" --pass 2 --bitrate 6500 --bluray-compat --level 4.1 --preset slow --keyint 48 --sar 1:1 --slices 4 --vbv-bufsize 15000 --vbv-maxrate 15000 --open-gop --weightp 0 --colorprim bt709 --transfer bt709 --colormatrix bt709 --me umh --merange 64 --deblock -1:-1 --qcomp 0.8 --stats ".stats" --output "G:\ESB_WPfile1.264" "K:\FINAL ENCODE\ESB WP 1.avs"
PAUSE
And here are some PNG screens showing the difference between the original lossless render from AE and the x264 encode made from it using the above script. The screens marked A are from the lossless render and those marked B are from the x264 encode. I uspcaled them to 1920x1080 because the difference is much more visible in full-screen 1080p. It is also much more intrusive in movement - any suggestions on how to alter the script to make it better? It doesn't matter if it makes the rendering slower - with the new CPU it should run quite fast anyway :-)
After I have the new script ready, I'm going to test it by rendering the whole first half of ESB v2.0 (which should be done soon) and post it as a workprint.