Originally Posted by Aegwyn11
First let me thank you for your efforts on this. Your guide has given me a pretty strong start in understanding the ideas behind all this so that I can customize it to my liking better.
Can you shed some more light? Your answers don't really help clear anything up for me. Details below.
Lip sync - you say its because my CPU is "weak". Its a Core i3 530 using the 'medium' settings. Are you saying that the i3 530 isn't enough to handle this (even though it hovers at 30-50% CPU usage)? If thats the case, perhaps you could post some minimum recommended hardware specs? Another idea would be to post your system specs and CPU usage for each batch of settings you recommend...that could be very helpful as well.
Raw video - Your answer doesn't make any sense to me. Based on what I've been able to find, it looks like the raw video setting allows FFDShow to be used to do post-processing when decoders other than FFDShow are used. If FFDShow is the decoder being used, it appears to me that raw video doesn't need to be checked. In fact, if you're using your system like I am (for both SD and HD content), enabling raw video causes complications (profiles would have to be configured so that HD doesn't get processed like SD, DXVA is disabled, etc). In my case, having raw video disabled and using the FFDShow decoder for Mpeg2 (with AVISynth, resize, etc) seems to work okay to me, I just want to know if I'm losing anything doing it this way.
RGB conversion - The best info I could find on this is that certain post-processing functions can sometimes be done faster in the RGB colorspace. Is this why you prefer RGB? Can you help us understand why you chose to disable everything other than RGB and force the conversion?
Just so you know. I have never, I repeat! NEVER! gotten the medium settings to work properly. The lipsynch is always screwed up for me with LSF config and I have an i5 750 @3.6Ghz. The only real reason to use Avisyth at all is for Motion compensation. Also LSF looks POOR on HDTVs. The standard unsharpmask sharpener in ffdshow works much better for this purpose.
About RGB conversion. I can tell a small difference in colors compared to NV12, but this goes away when I recalibrate duh..
Currently I prefer to use NV12 for everything and not force the RGB color conversion, as it is just not needed and gives extra processing overhead in FFdhsow. With NV12 you can now use EVR-CP renderer and you get vector adaptive de-interlacing for everything and hardware decoding for x264 and VC-1. That is If you have an ATI card or Nvidia with a G92 GPU or newer. That is most people! So, hardware de-interlacing for all material and hardware decoding and de-interlacing for most HD video and pretty much all bluerays.
You can't tell me that interlaced content never crosses your screen. Tons of disk menus are interlaced and most previews on the disks are interlaced. I guess if you don't mind the menus and previews being totally craptastic and unwatchable, then it is no big deal, because the movie it self is progressive. Matter of fact, there is a lot of animated material on DVD that is interlaced. The wife brought one home last week. Donte's Inferno. I didn't want to watch it
But she did. What was I to tell her? I'm sorry sweat heart. I can't fix the ugly lines because then I can't force RGB32. That would have gone over about like a turd in a punch bowl.
Originally Posted by 8:13
my ffdshow is custom built
It also happens to be just about the only one I can find that works properly! I wouldn't have any HTPC satisfaction if it wasn't for you sir. Thank you very much.
Originally Posted by Mnoe
madVR only works with YV12, fwiw.
Noted and fixed.