[boris]

Hi, I 've been looking at some threads on smooth panning and it seems there are a few different opinions on the forum. Most of them handle ntsc material and some of the favourites are 48 (2x film) 72 (1,5 x film), Others try to multiply videorates and combine multiple filmrates(eg 120 Hz 2xntsc 5x film) . I 've been trying to get an optimal result with windvd3 (gforce2mx0 trying 60 ,72,75 hz refreshrates at either 1025z768 and 1280x720 resolutions. At 60 Hz the results suck. Using a long slow pan (intro from the mummy) you can really see a constant judder which is unwatchable. APllying the film theory, I tried playback at 72 Hz: on a PC monitor this didn't look bad, but projecting it revealed stutters now and then. Then I changed over to 75 Hz. Now the pan was really smooth but still not perfect. Every now and then I had a slight stutter (like the 60Hz stuuters) during panning. I hear that theatertek software does very well on panning and zooming. Does anyone here have experience woth both theatertek and PAL resolutions



HTPC: Duron 900, MSI KT7pro, GF2mx, windvd3(dts), powerstrip, SBlive5.1

Barco BD708 (69Khz max)

 

[boris]

My projector can't handle 100 Hz scanrate at the resolutions I am using, the picture gets soft when I use a high resolution at higher scanrates.

Pootle , what resolution do you use at 100Hz?.

I find it strange that your picture quality improved so much from the processorupgrade. Are all other components the same or did you change other components as well? Maybe it's the high scanrates that are a heavy burdon on your system (more to scale)

 

[boris]

Havent tried 50, cause my gforce card can't do anything less than 60 with powersreip . Have to get a radeon for that. It actually doesn't make a difference if you playback on a 20" desktop monitor or a Barco projector. I 'm sure that all the panning problems come from algorithm problems of the decodersoftware icw resolution. When going to higher resolutions, and then not using the same aspect ratio there is a lot of calculating to be done. If the algorith isn't perfect you run into problems. I recently saw the TAW rock at a high resolution and it really was smooth as silk.

The microjudders I am talking about probalbly won't be seen by anybody not looking for them. They are really only taking place for a frame or so and I can't tell for sure where they come from. It happens more while testing the same scene over and over then when just playing the whole dvd. Maybe the memory gets fragmented while testing