CompuServe Messages

And to All: I'm getting a better handle on the artifact issue now and how the PAR handles compressing its files for use. Here is a brief explanation for the uninitiated so my evaluation will make more sense. The compression in the PAR is handled two ways, One method for animation renderings, ie. sequential TGA's and a slight variation to this method for real time frame grabbing using the TBC IV. There are two settings that the PAR uses. First, is the Q-factor and is adjustable to a maximum of 23 (highest quality). The other is Block limit which is the largest size allowable for a frame to be which is 508. I don't know how these numbers relate to compression ratio and byte size. The block limit is set for a particular hard drive performance. The faster the hard drive the higher the number allowed. I have been using 240 for a Micropolis 2210A but haven't really experimented with it yet. In animation compression from targa files the Q-factor is optimized for each frame file and will allow the file to grow to the block limit size for maximum quality on a given frame. In real time frame grabbing of video the block limit is also set but the Q-factor is manually fixed to a number for all frames in the grab sequence. The reason for this difference is that with frame grabbing the frames arrive on the PAR at 30 per second and apparently there is not enough time to determine the Q factor for each frame individually. Typically I would set it arbitrarily at 20 and try to record the video sequence. If the q-factor is too high then the PAR will issue an error message advising that the q-factor should be lowered. The lower you go the less quality the video will have. The number varies for each piece of video grabbed so each scene might take several run throughs to optimize the actual number. It would appear that the shorter the length of the grab the higher the Q-factor could be. If one wants to grab several minutes of video footage the Q-factor has to be down around 10 or fairly lossy. I did some 5 second grabs at a q factor of 20 but the same video linked together for 30 seconds required a Q-factor no higher than 16. All with a block limit of 240. What does all this mean? So far my opinion is that for animation renderings that are low in what I will call the "busy factor" which translates to a lot of movement and busy texture maps such as outdoor scenes and cloths etc. the animation can be output to an almost lossless quality. If the busy factor is great then there will be more noticeable pixelization and edge aliasing visable. Chromes and reflections seem fairly low loss but lots of live images maps tend to get lossy due to lower q factors. At this early point in time I say that I still don't have a strong grasp of the block limit settings but an early subjective evaluation is that the overall quality on animation output is better than wh at I anticipated but the frame grabbing is less than what I expected. The frame grabbing is still superior in sharpness and chroma integrety even at Q-factor 16 when compared to the targa+ using the diaquest action animator frame grabbing. A good sharp video picture shot at the beach with beta SP. produced a very nice still shot that I was able to mask the very slight edge aliasing in Photoshop and it was used for a print piece for a client today. He noticed the improved quality on the print piece comparing with my former work with the targa+. I have a bunch of these video stills to print to do and this is just one more way the PAR is improving my product quality. I would say that tinkering with the Q-factor for frame grabbing doesn't take long but is a trial and error process. Kinda like doing test renderings with IPAS settings. What is more disappointing is to achieve higher Q-factors one must grab shorter video sequences. To grab a couple of minutes and then go to work with it is to suffer higher losses than necessary with the PAR system. There are more trade offs to deal with than I anticipated. I hope the above detailed explanation made sense to you non PAR owners but keep in mind I'm still on a day 1 learning curve and just beginning to understand what makes this thing tick myself. My work load is somewhat heavy right now and the PAR just saved me 4-5 hours of time today with the real time frame grabbing job. Not to mention the client's excitement when I presented him the bill at less than half the quote and "twice the quality" (his words but I tend to agree) I'll keep you all posted with my findings, good and bad, as I continue to learn this new system.