From: Dedalus Lunardi
Hello. I'm back here.
My screen freq is at 75hz (85hz make it looks very bad also with still images).
I look at the fps of the original takes and the result.
And ehm..., maybe it's here the cause of the bas "bands" on the movie. The fps of the original files are at 15fps, the edited scene is at 29fps.
Do you think is the frame per seconds of the original files that can cause the distorted bands on the movie?
When you use a mixture of captured footage at different fps, the rendering engine software that you use to 'compile/render' the output file, would have to decide which frames to drop and which frames it can 'tween'.
For instance, at a rate of 29fps input, to 25fps output, there will always be 'leftover' frames which cannot be used, as these frames are dropped, this can cause the appearence of jerky movement in the movie.
Similarily, 15fps input, to 25fps output, the programme would have to condense the frames available and effectively speed up the movie, or try and 'tween' frames for those that are missing. If you add a music track, the 'tempo' of the music will dictate the bit rate (speed) of the movie, the music track will take preference over the movie fps rate. You can alter this priority using codex formats such as .rm but the results are less pleasing to the ear.
The human brain can handle a certain level of 'tweening' itself in merging 'flipbook' scenes into a fluid movement, as in TV, monitors, celluliod movies etc.
As you are using a CRT monitor I would suggest you best way forward, would be to use a high refresh rate (75htz or above) which is relative to your fps capture rate, using the calculation of whole number divisions (75htz divided by 3 = 25fps), given the inconsistancies of the rendering engine, you stand a better chance of reducing the scan lines you are experiencing. Another element to consider is resolution, if you try capturing the whole screen at say 1024x800, that amounts to a lot of data capture that is not only showing on your screen, but also being transfered onto your HD. My suggestion is that you use a screen capture programme such as Fraps, use region select to capture a smaller window of data, and combine that with saving your data to a seperate HD on its own channel controller (not a partition or slave HD on the same channel). Keeping that HD defragmented and clear of any caches/restore points/indexing/virus checking etc will help. Or if you don't want to use a screen capture programme, lower the client resolution to say 800x600.
When you come to post-production, always set the output movie resolution to the same aspect ratio of the captured data, the same with the input fps capture rates, mirroring the output fps. Most good post-production movie editing programmes will allow you to crop/zoom the output window (aspect ratio/resolution) to reduce borders.
I don't know if you shot the movie as a whole seqment, but you might try using cut scenes and blending them in the post-production software (don't use transition wipes, they can cause problems when applying codex), this will effectively reduce the fps inconsistancies experienced with making long shoots, which can effect the data flow to the HD.
What it all revolves around, is having a system that can cope with the capture of the movie raw data in a consistant manner. Combination of connection speed (packet loss, latency etc), clean flow of captured data to the HD, correct scan and fps rates that are maintainable and good post-production work.
There are no hard and fast rules that will make anyone with a PC into Alfred Hitchcock, it only supplies the means, much in the same way, having a car dosen't make you a world class rally driver.