Destructive Editing

No peas were harmed in the making of this clip.

(Hint: They were still mostly frozen)

Back to topic...I have been rendering to pjpg at 87.5%. Now rendering a clip at 91.2%, and besides being a lot slower, I'm looking to see if it's noticeably better. I might even render it out again exactly as is at 87.5% and use difference masks to see what that tells me.

Just sent my better machine back to Puget today for some diagnostics....so intensive stuff is slow....but whatever it takes....

I also have another problem with the .Mov containers. It's the "Gamma Shift" problem. A while ago i read there is a fix for Mac, but not for Windows.

Anyone got experience with this problem

I know little about that, but from what I've read it's not the container, it's the codec within it. I may be confused. There's a lot of lore, and understanding comes but slowly. I am no longer content to be told which way is right, but I need to understand why.

Next week I get my PIX220, and will be doing performance and quality tests between ProRes and DNxHD codecs, both on .mov wrappers. I would be interested if someday Sound Devices would offer a firmware upgrade to allow those recorders to record video with wavelet compression, like Cineform....if I understand the terminology right, the macroblock banding that looks so crappy in gradients, is eliminated, but the same circumstances that lead to it in the first place yield more noise, still not perfect, but more akin to film grain, and easier on the eye.

I performed the first experiment. I will be repeating it under more controlled circumstances and with more range, maybe even do a real article with frame grabs to be posted on my site.

If I ever express total idiocy in anything, or anyone notices a lack of understanding on my part that might lead to more confusion, please let me know!

Here are the preliminary results: Rendering a 57 sec. clip with lots of color correction, etc, at 91.5% yielded a good-looking result, but no better than the same rendered out at 86.5%. The difference mask test indicated NO difference. However, there is some slight difference between both the resultant files and the original clip with FX. Very slight, as in hard to see. The file size difference was as follows: 91.5% = 924 MB, and 86.5% = 839. I bollixed the render time comparisons because I was accessing one of the drives for something else at the same time during one of the renders.

I realized that this test is a waste of time and drive space. I am now embarking on a better test involving a shorter clip, with a blue sky/bright sun over snowfield and distant forest, which should be better test fodder than this:

@Eddie, as far as I know gamma shift problems Mac vs Win, was that older Mac os (leopard and older) used gamma 1.8 as default.
If you upgraded to mac os-x snow leopard or newer the default gamma is 2.2 which is the same as windows. So I don´t think this is no problem any more?

I'm using windows and i noticed this problem a while ago when i was rendering a musicvideo that was very heavy color corrected. The ouput in the most codecs had a flatter image compared to my picture in premiere.

So i think the problem is still there.

Im using a mac with snow leopard and the problem is definitely gone. The only problem is that when you open the mov file with quicktime it looks washed out because it displays it wrong (maybe 16-235 instead of 0-255)... Anyway the output when viewed on a tv screen is important or the upload here on pond 5. When I watch the same movie clip with VLC player or inside After Effects it displays it correctly. So it has something to do with quicktime displaying incorrectly.

OK....I did a test. Pretty exhaustive, and fairly small fries, as far as these tests go, but given the results, I can say the percentage slider does very little (in the "normal" range) .

If you are REALLY bored, go here, and let me know what you think. If you are not asleep...

http://mizamook.com/?p=110

I think you need a turkey break.

Tofurkey, thanks, and I need a break from that.

Here's a cool article in case you've not been rendered totally blotto by all this stuff:

http://www.soundonsound.com/sos/aug10/articles/video-compression.htm