@ Mihuel: Yes, I used his latest one. My results are a bit erratic, since the first try (which didn't span) played back, but the second one not. This may need some further testing.
Here's some GOP1 I shot last night with @driftwood's latest settings (first two shots only, the third shot was cbrandin's 66m):
I wanted to try the GOP1 because previously (in cbrandin's GOP12) I had poor motion quality in shots with lots of microbes. The problem manisfests as a subtle uniformity of motion in the microbes at a 1/2 second pulse. would all move at the same time (subtle, but definite.) The GOP1 seems to solve this.
If we think about it, Sir Driftwood's setting is better than Panny's AVC-Intra, which is considered professional.
I observed a real data-rate around 90 mbps, AVC-Intra is only 100, but has to squeeze more data into that, since it's 10 bit and 422. So, that's more compression. I admit, we need to be more careful with contrast and exposure from the start, since 8 bit and 420 is less flexible in post. But if you use this little camera carefully, you can stand up against camcorders costing many times as much.
we're interested here (this thread) in quality discussion with the primary purpose of discovering more about the GH2 hack settings, specifically those low GOP ones. This is "Personal View" forum, but there are other threads for such things. Here we need constructive testing, feedback and honest discussion, not unsupported opinion. I'm not saying you're wrong in your assessment . . . just please make some effort to support your claims.
Just using standard eyepiece projection method with a microscope and Nikkor 55mm lens on my GH2. The microscope magnification varies between the shots from 40x to about 400x. The GH2 adds some additional magnification.
I think these tests need to be better. 1) in focus. 2) reasonable detail 3) Reasonable motion and speed. Then since we are experimenting with GOP 1. A series of consecutive stills from the footage, side by side with regular stills. I think, only looking at individual frames can we see how accurately this hack is compared to a still. A lot of the out of focus, no motion test shots really don't tell the story.
@disneytoy We're doing a proper show this week. Those shots were quick and for the record from the night we cracked it, not brilliant, but proved to ourselves it could be done. Absolutely no concentration was put into the making of that clip :-) PS we're testing a big percentage rise in i frames right now with some new settings for GOP1
@driftwood I appreciate what the milestone of 1 GOP was able to carry out thanks to you.
@proaudio4 I tried the efforts keep the bit rate from falling as much as possible on my way. This is my 66MB 1GOP AQ3 setting. Although tested with swaying the Stray's crazy chart and ss=1/800, extream stress as your test cannot be applied in my environment. Would you find out how much tolerance over stress there are with 66MB 3GOP AQ3 stuck previously, supposing you are not disagreeable?
If the bit rate may fall when stress is low, I think it possible to enlarge highest frame size more by using AQ2 or AQ1 and the multiple of FL.
NOTE: I have not tested other than 24H. Moreover, I am not checking about the spanning, either.
@MbhuyKo first time ive heard of a problem with a panny lens on it. I dont own a 14-42 as its not brilliant but can you do a number of tests on it and then test it with my latest setting which will be up soon.
@driftwood Since it had not actually become half even if it was displayed that (actual frame rate is half) by Streamparser in my some 3GOP settings, I did not care about the display, but why is this or do you understand it?
Although I may have overlooked the important thing in somewhere --- :-(