@pop24 You have the option of unchecking the Scaling Matrix if you want Panny's own. This has been mentioned many times.

That indicates that Panasonic has changed the way the buffers / inner processing work in the latest firmware, maybe to make the HBR mode work properly. I have noticed the same thing.

Another good thing about Orion I don't think theres been enough said about, for me it seems to be a lot more stable over a range of class 10 cards than previous GOP1 settings.

People saying that they can record 720 60p with Orion patch in a card class 6 or 4 it´s not a good sign... @cinemon I also notice some noise not also in shadows, but midtones and highlights too where with previous patches I did not, but still have to do more testing. Like @mozes says with the new matrix you may have to relearn using the camera profile settings

@sohus don't hold your breath, unfortunately other comittments have creeped up on me meaning I won't be contributing anything seriously for a few weeks now. Not that I was ever that hot on scientific tests imo anyway. Plenty of other ppl are doing pretty good testing. I am overall happy with the Orion patch, I've not gone near pictoris cos my gut instinct is that that is really much too low a bitrate for a GOP 1 setting for me. I do know though that the new matrix would mitigate this, but I've not seen any output from it that changes my mind, yet. But, I say again, I haven't tested/used it. As far as changes related to the profiles are concerned, contrast settings are key as far as I'm concerned with this matrix, particularly for getting a good starting point for CC/Grading work.

The truths about film profiles re: shadow detail, highlight rolloff etc that we already established still hold on the whole. But, maybe its just me, but I think the overall response of Nostalgic and Cinema have altered positively. It is, as ever, shot, lighting and glass dependent. Noise does have a very different character with the new matrices, not surprisingly, and I'm really not sure about that yet. In camera NR is still totally arse and to be avoided imo. Personally the most interesting news I've seen recently is @stonebat discovering that the old @cbrandin 66M setting is now stable on this firmware without tweaking any of the 'for testers' settings. So... maybe I should try and get my old 88M setting to actually work with ETC mode ;), strip it down and start again. In terms of IQ, I do like the Orion Matrix, but I'd like to see a long GOP (or at least a 3 or 6) with some alternative matrix options (as I still believe strongly that it holds IQ better than short GOP settings). I believe that @cbrandin is working on that at present, and I look forward to it. More can be done, I think, in terms of maintaining quality/stability, rather than just softening (and the Orion matrix is doing more than just softening) with the control we now have over the scaling tables and other parts of the encoder. Don't look at me though, tweaking those is way beyond my focussing/learning skills in my current situation.

@mozes I am not that technical, simple as that. I can say that I like an image but not explain in technical terms why I prefer one over the other. For me, it is more of a subjective judgement while I am also interested in a more scientific judgement. I know @stray and @cbrandin are really good in the latter category.

Getting the shots the same is easy. Just save a camera setting to C3 and load it every time you do a comparison.

I think the following message by @driftwood should be reported on this thread too:

Ok, to all those with freezes, etc... we'll lower frame limits and buffers even more. @duartix do the same test on ANY testers setting for the GH2 and tell me the Q youre going to get on death charts. Its the SAME. Its the GH2! You can go massively long gop and then youre in other troubled waters. For CBR style you can also use IQ/Q=0. If you want the whole picture macroblocks to be coded with exactly the same Q then raise your IQ/Q to the high 20s. I'm against the same Q for whole picture theory... that means highlights and shadows get the same Q... not good in my book. Death charts can fail even on stock settings - we've noted this. If the GH2 had more memory, bigger buffers, faster cpu it'd be a different ballgame. For most real world filming situations - filming people, closeups, general docu makes these settings work very well. Other than that, buy a RED. Most importantly, get out and film REAL things now.

:-)

@sohus why not try yourself a comparison between the differences with the driftwood settings.?
i always save one clip from every setting, and compare them.<br. What i have learn is, they all have it, and not alone the driftwood patches.
But its more complicated then that.
I MUST always remember the iso bug, i ALWAYS must set the exact same Whitebalances.
i MUST Set exact the same profile with exact the same + and -.
And still, the light off the moment off shooting can ruining it all..
Then i must use the exact same settings in my editing program.

Thanks for this, im really stoked about these new settings to play with. I do events, so im really interested in the "everyday" patches, specifically "baby Quantum X". I saw others reported this as well, but im noticing tons of noise in most of my outdoor clips. Is this because its starting out with such a complexe scene already - Trees, bushes, and grass? I noticed that in scenes with less details, it seems to have very smooth and clean video. Im using the latest version, with the "deblocking" off. I tried both CBR and VBR at 60p/80%/24p l/h all with noisy results (smooth -2 all).

@mee The 720p of SpanMyBitchUp is soon to get an overhaul...

@jhero Thanks for the Cinema -2 to +2 tests. Excellent.

Thanks Nick.

*** NEWS ***

A new matrix and deblocking 'VY Canis Majoris' is being worked on to analyse the various stages of deblocking in INTRA settings. Issued 16th March 2012