@thomisus No, everything's fine with your cam. I think you're using Adobe CS6, right? Look here: http://www.personal-view.com/talks/discussion/6094/diagonal-rain-appears-using-gh2-premiere-pro-cs6-without-a-hack/p1

@bkmcwd Aashay said it right, get well soon! PM me for further PAL testings if you're right back again.

@bkmcwd. Dude get well soon.. Patches are one thing but your health is all that matters.

@bkmcwd: I hope you are better soon m8! Get well quickly! :)

Hi mate @towi

Now, since I am in the hospital, I cannot write in detail, but even if I see the result of your test, I do not necessarily think that Spizz is the best. It is because the lowness of qp is not top priority for me.

I think that the opinion of @Mirrorkisser also resembles me.

Of course, I do not deny your opinion. However, for some persons, since there are opinions various otherwise, even when Valkyrie and GOLGOP3-13 are the best, I am not surprised.

Spizz for me is considered to be a hybrid concept as already stated. It is because the matrix is scattering for every recording mode. Of course, I also support this concept. :-)

@spreeni

Isn't a 3-GOP setting for 50 interlaced pictures per second always better than 6-GOPs?

Always? I think it depends on the respective setting ("patch"). I feel motion (and also noise) in 25p mode looks a little bit better with 3GOP (from Valkyrie, Spizz, GOLGOP, FlowMotion) than with a longer GOP setting. On the other hand Nebula produces somewhat higher IQ on well lit scenes (the image looks a little bit clearer and maybe also sharper) and since you'll never have trouble with motion in 50i I would prefer the higher IQ. (Higher IQ is also helpful when you shoot 50i to render 50% slomo in a 25p timeline.) Then again ... we are talking about minor differences... IMO.

@Mirrorkisser

i shoot in 24p almost 95% of the time

I understood @thomisus' question was about HBR/25p ...

@spreeni i prefer the the 444 Matrix.

@towi i shoot in 24p almost 95% of the time. I simply prefer the gradation of valkyrie compared to the other 3 gop settings. Low QP is nice, especially on the paper, but in my opinion its not everything that defines a good picture. In the below 1000$ market i think its often the gradation that keeps footage from looking to videoish.

Still the clusterX settings are superb. I really like intramovement, IntraVenus is marvellous, too bad its not so good for wide and superdetailed shots. Moon 3 is so far my favourite allround intra setting.

@thomisus

What's the main difference between golgop3-13 and spizz?

@Mirrorkisser

valkyrie is the best 3gop setting out there

I like (and use) all 3 settings. In general use Spizz produces the lowest QP on identical scenes in HBR/25p mode. Attachment 1 shows Streamparser's JM-SP Decoder analysis of Spizz, Valkyrie444 and GOLGOP3; scene = high contrast stress chart. Attachment 2 shows Streamparser's JM-SP Decoder analysis of Spizz and Valkyrie444; scene = bright outdoor daylight scene.

Only in low key scenes with little detail to resolve (for instance low light scenes with narrow DOF) GOLGOP3 produces a lower QP (13) than Spizz (16). In low key scenes GOLGOP3 also produces higher bitrates than Spizz and/or Valkyrie. Attachment 3 shows Streamparser's JM-SP Decoder analysis of Spizz; scene = low key, narrow DOF. Unfortunately I don't have comparable scenes of GOLGOP3 and Valkyrie at hand right now … but for GOLGOP3 JM-SP Decoder would show 13 and for Valkyrie it would show 18 (with comparable low "Range" and "DC" values).

Technically, looking at the benchmarks, of the 3 settings Spizz should be the best all around setting in HBR/25 mode. GOLGOP3 is the best for low light scenes … although GOLGOP3 may also show more color noise sometimes (so it requires some more post work). Then again Valkyrie shows the smoothest motion in 25p (at least a little bit smoother than Spizz … I haven't had the time to test the latest GOLGOP3 thoroughly yet)… and, yes, I also like the look of Valkyrie very much.

You can't go wrong with either of the settings in HBR/25p mode … all 3 are really good.

@bkmcwd

Since "auto-quantisation" works correctly only in LongGOP according to Chris, I am not going to use it for 3GOP settings. These are only my personal views.

That's an interesting point. if I understand this correct it's not clear that SPIZZ (as a 3-GOP setting) uses this variable quantisation. So in the end it is just the different matrix driftweood uses in SPIZZ what makes a difference to GOLGOP3-13 or Valkyrie

I think that GOLGOP3-13-2 was completed in the BETA6.

I'm glad to know that the final version is near, but in my high-stress-test (ex-Tele on, auto-iris on, iDynamic on, cont. autofocus on) spanning does not work reliable. I know, noone uses so many automatic functions at the same time in real life, but for that it is a high-stress-test.

I tested the 1080i50 FSH mode, I will do a test in PAL HBR, too. What do you think, should we try to make GOLGOP3-13 'rock solid' in this mode or would it be better to say, it's finished now. For example FlowMotion (which isn't continued anymore) doesn't span reliable in my test, too - but nevertheless LPowls said it's development is finished. I can understand this attitude, one thing cannot be in beta status forever.

This is a sample in HBR of following Valkyrie 444 TZ2.

@spreeni

Hi mate! :-)

So the question is, is GOLGOP not up-to-date anymore? What do you think?

Thanks for your question. I think that GOLGOP3-13-2 was completed in the BETA6. Of course, if fault is found, I will correct it immediately.

Since "auto-quantisation" works correctly only in LongGOP according to Chris, I am not going to use it for 3GOP settings. These are only my personal views.