Cluster X: Series 1 - moon - ЅріzZ - nebula - drewnet - Slipstream

driftwood

Another thing, GOP Related patches in ptools if you look at flow motion - even though the bitrate reads 99000000 (nearly 100Mbps) this is also true of other settings including stuff Ive done in the past - where GOP Related is set at underwhelming IPB weights as in;- lets use 720p60 flowmo 202 as an example;-

720p60 Opt1 GOP Table=2, 15, 0, 0, 48, 0 = GOP 65 !!!!

Its doesnt physically look like that when you see it in Streameye or Streamparser but nevertheless the bitrate has been given a percentage ratio for weighting (more numbers here utilises less quant/bitrate). Its another way to factor rate control and stability.

Look at how much bitrate weight has been taken off the above. 2 for i frames, 5 for P frames and 28 for b frames for 1080i60 and 720p60 2 for i frames, 15 for P frames and 48 for b frames. This has had to be done to stabilise the setting to work at a peak bitrate on a death chart or very busy detailed scene but what actually also happens is that normal scene recordings receive an underewhelming use of the bitrate stated. Its just one way of going about spreading the bitrate but is not actually in theory the correct way.

Cluster X like moon is wanting to put the GOP Related balance back to how it should be - for example;

720p60 3 GOP would be as follows;- 1 for i frames, 2 for Top field p frames or 0 and 2 for Top field b frames as in;-

1, 0, 0, 0, 2, 0 = GOP3 with b frames 1, 2, 0, 0, 0, 0 = GOP3 using P frames.

And then use quantisation values and matrices along with the available bitrate to balance rate control. This way Ive found scenes with less detail get solid quant and use of bitrate.

This will be more apparent in the next Trial. Its already correctly being employed in 1080p24 during trials 3 and 4 :-)

kellar42 · Screen Shot 2013-02-15 at 8.00.17 PM.png

5D2RGB and yes a touch of grade but I'm seeing every inch of shadow and highlight from deep shade to sunny sky with MoonTrial3.

Can we rename Moon to 'Russian Meteor' to mix it's Russian roots with Astronomy :D?

WhiteRabbit

@driftwood, thanks for the clarity. So, of all the Cluster X/Moon patches, do they produce as sharp or sharper footage than AN C7 '444 Sharp2' I am running currently? Isco anamorphic attachements are about to go in the cupboard, again. Softer Lomos back on the test bench. Cheers.

WhiteRabbit

@QuickHitRecord, perhaps stop down, use the GH2 light Meter Mode to measure the centre point, from memory, that is the square with point in the middle. Perhaps if that is pointed at light on the wall and you lock down the tripod and not move anything between patch changes, you may perhaps match the GH2 exposure bars to match each patch. If the patches vary, you can adjust aperture to up/down you control point on the horizontal exposore bar +/- 0. Ideal is highlight near/below 100% on zebra, if light source can provide that. Also, better to have declicked aperture control, for fine adjustment. My method my be incorrect, however, there the difference between your Intravenus and Moon frames, and it begs the question as to why? gtg, thanks, and have a good one.

driftwood

@WhiteRabbit 444 sharp2 sharpens the existing soft matrix of chris brandin's original 444 matrix idea. So its actually a setting which is not as soft as chris's original - as I too wanted something a tad sharper for my FD and Noktons - but all in all isn't OTT sharp.

h264 uses perceptual weighting whereby the low frequencies (top left of a matrix) are generally coded with better resolution than the bottom right part of the matrix (the hi frequency range) because the human eye is not brilliantly susceptible to hi frequencies and anything OTT is wasted to the eye. Quantising scaling matrices are used to weight the frequency range in a similar way to a low pass filter.

WhiteRabbit

@driftwood, when I finish this project, I am going to have to play catch up with these new patches. My question remains, been asking it a lot, however it seems to get passed over as these threads are full of so much footage and evolution of the patches. I am looking for a sharper type of patch for my old lenses which are, perhaps, on the softer side. Does AN C7 Nebula '444 Sharp2' actually sharpens the image or is that a matrix internal thing? Does Moon cater for the legacy softies? Thanks again, you're effort and contribution is appreciated.

QuickHitRecord

@WhiteRabbit Perhaps. I have some fast primes that I can try out. Maybe this weekend.

Another thing to note that I may not have mentioned before was that all shots were at 640 ISO, which would decrease dynamic range. In a brighter setting, I'll have to try out 160.

WhiteRabbit

@QuickHitRecord , if all shots were with the lens wide open, then perhaps with Moon you need to little more 'open', to match the 100% peak of the other patches? Perhaps this is the contrast difference?

driftwood

moon is using GH3 technology and rate control. Ive spent an age analysing the GH3 (I have two) on elecard software and VEGA h264 software - the GH2 is doing exceptionally well with 4x4 transform, exceptionally well (albeit requires higher bitrate) :-)

@WhiteRabbit No. No offence at all. You have always had the respect here.

WhiteRabbit

@GravitateMediaGroup, no, I am stuck in a client project, shot on F3, 5D, Go-Pro and another camera (name forgotten), talk about different footage look to match.

@driftwood, if I read your comment above correctly, I did not mean to offend. The question for me remains, was there any exposure adjustment between the 3 patches applied above, it may vary the result. That is a complicated shot, high DR between black and white, and gradient/banding issues with the light source off the wall. Intravens image, after colourist makes his/her choices, may result in same image as Moon, so like I said, it appears to be a great patch to shoot and bascially just edit, if the WB is correct. Again, I may have it all wrong. Nick, you rock, man!

QuickHitRecord

And as always, thank you @driftwood for never being satisfied with 'good enough'.

kellar42

@WhiteRabbit I know what you mean from @QuickHitRecord 's test, and I don't know why it's like that but my experience is the opposite with Moon. That video I posted a few pages back of my girlfriend running around is getting me PMs on Vimeo of people asking about 'increased Dynamic Range' and stuff like that which is of course unlikely, but still. The file that comes off the camera really quite 'flat', I would say, though a pleasing flat, not one that is ugly without a grade. Not sure why the discrepancy, and I haven't used moon a lot in low light, but I would call it the opposite of baked in.

@driftwood, excited and standing by to test Moon trial 5.

GravitateMediaGroup

@vicharris I got you beat, I'm gonna buy a vx1000 off eBay and go full throttle

QuickHitRecord

@WhiteRabbit My methodology was to find a that had a wide dynamic range and then to cycle through the apertures. The tests above are all with the exact same picture profile (I didn't change anything, and then I double checked when Moon came out looking higher-contrast) and all shot wide open with my Nikkor 28-50mm f/3.5. Nothing was changed. I am just as surprised as you are, but also excited to see a setting that seems to minimize banding a bit.

Based on what I am seeing Moon has some built-in contrast and is not as flat as the others I tested. Other members are saying that it is a good setting for grading and my test seems to go against that, but I haven't tried grading any footage yet so it's hard to say for sure.

driftwood

LOL @vicharris Best thing you've said.

vicharris

That's it, I'm giving up and dusting off my HVX200! I can't take it anymore!! Too many choices!!!!!!!!

GravitateMediaGroup

@whiterabbit have you graded anything shot with moon?

driftwood

Anyways if you like the look of a setting stay with it. You cant please everybody - HVS (human visual system) is different from person to person). I can only go on what measures better in the Lab.

WhiteRabbit

@QuickHitRecord, imo, of the three frames posted, above, Moon looks like more of a 'baked in' look, whereas I prefer the flater images like Intravenus and FM, where I can make more post choices to colour correct/grade. Did you change exposure settings for each of the patches? I ask, because Moon seems to have more contrast and perhaps is a little darker, did you expose each take by adjusting exposure to just before the zebra hits 100%? Just an idea, as it may change the result. Not sure if this is the correct method, it is what I would do. Based on your image, Moon appears to be a graat way shoot and post without much post tweaking, which will suit certain projects and client requirements, etc. I might have it all wrong...

GravitateMediaGroup

@driftwood I didn't do ANY test and knew it was better lol I gave FM up months ago. for lack of a better word it's "Played out"

vicharris

I'm sure it is I just can't see that big of a difference in the banding here. Like I said though, like I'm some sort of expert...I just like it and that's good enough for me! :)

driftwood

Guys I already know moon is better - rest assured. Ive measured the DC and QP results. Take it from me. Knew this ages ago.

GravitateMediaGroup

@matt_gh2 yeah, there is a slight grade on a couple clips to see how well the "yellowness" of the sun at that hour could be removed

matt_gh2

@GravitateMediaGroup Yeah, some of those shots had the look I was thinking about. Looks like a slight grade was done at 35 seconds on the roof...and that's kinda how I thought those lenses would play with Moon. Cool. Thanks.

GravitateMediaGroup

what's the link to the youtube video getting commented? lol PM it to me

Howdy, Stranger!

Categories

Tags in Topic

Top Posters