@BurnetRhoades. You know about the GH2 ISO bug, right?
@BurnetRhoades would be interesting this test at different white balance.Because i noticed that if you force digital sensor to increase one channel more than another, then, noise is amplified.For example in tungsten light where is more red, sensor is forced to artificially raise the blue channel to compensate for clinical white balance(You must find the best compromise between what can sensor and perfect white balance) .If you understand how the sensor reacts in different circumstances you can substantially reduce noise. For me noise is not a problem since I figured out how to control it and the last series of CLUSTER X patches have the finest noise .
Guys there is a reason there is always a scramble on high-end cameras to find the right profile settings, all these cine-gammas and log functions...the DR of a camera is the DR but along the lines of what @BurnetRhoades says it's about making as much of it usable as possible or which sections we are using.
And getting back on Topic, yes, as others have also noticed, the MOON hack, run through 5D2RGB, in my case shot with smooth dialed down delivers a nice starting point for a grade, quite flat, but not like Cinestyle which I can't quite like on my 5D. I have not previously noticed the ability to get nice, well exposed skin tones and perfect highlights in the sky behind in the middle of a tropical sunny day, and now I can. It's no more stops of DR than whatever the GH2 delivers, 8,9,10...I don't know, but its closer to the full range than a stock GH2 with FCPX transcoding running dynamic (to make the extreme point) would be.
@BurnetRhoades Based on this topic I have just ran this quick test.
The first 10 seconds is ISO 160 and the other 10 seconds is ISO 640. In my opinion 640 looks good but its alot 'busier' noise than that of 160...I turned up the gain and gamma to show the noise more clearly on the PC unit. This test was carried out with the GH2 ISO Bug in mind, meaning that I selected each ISO by first selecting the ISO an increment above it first and then selecting downwards.
Quick and crude grading test out of my office window testing Moon Trial 5...setting on smooth all at -2
@lmackreath really nice grading! what tool did you use?
@danielvy Sony Vegas on my work pc (normally use fcp x on my mac).
Just used Vegas's in built colour corrector and colour s curve.
IMO the grading ruined the look which became totally videoish. Nothing personal, friendly said.
@producer totally agree. I wanted to do a quick test to see how quick and easy it was to push the flat smooth image from the moon hack without losing detail, blowing highlights or crushing blacks.
Never had footage this open to grading before, it's like playing with a raw photo in Lightroom!
I always try to shoot as flat as possible, but it is mainly for grading possibilities purpose. I hate the "flat look" for the final output that seems so trendy nowadays as it was the holy grail of beautiful image. Thank god the movies I watch and like are worked on to get a specific look. Vintage-overwashed looking movies? No thanks. Just my 2 cents.
would be interesting this test at different white balance.
This was using the tungsten WB. I only ever shoot with this or the daylight balance presets. If I have time today I'll run it again with daylight.
Seeing this, along with previous tests done with other patches, earlier posts in this thread make sense where test footage shot at ISO 500 immediately drew comments regarding visible noise. That's just a bad ISO for this camera, across multiple patches it seems. I'd been curious on a different patch why my ISO320 footage was so noisy in the shadows but I hadn't seen the noise tests. I'm glad Moon appears to have made that setting more useful.
I'm going to do this sort of thing for each patch that I shoot now to have a catalog of their noise. It'll be useful for both NR but also as a layer to add to CGI to even better integrate with the GH2 footage.
Vintage-overwashed looking movies? No thanks. Just my 2 cents.
Yeah, going back and watching some of my favorite films from the '70s and '80s, before DIP, you could really appreciate the looks of the stocks they were using at the time. There's so much more color than some of these homogenized contemporary films and grading styles. I mean, DIP opened up all kinds of possibilities that simply couldn't have been done in the purely photo-chemical days and the prevailing look is all teal and orange :(
One step forward, two steps back.
ok cool, just did a quick test comparing intravenous 2, moon 5, drewnet 6, and finally gh3.
drewnet 6 and the gh3 footage were a fantastic match. i used to use flowmotion, then switched to intravenous 2 because of the flowmotion rain. now i'll be switching over to drewnet 6 because i can record more footage to each card (i do interviews) and the color matched so well with the gh3.
the intravenous 2 footage and moon 5 footage looked very similar, but i like the moon 5 picture slightly more. the problem with both of these was one, the high bitrates fill up a card quick (my interviews can run 3 hours) and two, they were way green compared to my gh3 (moon was more slightly more green than intravenous). drewnet 6 was way more red than either of these 2 and it made it way more similar to the gh3. as far as noise goes, intravenous 2 and moon 5 were only very slightly better than drewnet 6.
if i were to rank the hacks in order of image quality i'd go 1 - moon 5 2 - intravenous 2 3 - drewnet 6
but seriously folks, it was like splitting hairs, i'd say 2-3% max difference in image quality between them. little enough for me to choose drewnet 6 as my hack of choice from now on.
i have to admit though, the gh3 trumped all 3. the tiny noise, fine detail, and dynamic range are fantastic.
also, i pushed the rgb curves around on all the footage and in my quick test, they could all be pushed around quite a better. a helluva lot better than my old AF100. so glad i sold that thing.
oh and a funny thing i noticed, i did all these tests with the same iso, lighting, and apeture, whether hacked gh2 or gh3. the gh3 exposure diagram said that i was shooting at 4 bars under proper exposure, and all of the gh2 hacks showed 2 bars under proper exposure. the gh3 seemed way better exposed than the gh2, the shadows were brighter and there was less noise than any of the gh2 hack, despite the different exposure displays.
@SeekingHeartwood @Driftwood just a question but for long-doc intvs, has anyone tried 24L setting? Would it still use the same excellent matricies, etc just encode at a lower bit rate? Might be a good solution if you're trying to get 3 hours on one card....
@JDN Yes it does use same matrix. The L setting is good for mainly static images but if youre doing a lot of fast action go for the high (H) setting.
@SeekingHeartwood The sensor on the GH3 is defo better. In visual terms its got more of the 'hit between the eyes' look straight outta the box. Not forgetting slightly more dynamic range. My point being is when we get comparable settings onto the GH3 things are going to be very sweet indeed.
@SeekingHeartwood there's no such thing as "flowmotion rain".
@burnetrhoades how do you know that the settings he likes to shoot with isn't causing some kind of odd noise?
So i tried the new Moon settings and my 16Gb sdxc 95mbs card still says its not fast enough, i thought FSH mode was improved
@jclmedia you must have a bad camera
? really how is that?
maybe you don't?
Oh ahah i dont think so , i mean if theres someone else with same Sandisc Extreme 95mbs card 16gb , Give it a try and see if same thing happens,happened to me slightly after 1 min of rec. FSH mode. 24H didnt seem to give me any problems.
till then im back to flowmotion 2.02
maybe stick with flowmotion? this is the 2nd if not 3rd time you have said this?
@jclmedia @driftwood I managed to record 12 minutes with Moon Trial 5 using FSH and Sandisc Extreme 16GB 95mbs card. Then it stopped. I normally use HBR and I am desperately waiting for the next Moon update for HBR 25p spanning fix...
It looks like you're new here. If you want to get involved, click one of these buttons!