massive shift in green color on BMPCC4K

The place for questions about shooting with Blackmagic Cameras.
  • Author
  • Message
Offline

John Paines

  • Posts: 5787
  • Joined: Tue Jul 28, 2015 4:04 pm

Re: massive shift in green color on BMPCC4K

PostWed Jun 24, 2020 10:48 pm

Ryan Earl wrote:I think that is disingenuous John. You've had to alter your previous comment to include "at least none that can't easily be compensated for in the grade."



"Disingenuous"? I just wanted to be precise -- and I made the change before anyone responded, btw, not in response to anyone's objections. As you can see, there's no record of a revision.

Based on the samples you've provided, the difference between v3 and v4 is the least of it. It's a wrong approach, which is wrong for both, and may be amplifying the differences in your case.
Offline

John Paines

  • Posts: 5787
  • Joined: Tue Jul 28, 2015 4:04 pm

Re: massive shift in green color on BMPCC4K

PostWed Jun 24, 2020 10:50 pm

Mike Potton wrote:I think previous examples clearly show its a difference between gen3-4 color science, its seems BMD went a little overboard trying to correct skin tones and nuked the greens in the process.


Throughout the thread, I've offered maybe 3 samples of green, v4. Where were the "nuked greens"? I'll try Ryan's again, with v4, just to scrupulously fair.

Here we are, this time with 50 default saturation, extended video v4 LUT:

v4_1.1.1 copy.jpg
v4_1.1.1 copy.jpg (509.7 KiB) Viewed 7827 times
Last edited by John Paines on Thu Jun 25, 2020 12:28 am, edited 3 times in total.
Offline
User avatar

Dmytro Shijan

  • Posts: 1760
  • Joined: Wed Sep 17, 2014 7:15 pm
  • Location: UA

Re: massive shift in green color on BMPCC4K

PostWed Jun 24, 2020 10:57 pm

Note that Rawlite In addition to Moire/aliasing filtration and IR pollution filtration adds color temperature shift to improve colors and provide less noise under tungsten light. Similar to 80B filter but less stronger. So it is really recommended option. It may help to produce better greens.
BMMCC/BMMSC Rigs Collection https://bmmccrigs.tumblr.com
My custom made accessories for BMMCC/BMMSC https://lavky.com/radioproektor/
Offline

Ryan Earl

  • Posts: 517
  • Joined: Sun Jan 17, 2016 6:56 pm

Re: massive shift in green color on BMPCC4K

PostThu Jun 25, 2020 12:51 am

John Paines wrote:
Mike Potton wrote:I think previous examples clearly show its a difference between gen3-4 color science, its seems BMD went a little overboard trying to correct skin tones and nuked the greens in the process.


Throughout the thread, I've offered maybe 3 samples of green, v4. Where were the "nuked greens"? I'll try Ryan's again, with v4, just to scrupulously fair.

Here we are, this time with 50 default saturation, extended video v4 LUT:

The attachment v4_1.1.1 copy.jpg is no longer available


John, I tried to match your contrast and saturation with V3 without LUT or CST, just contrast and saturation adjustment. And it was raining in the image where greens will really saturate.

I think your result looks like Mike's salad image before hue adjustment. The wall also gets really blue there.

backyard-raining_v3.jpg
Ursa Mini 4.6K Color Science 3 Saturation & Contrast Adjustment, no lut, no cst
backyard-raining_v3.jpg (789.8 KiB) Viewed 7821 times
Offline

Ryan Earl

  • Posts: 517
  • Joined: Sun Jan 17, 2016 6:56 pm

Re: massive shift in green color on BMPCC4K

PostThu Jun 25, 2020 1:37 am

dusk-portrait_v3.jpg
Dusk Ursa Mini 4.6K SLRMAGIC 50mm V3
dusk-portrait_v3.jpg (482.34 KiB) Viewed 7810 times


My own take on this image, again without lut or cst. I increased contrast and saturation and pulled some red out with the offset wheel. It was a reddish sunset, so here I left more red in the image compared to Dmitry's versions.
Offline

John Paines

  • Posts: 5787
  • Joined: Tue Jul 28, 2015 4:04 pm

Re: massive shift in green color on BMPCC4K

PostThu Jun 25, 2020 2:01 am

Ryan Earl wrote:John, I tried to match your contrast and saturation with V3 without LUT or CST, just contrast and saturation adjustment. And it was raining in the image where greens will really saturate.

I think your result looks like Mike's salad image before hue adjustment. The wall also gets really blue there.


Without reality for a reference, there's no way to tell what it "should" look like. But yours is not a realistic green. It's unnaturally uniform and over-saturated, and the hue is one you won't see in nature. Real green in a real backyard is dozens of shades and tones. And yellow.

The "original" Mike salad looks like a post-production error to me. At least, I can't imagine that happening with the BMPCC 4K, unless the salad was actually that color. But, whether it is or not, the correction present its own issues. That look may be pleasing in an ad, but actual salad is nowhere near that uniform or uniformly saturated.

So we're no longer talking about image accuracy. I can make the v4 version into what I think is a cinematic green. But that's grading.
Offline

Mike Potton

  • Posts: 193
  • Joined: Fri Nov 25, 2016 11:19 pm

Re: massive shift in green color on BMPCC4K

PostThu Jun 25, 2020 2:04 am

John Paines wrote:Throughout the thread, I've offered maybe 3 samples of green, v4. Where were the "nuked greens"? I'll try Ryan's again, with v4, just to scrupulously fair.


I can see the nuked greens right there in that example, they are yellow. It is subtle but when you compare it to something (I could see the salad in real life) it's obvious something is up.

The gang at 4k shooters found the same when comparing a UMP G2 to the c200, look at the greens in the image below. It's happening under sun, under lights, it's happening on Fairchild (UMP) and Sony sensors (p4k), with and without IR filters. There is definitely something going on and gen 4 seems to make it worse.

C200_UrsaG2.jpg
C200_UrsaG2.jpg (119.39 KiB) Viewed 7794 times
Offline

John Paines

  • Posts: 5787
  • Joined: Tue Jul 28, 2015 4:04 pm

Re: massive shift in green color on BMPCC4K

PostThu Jun 25, 2020 2:15 am

Mike Potton wrote:I can see the nuked greens right there in that example, they are yellow. It is subtle but when you compare it to something (I could see the salad in real life) it's obvious something is up.


In the end, I can't answer for other's people results. Wasn't there, didn't shoot it. I posted BMPCC 4K examples which, in my view, disproved the "massive shift" claim. That's as far as I can go in confidence. The Ursa 4.6s, no experience.

However, knowing nothing about the post-production procedures, look at those two shots. Nothing is the same, between them. Not her facial skin tone. Not her top. Not her pants. And not the greens, either.

What can we really conclude, other than no one bothered to match them?
Offline

Ryan Earl

  • Posts: 517
  • Joined: Sun Jan 17, 2016 6:56 pm

Re: massive shift in green color on BMPCC4K

PostThu Jun 25, 2020 2:59 am

John Paines wrote:
Ryan Earl wrote:John, I tried to match your contrast and saturation with V3 without LUT or CST, just contrast and saturation adjustment. And it was raining in the image where greens will really saturate.

I think your result looks like Mike's salad image before hue adjustment. The wall also gets really blue there.


Without reality for a reference, there's no way to tell what it "should" look like. But yours is not a realistic green. It's unnaturally uniform and over-saturated, and the hue is one you won't see in nature. Real green in a real backyard is dozens of shades and tones. And yellow.

The "original" Mike salad looks like a post-production error to me. At least, I can't imagine that happening with the BMPCC 4K, unless the salad was actually that color. But, whether it is or not, the correction present its own issues. That look may be pleasing in an ad, but actual salad is nowhere near that uniform or uniformly saturated.

So we're no longer talking about image accuracy. I can make the v4 version into what I think is a cinematic green. But that's grading.


You have my DNGS and one does include a color checker chart and face, I don't agree with your conclusions here. Come down to Philly, lets do a test! I've got plenty of scenery here.
Offline

Bunk Timmer

  • Posts: 171
  • Joined: Sat Oct 15, 2016 8:14 pm

Re: massive shift in green color on BMPCC4K

PostThu Jun 25, 2020 9:46 am

John Paines wrote:I would look again at your workflow. This is the backyard shot, with RCM and v3, and saturation lowered from the default of 50 to 30. Sorry about the bars, incorrect export settings.

The attachment 2_1.2.1 copy.jpg is no longer available
RCM, v3 saturation to 30. Slightly different result. What did I miss?
different-look.jpg
different-look.jpg (178.97 KiB) Viewed 7724 times
Offline

John Paines

  • Posts: 5787
  • Joined: Tue Jul 28, 2015 4:04 pm

Re: massive shift in green color on BMPCC4K

PostThu Jun 25, 2020 12:43 pm

Bunk Timmer wrote:RCM, v3 saturation to 30. Slightly different result. What did I miss?
different-look.jpg


I didn't save the project, but I would have brought the blacks down to zero and added a bit of contrast, as a matter of course. This isn't cheating: the initial RCM result will depend on the camera exposure, as would also be the case with a LUT. One size doesn't fit all.

Your image above could pass as log, so it could be your RCM output setting, as well. Rec. 709 2.4?
BTW, if you want to access raw controls, you need to enable it in the first two fields. Choose the debayer quality and then "clip" in the second field.
Offline
User avatar

carlomacchiavello

  • Posts: 2587
  • Joined: Tue Aug 28, 2012 6:04 pm
  • Location: italy

Re: massive shift in green color on BMPCC4K Blind Test

PostThu Jun 25, 2020 3:24 pm

green develop.jpg
green develop.jpg (642.7 KiB) Viewed 7710 times

Yesterday i ask to my 7 years old nephew to help me.
We collect different green shade and put myself in front of camera to have skint tone visible.
i decide to be near to overexpose picture, i observe that most of problems are with raising of luminosity of many shade of lime green.
In this first two picture you can see like light green of plastic box and football seems near to be yellow on picture, not a light green, but i see also in the reality that are near to yellow when sun is bright, also my iphone had trouble to reproduce it correctly, to name a different branded camera.

I start resolve, standard RCM rec709 color space, load braw without change develop from standard.
Apply first a CTS node to normalize with right color space, Apply contrast node to manage two different contrast of picture. +10 saturation in one of picture.

On both picture skintone stay the same, some green are the same, same not... but not both to yellow.
to my eyes the only great shift is on plastic broom, ones is a vivid green, other is less vivid, but change angle and light of shooting.

here a direct cut between two different picture
green develop match.jpg
green develop match.jpg (420.23 KiB) Viewed 7710 times

there are some difference but so bg like i saw in other picture, why?
later i can upload a braw frames and dra of resolve.

I could add a small curve and obtain perfect matching, but i prefer to avoid.
green develop match +curve.jpg
green develop match +curve.jpg (789.02 KiB) Viewed 7699 times
.
Offline

Ryan Earl

  • Posts: 517
  • Joined: Sun Jan 17, 2016 6:56 pm

Re: massive shift in green color on BMPCC4K

PostThu Jun 25, 2020 5:23 pm

carlomacchiavello wrote:there are some difference but so bg like i saw in other picture, why?


I think we just have too much time on our hands! In the end we're splitting hairs and any camera requires adjustments. I've posted examples of V3 to V4 really to explain the color differences that people are seeing when they see a shifted green, that's my thesis, that additional correction is required to bring the greens up if you have a problem with what you're seeing. Once you put in another reference point, like from another camera, while not perfect is still a relevant way to judge color science. That's really what's being picked apart here.

I've been photographing paintings, products, portraits, landscape etc for years with the Pentax 645Z, then added the URSA Mini 4.6K when it was released. I also have a large format Epson printer in the studio where I can make prints to get them off the monitor. I'm working primarily in advertising, product and artwork reproduction commercially. So it's hard to imagine that my workflow would suddenly be incorrect and that the de-saturated CST would be more correct (when I can see it's too yellow) over doing a custom hue shift with the pocket 4k. And again I'm open to changing workflow, just haven't been convinced to do so.

So. . . I've done one final test in my studio for forum consumption where I set the white balance according to the Cine Meter 2 app then developed the pictures with Photoshop raw converter. I'm not convinced my workflow is terrible or that the images are too saturated in these examples. I think I have seen a lot of agreement between the 2 when URSA Mini is in V3, so I'm including them for comparison. The Pentax and Ursa Mini have the white balance set to match each other 3250K, then I cooled off the Pocket 4K to 3000K +4Magenta to get the greens greener but then reversed that decision when the whole image became too cool. Pocket 4K is shown here 3250K. Tungsten lights are producing better results all around that I wouldn't feel bad using.

Pentax-645Z.jpg
Pentax 645Z and 90mm Macro 2.8
Pentax-645Z.jpg (936.36 KiB) Viewed 7682 times


A105_06271438_C002_000002_URSA_Mini_4.6K.jpg
URSA Mini 4.6K SLR Magic 50mm APO
A105_06271438_C002_000002_URSA_Mini_4.6K.jpg (755.77 KiB) Viewed 7682 times


A008_06290116_S002_Pocket_4K.jpg
Pocket 4K SLR MAGIC 50mm APO
A008_06290116_S002_Pocket_4K.jpg (737.59 KiB) Viewed 7682 times
Offline

Ryan Earl

  • Posts: 517
  • Joined: Sun Jan 17, 2016 6:56 pm

Re: massive shift in green color on BMPCC4K

PostThu Jun 25, 2020 5:50 pm

Here again is the Pocket 4K with the White Balance Set to 3100K along with the iPhone. The brightest greens improve but the violet flower vase in the center goes blue over the iPhone and other cameras.

iPhone_SE_2020.jpg
iPhone SE 2020
iPhone_SE_2020.jpg (906.25 KiB) Viewed 7666 times


A008_06290116_S002_Pocket_4K_3100K.jpg
Pocket 4K 3100 WB
A008_06290116_S002_Pocket_4K_3100K.jpg (825.49 KiB) Viewed 7666 times
Offline
User avatar

carlomacchiavello

  • Posts: 2587
  • Joined: Tue Aug 28, 2012 6:04 pm
  • Location: italy

Re: massive shift in green color on BMPCC4K

PostThu Jun 25, 2020 7:03 pm

Ryan Earl wrote: I think we just have too much time on our hands!

right :-P :-P
i'm stopped at home due Covid reason, i'm without most of works did in past, i'm study and research.
I try to use at best my time :-D

I saw this thread and i did some test with cameras that i had, be cause in past i never noted this yellowish in green and i continue to not see this shifting on my cameras...
this very similar shooting are a side to side canon camera and bmpc4k with same lens (shoot in two different time obviously), and where fail bmpc4k fail also canon to go on yellow on football and plastic box, and to be honest... canon saturated too much broom, which had a more realistic color on bmpc4k... i had here in my hand and broom and picture on my calibrated monitor are more precise on bmpc4k, less saturated less pure green than Canon picture.
same iris, same color temp, same shutter, same iso.

Schermata 2020-06-25 alle 20.50.15.jpg
Schermata 2020-06-25 alle 20.50.15.jpg (227.25 KiB) Viewed 7650 times

here a link for original braw and jpeg where canon burn it its color science. Everyone can download and see that i not lie.
http://www.macchiavello.com/tmp/greentest.zip
Not a memory imprinting, not a better canon color (like many people think, i tell unbalanced on magenta and other color, often too much saturated
it's different color, side by side canon and bmpc4k are weak on both place, but canon are less accurate than bmpc4k if we talk to reproduce reality.
i decide to do a very simple test, same condition, same lens, very very simple develop situation like a newbie should do with his first camera and resolve.
i often mix pocket4k, 6k, ursa, gh5, and easely with ACES match correctly without problems (in that time i use gretag chart video to speed up matching if i had problems).

corret color.jpg
corret color.jpg (362.37 KiB) Viewed 7644 times

here the right color, with simple change on CTS, one stop down on exposure, NO COLOR CHANGE
and pocket show correct nuance of light lime green of football and plastic box.
Offline
User avatar

Dmytro Shijan

  • Posts: 1760
  • Joined: Wed Sep 17, 2014 7:15 pm
  • Location: UA

Re: massive shift in green color on BMPCC4K

PostThu Jun 25, 2020 7:56 pm

Image
Image
Image
Image
BMMCC/BMMSC Rigs Collection https://bmmccrigs.tumblr.com
My custom made accessories for BMMCC/BMMSC https://lavky.com/radioproektor/
Offline

Ryan Earl

  • Posts: 517
  • Joined: Sun Jan 17, 2016 6:56 pm

Re: massive shift in green color on BMPCC4K

PostThu Jun 25, 2020 9:51 pm

Carlo_CST_Instructions.jpg
Carlo with CST Instructions - 1 Exposure in the Raw Tab
Carlo_CST_Instructions.jpg (880.4 KiB) Viewed 7605 times

Carlo_Melara_P4K-Powergrade_ArriREC709_Lut.jpg
Carlo, Melara P4K Powergrade - Arri to Rec 709 Lut
Carlo_Melara_P4K-Powergrade_ArriREC709_Lut.jpg (940.35 KiB) Viewed 7605 times


Carlo I followed your instructions for your portrait above and included the "Juan Melara Powergrade" with the REC709 Lut. Along with Dmitry's Velvia 100 version it feels like an IR cut is being added and removed from the lens as I switch back and forth, despite that I feel like skin tone is natural in your version. I think it lines up well with my previous examples.
Offline
User avatar

carlomacchiavello

  • Posts: 2587
  • Joined: Tue Aug 28, 2012 6:04 pm
  • Location: italy

Re: massive shift in green color on BMPCC4K

PostThu Jun 25, 2020 10:48 pm

Hi Dmitry you touch my heart, i'm old lover of velvia and kodakrome :-D

Hi Ryan, yes my result meet your, often if we not a parallel shooting is difficult do understand if it's real different or feel different.

No, i not put ircut filter, i tried to be consistent on two shooting (i leave at home NDIRcut) i decide to shoot as simple as possible like generic user of both cameras.

jpeg is from canon 1300d, braw is from pocket4k, may be that canon put an ir cut on their sensors, i don't know, the camera is from my seven year nephew (he start early with a reflex, he's a pride of his uncle Carlo :-D)
I used for both cameras the same canon 55-250 at 7 iris, 1/250 of shutter, 5200 wb fixed for both, i used a metabone adapter (no speedbooster) for pocket.
Just to test (i'm the first to put myself in dubs) real difference of parallel shooting in the exact same condition, same lenses and same setup when possible oviously, to verify if... Canon is so realistic reproduction or not.
At the end of day i think they are near, with some difference obviously, no so big dominance of canon color like many people want to think, but more important not so realistic or lifelike... if is important for someone...
The key (to me) to manage shifting is the exposure, feed more light mean more yellow (sun light... is not white), correct exposure (god bless the raw) also in post, put color in the right place for my shooting.
Offline
User avatar

Dmytro Shijan

  • Posts: 1760
  • Joined: Wed Sep 17, 2014 7:15 pm
  • Location: UA

Re: massive shift in green color on BMPCC4K

PostThu Jun 25, 2020 11:03 pm

BRAW (as well as Canon) was overexposed. Some blown highlights in brightest parts that impossible to recover. In my examples with film emulation LUTs i decrease Exposure 2 stops.
BMMCC/BMMSC Rigs Collection https://bmmccrigs.tumblr.com
My custom made accessories for BMMCC/BMMSC https://lavky.com/radioproektor/
Offline

Ryan Earl

  • Posts: 517
  • Joined: Sun Jan 17, 2016 6:56 pm

Re: massive shift in green color on BMPCC4K

PostThu Jun 25, 2020 11:17 pm

carlomacchiavello wrote:At the end of day i think they are near, with some difference obviously, no so big dominance of canon color like many people want to think, but more important not so realistic or lifelike... if is important for someone...


+1

Concerning color science bias, if there is one, I'm not hoping as much for an accurate "fix" as much as additional color science versions that you could swap in camera. I don't think you can easily strip away the look just by shooting raw and applying a different lut. I'm happy to use the pocket 4K as-is.

Film's naming structure isn't hierarchical; Astia, Provia, Velvia, or Kodak Ektar, Portra maybe they could name color science versions instead of numbering them.
Offline
User avatar

Dmytro Shijan

  • Posts: 1760
  • Joined: Wed Sep 17, 2014 7:15 pm
  • Location: UA

Re: massive shift in green color on BMPCC4K

PostThu Jun 25, 2020 11:29 pm

It is not a color science versions :) , it is a G'MIC and RawTherapee Film Simulation LUTs made from real 35 mm film stock sources. viewtopic.php?f=21&t=77553&p=637455#p569634
BMMCC/BMMSC Rigs Collection https://bmmccrigs.tumblr.com
My custom made accessories for BMMCC/BMMSC https://lavky.com/radioproektor/
Offline

Ryan Earl

  • Posts: 517
  • Joined: Sun Jan 17, 2016 6:56 pm

Re: massive shift in green color on BMPCC4K

PostFri Jun 26, 2020 1:26 am

Dmitry Shijan wrote:It is not a color science versions , it is a G'MIC and RawTherapee Film Simulation LUTs made from real 35 mm film stock sources. viewtopic.php?f=21&t=77553&p=637455#p569634


Yes! You've got me, the Velvia looks so good!
Offline
User avatar

Dmytro Shijan

  • Posts: 1760
  • Joined: Wed Sep 17, 2014 7:15 pm
  • Location: UA

Re: massive shift in green color on BMPCC4K

PostFri Jun 26, 2020 2:16 am

Velvia 100 produce very "happy" look. It is very high contrast and high saturated. Usually if i want less pronounced film look touch, i use Astia 100 or Provia 100. These LUTs produce very subtle changes, but same time looks way less boring than basic Log to Rec709 transforms.

Image
Image
Image
BMMCC/BMMSC Rigs Collection https://bmmccrigs.tumblr.com
My custom made accessories for BMMCC/BMMSC https://lavky.com/radioproektor/
Offline
User avatar

carlomacchiavello

  • Posts: 2587
  • Joined: Tue Aug 28, 2012 6:04 pm
  • Location: italy

massive shift in green color on BMPCC4K

PostFri Jun 26, 2020 4:07 am

Dmitry Shijan wrote:BRAW (as well as Canon) was overexposed. Some blown highlights in brightest parts that impossible to recover. In my examples with film emulation LUTs i decrease Exposure 2 stops.
Yes, I try to emulate a common shooting without filter or other, I use lower iso, higher shutter and quite closed iris to avoid diffraction, but no filter.
In this situation I ever used a variable Nd, or better ndircut, but with test I would like to avoid everything in front of camera to avoid shifting from filter, and test how both sensor (canon and Blackmagic Design ) manage overexposure on light green.
Also my eyes see more yellow in that lime green.
For my taste in a working situation I would add diffusion reflection to fill shadow and probably I would used a low contrast filter.
But in this thread most of test seems oriented to verify a simple setup as possible on shooting and post.


Inviato dal mio iPhone utilizzando Tapatalk
Offline
User avatar

Robert Niessner

  • Posts: 4948
  • Joined: Thu Feb 21, 2013 9:51 am
  • Location: Graz, Austria

Re: massive shift in green color on BMPCC4K

PostFri Jun 26, 2020 8:42 am

This is my version of Carlo - just my LUT LBK-Neutral-Hi-DeSat v10 applied:

Carlo - LUT LBK-Neutral-Hi-DeSat v10_1.1.1-50%.jpg
Carlo - LUT LBK-Neutral-Hi-DeSat v10_1.1.1-50%.jpg (921.18 KiB) Viewed 7564 times
Saying "Thx for help!" is not a crime.
--------------------------------
Robert Niessner
LAUFBILDkommission
Graz / Austria
--------------------------------
Blackmagic Camera Blog (German):
http://laufbildkommission.wordpress.com

Read the blog in English via Google Translate:
http://tinyurl.com/pjf6a3m
Offline

Steven Abrams

  • Posts: 275
  • Joined: Tue Feb 24, 2015 12:43 am
  • Location: LA La Land

Re: massive shift in green color on BMPCC4K

PostFri Jun 26, 2020 10:51 am

John Paines wrote:But yours is not a realistic green. It's unnaturally uniform and over-saturated, and the hue is one you won't see in nature. Real green in a real backyard is dozens of shades and tones. And yellow.

I completely agree. I wonder about Ryans and others monitoring calibration. I've seen a lot of SONY greens that look like neon single hue garbage posted here as what should be a reference, and it's horrible. Please no.

Ryan if you want grass to be a single hue of neon green, get a sony and enjoy video looking colour.

Real grass is more yellow than green FYI.
Offline

Steven Abrams

  • Posts: 275
  • Joined: Tue Feb 24, 2015 12:43 am
  • Location: LA La Land

Re: massive shift in green color on BMPCC4K

PostFri Jun 26, 2020 10:55 am

Ryan Earl wrote:Image

This is horrible. Please get your monitor calibrated. I can't explain how someone could like this otherwise. Unless they really do like SONY video looking neon green grass (I say again, real grass has more yellow than green).
Offline

Ryan Earl

  • Posts: 517
  • Joined: Sun Jan 17, 2016 6:56 pm

Re: massive shift in green color on BMPCC4K

PostFri Jun 26, 2020 2:06 pm

Steven Abrams wrote:This is horrible. Please get your monitor calibrated. I can't explain how someone could like this otherwise. Unless they really do like SONY video looking neon green grass (I say again, real grass has more yellow than green).


It's not about "real life" or calibrating monitors, it's about matching cameras and color science. As you compare sensors some will get a specific hue more accurate than another. You're welcome to download my dngs. The 2nd version is the Melara Powergrade and Arri Lut. I didn't make any custom changes to the color. I think Carlo's image looks more natural too.

It just illustrates the objections people are raising where they want less brown in the objects and grass. Dmitry's examples are also manipulating the color of the grass to show the differences. If somebody wants one look or the other fine, but it's helpful to know how the sensor and color science behaves in relation to others.
Last edited by Ryan Earl on Fri Jun 26, 2020 2:21 pm, edited 1 time in total.
Offline

Ryan Earl

  • Posts: 517
  • Joined: Sun Jan 17, 2016 6:56 pm

Re: massive shift in green color on BMPCC4K

PostFri Jun 26, 2020 2:16 pm

Steven Abrams wrote:Ryan if you want grass to be a single hue of neon green, get a sony and enjoy video looking colour.


And you're just not reading my other posts carefully where I gave examples of my own portraits with color science V4 and V3 and preferred the natural look in the image and the skin tones. I'm monitoring out through an intensity Pro 4K with calibration. For me, I've narrowed the issue down (for my own work) to an individual's preference for color science between cameras or film. Kodak Ektar vs Portra, Fuji Provia vs Velvia, Canon vs Sony. Ursa Mini 4.6K V3 does get the Ryobi green accurately out of the box. It's strange to deny that bright green exists in plants in nature, I'm not denying that burned out grass wouldn't be brownish.

It's not productive to slam me for having bad taste, when I'm agreeing with wanting the natural look from the Blackmagic (V4) camera, just not the camera I'd choose to photograph a Ryobi drill or bright green salad or if I did why not just tweak the hue? I can make that decision from my own tests and workflow.

If you find issues, dig it out for yourself.
Offline
User avatar

carlomacchiavello

  • Posts: 2587
  • Joined: Tue Aug 28, 2012 6:04 pm
  • Location: italy

Re: massive shift in green color on BMPCC4K

PostMon Jun 29, 2020 7:37 am

Me too I not like too bright color with exception of some stylish video. But...

The goal of my first test is to show that too often the color perception is subjective and color of canon 1300d, and pocket4k in the exact situation, same lenses, same exposure, same iso, same wb will be perfectly overlappable, without big effort, or big knowhow, with very similar reproduction of most of green (the theme of this thread).

I wann’t start a flame with my photos :-)

I used canon color apparence like base because Eugenia told that she prefer canon color, and she observe that some green shift on yellow on Bmpc4k against canon.

To be honest I never observed it, but after her interesting observation I test myself and see that with up the exposure bmpc tend to shift in some condition to yellow, and some green could be shifted.

Under my test I observe that not light green is weak point but some nuance of green, see the groom between canon and Bmpc4k.
Different green Light green of box and football Are wrong on both cameras with i overexposed camera but correct with correct exposure (I had both near me to check with my eyes).

Color taste is subjective, like contrast, some one prefer crushed shadow I prefer silked detailed shadow, but sometime me too I like to do a crushed shadow picture like this on shooting did last summer.
Image

Everyone should can have his taste, and it’s, if we would have all the same taste will be boring :-)
For example I like Fuji Eterna and Blackmagic Design for their delicacy of contrast and color, both are near iso invariant on every flavour or raw (Fuji foto, Blackmagic Design video shooting), and that allow me to manage correctly on my taste.


Inviato dal mio iPhone utilizzando Tapatalk
Offline

JH_7900

  • Posts: 1
  • Joined: Fri Sep 24, 2021 6:44 pm
  • Real Name: Jeremy Harrison

Re: massive shift in green color on BMPCC4K

PostFri Sep 24, 2021 7:30 pm

IvanTheEditor wrote:I've recently started filming with my pocket 4K more and I'm loving it. However, on a recent product shoot where the product was all different shades of green I noticed that the greens are severely pushed towards yellow to the point that some colors which are 100% green in real life looked 100% yellow in my bmpcc4k. I've been searching for some LUTS that correct this but no luck so far. Has anyone else encountered that before? I know how to properly white balance, I use my gray and color charts all the time, but nothing seems to fix the green back to where it should be. It's like it just slips through the cracks. What can I do to make sure I get true, real-to-life colors with my blackmagic 4k even after I've properly white balanced?


I'm coming very late to this thread but it's been very helpful to read.
I use basically the same workflow as you, Ivan, and here's something I've stumbled on that could be helpful. In lumetri color, I tried a hue vs hue curve adjustment of the yellow only and was very pleased with the result. I shoot using the "video" setting in the camera, not the "film". I'll post pictures.
Image
Image
Image

Of course, from here I'd adjust other parameters...exposure, saturation, etc.

Hope this helps.
JH
Offline

John Brawley

  • Posts: 4267
  • Joined: Tue Aug 21, 2012 7:57 am
  • Location: Los Angeles California

Re: massive shift in green color on BMPCC4K

PostMon Sep 27, 2021 9:20 pm

This is also very easy in Resolve in the same Hue s Sat page ( i think it’s the second one )

You can sample a green and tweak it very quickly .

JB
John Brawley ACS
Cinematographer
Currently - Los Angeles
Offline
User avatar

carlomacchiavello

  • Posts: 2587
  • Joined: Tue Aug 28, 2012 6:04 pm
  • Location: italy

Re: massive shift in green color on BMPCC4K

PostTue Sep 28, 2021 6:55 am

JH_7900 wrote:
IvanTheEditor wrote:I've recently started filming with my pocket 4K more and I'm loving it. However, on a recent product shoot where the product was all different shades of green I noticed that the greens are severely pushed towards yellow to the point that some colors which are 100% green in real life looked 100% yellow in my bmpcc4k. I've been searching for some LUTS that correct this but no luck so far. Has anyone else encountered that before? I know how to properly white balance, I use my gray and color charts all the time, but nothing seems to fix the green back to where it should be. It's like it just slips through the cracks. What can I do to make sure I get true, real-to-life colors with my blackmagic 4k even after I've properly white balanced?


I'm coming very late to this thread but it's been very helpful to read.
I use basically the same workflow as you, Ivan, and here's something I've stumbled on that could be helpful. In lumetri color, I tried a hue vs hue curve adjustment of the yellow only and was very pleased with the result. I shoot using the "video" setting in the camera, not the "film". I'll post pictures.
Image
Image
Image

Of course, from here I'd adjust other parameters...exposure, saturation, etc.

Hope this helps.
JH

uhm... on my eyes (on my calibrated monitor) what i see it's not the green of foliage, but the strange color of sky.
Are you used a strong ND filter?
Are you did a WB before shooting?
Offline

Joe Bond

  • Posts: 2
  • Joined: Thu Jan 06, 2022 3:14 pm
  • Real Name: Joe Bond

Re: massive shift in green color on BMPCC4K

PostThu Jan 06, 2022 3:28 pm

I'm also coming to this *very* late, but I wanted to just comment that both myself and the company I used to work for avoided using Black Magic cameras (up to the pocket 4k) for green screen shoots, because we ran into problems keying due to the fact that the green screen did look too yellow which meant it was harder to key when the subject's skin tones were a similar luminance. It could be helped a little by separating the colours with hue before the key, but the fact is that there just isn't the separation of colour in the raw data.

I agree it can easily be 'fixed' in post if you're just trying to achieve a nice image, but there definitely is a problem with the cameras, at least up to the pocket 4k. Judging by some tests with the pocket 6k it seems they may have finally fixed this problem, so I intend doing some tests when I can with that camera and a reference camera (Red).
Offline
User avatar

Uli Plank

  • Posts: 21281
  • Joined: Fri Feb 08, 2013 2:48 am
  • Location: Germany and Indonesia

Re: massive shift in green color on BMPCC4K

PostThu Jan 06, 2022 4:43 pm

How did you light your green screen at the time? If it was Tungsten, this might be IR pollution.
Only recent cameras by BM with ND filters have some IR filtering, early ones were very sensitive to IR.
No, an iGPU is not enough, and you can't use HEVC 10 bit 4:2:2 in the free version.

Studio 18.6.5, MacOS 13.6.5
MacBook M1 Pro, 16 GPU cores, 32 GB RAM and iPhone 15 Pro
Speed Editor, UltraStudio Monitor 3G, iMac 2017, eGPU
Offline

Joe Bond

  • Posts: 2
  • Joined: Thu Jan 06, 2022 3:14 pm
  • Real Name: Joe Bond

Re: massive shift in green color on BMPCC4K

PostThu Jan 06, 2022 7:54 pm

Uli Plank wrote:How did you light your green screen at the time? If it was Tungsten, this might be IR pollution.
Only recent cameras by BM with ND filters have some IR filtering, early ones were very sensitive to IR.


It was likely either tungsten or HMIs, though more likely tungsten, so this could have been the issue. We also had the purply black problem too which was another reason we stopped using the camera, so presumably that was the same problem? As we were mainly shooting commercials and clients were generally on set, this became a real issue, but the company had the budget to switch to Arri cameras so we didn't really go back to Black Magic cameras after that.

It does look like they've gone a long way to fixing this problem and everything else about the camera makes it a great choice for shooting green screen, so I'm going to try to set up some tests when I get chance.
Offline
User avatar

rick.lang

  • Posts: 17175
  • Joined: Wed Aug 22, 2012 5:41 pm
  • Location: Victoria BC Canada

Re: massive shift in green color on BMPCC4K

PostThu Jan 06, 2022 8:31 pm

Joe Bond wrote:… likely tungsten, so this could have been the issue. We also had the purply black problem too which was another reason we stopped using the camera, so presumably that was the same problem?...


Purple blacks are a dead giveaway, you were suffering from infrared contamination. Shame this wasn’t recognized at the time. I have some public posts on Vimeo illustrating the problem.

I usually shoot with B+W IR Cut 86mm screw-on filter or Schneider 715 IR Cut rectangular filter. I’ve had those filters for five years and I believe there are updated equivalent filters available from Schneider and other manufacturers.
Rick Lang
Offline
User avatar

Uli Plank

  • Posts: 21281
  • Joined: Fri Feb 08, 2013 2:48 am
  • Location: Germany and Indonesia

Re: massive shift in green color on BMPCC4K

PostFri Jan 07, 2022 12:24 am

While the cameras with built-in ND filters are better regarding IR filtering, it's still a good idea to add some when you have a lot of IR in the light. The ideal solution for me was the RAWlite filter, which cures both of the biggest flaws of these cameras: moiré and IR! Now my UMP is close to an Alexa ;-)
No, an iGPU is not enough, and you can't use HEVC 10 bit 4:2:2 in the free version.

Studio 18.6.5, MacOS 13.6.5
MacBook M1 Pro, 16 GPU cores, 32 GB RAM and iPhone 15 Pro
Speed Editor, UltraStudio Monitor 3G, iMac 2017, eGPU
Offline
User avatar

Mark Foster

  • Posts: 2089
  • Joined: Tue Oct 27, 2015 10:59 am
  • Location: austria - no kangaroos +g*

Re: massive shift in green color on BMPCC4K

PostFri Jan 07, 2022 3:40 pm

I can confirm that rawlite is the best option
hans does a really good job!

https://rawlite.com/
cMP 5.1 2x3,46/96GB/2x2TB SSD/4x4TB/7101A 4x2TB 970evo+/HP1344/BMD4k/RadeonVII
macOS 12.6.3
BMPCC 6k pro (7.9.1)
meike s35 cine 25mm, 35mm, 50mm, 75mm
resolve studio 18.1.4
mini panel
speed editor
desktop video 12.1
intensity pro 4k
atem extreme (8.6.1)
Previous

Return to Cinematography

Who is online

Users browsing this forum: Google [Bot] and 97 guests