massive shift in green color on BMPCC4K

At the end of the day, I should be able to point my camera at a green object, white balance my camera, and get footage straight out of the camera where the green object looks green - should I choose to do so. I currently cannot do that. If anyone can tell me how I can achieve that with ONLY my camera, my lens, a 5500K LED light, and a gray card for white balancing and nothing else - I'm listening. That should be all that is required for a simple task such as taking a shot of a green object. If I cannot achieve something as simple as that using just those items - then it's time to accept the fact that this camera has a major flaw.

Isn't that what the 'video' setting is for in the Dynamic Range menu so you can get out of the box ready to go footage? It's not really rational to forget that Resolve exists as the camera and software go together. It's a 'cinema camera' so it captures as much information in the scene as possible in terms of tone, colour with the expectation that the end user will want to create their own look and not have it dictated to them. PLenty of people use these cameras as they don't like the baked- in looks from other cameras (apart from Canon - Obviously).

At the end of the day, I should be able to point my camera at a green object, white balance my camera, and get footage straight out of the camera where the green object looks green - should I choose to do so. I currently cannot do that. If anyone can tell me how I can achieve that with ONLY my camera, my lens, a 5500K LED light, and a gray card for white balancing and nothing else - I'm listening. That should be all that is required for a simple task such as taking a shot of a green object. If I cannot achieve something as simple as that using just those items - then it's time to accept the fact that this camera has a major flaw.

Isn't that what the 'video' setting is for in the Dynamic Range menu so you can get out of the box ready to go footage? It's not really rational to forget that Resolve exists as the camera and software go together. It's a 'cinema camera' so it captures as much information in the scene as possible in terms of tone, colour with the expectation that the end user will want to create their own look and not have it dictated to them. PLenty of people use these cameras as they don't like the baked- in looks from other cameras (apart from Canon - Obviously).

That's a pretty poor excuse to justify an actual problem.
The issue should be solved by either having a preset in resolve, or a firmware fix that deals with the issue at the in-camera level.

As soon as I get back to my cameras I’ll take a bunch of photos and push them through the various stills processors out there

One of the reasons I love capture 1 for stills over Lightroom / PS is the way they “interpretative” my stills.

They read and Identify the camera by the RAW file and then apply their own internal
Profile to generate a default “look”

Nine times out of ten that look is better or nicer or more accurate or pops or whatever more than the way other image engines generate their starting point.

And that’s the thing. It’s a STARTING POINT. I then expect I’ll go and make grades over the top of that starting point.

The default VIDEO is an interpretation of a default look.

You can just as easily make your own one encapsulated as a LUT.

It’s just a starting point that conforms to a few basic gamut and contrast standards with someone’s SUBJECTIVE judgement of colour.

If you can do better then knock yourself out.

All the default / auto / out of the box looks are variations of this.

I think this thread has shown that workflow is very important to getting the most out of the images from any sensor and that there are several ways to achieve a similar result.

And none of them are “right” or “wrong”.

JB

Would it be OK if you bought a new car, and were told you have to get a tune up first, before it's ready for standard use?
And the excuse was that the reason it's not tuned properly, is because everyone uses their car differently, so everyone requires a different tuning.

Same thing.

Again. Everyone here has pretty much admitted that out of the camera there's already an issue.
You should have to do correction, just to arrive at a proper base starting point m
That's just simply wrong.
So stop making excuses for it.

Would you start driving a car without getting a few lessons and getting a test done to make sure you understand how to drive ?

JB

Would you start driving a car without getting a few lessons and getting a test done to make sure you understand how to drive ?

JB

Knowing how to drive will have zero impact on the lack of a proper, basic tuning.

And that's my point.
We shouldn't have to be dealing with fixing an issue, befire we can start working on the proper grading

At the end of the day, I should be able to point my camera at a green object, white balance my camera, and get footage straight out of the camera where the green object looks green - should I choose to do so. I currently cannot do that. If anyone can tell me how I can achieve that with ONLY my camera, my lens, a 5500K LED light, and a gray card for white balancing and nothing else - I'm listening. That should be all that is required for a simple task such as taking a shot of a green object. If I cannot achieve something as simple as that using just those items - then it's time to accept the fact that this camera has a major flaw.

Isn't that what the 'video' setting is for in the Dynamic Range menu so you can get out of the box ready to go footage? It's not really rational to forget that Resolve exists as the camera and software go together. It's a 'cinema camera' so it captures as much information in the scene as possible in terms of tone, colour with the expectation that the end user will want to create their own look and not have it dictated to them. PLenty of people use these cameras as they don't like the baked- in looks from other cameras (apart from Canon - Obviously).

My point exactly!! I said that exact same thing in my last post:

"Even when applying the BMD LUTs that come with the camera which are supposed to give you fully graded footage straight out of the camera don't even come close to what the eye is seeing."

Of course it's not rational to forget that Resolve exists and of course most people will use this camera to get some really great cinematic footage which will be put through some intense color grading workflows. But what about people who don't want to use Resolve and use another NLE? What about people who don't want to have to color grade their footage just to get a simple shot of a green object? Or what about those situations when a client is sitting next to me and is looking at my footage which looks absolutely nothing like the product that they hired me to film?? Or when I have to deliver the raw files to an editor and I have no say in how they're color graded and that editor gets the files and they're unusable?? Or when livestreaming? Or if I don't want to use BRAW and instead want to save space and shoot ProRes Proxy? There are soooo many situations where this camera completely fails us. Just because it can do some complicated stuff does NOT mean that it should completely ignore something as simple and basic as properly displaying the color green!

Currently, the BMP4K claims that it can do that. That's why there is video mode. That's why there is ProRes proxy. That's why there is a built in LUT. Well, it can't. It cannot even produce a file that is remotely useable in the situations I described above.

Would you start driving a car without getting a few lessons and getting a test done to make sure you understand how to drive ?

JB

Knowing how to drive will have zero impact on the lack of a proper, basic tuning.

And that's my point.
We shouldn't have to be dealing with fixing an issue, befire we can start working on the proper grading

If you don’t know how to drive then how do you know what to tune ?

We can do this all day.

You’re not happy with the default look. I get it.

I’d never use the default REC 709 transform either.

But it’s grading 101 to address this. You can even make it easier by tweaking the look and loading your own LUT. This is pretty basic stuff we’re talking about here.

JB

Would you start driving a car without getting a few lessons and getting a test done to make sure you understand how to drive ?

JB

Knowing how to drive will have zero impact on the lack of a proper, basic tuning.

And that's my point.
We shouldn't have to be dealing with fixing an issue, befire we can start working on the proper grading

If you don’t know how to drive then how do you know what to tune ?

We can do this all day.

You’re not happy with the default look. I get it.

I’d never use the default REC 709 transform either.

But it’s grading 101 to address this. You can even make it easier by tweaking the look and loading your own LUT. This is pretty basic stuff we’re talking about here.

JB

Completely non valid comparison.
And having to first fix at a base level, cis pretty unacceptable in my books.

I actually love the overall look of Blackmagic cameras.
But why they don't fix this very base level issue even in camera, is a bit baffling to me .

Just as baffling as how people will make excuses for the existing problem.

Knowing how to drive will have zero impact on the lack of a proper, basic tuning.

And that's my point.
We shouldn't have to be dealing with fixing an issue, befire we can start working on the proper grading

What, exactly, is "proper, basic tuning." Log footage assumes there's going to be a grade, otherwise why use log? And log requires a transformation, into an accepted color space, for grading.

The transformations the companies provide (LUTs, or the embedded math in Resolve Color Management or ACEs) bring the footage (for most of us) from log into rec. 709 space. But they all do it differently, and the result is not supposed to be final, or even necessarily pleasing. (And BTW, some people choose to do this normalization manually, with curves and other grading tools. Nobody stopping you....).

In your case, I'd suggest ACEs, because ACEs tends to do the most work on the image. BMD LUTs and Resolve Color Management do relatively little, but with the advantage that the Resolve color grading controls work as they would when using DaVinci YRGB color -- they work as expected.

Now, the issue Eugenia is raising is something else. She claims that BMD normalizing math stinks. But for the life of me, I don't understand Eugenia's position, mainly because the grades she's posting, including samples which she says represent the traditional workflow, look all wrong to my eyes. No idea what's going on here.

Would you start driving a car without getting a few lessons and getting a test done to make sure you understand how to drive ?

JB

Of course not. That's why I'm not just turning the camera on and expecting to get a good result. First, I use a VERY good light such as the Apurture LS 300D. It IS a professional light. Is it 100% correct? No. But it is 98% correct. After that, I also properly white balance my camera and properly set my recording format and LUT. I have taken way more than a few lessons at this point. I should be able to start the car and drive it in a straight line for 100 feet on an empty road. Well, I can't. This "car" will throw me off the cliff the second I turn it on - even after it's been properly tuned up according to the manufacturer's standards.

Would you start driving a car without getting a few lessons and getting a test done to make sure you understand how to drive ?

JB

Of course not. That's why I'm not just turning the camera on and expecting to get a good result. First, I use a VERY good light such as the Apurture LS 300D. It IS a professional light. Is it 100% correct? No. But it is 98% correct. After that, I also properly white balance my camera and properly set my recording format and LUT. I have taken way more than a few lessons at this point. I should be able to start the car and drive it in a straight line for 100 feet on an empty road. Well, I can't. This "car" will throw me off the cliff the second I turn it on - even after it's been properly tuned up according to the manufacturer's standards.

Did you see the comments above concerning what happens to that peculiar color when the brightness is raised -- as it is in the sample you posted a few miles back? Whether this satisfies you I don't know, but no sensor/workflow is perfect. You get strange artifacts with all of them, and you hit on a very strange one, with this odd color

Would you start driving a car without getting a few lessons and getting a test done to make sure you understand how to drive ?

JB

Of course not. That's why I'm not just turning the camera on and expecting to get a good result. First, I use a VERY good light such as the Apurture LS 300D. It IS a professional light. Is it 100% correct? No. But it is 98% correct. After that, I also properly white balance my camera and properly set my recording format and LUT. I have taken way more than a few lessons at this point. I should be able to start the car and drive it in a straight line for 100 feet on an empty road. Well, I can't. This "car" will throw me off the cliff the second I turn it on - even after it's been properly tuned up according to the manufacturer's standards.

Amen to that.

Would you start driving a car without getting a few lessons and getting a test done to make sure you understand how to drive ?

JB

Of course not. That's why I'm not just turning the camera on and expecting to get a good result. First, I use a VERY good light such as the Apurture LS 300D. It IS a professional light. Is it 100% correct? No. But it is 98% correct. After that, I also properly white balance my camera and properly set my recording format and LUT. I have taken way more than a few lessons at this point. I should be able to start the car and drive it in a straight line for 100 feet on an empty road. Well, I can't. This "car" will throw me off the cliff the second I turn it on - even after it's been properly tuned up according to the manufacturer's standards.

You have your basic assumptions wrong. Light + WB + LUT is just a starting point if you have a BM cinema camera.

Knowing how to drive will have zero impact on the lack of a proper, basic tuning.

And that's my point.
We shouldn't have to be dealing with fixing an issue, befire we can start working on the proper grading

What, exactly, is "proper, basic tuning." Log footage assumes there's going to be a grade, otherwise why use log? And log requires a transformation, into an accepted color space, for grading.

The transformations the companies provide (LUTs, or the embedded math in Resolve Color Management or ACEs) bring the footage (for most of us) from log into rec. 709 space. But they all do it differently, and the result is not supposed to be final, or even necessarily pleasing. (And BTW, some people choose to do this normalization manually, with curves and other grading tools. Nobody stopping you....).

In your case, I'd suggest ACEs, because ACEs tends to do the most work on the image. BMD LUTs and Resolve Color Management do relatively little, but with the advantage that the Resolve color grading controls work as they would when using DaVinci YRGB color -- they work as expected.

Now, the issue Eugenia is raising is something else. She claims that BMD normalizing math stinks. But for the life of me, I don't understand Eugenia's position, mainly because the grades she's posting, including samples which she says are the traditional workflow, look all wrong to my eyes. No idea what's going on here.

Let's leave Eugenia out of this. Let's also stop talking about your color grading workflows or LOG. As I have explained now multiple times - I love that I can shoot LOG style on this camera and I think the BMP4K is a great camera that gives us some amazing footage to grade. But there are times when shooting RAW or LOG is not ideal or practical. I should be able to apply the video LUT, choose to record it directly to my file, and be able to deliver a file that looks at least 95% like the real thing. And I currently cannot do that. How are you not getting that simple concept?

But there are times when shooting RAW or LOG is not ideal or practical. I should be able to apply the video LUT, choose to record it directly to my file, and be able to deliver a file that looks at least 95% like the real thing. And I currently cannot do that. How are you not getting that simple concept?

So your beef is that the camera fails as a point and shoot, in video mode? Who knew.

This time I truly do give up.

Would you start driving a car without getting a few lessons and getting a test done to make sure you understand how to drive ?

JB

Of course not. That's why I'm not just turning the camera on and expecting to get a good result. First, I use a VERY good light such as the Apurture LS 300D. It IS a professional light. Is it 100% correct? No. But it is 98% correct. After that, I also properly white balance my camera and properly set my recording format and LUT. I have taken way more than a few lessons at this point. I should be able to start the car and drive it in a straight line for 100 feet on an empty road. Well, I can't. This "car" will throw me off the cliff the second I turn it on - even after it's been properly tuned up according to the manufacturer's standards.

You have your basic assumptions wrong. Light + WB + LUT is just a starting point if you have a BM cinema camera.

We are literally saying the same exact thing at this point. Light + WB + LUT SHOULD give me a starting point. Well it doesn't. My brightness was very much in the correct ball park to give me something usable. Also, please stop talking about that specific Ryobi color because as we've seen from the MANY examples on here already this is the case with ALL green, ALL brightness situations, BOTH indoors using LED's and outdoors using natural sunlight.

It’s impossible to create a look out of the box that will look fantastic for all shots and be perfect in every situation.

Could it be better ? Yes.

Should it be better ? Yes.

Is it so easy to fix that I can’t believe we’re still talking about it ? Yes.

Would you ignore the look anyway because you’re going to be doing your own colour correction ? Yes

Can that extra work be something that’s simple to apply ? Yes.

The whole reason the after market LUT market exists evacuate there’s a zillion different ways a shot can be “better”.

Personally I’d rather BMD engineers keep making more meaningful improvements than working on making the generic look better.

If you want a baked in look out of the box you have two choices.

Do the same hard work you expect BMD to do and make your own look that satisfies you.

Buy a camera that does give you that look out of the box.

No matter what camera I’m using, I’m always going to be grading that image in some way so the starting point, once you’ve spent some time with the files, is 100 irrelevant.

If you want to shortcut that, then maybe consider a different camera.

JB

But there are times when shooting RAW or LOG is not ideal or practical. I should be able to apply the video LUT, choose to record it directly to my file, and be able to deliver a file that looks at least 95% like the real thing. And I currently cannot do that. How are you not getting that simple concept?

So your beef is that the camera fails as a point and shoot, in video mode? Who knew.

This time I truly do give up.

I do think that’s what’s been asked for.

A better baked in no grade look.

JB

But there are times when shooting RAW or LOG is not ideal or practical. I should be able to apply the video LUT, choose to record it directly to my file, and be able to deliver a file that looks at least 95% like the real thing. And I currently cannot do that. How are you not getting that simple concept?

So your beef is that the camera fails as a point and shoot, in video mode? Who knew.

This time I truly do give up.

You can belittle the issue all you want. But I invested A LOT of time and money in building my BMP4K rig. So yes - I am expecting to be able to do something as simple as take a shot of a green object if I needed to and perform a basic "point and shoot" shot. I'm also expecting to be able to do all the things that all you "pros" are doing with this camera. That's what this camera claims it can do. Well it can't. And I'm disappointed that BMD has not made that clear from the beginning or haven't provided a solution to this problem. They shouldn't be making claims that they can't back up.

Ivan - why exactly did you choose the P4k for product photography and where did you get the information from that it would be suitable for this kind of work because even if you ignore the color side of things it would seem a rather unsuitable camera for other reasons as well. The 1:1 subsampling sensor without an OLPF is a terrible choice for product shots where fine detail will be rendered with aliasing and textures with moire. If you are filming printed packging it's even worse with regards to moire on the print dot patern.

Ivan - why exactly did you choose the P4k for product photography and where did you get the information from that it would be suitable for this kind of work because even if you ignore the color side of things it would seem a rather unsuitable camera for other reasons as well. The 1:1 subsampling sensor without an OLPF is a terrible choice for product shots where fine detail will be rendered with aliasing and textures with moire. If you are filming printed packging it's even worse with regards to moire on the print dot patern.

Now we're talking! You are starting to get it. I did not purchase this camera for product videography. I purchased this camera to be my main camera to shoot MANY different videos with - not just one type of video. I researched this camera and its workflow for a long time before I pulled the trigger and bought the camera along with $1000's more in accessories. I thought to myself "Great, I can film all these amazing projects with it and when I need to do something simple or when I don't want to fill up my hard drive I can use ProRes proxy and use the video LUT to get the shot quickly. Then when I'm ready to dive in deeper and create more complicated stuff I can grow with this camera" I've shot some really great footage with it. I was happy with it until I needed to do something as basic as what I described above. Then reality hit me hard in the face.

When you buy a fully spec-ed out MacBook Pro you still expect that you'd be able to use it for email and web browsing, right?
When you buy a Maserati, you still expect to be able to run a quick errand with it if you needed to, right?
So why can't I take a quick shot of something with my P4K and be on my way which is what the camera and its features suggest I should be able to do?

So why can't I take a quick shot of something with my P4K and be on my way which is what the camera and its features suggest I should be able to do?

But what's stopping you from making your own grade that you'd be happy with ? It takes such little effort ?

JB

So why can't I take a quick shot of something with my P4K and be on my way which is what the camera and its features suggest I should be able to do?

But what's stopping you from making your own grade that you'd be happy with ? It takes such little effort ?

JB

Because you shouldn't have to.
Again, the base level color output should be correct.

Period.

Man, why people keep making excuses is truly beyond me.

So why can't I take a quick shot of something with my P4K and be on my way which is what the camera and its features suggest I should be able to do?

But what's stopping you from making your own grade that you'd be happy with ? It takes such little effort ?

JB

Because you shouldn't have to.
Again, the base level color output should be correct.

Period.

Man, why people keep making excuses is truly beyond me.

Because I shouldn't have to. Because I was told I wouldn't have to. Because as we've seen from all the other people who have attempted to do this before me - the results are just "meh". So it's not that simple to do.

I do think that’s what’s been asked for.

A better baked in no grade look.

I'm curious: does any cinema camera on the market offer a usable "baked-in no-grade look?"

Because you shouldn't have to.
Again, the base level color output should be correct.

Period.

Man, why people keep making excuses is truly beyond me.

I don't think any camera will give you this out of the box in all situations for all colours.

Is that making excuses ? I've already said it could be better. I don't use it because it's not good. But if I put in a TINY but of work i can THEN get something better than I can get from a lot of other cameras.

Still making excuses ? Why does someone trying to explain a difference in workflow turn into attacking the person rather than focusing on the issue ?

But if you want to get the most out of your images, shot by any camera, you have to put in some effort.

If you want something that looks good from the get go and you have no desire to learn how to get more from it then I don't think you're going to find any satisfaction here.

JB

I do think that’s what’s been asked for.

A better baked in no grade look.

I'm curious: does any cinema camera on the market offer a usable "baked-in no-grade look?"

Many are better at it, but none of them are perfect.

Again, presuming we're comparing the basic REC 709 transform.

A lot of people like how the Arri one works, but it also has issues on blues and greens. It also doesn't stop a lot of people from using ARRI processes on BMD cameras in an ACES environment.

It's funny because the footage that Malera did that was posted earlier was some of the first ever material shot with the then new 4.6K sensor and when we took it into post, the colourist kept on wanting to apply an Arri based transform because for him it looked better.

Except, we started to notice it was causing other out of gamut issues. So we had to go back to square one and build the grade manually. No short cuts.

This is pretty normal for any new sensor.

I've got a few generations of Leica M and S cameras and everytime there's a new one I have buyers remorse for about 3 weeks while I try all my old grading techniques that are muscle memory on the "new" sensor and they don't work as expected. Eventually you find other ways to "process" the image and make it look nice. Often times, you're doing this blind because it takes a few months for Adobe and Capture 1 to make their profile and add it into the apps.

As an Olympus Visionary i often get access to new Olympus cameras very early and it's very frustrating to have to then go back and use Olympus software to "grade" in because the files can't even be opened by the regular image processing apps because the camera's not announced. About three months after announcement the profiles appear in the usual apps and I can open the photos. I find I can do something in 2 mins that took me 30 mins to previously in the app that I have no experience in.

This is the wrong camera to expect an instant gratification look from.

Maybe go get a Fuji ? They have some great LOOKS built into the camera. Again they're not "accurate" either, but they are a LOOK.

If you really want that you CAN make your own, but it's kind of got to be done by you, because you'll find, nothing is ever going to work in all situations.

JB

Here's some perspective from another forum (and lauding BMD's red)

http://www.dvxuser.com/V6/showthread.ph ... 1986827046

JB

As soon as I get back to my cameras I’ll take a bunch of photos and push them through the various stills processors out there

I appreciate the tests you do on your blog, a lot of time invested there and you are making something public that's rare to see.

My quick test today between the pocket 4k developed in Resolve and Pentax 645z out of the camera JPEG cropped a bit and scaled to 4K, auto wb with a cheap 5600k led and a window.

To use the car analogy, more like automatic transmission vs manual?

pentax_645z_auto_wb.jpg (972.37 KiB) Viewed 12469 times

pocket_4K_prores422hq_film.jpg (900.29 KiB) Viewed 12469 times

As soon as I get back to my cameras I’ll take a bunch of photos and push them through the various stills processors out there

I appreciate the tests you do on your blog, a lot of time invested there and you are making something public that's rare to see.

My quick test today between the pocket 4k developed in Resolve and Pentax 645z out of the camera JPEG cropped a bit and scaled to 4K, auto wb with a cheap 5600k led and a window.

To use the car analogy, more like automatic transmission vs manual?

...

Pretty darn close! I couldn't find the Ryobi green color specs online but LulzBot Green (3D Printer company's color) has been described as being quite close. Color specs for that are:

LulzBot Green
RGB: 193/216/47
CMYK: 29/0/100/0
HEX: #C1D82F
PMS: 382c

Steve

My quick test today between the pocket 4k developed in Resolve and Pentax 645z out of the camera JPEG cropped a bit and scaled to 4K, auto wb with a cheap 5600k led and a window.

Not that close. If you look at the green tapper box and the green plants (and not the Ryobi device), you will see that the Pocket4k has shifted these greens to yellow-green. In other situations, where the light is different, the Ryobi device itself would shift to yellow too (as the OP has found out). The greens in the Pentax are proper greens, while in the Pocket4k are yellow-greens.

My quick test today between the pocket 4k developed in Resolve and Pentax 645z out of the camera JPEG cropped a bit and scaled to 4K, auto wb with a cheap 5600k led and a window.

Not that close. If you look at the green tapper box and the green plants (and not the Ryobi device), you will see that the Pocket4k has shifted these greens to yellow-green. In other situations, where the light is different, the Ryobi device itself would shift to yellow too (as the OP has found out). The greens in the Pentax are proper greens, while in the Pocket4k are yellow-greens.

And the Pentax is overal more vibrant

I still cant see whats the conclusion of all this debate. So far, it only looks like kids fighting over a candy saying "your candy is sweeter than mine".
Please, someone enlighten me.

My quick test today between the pocket 4k developed in Resolve and Pentax 645z out of the camera JPEG cropped a bit and scaled to 4K, auto wb with a cheap 5600k led and a window.

Not that close. If you look at the green tapper box and the green plants (and not the Ryobi device), you will see that the Pocket4k has shifted these greens to yellow-green. In other situations, where the light is different, the Ryobi device itself would shift to yellow too (as the OP has found out). The greens in the Pentax are proper greens, while in the Pocket4k are yellow-greens.

The Pentax set an auto WB of 4900K and tint of 29 magenta and cooled off the scene removing the yellows. I could have alternatively manually set the white balance to match the fill light of 5600K and it would have made the scene more yellow in the greens.

The Pocket 4K has the white balance set to 5400K and 14 magenta. I then did an input transform on the color space to "Blackmagic pocket 4K V4." What about lenses? Don't they all have a tint to them? I used a Nikon 50mm 1.4g on the Pocket 4K and Nikon lenses are known to be a little yellow. So it's building up to an unfair comparison, no BRAW, ProRes in YRGB REC709.

The light in the scene is yellowish and I think the Pentax is just making an automatic effort to correct that for you so you don't see as much yellow in the picture. What would happen to the greens if I manually set the white balance of the Pentax to 5600K?

I think I've driven the Pocket 4K manually to a more natural result while maintaining the Ryobi Green. I would expect there to be a yellow - brown cast from the floor and the fill light on the plants and the bright green lid. The Pentax took the reds too blue. The red disk and on/off sticker on the Ryobi are more accurate on the Pocket 4K. The Pentax has the reds too blue. The purple bottle? I'm not sure it is as red as in the Pentax either.

Correct hue is variable according to the light reflected on the object being photographed. DSLRS and the Pentax in auto mode aim to change this to represent objects with a specific or known hue but also have settings to leave warm lights warm as you are auto balancing. What do people prefer?

If I put a 5600K pocket led on a face it should look yellow as its yellow to my eye and if you attempt to correct it with a camera to a specific hue well . . it's up to you to make a decision about what you like for a given scene.

It would have been more fair to test between two different cinema cameras with the same lens and to set them at the same Kelvin. But showing the previous pictures is more a point about correcting a color to specific hues as in the mind of the creator vs the displaying hues accurately as your eyes see it.

Am I saying that right? I don't know I went to art school and I'm not an engineer, most of my ideas about color come from this guy: http://www.josephsantore.com/paintings-1999-2009/

My advice specifically to Ivan is to take the time to read and learn what people are presenting here; John Brawley, Eugenia, all presenting in civil debate and giving away what I would otherwise think would be trade secrets for free!

I guess even if there is no strong IR pollution, IR filter will produce additional color tint and so will always capture better greens. Earlier i done some tests with and without IR filter to compare IR pollution, but i never tested it to compare green colors quality.

There are some differences between ACES and BMD color transform compression. ACES provides more saturated and more contrast look. Some people like it, some don't. The good news - there is no need to switch to ACES project setting. ACES transform Node provides you ACES engine inside YRGB project. Quick and easy.

Don't forget that when you switch in RAW options from BMD film to different color space, you start to use simple matrix-based input profile transform instead of factory calibrated LUT-baced color profile. Without factory calibrated LUT-based color profile greens became even less green. Some examples:

BMD film -> CST to Rec709


BMD film -> ACES -> ACES Transform to Rec709


ACES -> ACES Transform to Rec709

If your expectation is that all cameras would or should look the same then your expectation is not based on education or experience. A logical conversation based on whatever is left seems challenging but who's working these days?

On occasion I match cameras for a living, I'm not saying you should believe that appeal to authority. More directly, that function's very existence is predicated on the fact that cameras don't match. Hundred thousand dollar broadcast camera chains don't match each other out of the box, let alone other cameras. Are Video Engineers unicorns?

The concept of color matching multiple discrete camera systems with varying manufacturers, sensors, debayer algorithms, gammas, gamuts, codecs, and lenses with one-click? Is this the unicorn?

Are unicorns real?

Good Luck

It's a tool...an instrument.
The more you practice with it the better your results.

There are basic tools and instruments that are designed for those who don't want/need the flexibility, control and options to do the more detailed and high end work of some of their counterparts.

There are moe complex instruments that are more fiddly and require more study and skill to master but offer more control of the ultimate performance.

You can say that about everything from synthesizers to lawn mowers.

Don't get why this (other than perhaps a covid lock down) is such a difficult subject to grasp from either type of user? I don't think it is actually...it's just a way to spend some time.

We've gotten the best overall reponse and controllable look from our BM cameras, including the P4K, than our Canons and Sonys - and with surprisingly little learning curve using Resolve. That's my subjective opinion.

The opinions as to what looks "right", are the definition of subjectivity.

The idea that different instruments and tools are designed to be more simple or more complex isn't though.

It's a tool...an instrument.
The more you practice with it the better your results.

.

Except in thus case, the instrument is being sold with pickups that are wired incorrectly

In Cinema5D underexposed tests Pocket4K also looks very different to Pocket6K. More IR magenta/less greens.

Great points.

It is a shame that one is denounced as an apologist or some kind of rabbid fan when those that have been around and have decades of experience are sharing what they know.

I’ve had four Alexa’s lines up and none of them match. It’s routine to have to dial in slightly different white balance. It’s a standard part of camera prep on a big show. It’s often down to even pairing with NDs because, surprise surprise, NDs often don’t match either, Especially at 5+ stops. So you’ll often see a set of NDS that are paired with that camera and each filter will have a G—>M offset written on the tab which the AC dials into the camera when that ND is used.

Every camera is slightly different. Under different lighting conditions different colours are reproduced with different hues that are subjectively what the camera engineering team thinks is “good”.

No one look works for everything.

If you don’t like it, then you do what those that have been doing this for a while do. You make a better one for yourself. You learn your tools and you improve it from the default.

This thread as I think shown that it is indeed possible to get the subjectively “right” colour when the footage is shot properly and that most subjective differences are more to do with workflow than something inherent to the camera itself.

The default VIDEO look as been very poor since day 1 on the BMCC. I don’t think anyone would disagree with that.

I think most users get on with learning what they need to do to make it better.

JB

In Cinema5D underexposed tests Pocket4K also looks very different to Pocket6K. More IR magenta/less greens.

I wouldn’t be surprised if that was a copy to copy variation rather than something indicative of P4k Vs P6k. Did they test more than one copy ?

I’m sure you’ve seen this in your own extensive testing where the same camera can have quite different base looks ?

JB

Interesting question. I have only two BMMCCs. I can try to test them both side by side with same lens and with/without IR filter to see color variation.

IThe opinions as to what looks "right", are the definition of subjectivity.

You probably mean me.... It would not have been difficult to raise specific objections to the footage in question, but the argument was over process, not anyone's actual work. Going "there" seemed pointlessly provocative, even more so when it wasn't clear if what was posted was meant as a final result or a preliminary one.

The alternative was saying nothing at all, which was frustrating because nobody appeared to want to state the obvious: that whatever the merits of the proposed new method, the examples provided didn't measure up to the claims being made for it. Subjectively, anyway.

Here's a fun one. I couldn't coax any more accuracy from the Melara method. Better gamma curve for sure but it wasn't swinging the vectors one way or another. So.....

What would it look like if you transformed SMPTE bars into Log C gamma/gamut. It would look like this.

709 to Log C.png (92.62 KiB) Viewed 12257 times

You can see what's going on under the hood. If you don't know what you're looking at essentially anything over 584 is going to clip once you apply the Log C transform.

Now what if instead of matching to 709 targets you matched to the Arri Log C targets generated by the Log transform, and then used the Arri transform? You'd get this.

Log C.png (864.37 KiB) Viewed 12257 times

These were all done using only the chart and never looking at the Ryobi box. First 2 no surprises, I took a shot at manual correction, close but with such a loosely defined gamma the colorchecker charts aren't accurate. The last sample is manually matching the P4k Log to Log C then using the Arri transform.

Good Luck

The last sample is manually matching the P4k Log to Log C then using the Arri transform.

That's really good.

In Cinema5D underexposed tests Pocket4K also looks very different to Pocket6K. More IR magenta/less greens.

I'm finding that the Hoya IR Cut that I have and used quite a bit with the Micro Cinema camera is not really making much difference on the Pocket 4K. I'm really amazed at the IQ of the 4K sensor without having tried the 6K. I was used to seeing some IR pollution on the edges of the color chart in the Micro cam.

Here's another BRAW clip indoors if you want to download it with a DNG, I'm shooting it off my desk, so a little shaky, but I can then look at the screen and the scene at the same time. Maybe more fun to look at than the Ryobi?

https://we.tl/t-iEnexeJjmF

Pocket_4K_Poo_Doo.jpg (951.99 KiB) Viewed 12245 times

I'm finding that the Hoya IR Cut that I have and used quite a bit with the Micro Cinema camera is not really making much difference on the Pocket 4K.

Speakign of IR Cut filters.
Did a test with one yesterday I bough on ebay.

It literally went staright into the garbage when I compared the footage.'
Hahahahahaha

Same results here. Uncorrected BM input profile is weakest greens. Color Checker correction makes greens look better. Film emulation LUT makes greens even more vibrant and nice looking. Note that Provia100 is very "lightweight" LUT that adds very minor correction. So greens are always off. Have no idea why over all 7 years of development and camera generations it was so complicated for Blackmagic team to shoot proper larger sized test target like IT8.7 or X-Rite ColorChecker SG and build better looking input color profile for DNG and BRAW files?

The opinions as to what looks "right", are the definition of subjectivity.

The key word here is opinion, a proper (Chroma DuMonde) chart is display referenced, "right" and wrong aren't subjective within a defined gamma/gamut, It's correct or it's not.

Good Luck