C300: Share Camera Profile Between C300mk2 and C200?

[Mark Smith]

I have a shoot coming up where I may have to match a 300 Mk2 and a C200. CP sharing between the 300Mk1 & Mk2 wasn't possible, and I am wondering if I can take my 300mk 2 profile which I like very much and read it into a C200. Does C200 even have a 'production camera' matrix setting?

 

[puredrifting]

I haven't tried it but I don't think it will work. Maybe someone who has access to both can try it? Attached is an excerpt from the manual, yes, it does have Production Camera Matrix.

 

[Mark Smith]

Well Maybe I can manually get the 200 into the zone by replicating my 300 Mk2 CP with the 200 Menu choices. They have the same sensor.....

 

[Run&Gun]

There's more to it than them "just having the same sensor". A setting of "5" on setting X on one camera doesn't necessarily mean the same thing(net the same result) on another camera if you set setting X to "5" on it. You can try to manually port over the settings, but there's no guarantee that it will give you the same results. When I bought a P2 VariCam I wanted to set it up with the exact same scene files that I used in my tape Vari, but it wasn't possible to port the settings, they were different cameras. I did manually enter each setting that had a corresponding setting on the other camera by hand to give me a baseline of sorts, but it ended up taking me many, many months of trial and error to get the P2 Vari to visually match the original.

 

[Mark Smith]

Looks like the 200 lacks some of the gamut and gamma choice that the 300 has but I think there is enough commonality between camera menus that it will be possible to dial in common menu settings.

 

[jcs]

I just tested stock settings of the C300 II against my settings optimized for skin tones (Filmic Skin) for our latest green screen shoot (haven't done one in a while): I no longer recommend the ARRI settings (Canon Log 2, Cinema Gamut, Production Matrix), not even for green screen. It has a magenta bias, and skin tones are compromised, with no advantages in highlights, even when using 12-bit RGB 444. Perhaps related to something PP CC is doing with their ARRI LUTs? While I'm sure there are applications where wide color gamuts are necessary, I've realized that sticking with Rec.709 provides the most predictable results. I think to really use wide color gamuts correctly, we need to use something like Resolve, where after color grading, a (correct) wide-gamut to Rec.709/2020 node is applied at the end of color correction/grading graph. If doing HDR, Rec.2020 would be fine (which YouTube now supports- in order to watch you'll currently need an HDR HDTV or certain mobile devices (desktop support now available with Windows 10).

The closest stock settings that I've found provide pleasing skin tones is: <Take your pick of gammas, WideDR is good for little or no post needed>, Cinema EOS Original, and Rec.709. If you set both cameras this way, I'd expect the cameras can be matched fairly well in post. You could explore color matrix changes, however since they are different camera systems, that could take a long time to verify that your changes are bringing the cameras closer together vs. farther apart (comparing recorded outputs with scopes and color charts to really be confident of no surprises in post: comparing live HDMI/SDI output (even with scopes) is not sufficient).

 

[Mark Smith]

One Edit I made in my camera profile was in other settings - white balance . I dialed down red and blue gain a couple of points - minus 2 or minus 3 . I think minus 3 could be too much, but I have shot with that profile and was gnerally pleased the results. My top level choices were EOS Gamma, production matrix, cinema gamut. The point was to arrive at a ready to edit CP that could rival C-log3, color wise but without the need for additional post work. If one were going for maximum DR the EOS gamma wouldn’t be the choice, and that wasn’t my goal. I think DR has become something of a fools errand and a point of argument with little value, because you know what ? Almost no one can see that 14th and 15th stop of DR in most delivery methods .

 

[jcs]

One Edit I made in my camera profile was in other settings - white balance . I dialed down red and blue gain a couple of points - minus 2 or minus 3 . I think minus 3 could be too much, but I have shot with that profile and was gnerally pleased the results. My top level choices were EOS Gamma, production matrix, cinema gamut. The point was to arrive at a ready to edit CP that could rival C-log3, color wise but without the need for additional post work. If one were going for maximum DR the EOS gamma wouldn’t be the choice, and that wasn’t my goal. I think DR has become something of a fools errand and a point of argument with little value, because you know what ? Almost no one can see that 14th and 15th stop of DR in most delivery methods .

Pulling R+B = magenta, which would certainly help the magenta bias issue. In the past a lot of light sources had a green bias (fluorescent, LED), though recently quality is much better, and no more green bias (some LEDs went too far and have a magenta bias; again recent designs don't have any major biases). So perhaps that explains why Canon defaults to a magenta bias?

Regarding DR- I was suggesting Cinema EOS Original for the Color Matrix, and WideDR for gamma, along with Rec.709 color space (or Rec.2020 if needing HDR). I haven't done extensive testing, however Filmic Skin (and presumably the recommended settings here) works well with Canon Log 2 when one needs max DR, and a simple 1D LUT or manually creating a custom curve works well. I agree regarding max DR- it's nice to have but sometimes it's necessary to let the sky/windows/lights blow out as the subject is more important. Skin tones >>> preserving highlights if we have to chose between the two.

 

[Mark Smith]

I’m less concerned about the specific characteristics of a particular light and more interested in a general approach to a color scheme because I shoot in all kinds of situations, daylight , tungsten and every other kind of lighting so I shape my CP’s in that direction. I deally the camera will see a lighting situation the way my eye sees it - that’s a good place to start.

 

[puredrifting]

Looks like the 200 lacks some of the gamut and gamma choice that the 300 has but I think there is enough commonality between camera menus that it will be possible to dial in common menu settings.

I've done it, they looked very different IMHO with identical settings (as close as they could be made). The C300 MKII has punchier, more contrasty colors and in comparison, my C200 looked pretty different, I've tried this several times with different setups. Sensor may be the same but I am convinced that the DSP and other processing circuitry aren't. Not saying they couldn't be matched very closely but setting them up the same way won't give you that necessarily. FWIW, I used to shoot a lot with the C100 MKI and when I rented C100 MKIs for two and three camera shoots, even sharing identical setups, the cameras looked pretty different at times, I was surprised.

 

[jcs]

Sunlight, tungsten (black-body emitters) are the easy cases for cameras (continuous spectrum). CFL, LED etc. are spiked spectrums (LEDs are getting smoother), which are quite challenging for cameras.

I do the same- develop generalized solutions for all lighting conditions: predictable and accurate results as the eye sees is the goal. The mirror test where you compare the camera and settings to a mirror can be quite revealing (e.g. when using yourself as a test subject for skin tones). Part of the challenge is the camera LCD doesn't match the HDMI output, and there is currently no way to calibrate all these displays together.

Someday cameras will be much easier to match: I'm sure it's on a future roadmap. As will be self-calibrating display systems and HDTVs... Until then we get to fix everything in post (though I can fairly closely match HDMI output on the 1DX II and C300 II).

 

[cpreston]

I've gone through and created custom color matrix matches for the different cinema cameras that I have owned using picture by picture on a production monitor while shooting a color chart. It is a pain. I don't have a C200 so I have no idea how that compares to the C300 II. The biggest issues I have found is that manual white balance is not the same between cameras. 3200K on one camera matches 3000K on another for white balance. The next big problem is that trying to match the magenta bias and the transition from blue to red is near impossible under different LED and CFL lights. The cameras are just making a guess at how the eye interprets these colors and C300 II always guesses that we see purple when we are actually seeing blue. If you ever get stuck shooting an event with blue LED lights, the problem becomes really apparent.

 

[jcs]

I've gone through and created custom color matrix matches for the different cinema cameras that I have owned using picture by picture on a production monitor while shooting a color chart. It is a pain. I don't have a C200 so I have no idea how that compares to the C300 II. The biggest issues I have found is that manual white balance is not the same between cameras. 3200K on one camera matches 3000K on another for white balance. The next big problem is that trying to match the magenta bias and the transition from blue to red is near impossible under different LED and CFL lights. The cameras are just making a guess at how the eye interprets these colors and C300 II always guesses that we see purple when we are actually seeing blue. If you ever get stuck shooting an event with blue LED lights, the problem becomes really apparent.

So true. I've got to say I found the old 5D2 and 5D3 produced much better color straight from the camera (with minor tweaks to the built in profiles), and 5D3 ML RAW was exceptional: first with Adobe CR processing, then later just using MLRawviewer to convert to ProRes. After "upgrading" to a "Pro" camera, the FS700, I found skin tones were much harder to deal with. GH4, A7S I, A7S II (better): all more challenging skin tones vs. the old 5D3. Stock, the 1DX II doesn't look as good as the 5D3 either. The C300 II (also tweaked) can look pretty good, but to be honest I would prefer the 5D3 look (tweaked settings or even better ML RAW in ProRes look). The C200 footage I've seen online so far also doesn't have color for skin tones that matches the 5D3. Perhaps folks haven't mastered how to get color right yet using RAW; limited tools etc. This all goes back to the point that cameras, especially at this level, should come stock with color systems that just work, and don't require endless hours of testing, tweaking, searching the internet for solutions, etc. Cameras are now very sophisticated computer systems- they should be able to provide perfect white balance, in the same way pro color meters can sense color. I'm sure built-in color meters are on the roadmap for future releases Maybe I should write an iPhone X camera app which uses the depth sensor to simulate shallow DOF and also provides full control over color and just shoot with that ? Imagine if Canon or Sony provided SDKs which allowed us to make vastly improved camera systems?

 

[Run&Gun]

I've done it, they looked very different IMHO with identical settings (as close as they could be made). The C300 MKII has punchier, more contrasty colors and in comparison, my C200 looked pretty different, I've tried this several times with different setups. Sensor may be the same but I am convinced that the DSP and other processing circuitry aren't. Not saying they couldn't be matched very closely but setting them up the same way won't give you that necessarily. FWIW, I used to shoot a lot with the C100 MKI and when I rented C100 MKIs for two and three camera shoots, even sharing identical setups, the cameras looked pretty different at times, I was surprised.

Yep. I've seen it most of my career. Sometimes you can get same model cameras to match with the same settings and sometimes they can be "off". Lots of factors/variables. 'Back in the day', I had my Betacam set-up by Roger Macie and I shared my set-up with a friend who had the same camera, because we did a lot of shooting together. They matched(until he had a board replaced). Did a two-cam live show with my tape Vari years ago. Shared my custom scene file with the other camera and and they matched dead-nuts-on(same monitor, split-screen). They couldn't have looked any closer if you had had an engineer painting them. But my two P2 Vari's... I created a set-up in one and then put it into the other and they are REALLY REALLY close(most people would never see it), but I can see a slight difference. I've also seen variations in the image with my scene file in other guys P2 Vari's that I've shared it with. And to go with the Canon theme, I have seen variations in C300's running the same set-ups(usually with custom set-ups, more-so than the built-in scene files).

Something I've tried to start doing to help them in post on multi-cam shoots is to have all cameras shoot the same "chip chart" at the same time in the same position under the same light before we start. I even try to shoot the chip chart on single cam shoots, if I'm shooting log.