Does camera color science really matter?

We are always talking about camera color science, but does it really matter? This video may or may not debunk camera color science but it DOES feature my cat Tini, so that's pretty cool.

***DISCLAIMER: I wanted all footage to feel like tungsten house lights. The color checker will do correct white balance but I intentionally wanted the warmer look.

My Gear:

#ZCAMF6 #BLACKMAGIC #FUJIXT3

[robertulrich]

color checker won't work if you aren't shooting for the same exposure of grey the card wants. also, color checker can't fix shadows and highlight's hues because it doesn't have enough swatches. you can kinda do a hack by racking exposure while shooting the exposure racking in real time and then in the vectorscope, use a luma matte to control the tonal range of hue vs hue to get perfect linear response to match all the cameras.(i.e. the vectorscope will have straight lines across the hues, not curved.

[BryceMcNabb]

Kinda funny to me that we are still having conversations like this even after Zacuto’s camera shootout of 2012. That was when the majority of people couldn’t pick a favorite, the GH2 beat out the Arri Alexa, and Coppola favored the Canon 7D image above the rest. That we keep harping on minute image quality variations tells me one thing: marketing works.

[TonyDae]

If you have enough data and a proper color checker you can make any color science match any other color science with the only limitation being your technical skill in post production. 8 bit recordings will require a proper look and color baked into the file in order to ensure "good" color whereas 12bit log encoded color doesn't.
---
One of the major issues in my opinion is the way comparisons between different cameras are done and often what people compare are the rec709 initial transforms without trying to match the colors or tone mapping between two cameras. The color checker point here is very important as so many (especially youtubers) tend to ignore the importance of color checkers. Along with this is the A/B wipe comparison where the goal is to wipe back and forth between two images specifically for the purpose of spotting differences instead of seeing the image for what it actually is. There are always going to be some kind of difference between two images, however that should not in my opinion be the most important thing. The important thing is how the images look in sequence WITHOUT comparing with the intent of specifically trying to find problems or differences. Does the image fit the mood/feel/look intended or doesn't it? Can you see a problem when watching the sequence on its own merits or not? Will a regular viewer of your content ever see the "problems" in the image that you see or not?
---
In the VFX field we called the incessant need to A/B wipe with an eyeball 3 inches from the screen to spot problems "pixel f***ing". Often if you have to A/B wipe to see anything then there is not a relevant difference and the result is often ballooned cost for a placebo that pleases a supervisor. Buying cameras for color science when shooting 10 bit log or raw codecs is often one of these placebos and it can get very expensive.

[jasonnowwhat]

It’s not whether or not they can match, I’m most concerned with how long it takes to match in post. Shooting 3 cameras (canon) that shoot c log surely must be easier to match than 3 separate cameras, no?

[DirectedbyJalen]

What I appreciate most about this channel are the much more experienced cinematographers and filmmakers all brought together having a discussion. There's soooo much to learn and this channel is the best to just sit back and be a full-time student! Keep it up, Spenser!

[AxTechs]

before I clicked on this video I said audibly "that's a great question!"

[VO_Blather]

Spotted the intro scene camera color science differences on 1st viewing - in no particular order: Beer Colored, Less Beer Colored, Very Beer Colored. Not crazy about the hoppy LUT of your channel, but absolutely love what you're doing overall.

[TheFilmGuyOfficial]

it's not about colour science per say it's about
1# How far you can push the colours in the post to get the look you want.
2# How much work you need to work it is to get the colours you want.
People like canon, black magic, and Arri colour science because they look rather good in-camera with the standard colour profiles, with Canon looking very flattering yet natural and BM/Arri looking very filmic

[GlenReed]

I do a lot of multi-angle interview style videos and the *Color Checker Passport* and *Davinci* have definitely saved me a ton of work. Great video man!

[NirnBootMod]

I dont think color science matters anywhere near as much as people think. If you got a good eye for art and learn how to color correct and edit... you can make it look great. EDIT: Oh yeah, I do need a color checker.

[ChrisdeCastro]

And if the DP actually reslates with a color checker every time there’s a lightning change and DIT is on the ball building a LUT for every take and the offline online editors remember to bake in those LUTs and the edit gets reconformed for the color session with the LUTs intact... color science is solved.

[jan-martinulvag]

If you photograph your own paintings with different cameras you find that its easier to detect the cameras that are flawed than when you photograph nature or portraits. The Samsung nx mini is weird. It does the best of all cameras I have tested. Next best is the fuji X-T100

[squidward]

Not realistic story. Nothing has been knocked down on the floor.

[christopheririonfilms]

Sorry, maybe it just the covid, but I simply could not watch The Whole thing. When you finally explained that you had shot the whole clip with three different cameras I just had to give up. There’s absolutely no way you can evaluate skin tone differences with a heavy beer colored correction over everything. It would help the rest of us if you graded every camera to its most accurate skin representation and THEN see how well the different sensors match. There is a reason why everyone loves Arri. Push it anyway you want-pale, saturated, contrasty, dark, light it’s stunning how beautiful it can be.

[FilmConvert]

Camera matching between different models and manufacturers is a key case where color science differences will show up in subtle ways. We're working on a new project called CineMatch which is designed to speed up the process of camera matching by using sensor match data to enable you to effortlessly match the color science of many different cameras, including ZCam, Blackmagic, Fuji and more. Check out www.cinematch.com to sign up for the beta release

[JoshuaMKerr]

Loving this video. I thoroughly reccomend a colour checker for speed especially useful for a-b cam interviews

[DatoAliffAlex]

Isn’t that why ACES color spaces is invented to solve the different color science on each camera manufacturer?

It’s a real problem if say, you want to match a Sony 8-bit and Canon 8-bit and the shot isn’t an ideal situation.

Even with white balance adjustments and so the lack of shades can make things a bit ugly.

But with 10bit all round it’s a lot less of an issue.

Mainly it’s just about time spent though.

[jonahtieman]

hey when I shoot in low light for some reason my camera has these weird vertical lines going down it, any idea how I can fix this?

[floriankochvisuals]

Hey Spenser, did you shoot some of the clips anamorphic? At 9:18 the color checker looks somehow not rectangular.