FS7: Sony FX9 v Canon C500 mkII

My own opinion has changed over the years and currently I find myself leaning towards a bit more green (not too much) than magenta. And I'm sure that will change again.

This doesn't surprise me. I remember way back attending a seminar on the physiological difference in human eyes regarding the perception of color. This seminar was given by a couple of people from what was then Agfa film. Specifically they were discussing the visual difference back then between Agfa film stocks which were German developed and manufactured and the various Fuji film stocks available at the time that were developed and made in Japan. People were commenting that Agfa color in general had a more "earthy," warmer look and very slight bias towards green versus the slightly more blue/red bias of the Fuji film stocks. We were told that after years of study of the physiological difference in human eyes there was plenty of evidence indicating that Caucasians perceive color slightly differently to Asian due to the fact that there are genetic difference in the cones of our eyes which causes them to react differently to color depending on the wavelengths involved.

"The reason for this is that photoreceptors vary in the wavelengths at which they are most sensitive. For example, one person may have "red" photoreceptors, or "long-wavelength cones", that are most sensitive to wavelengths of light around 564 nanometers. Another person may have long-wavelength cones most sensitive at 568 nanometers. This may seem like it isn't much of a difference. However, this slight shift in photoreceptor peak sensitivity can make all the difference in the world when it comes to perceiving colors as "red" versus "orange", or other colors."

How many of you have heard the comment that ARRI color has a very slight leaning towards green. There are numerous LUTs around that are modified ARRI LUTs that have this very slight green bias dialed out. Bert Degraeve's popular Slog3_Arrilook709_standard vs his Slog3_Arrilook709_balanced LUT that has less green for example. Alister Chapman put out a series of "Venice" LUTs because of requests to have the slight green bias taken out. They were Venice Look -G1, Venice Look -G2 and the standard Venice look which has the green bias typical of the original Venice "look" where it is closer to the ARRI look with its inherent leaning towards green. There have been posts on this forum where the original s709 Venice LUT was considered to be "too green" for some. The generally accepted reason for this green/yellow bias is that most skin tones tend to have a have a higher yellow content than magenta and it's no coincidence that yellow is on the green side on the vector scope. Very subtle negative adjustment to yellow pushes towards green.

"Most Caucasians fall in the range of 5%–20% more yellow than magenta. A fair-skinned Caucasian adult could be as low as 20% magenta, 25% yellow. A bronzed Caucasian could be as high as 45% magenta, 62% yellow."

It has also been hypothesized in certain studies that over the many years of western cultural influence in Asia, starting with Japan specifically since the war that for the Asian eye to perceive color in the same terms as what Caucasians call "correct values" for skin tones Asian eyes need skin tones to have a higher magenta content. In other words a lot of what we perceive as the "correct colors" for skin tones can vary greatly depending on our genetic makeup and then again can vary quite considerably within our own genetic grouping. All this really came to the fore for me when I did eighteen months of a weekly current affairs TV show for TVB in Hong Kong. I was working with two resident Hong Kong producers based here in Sydney and time and time again in post they would request more magenta into interview shots before they were happy. I just had to accept that that is what they saw as normal skin tones whereas for me and my editors at the time we would shake our heads and think "Way to piggy pink" as one of my editors used to phrase it. Maybe just maybe that is why Sony cameras for so long have been accused of a magenta bias. I have no evidence to that but I lean towards thinking that way, just a hunch based on experience.

So this whole talk of what are the "right" skin tones and color science from cameras can only ever be subjective. Two of us sitting side by side in an edit suite can disagree on what we think is a nice skin tone grade just because your "red" photoreceptors, or "long-wavelength cones" are more sensitive to wavelengths of light around 564 nanometers whereas my long-wavelength cones are more sensitive at 568 nanometers. One of us will see more red in the skin tones and the other will see more orange as pointed out before.

I can find no reference to any of the Agfa studies on the physiological differences in human eyes. That seminar was was many moons ago. My short extract quotes come from the following links which may give readers a more in depth look at color perception in general:

http://scienceline.ucsb.edu/getkey.php?key=719

https://help.smugmug.com/correct-skin-tones-for-print-rkUgv14rM

Chris Young

EDIT:

Get your head around this one. A truly deep subject matter.

https://color.psych.upenn.edu/brainard/papers/Brainard_2019_CurrOpp.pdf
 
Last edited:
We do all indeed see color differently. That said we can test color perception: https://www.xrite.com/hue-test . First quick try on a non-calibrated MBP got a 4 (perfect is 0, bad is >8000). Tried again more carefully and got 0.

XriteColorTest.jpg

Regarding comments the ARRI example wasn't super saturated color (+ thread is 'weird' regarding pushing color saturation and accuracy?): it's accurate, and many others see the same thing (see comments on YT). ARRI is the top camera, #1 in awards, not sure why someone would think these lower end cameras can hang with the top cameras (someday, but not yet). The Venice example is saturated, as are all the examples posted that I shot and graded myself. Haven't seen any super saturated C500 II footage yet that looked good: certainly possible- nothing found so far.

After reviewing many online examples, really impressed with the mojo of S-Cinetone and the accuracy of SLog3 on the FX9. The C500 II in terms of color looks similar to the C300 II (owned one) and the C700 FF (tested one). The C300 II had the edge in color vs. the FS7x, now the C500 II and FX9 are much closer and I'd expect the FX9 to have more usable DR (won't know for sure until tested). In terms of mojo, the FX9 is looking like a more fun tool for SOOC camera looks. That said, since the C500 II now supports 3D LUTs: there's much more flexibility to tune color in-camera (vs. the C300 II which didn't support 3D LUTs). My comments have been focused mostly on image quality, color, and skintones. These cameras have other features that will make or break a decision to purchase/rent based on those differences.

If our schedules line up, I'll try to be involved in a local C500 II + FX9 side-by-side comparison.
 
FX9 vs. C500 II, FS7, Venice:

FX9 does significantly better than the C500 II in under/over color recovery, as expected (more usable DR). Both cameras are manufacturer rated at 15 stops; C500 II is clearly less than 15. Maybe they're around 14 with better noise performance to get near the ARRI cameras (14 stops usable DR: the real deal). Then again, topping out at +5 and -5 for usable color recovery (barely), that's 11 stops for the FX9; C500 II is -4, +4 for 9 stops. Color recovery is a lot easier to judge vs. e.g. a Xyla chart.

FX9 does surprisingly well vs. the Venice (which not surprisingly did the best in these tests).

This test shows the limited usable DR of Canon sensors vs. Sony (early loss of recoverable color as exposure increases): I see similar behavior in the 1DX II and EOS R (C300 II was similar). Par for Canon sensors.
When shooting in controlled conditions and/or just letting the highlights blow out to protect the subject(s), these differences aren't as important (that's what I do with current cameras: 1DX II & EOS R). Though these results do mean that color near the highlights in properly exposed footage will have less accurate color (blow to white early as seen in Geoff's example with the C200 and C700 FF).

For pure IQ, Sony wins this round with the FX9 in both color (IMO) and DR (fact). Kudos Sony for catching and passing Canon! Canon- hope you're giving the EOS R 2 10-bit (and less RS) else will wait for the Unicorn Cam (A7S Next). Or maybe should just get an A7 III or A7R IV and start collecting Sony lenses and selling off Canon L lenses (yeah, I'm that impressed with Sony). Need to see if I can match 1DX II / EOS R raw stills with Sony's latest (wasn't possible with the A7S II (was good, just not as good as the 5D3 at the time)).
 
Last edited:
Don’t you just love the internet. Populated by (self-certified) “experts”, most of them never having seen the cameras they are evaluating in minute detail, spending significant amounts of time compiling detailed posts full of BS.

The only way to compare cameras is to shoot them yourself on subjects of your choice. The best you can hope for from a forum is to learn of major faults but that’s rare these days. At worst you get meaningless discussions/arguments about skin tones, blah, blah, bloody blah.

The only people I bother listening to on this forum are Liam, Barry G, cyvideo, Doug and Sam MM. Mostly they go way back to Betacams and before (sadly so do I) and, more importantly, actually use the cameras they talk about. Some of the others (not all) just need to stop claiming knowledge they clearly don’t have.

Withdraws to a safe distance, waits for the barrage from people in need of a life...
 
Man I'm surprised. Score '0.' The old eyes still cut it! With this test anyway. Well I know they work... how did the song go?

There she was just a-walkin' down the street, singin' "Do wah diddy diddy dum diddy do"
Snappin' her fingers and shufflin' her feet, singin' "Do wah diddy diddy dum diddy do"
She looked good (looked good), she looked fine (looked fine)
She looked good, she looked fine and I nearly lost my mind :)

Chris Young
Chris The X-Rite Color Challenge and Hue Test.jpg
 
Tough to admit the FX9 is indeed starting to look like the better choice for my shooting operations. Its also nice to keep the dual charger, batteries, and cards, as well as both Sony lavaliers as well.

Has anyone tried the autofocus with Metabones adapter? I dont shoot with Sony lenses, so the only one i own is the Sony E 24-70 f/4 with Zeiss glass, for lightweight photography.

Can you use the hotshoe mounted XLR extension from FS7 with the FX9 btw?
 
Man I'm surprised. Score '0.' The old eyes still cut it! With this test anyway. Well I know they work... how did the song go?

Chris Young

Did the test, found out that my spectacles had to be cleaned. Score 2 though. Physical age 78, real age 55
 
Don’t you just love the internet. Populated by (self-certified) “experts”, most of them never having seen the cameras they are evaluating in minute detail, spending significant amounts of time compiling detailed posts full of BS.

The only way to compare cameras is to shoot them yourself on subjects of your choice. The best you can hope for from a forum is to learn of major faults but that’s rare these days. At worst you get meaningless discussions/arguments about skin tones, blah, blah, bloody blah.

The only people I bother listening to on this forum are Liam, Barry G, cyvideo, Doug and Sam MM. Mostly they go way back to Betacams and before (sadly so do I) and, more importantly, actually use the cameras they talk about. Some of the others (not all) just need to stop claiming knowledge they clearly don’t have.

Withdraws to a safe distance, waits for the barrage from people in need of a life...

If you are referring to the CVP video above, CVP is a major retailer of professional broadcast and cinema cameras located in England. Their camera reviews are no less informative than those provided by B&H or or other retailers. www.cvp.com
 
Hot dog. Scored a zero.

Also would like to say cyvideo scores MVP of the day for that post with all the color/retina info. Maybe the most valuable post in this entire thread.
 
Good job Team Zero! Any low number is excellent too. Looks like a few humans can agree about color and what looks right after all.

Now that it’s clear the FX9 is the winner in IQ (C500 II may still be a better tool for some folks), we’ve moved into the Disqualify the Messenger stage of our show. If you haven’t used the camera, your comments, evidence, facts, and logic are irrelevant. If you’ve used and thoroughly tested the camera and you’re a retailer, all results must be ignored! What will you say after I’ve tested both cameras? “He’s obviously a spy working for someone I disagree with. He can’t be trusted, only people I agree with can be trusted!!” Lol.

Seriously, if you disagree, present your case: provide evidence, facts, and logic (proof) to show how you arrived at your point of view (conclusion). Ad hominem really gets boring- it’s lazy and not very creative (crap did I just diss Ad Hominem lol).
 
I think my eyelashes add to the chromatic aberration.

I finally closed one eye and realized, i had been seeing in HFR 3D gimmick this whole time. I now walk around with an eye patch. Still trying to get more strobe and motion blur.
 
I scored 0 on first/only try. But it still doesn't mean we can agree on what's right until we agree on a name or number for each slice and be in agreement that two slices have the same name.
 
A) Now that it’s clear the FX9 is the winner in IQ.....

B) What will you say after I’ve tested both cameras?

C) Seriously, if you disagree, present your case: provide evidence, facts, and logic (proof) to show how you arrived at your point of view (conclusion).

A) Priceless.

B) Lets hope your involvement in the test is more extensive than a few hours. Lets hope it's more involved than choosing the lens used. Lets hope it is more specific than "can be made to look like Arri". After that, I'm all ears.

C) Perhaps you should do the same. Your evidence is based almost entirely on what seems like very limited experience with other cameras, viewing youtube videos, returning to the same videos and discussions that were already had in this thread over a month ago. The only side by side we've seen is the CVP review of the FX9, which was mostly a review of the FX9 (!), with a few, undocumented, clips of the C500II thrown in(to the review of the FX9, by a retailer of the FX9(!). If that test is all you need to prove your point, then you should stop right there. Others, including me, have roundly questioned the methodology of that comparison, bought and used the camera, and have reported our evidence. You on the other hand, have used incredibly circular arguments that seem to fall on all sides of every equation, peppered with paternalistic advice for everyone and anyone, and yet you claim the moral and intellectual high ground of logic and proof.

So back to A -- Do you really think you are qualified to state that conclusion? Given your interest in C)...wouldn't completing B) be generally an important prerequisite to stating A)? A yes or no answer would be just fine.

______________________
-------------------------------
 
Back
Top