FS7: Sony FX9 v Canon C500 mkII

Camera 1 always did something right.

Camera 4 was my least favourite.

Camera 7 averaged pretty good on the first watch through for me.
 
TheDingo said:
Not a good sign that the test was done properly. Some of these cameras need the C-AF settings to be set correctly to work properly.
Click to expand...

some of them don't have AF at all.. so the guess might be what/who was not working properly: AF or AC? :laugh:
 
camera 1 is alexa, camera 4 is sigma more difficult to found other cameras: by deduction 8 is gemini and 9 helium still 2, 3 56 and 7....
by AF camera 5, 6 sony FX9 and 2 canon.
3 ?
7 Venice.
 
rudo_fr said:
camera 1 is alexa, camera 4 is sigma more difficult to found other cameras: by deduction 8 is gemini and 9 helium still 2, 3 56 and 7....
by AF camera 5, 6 sony FX9 and 2 canon.
3 ?
7 Venice.
Click to expand...

Well, you only have the Monstro left. :)
 
Well after looking at these tests all I can say is that by rights the FX9 should be dead to rights dead in the water as it was shot SLOG3 with a 709 LUT for one pass and with the new baked in "S Cinetone" 709 preset for a second pass... or vice a versa. Up against RAW recording cameras on paper the FX9 should be way behind. All I can say is that there are some pretty horrid images coming out of some of the cameras at 6 over and 6 under. That being the case if the worst images are coming from some of the RAW cameras and some of the better images are coming from the FX9 not shooting RAW it really does go to show that RAW is not the holy grail a lot of people paint it to be. As is demonstrably shown there are a number of cameras showing that their RAW files don't have the necessary information in them to be recovered. Sure nobody is going to intentionally shoot 6 over or under but often you have scenes where the critical part of the scene is exposed and captured correctly whilst other parts of the scene can be a number of stops over or under. Look into some of the worst images then dive into the shadows of some of those images and the story is quite ugly.

Chris Young
 
You are right Chris, raw examples look like not the gap we expected on this test. It would be better for our knowledge to have original cameras files to grade them our self and compare share our results with others people here.
 
cyvideo said:
As is demonstrably shown there are a number of cameras showing that their RAW files don't have the necessary information in them to be recovered.
Chris Young
Click to expand...

Indeed interesting! Im looking at the Canon C500 for raw particularly shooting landscapes and having blown highlights and noise-washed shadows. Getting the cheaper FX9 might not be a bad solution.

It would be nice with a comparison between FX9 and FS7 mk1 to check images when post-produced similarly. How big is the difference exactly?

Ive personally always felt the FS7 is a little dull and unsharp. Others disagree. But in landscape photography i absolutely love details, and sharp images benefit nature shots alot more than feature or human artsy films.
 
Martin.G said:
Indeed interesting! Im looking at the Canon C500 for raw particularly shooting landscapes and having blown highlights and noise-washed shadows. Getting the cheaper FX9 might not be a bad solution.

It would be nice with a comparison between FX9 and FS7 mk1 to check images when post-produced similarly. How big is the difference exactly?

Ive personally always felt the FS7 is a little dull and unsharp. Others disagree. But in landscape photography i absolutely love details, and sharp images benefit nature shots alot more than feature or human artsy films.
Click to expand...

Totally on board with your comments here. Films like “Lawrence of Arabia” or “Playtime” were shot with extra resolution in mind. Many of the scenes play out in wides that maybe other formats would go for. Also, the wide shots of “Revenant” in the 4K bluray feel better. Although, the movies as a whole can be enjoyed in any resolution, those are some movies that use resolution to good effect.
 
cyvideo said:
Well after looking at these tests all I can say is that by rights the FX9 should be dead to rights dead in the water as it was shot SLOG3 with a 709 LUT for one pass and with the new baked in "S Cinetone" 709 preset for a second pass... or vice a versa. Up against RAW recording cameras on paper the FX9 should be way behind. All I can say is that there are some pretty horrid images coming out of some of the cameras at 6 over and 6 under. That being the case if the worst images are coming from some of the RAW cameras and some of the better images are coming from the FX9 not shooting RAW it really does go to show that RAW is not the holy grail a lot of people paint it to be. As is demonstrably shown there are a number of cameras showing that their RAW files don't have the necessary information in them to be recovered. Sure nobody is going to intentionally shoot 6 over or under but often you have scenes where the critical part of the scene is exposed and captured correctly whilst other parts of the scene can be a number of stops over or under. Look into some of the worst images then dive into the shadows of some of those images and the story is quite ugly.

Chris Young
Click to expand...

I think where the fx9 is going to suffer isn’t in stops of highlight, but in detail loss, and unrecoverable color noise in shadows. So far I don’t see any additional highlight detail in raw on the c500 mark i, but I notice a fair bit more detail on the shadow end, (and more detail over all) than the xfavc. There is a lot of noise in the raw shadows, but it’s a detail preserving scheme, and when properly denoised produces a remarkably detailed image even at high iso. The fx9 starts to lose its lunch in terms of detail when under exposed.

I have to question the importance placed on looking at 6stop overs for skin tones. I realize some cameras are able to pull this off, but that’s just not where skin tones are ever placed in a real exposure. The extreme stops above middle grey are for headroom — skys, windows,etc— stuff brighter than Matte white tones, 2,3,4 stops brighter than the brightest skin highlight. If you are clipping or close to clipping skin highlights, you have many other problems with those brighter tones unless you are shooting someone in dark clothing against a dark background. Sure, extra highlights stops are better, but i think they are being overrated in this discussion. More interested in how these cameras compare at proper exposures (there are notable differences) and i really found the night exterior comparisons illuminating.
 
Last edited:
Barry G said:
I think where the fx9 is going to suffer isn’t in stops of highlight, but in detail loss, and unrecoverable color noise in shadows. So far I don’t see any additional highlight detail in raw on the c500 mark i, but I notice a fair bit more detail on the shadow end, (and more detail over all) than the xfavc. There is a lot of noise in the raw shadows, but it’s a detail preserving scheme, and when properly denoised produces a remarkably detailed image even at high iso. The fx9 starts to lose its lunch in terms of detail when under exposed.

I have to question the importance placed on looking at 6stop overs for skin tones. I realize some cameras are able to pull this off, but that’s just not where skin tones are ever placed in a real exposure. The extreme stops above middle grey are for headroom — skys, windows,etc— stuff brighter than Matte white tones, 2,3,4 stops brighter than the brightest skin highlight. If you are clipping or close to clipping skin highlights, you have many other problems with those brighter tones unless you are shooting someone in dark clothing against a dark background. Sure, extra highlights stops are better, but i think they are being overrated in this discussion. More interested in how these cameras compare at proper exposures (there are notable differences) and i really found the night exterior comparisons illuminating.
Click to expand...

I think the most common issue people will run across is when they're filming someone outdoors with a shiny, bald head. :)

The ability to drop highlights on a patch of shiny skin and have something look nice rather than a white patch of zero information is valuable. In an ideal world we shooters would never have this problem, but the world, shooting conditions, and hair loss is not always kind to us.
 
  • Like
Reactions: jcs
theraptor56 said:
I think the most common issue people will run across is when they're filming someone outdoors with a shiny, bald head. :)
Click to expand...

Don't we all know that one! No make up artist to dull the dome? Digital Beauty Box has got me out of jail a number of times as it has a "de-shine" tool in its kit box. It's a bit CPU/GPU intensive so slows down even a fairly powerful workstation but it does do the job.

https://digitalanarchy.com/beautyVID/main.html

Chris Young
 
When shooting raw, the sensor has only the chosen base iso (ex. 800/3200) to operate from, the user keeps control over the broader exposure indices (EI) (ex.100-12800) and iris/shutter. Each setting of the EI is its own curve. What this means is the amount of latitude into the highlights varies with EI. If you shoot EI400 , the image is darker so you compensate by adding more light. This reduces highlight headroom but gives clean image. If you shoot EI3200 the image is brighter so you close down the iris giving less light to the sensor. This increases highlight headroom but gives a noisier image. For a given camera the amount of highlight latitude at EI400 might be 4 stops but could be 6 stops at EI1600. Those are just example numbers, the larger point being that highlight latitude (with raw) is under operator control. When you shoot other codecs like XAVC, changing ISO changes actual gain, so you do not have the same control over latitude.

Then when it comes to post, and you need to brighten or darken a highlight (bald head was example cited), you're just manipulating metadata, not the gain. It's a range of lossless adjustment to reveal buried detail from the shadows or the highlights.
 
powervideo said:
Polariser 4x4 can help the bald head shot too.
Click to expand...

Well blow me down Pete. That thought never crossed my mind for an indoors on the run no make up interview shoot. Must test that out. In the kit I always keep a pancake makeup dispenser plus a pack of those disposable applicator discs to de-shine subjects. Also a small pack of wet wipes to remove the pancake afterwards. You never know when this mini makeup kit is suddenly going save your butt. Good one Pierre!

Chris Young
 
cyvideo said:
Well blow me down Pete. That thought never crossed my mind for an indoors on the run no make up interview shoot. Must test that out. In the kit I always keep a pancake makeup dispenser plus a pack of those disposable applicator discs to de-shine subjects. Also a small pack of wet wipes to remove the pancake afterwards. You never know when this mini makeup kit is suddenly going save your butt. Good one Pierre!
Chris Young
Click to expand...

I didn't believe it myself until another DOP showed me on a shoot. :)
 
Teddy_Dem said:
There is a Full Frame Cine Camera Shootout on Reduser right now. Results will be revealed on Thursday, Feb 6th:

http://www.reduser.net/forum/showthread.php?180359-Full-Frame-Cinema-Shootout-2020

https://vimeo.com/388857434

4K 10 Bit Prores version can be downloaded from here:
https://vimeo.com/388805758

And to copy and paste the details:
"This comparison is between the following cameras (not in this order)...

Red Monstro
Red Gemini*
Red Helium**
Sony Venice
Sony FX9 S-Cinetone
Sony FX9 SLog-3
Arri Alexa Mini LF
Canon C500mII
Sigma FP

*Gemini used for studio and low light test only
**Helium used for studio only

All cameras were shot in their RAW codecs except the FX9 which was shot in XAVC-I 10bit. All cameras were lightly corrected using RAW controls in Resolve on a single timeline and split screened on a common calibrated Rec709 monitor for a neutral rendering. The AF capable cameras were shot with AF set to facial tracking.

The whole test was shot on Sigma 50mm T1.5 cine and 50mm 1.4 ART AF lenses except for the Canon C500mII which was shot with a 50mm 1.2 EF since there was no PL mount or EF Sigma available at the time of shooting.

Internal ND's or Firecrest NDs were used for exposure control. All exposures were set using a common waveform and native low/hi ISOs.

We intend this test to explore the subtle tonal and color differences between these cameras. Please watch and vote on your favorite in the poll, in a few days we'll reveal the actual cameras and see if your preferences line up with our expectations."
Click to expand...

Surprised more people here aren’t guessing.

1. LF, i mean, it’s gotta be, right?
2. Canon C500ii, i assumed it would look warmer, but it is over cast or late in the day, could be a white balance choice. Also, this is the only one that seems to be rocking a Canon 50mm f1.2.
3. Monstro, at first i wasn’t sure what to think of this one, then i saw how insane the -6 under exposure recovery was, and i realized this camera is comparatively being over exposed. So, for better or worse, that may be the reason for the sort of glow in the first day light shot. And may also explain the extra saturation, as a pulled down expaoure also tends to increase in saturation.
4. Hoping this is the cheapest camera. Although, i still kind of like it. Come a ways since the T2i we have.
5. Fx9 s log with the Alistar Chapman lut.
6. Fx9 s cinetone
7. Venice, Noisy in the shadows, but very consistent colour. And the only one that held onto the highlights in +6
8. Haha, this is odd, as the Gemini is cleaner than the Alexa above 1600iso, in my comparisons, so i was surprised to see such a magenta shift. Noise and detail look ok, considering that the thing is also still almost holding onto +6. In terms of dynamic range. Pretty impressive.
9. Helium, When exposed correctly this camera looks pretty awesome. Seems to be the same case in this comparison.
 
You must log in or register to reply here.
Back
Top