FS7: Sony FX9 v Canon C500 mkII

Actually, one of the charts has some green CA going on. Either the lens wasn't quite focused or the debayering is leading to some bizarre green fringing. The Zeiss does have some CA, so that could be it but I doubt it since it is mostly only seen when the lens is wide open. The other chart appears softer, which could either be focus or a better AA filter. Either way, the other chart doesn't have any of the green CA so it is inarguably the winner for image quality.
 
Assuming the lens is not perfectly square with the chart and that a lens’ focal plane is not perfectly flat, that some areas in the chart would show softer focus. The center looks adequately in focus. Considering this is camera sharpness and not a lens test, finding the best focus in the finest details of the chart might be all that matters.

Just in case you want to see the slashcam results:
https://www.slashcam.de/artikel/Tes...Bildqualitaet-fuer-Broadcast-und-Cinema-.html
 
Remember viewing fine detail is often limited by the monitor you're using. I would often use catalysts browse and punch way in on the chart to really see what's going on. Sometimes 2 lenses would look equally sharp until I punched in.
 
If you’re on a phone you can pinch zoom and it works great to see the winner.

Some comments are right on re: conclusions. There’s a clear winner in the pixel peep world; in the real world not an issue so much though a FW update might be able to address now that we’ve clearly exposed some issues.

Lens was at F8.
 
Hmm... Now I am actually kind of intrigued now that I know it isn't a lens issue. C1 seems to be showing more fine detail. C2 has some color artifacts or sharpening going on at high contrast regions. I would have guessed that C2 was the C500 II because I've seen some odd behavior with Canon RAW, but that slashcam test seems to show the FX9 with the reduced detail and artifacts. It really just looks like sharpening is causing the issue, though.
 
jcs said:
If you’re on a phone you can pinch zoom and it works great to see the winner.

Some comments are right on re: conclusions. There’s a clear winner in the pixel peep world; in the real world not an issue so much though a FW update might be able to address now that we’ve clearly exposed some issues.

Lens was at F8.
Click to expand...

Is there a way we can see the 4k image? I'm only seeing a sub 2k image (1922px wide) when I try to download or view it. Given the forum software, it seems like a Vimeo upload is the best way to see the tru 4k, even with their compression.

Also..why does one image look sharpened...and the other not. If this was a native file, I'd say it was a camera setting, but this has been downsized...so it's as if they've been downsized with different settings, or something else.
 
Barry G said:
Is there a way we can see the 4k image? I'm only seeing a sub 2k image (1922px wide) when I try to download or view it. Given the forum software, it seems like a Vimeo upload is the best way to see the tru 4k, even with their compression.

Also..why does one image look sharpened...and the other not. If this was a native file, I'd say it was a camera setting, but this has been downsized...so it's as if they've been downsized with different settings, or something else.
Click to expand...

No post sharpening or downsampling: center crop from 4K 1:1. Files are externally hosted to prevent forum downsizing. The softer example has slightly more detail while the example that looks oversharpened with color issues on edges is likely a deBayer issue. DeBayer methods can effectively do sharpening at the same time. In any case the camera with artifacts and oversharpened results would ideally be adjusted with less Sharpness / Detail in camera (checked manual: it's possible). The aliasing is minimal (very high frequency), and I haven't seen any in either camera's footage so far.

If I was performing QA and responsible for the product, I'd file a bug for the green/color artifacts (again, suspect deBayer issue). Highly unlikely anyone would ever see outside a test chart (if the manufacturer wants to win the test chart contest, they'll have to fix this issue). If after turning down Sharpness/Detail the color artifacts go away, that would show (at least) partial deBayer tuning (vs. sharpening RGB post-deBayer, which can't cause any color artifacts with grayscale (sharpening is applied equally to the RGB color planes)).

I'd call the softer more detailed example pretty close to ideal: lots of detail and not too sharp: keeps the needed 'antialiasing' pixels (which oversharpening removes).

One cameras wins the resolution chart test, while for real world images, they are a tie, and the oversharpened example can certainly be softened in camera by making the necessary adjustments. The softer example can be sharpened in camera and the sharper example can be softened in camera, so they can be matched if desired. Note the gorgeous Alexa (+mini) LF would be comparatively soft / lower detail for this test with only 4.4K horizontal pixels yet leads the market in IQ.

Also have model / skintone / highlight / shadow tests (and a few ramps). Will take some time to go through and edit.
 
James0b57 said:
That slashcam test for the Fx9 looked pretty bad. Kind of makes me laugh to think of them publishing analysis based on an out of focus test.
Click to expand...

I don't think that's at all what you're looking at. Not totally sure their method, but it would appear to be a small chunk of an older res chart shot from some distance (we are looking at a small center crop of the frame, probably enlarged to 200 or 400%). If it were out of focus you would not be seeing all the moire artifacts in the various samples. They are arguing that the FX9 has a pretty aggressive OLPF, and I've seen a fair number of samples that would back this up. The Reduser FX9 chart and clothing "looks" out of focus too. Maybe there's a conspiracy. :-) Anyone who conducts these tests is opening themselves up to a lot of criticism for how their tests are run, and the conclusions they make. I'm not sure it's fair to make any kind of assumption as to what intern did this, or how they can't focus on a res chart. (about the easiest thing in the world to focus on). All in, though, I don't think there's enough difference between these cameras, at least in the 4k output, to get upset about either way. Apparently the tester isn't going to tell us which is which, so we'll probably never know anyway, unless somebody wants to drag their FX9 up to central coast wine country for some vino, and chart busting. :-)
 
Barry G said:
unless somebody wants to drag their FX9 up to central coast wine country for some vino, and chart busting. :-)
Click to expand...

I’ll be right there!


And you’re probably right about the slashcam test. It just wasn’t something i saw in my personal observations with the camera.
 
Last edited:
Donny 123 said:
so we are agreed the fx9 is the better camera at half the price ..?
Click to expand...

Try shooting FF 60fps on an FX9 or shooting anamorphic or using PL mount lenses s with i-data or shooting RAW internal or using AF with your collection of EF lenses or putting it on a gimbal, but sure it is cheaper.
 
Liam Hall said:
Try shooting FF 60fps on an FX9 or shooting anamorphic or using PL mount lenses s with i-data or shooting RAW internal or using AF with your collection of EF lenses or putting it on a gimbal, but sure it is cheaper.
Click to expand...

Yes but the average shooter doesn't need all that fancy stuff... but yes the EF lenses just don't seem to fit at all.. what was Sony thinking of .. but the AF does at least seem to work with E mount lenses which fit quite easily .. well done Sony ..
 
Donny 123 said:
Yes but the average shooter doesn't need all that fancy stuff... but yes the EF lenses just don't seem to fit at all.. what was Sony thinking of .. but the AF does at least seem to work with E mount lenses which fit quite easily .. well done Sony ..
Click to expand...

Can purchase a few RnG AF lenses from Sony for the cost difference.
 
Donny 123 said:
Yes but the average shooter doesn't need all that fancy stuff...
Click to expand...

Who wants to be an average shooter...

J/k

To be fair, and I have said this many times in this thread, I like both cameras but given I own a lot of EF lenses, occasionally do anamorphic work, occasionally use PL mount lenses, do an awful lot of slow-mo and would like my A cam on a gimbal rather than my B cam, the C500MKII is a better fit for me at this moment. I think Sony have done a stellar job with the FX9, they have improved a lot of things over the FS7.
 
You must log in or register to reply here.
Back
Top