FS7: Sony FX9 v Canon C500 mkII

[jcs]

I didn't write that. Focus on what I write...

Thanks for the downvotes- now we’ve got more drama, well done for entertainment value; not logic or evidence though.

That’s what I’ve been repeating in this thread that’s up for debate. If you agree with my conclusions, all good. In cases where there’s disagreement, let’s debate the evidence. If in your view human perception makes it impossible to debate the evidence, why are you wasting time in this thread?

 

[Barry G]

If one believes Canon’s sensors have more DR than Sony- show us the evidence to support your assertion. All the examples I’ve posted or discussed show Sony has the DR advantage.

So, in the holy grail video, when you look at the Sony image at 6 stops under, you think that's more usable than the canon image at 6 stops under? Is there even an image left at 6 stops under in the Sony? It looks like nothing but noise(most of it color noise), and a ton of noise reduction sitting right underneath all that noise. Sure it has the color, but is there any detail left? Now the canon, affected by the LUT they are using, is seeing a loss of color information, but look at how much more detail, and how much less noise there is...that's a neat trick. Here's the thing...change the LUT, or don't use one, and the Sony's not going to gain any detail, or lose any noise. the canon on the other hand is going to get back that color, and not lose any noise or detail. And if you shot in raw...well I know that's not fair to the Sony...but would you say that in your opinion, the camera that clearly has the best IQ (your term) is the one that doesn't shoot RAW?

 

[jcs]

Barry- let’s focus on the highlights and color behavior, NR can deal with noise to some extent, no solution when color blows to white. Agree both cameras look similarly less useful when noisy.

Raw: sorry I thought the FX9 shot raw too, though kinda crippled by add-ons and file size. Comparing compressed makes sense as that’s what most will use. In any case, let’s see FX9 XAVC vs C500 II raw! Also it’s possible Sony is using the same/similar sensor in the Venice? If so, we can learn something about the sensor by studying Venice raw too.

 

[Liam Hall]

Thanks for the downvotes-

LOL I didn't downvote you, but I now know who downvoted me! (I did downvote the above comment though - just so you know)

If you agree with my conclusions, all good.

I don't agree with your conclusions. I don't see how you can make conclusions without doing your own tests.

why are you wasting time in this thread?

I started this thread...

 

[Guest]

Looking at the specs of the FX9 and C500 mkII, is it time to jump back to Canon?

Don’t make decisions based on video reviews, make them based on specs.

Both cameras look great. JCS may prove to be right, but until you or i do our own testing, who are we to argue against his views?

 

[jcs]

LOL I didn't downvote you, but I now know who downvoted me! (I did downvote the above comment though - just so you know)


I don't agree with your conclusions. I don't see how you can make conclusions without doing your own tests.



I started this thread...

Same question- if you think it’s a waste of time and perception will never match between people, why create a thread like this and tell folks presenting evidence with logic that they don’t know what they are talking about sprinkled with ad hominem?

If you disagree: counter with evidence, logic else perhaps it’s just fun to argue about nothing and everything, with argument being the sole entertainment draw to these kind of threads: drama.

 

[Liam Hall]

Don’t make decisions base one video reviews, make them based on specs.

Both cameras look great. JCS may prove to be right, but until you or i do out own testing, who are we to argue against his views?

Cheeky! That was written in September before either camera was released. I tested both cameras before making my decision.

And just so you know, I downvoted your comment - just for shits and giggles.

 

[Liam Hall]

evidence with logic

Apologies if I missed it, but I haven't seen any evidence or logic from you, just conclusions drawn from a flawed test.

 

[Guest]

Cheeky! That was written in September before either camera was released. I tested both cameras before making my decision.

And just so you know, I downvoted your comment - just for shits and giggles.

All good. No worries.

 

[Barry G]

Barry- let’s focus on the highlights and color behavior, NR can deal with noise to some extent, no solution when color blows to white..

No.. How bout lets not. If you want to talk about "IQ" and DR...you have to look at all aspects of the image. A camera that can produce color but not detail at a certain exposure isn't useful at that exposure. Highlights are only half of the image, you can't ignore the other half just because you prefer pale women with blonde hair. (that was mean, but possibly even true).

Applying NR to noise you're seeing on the FX9 at 6 stops under is not going to add detail that's already been lost to NR in camera. As for the highlights, if you take the color issue out of it, you'll see that the canon and FX9 break at exactly the same spot. Again, the color will come back if the processing is done correctly (this loss of color is absolutely a product of the LUT).

Finally, if you'd like to take the time to read Geoff Boyle's comments on their last camera tests, something glaring should stick out to you. All of the top cameras Arri Canon RED Sony achieved about 7 stops of latitude with varying levels of over under (Arri was unique in that it had less under and more over than the others, probably explains its more limited ISO range, but otherwise had the same overall latitude, and interestingly, the same clipping characteristics as the C700FF). The important point is that the best cameras all had about 7 stops of Latitude, yet the CVP test is looking at a range of 12 stops of latitude. This means that the CVP test is showing us at least 5 stops of latitude beyond what's usable for any of the cameras. Focusing on the extremes here is a big waste of time as it's outside how any of the cameras can be used. (FWiW, GB's tests showing 7 stops of Latitude were conducted in Raw not Xx-AVC)

https://cinematography.net/CineRant/2018/07/30/personal-comments-on-the-2018-cml-camera-evaluations/

Further, And I've posted this before, if you take a look at the ISO test that ProAV posted, you'll notice how well the Canon holds color and detail at the extremes of it's ISO range. That tells you a lot. ISO tests and Over/under tests aren't the same..one looks at the camera processing, and the other looks at how the image (codec and LUT) is affected by post processing. What you'll see in that ISO test is that at 7 stops under base (with the camera gained to 25,600), the canon easily holds all the colors in that chart, which supports the idea that the LUT is causing the loss of color. If you don't want to see that, don't look. I'm fine with that, but it tells you something about the camera that you are ignoring.

 

[Liam Hall]

Also it’s possible Sony is using the same/similar sensor in the Venice?

Alister Chapman said it's a new sensor design but shares some of the colour science from Venice.

 

[TheDingo]

Question: Would you base your main buying decision on how good a camera looks at 5 stops under, or how the footage looks when properly exposed ?

I don't think I've ever shot footage that was much more than 1 stop over or under exposed, so I've never really had to deal with having to "rescue" footage.

I find these maximum DR tests a little amusing because I'm never ever going to see footage that looks like that, nor will I ever have to fix footage like that. Pretty much all of the time I'm going to be dealing with good footage that requires minor corrections in post. Any decent camera monitor or EVF on the market is going to prevent you from making serious exposure or focusing errors.

Compared to the bad old video days (long before I got into video) we are all spoiled babies with the kind of gear we get to shoot with.

 

[Barry G]

Question: Would you base your main buying decision on how good a camera looks at 5 stops under, or how the footage looks when properly exposed ?

I don't think JCS is interested in buying either. His EOS R is barely a year old. bah.dump.dump

 

[Guest]

...Finally, if you'd like to take the time to read Geoff Boyle's comments on their last camera tests, something glaring should stick out to you. All of the top cameras Arri Canon RED Sony achieved about 7 stops of latitude with varying levels of over under (Arri was unique in that it had less under and more over than the others, probably explains its more limited ISO range, but otherwise had the same overall latitude, and interestingly, the same clipping characteristics as the C700FF). The important point is that the best cameras all had about 7 stops of Latitude, yet the CVP test is looking at a range of 12 stops of latitude. This means that the CVP test is showing us at least 5 stops of latitude beyond what's usable for any of the cameras. Focusing on the extremes here is a big waste of time as it's outside how any of the cameras can be used. (FWiW, GB's tests showing 7 stops of Latitude were conducted in Raw not Xx-AVC)

https://cinematography.net/CineRant/2018/07/30/personal-comments-on-the-2018-cml-camera-evaluations/

Is Geoff Boyle rating the C200 better than the Red Monstro? Am i reading the comments in that link correctly?

 

[jcs]

Motivating an Internet forum to focus on evidence and facts and not making things personal is like herding cats lol.

Let’s play the opposite game! Anything you say I’ll say the opposite. It’s drama and requires less effort since no need to back anything up with evidence or logic or do any work.

I never focused on noise, mostly highlights blowing out color with Canon cameras. Indeed I’ve owed a few (all brands), still own only Canon, posted many examples throughout this thread (my own work as well as others). It’s clear to me based on my experience with all these cameras and examining footage shot by others the FX9 wins for image quality (IQ). Im very familiar with Canon color science and prefer more useable color DR of ARRI and Sony vs. Canon. Even in perfectly exposed shots, Canon loses color too soon and goes white in skintones.

Agree that in controlled light with everything set up correctly these cameras are similar, and color comes down to preference. Sometimes we can’t control the environment (my use case outside the studio), and prefer the most flexible tools.
If we want to know which is technically better when pushed to the limit we do these tests. And again the FX9 appears to do better with evidence presented so far. Disagree? Provide evidence + logic showing the C500 II doing better.

Thx Liam for FX9/Venice sensor info.

 

[TheDingo]

It’s clear to me based on my experience with all these cameras and examining footage shot by others the FX9 wins for image quality (IQ). Im very familiar with Canon color science and prefer more useable color DR of ARRI and Sony vs. Canon. Even in perfectly exposed shots, Canon loses color too soon and goes white in skintones.

Are we comparing Canon C500 Mk2 2 Gbit RAW-Lite against Sony FX9 XAVC-I 4:2:2 10-Bit footage ?

With footage tweaked in post ? ( I don't care what comes out of the camera as long as I can get what I want in post )

 

[cpreston]

In general, I've found Canon Log 1 and 3 footage easier to work with in XF-AVC if it is one stop overexposed based on Canon's exposure meter. As Barry has pointed out, Canon reduces saturation in the superwhites and blacks and I regularly find myself trying to lift some detail in the shadows when I am working with unlit b-roll footage. The built in NR also reduces detail on the shadows. Of course, I only overexpose when I have extra head room in the whites since the desaturated whites also be difficult to recover, especially if you automatically apply a LUT. None of this much applies to the RAW footage.

The problem with the CVP footage is that I have absolutely no idea what picture profile, exposure level, or LUT that they were basing their tests on. I have no idea how to compare that to my previous experience with Canon cameras.

Now, here is my big question. Does the FX9 have more headroom in the audio preamp than the FS7 and FS5? I always find it slightly annoying that Sony's didn't have enough gain so that I could properly gain stage my Lectrosonics camera hops in the line level setting. I would have to set it the Sony to mic level and pad and adjust the receiver. Canon allows me to just set the camera to line level with a preamp setting around 8 and have a minimal of additional preamp hiss. I guess that makes the Canon indisputably better.

 

[Barry G]

Is Geoff Boyle rating the C200 better than the Red Monstro? Am i reading the comments in that link correctly?

He's reading the results of the myriad of charts in the test. He's provided a link to the raw camera files, and everyone is invited to download and form their own conclusions. (if thats too much, the pR output is available on Vimeo). Generally RED cameras don't hold up all that well in these types of tests, although I believe he really liked the Helium.

 

[Barry G]

Disagree? Provide evidence + logic showing the C500 II doing better.

uhh..whenever we do, you say it's just not important to you and change the subject.

 

[jcs]

uhh..whenever we do, you say it's just not important to you and change the subject.

Have been repeating the same thing over and over regarding color blowing to white and you brought up shadow noise. I explained how’d I’d deal with noise, then back on topic of highlights and color. I also pointed out this Canon issue in one of Geoff’s tests you shared against the Alexa (which did much better in color (WB easily fixed in post): both Canons blew to white in the neck and the footage looked correctly exposed!).