Definitely to each his own on this. For me AF has become increasingly important, so the Eva was never a consideration. I happily charge my clients extra for combo shoots (more gear at risk, more to load and take home, and the days are just harder) Although, generally we are booked for one or the other more often. (and my video day is 25% more than my photo day.) I think if I was sitting with a c200, or happily with an EVA, I would probably put the money into workflow enhancements rather than the c500II. But that wasn’t my situation. 2 weeks in with the camera, I feel it was a good investment, for me.
C500: C500 Mark II: Observations from a C300MkII point of view
- Thread starter Barry G
- Start date
Definitely to each his own on this. For me AF has become increasingly important, so the Eva was never a consideration. I happily charge my clients extra for combo shoots (more gear at risk, more to load and take home, and the days are just harder) Although, generally we are booked for one or the other more often. (and my video day is 25% more than my photo day.) I think if I was sitting with a c200, or happily with an EVA, I would probably put the money into workflow enhancements rather than the c500II. But that wasn’t my situation. 2 weeks in with the camera, I feel it was a good investment, for me.
Clermond
Veteran
I think it could be a good investment too. I was always missing AF but so far I put 10bit XAVC and 60p over it.
May I bother you with questions?
With what color settings did you shoot the street scenes?
As far as I understood the Nellie Rodeo scene was shot in RAW light?
Did you grade the CF edit directly in FCPX? from RAW lite?
How did you transform then to the ungraded footage? Which kind of log is that after converting?
I have only little experience with Canon Log. As some stated that the
genuine Canon LUTs are crap which one would you recommend as a
good starting point? Thanks much again!
G
Guest
Guest
I’d just say, avoid buying an expensive camera if funds are tight. It can take a few months before the camera starts paying for itself. Skill makes a bigger difference than cameras these days. Although, clients still don’t know that yet.
Clermond
Veteran
that's right - and my focus pulling skills are fare behind the camera AF capabilities.
I don't see other camera options except FX9 with AF + 60p + 10bit XAVC
From the guts I'd say clients will talk about full frame soon. As soon as one of
the competitor starts promoting it. It doesn't mean it would make a difference for
their videos but clients need to have the feeling their supplier is up to date
Last edited:
G
Guest
Guest
If i still shot a lot of interviews, and had moved to FF35, i’d be sure to get AF. Sitting for 45min, or more, micro adjusting focus can be quite the task. Back when interviews were shot on 2/3”, we had the opposite problem, and were looking to reduce dof. Even then, dof was pretty forgiving, and sitting through long interviews, focus was rarely an issue.
puredrifting
Major Contributor
Actually, the C200 outputs 4:2:2 10-bit FHD and 2K via SDI or HDMI. The Ninja Blade is a 1080 only recorder. Most clients, at least our clients, specifically DON'T want 4K. Yes, in 2020. 1080 is still the majority. Interestingly we have two clients though who now only want 4K DCI RAW. So it goes both ways.
Last edited:
Clermond
Veteran
thanks for the information. I didn't have 1080 in mind as 90% is requested in UHD.
Barry - first of all - this is nice work. Secondly, although I'm not really in the market for a new camera - your "tests" & methodology here are very much appreciated and certainly help me understand what the camera is capable of. And totally off topic - but am curious about CFP 2.0 and what's to hate?
I'm a CFP 1.0 user and agree with you that it can roll highlights better than FCP - and have been thinking up updating / upgrading.... so would certain welcome more info. Will be happy to discuss out of this thread by pm, etc if that's preferred... Thanks!
Clermond
Veteran
I was involved in early stage testing and I can say that it’s a great step from CF1 to CF2. Just one note if you decide to test it: make sure that you have a backup of your library before working on it with CF2. CF2 does transform your settings smoothly in the background and from there it’s no way back to CF1 once that it’s done. If you decide to stick with CF1 you wil have to use the old backup library. Send me a pm if you have questions or join the fab group https://www.facebook.com/groups/2876015042408628
Color Street scenes: Canon Log3, Neutral matrix, Not sure White balance, but it was probably the daylight preset or a 5600k setting.
Nellie -- shot in CRL. I've done several grades...some with FCPX tools, although the "final" edit was with color finale 2 Pro plugin inside FCPX. I also did some white balance adjustments on one clip in Canon Raw Development, and that's probably the most accurate workflow, but I wanted to work directly with the raw files in FCP just to see.
CRL comes into FCPX by default as Canon Log 2, and I don't think it matters what gamma you shoot with as it will display as CLog2 with a LUT automatically applied. For the Ungraded, I just turned off all the applied settings, and the LUT, so I'd guess that upload is in clog2. If you process the files in Canon Raw Development, you can convert to a range of output Gammas and color spaces and output formats. If you are trying to adjust white balance, it's best to do in Canon Raw Develop, as it's really a metadata workflow, and gives a very precise result. FCP doesn't have this ability as far as I can tell. Not sure if exposure is handled this way as well.
I wouldn't say canon's LUTS are crap...the new 2.0 version is better than the v1.1 that I have..they just make choices that I wouldn't make. If you were looking for a more modern "baked" neutral skin rendering, you might find that the LUT is a very good starting point for that. I've found that It's really hard to get the same results when comparing "LUT developed" versus "Tools Developed" --there is definitely some secret sauce going on there. They are probably pretty accurate for a broadcast spec Rec709, but I find they make skin tones too red in the shadows and too desaturated in the highlights, and the blues and greens are off more that they should be. I'm working on developing some settings that will allow me a good starting point without using a LUT, and may experiment with lutcalc to develop something I like, but I'm aways from that.
Generally, canon Log 3 is pretty easy to grade by simple adjusting contrast and Saturation. Canon Log 2 can be graded that way, but is in general a fair bit trickier, and a LUT can be helpful.
Ha...I brought it up.
Let me say this. there are some really great things in CFP2. I love the sharpening, the color wheels seem more responsive and the curves produce a result that is more expected than FCP's curves. The vector control is simpler than apples tool. The highlight thing is big. I might be convinced to buy it all but for the last thing on my ***** list below...although all of them add up to a program that simply isn't ready. FWIW, I'm working on a mid spec iMac PRO, which should be able to handle anything this plug in throws at it.
I think the "hate" started when I went to experiment with the Chart calibration feature, as I thought this was a great idea. Compare the image of a chart with some known numbers and you should be able to wrestle a cameras "color science" or lack of into something bordering on accuracy. The fact that they had added the Color Checker Video to the new version made this an interesting diversion for me. After a number of attempts at using it, under a whole variety of inputs and settings choices, I got wholly inconsistent results none of which were anywhere near as accurate as simply applying a LUT and calling it a day ( some results were just damn weird). This was on a perfectly exposed and balanced chart, not some dslr shot at 6400 in the forest. After some chit chat with the developer via email, it became clear that this tool doesn't "do" anything (he questioned why, if my LUT-ed result was good, would I need to apply this feature, and bluntly said that calibration was simply not within it's ability, that no software could correct that many colors..."uhhh...really," I said..."so that's not what X-rite's whole business is"), and went on to say that the feature was meant to match various cameras (this although the workflow is specifically "match chart" not match camera and if it can't get one camera remotely accurate, how is it going to get "two" cameras to be anything but a mess.) So this feature doesn't work in anyway like a normal person would assume.
Next the tracking -- I tried the tracking feature on some of the early "underpass" shots of nellie. In these there were a few times when she moved across the frame at a slightly higher rate than at other times. When this happened, the mask tracked with her just fine, but for some reason the "correction tile" (my term)...that is the masked area that contains the image corrections would lag behind, leaving a rather prominent edge visible in the shot. I tried this several times with the same result. I sent the developer this note, and I didn't receive a response. (I have done a few other tracking tests that did work, but for now, I'll say that this feature doesn't work.).
Next the "make LUT" feature. I thought as I'm probably not going to buy the program, I should save the setting I had on that grade I posted as a LUT. this is a really nice feature. Unfortunately, the LUT that written by CFP2 bore very little resemblance to the settings used to produce it. Even when applied to the same clip. So I'd say this feature doesn't work.
If you make an adjustment in a layer, and want to reverse it, hit Command-z, it removes the plug in from the inspector, and you lose all the work you just did.
While CFP2's all in one approach makes a lot of workflow sense, there are times when I want to remove specific settings from a grade for some reason. Final Cuts "departmentalization" of this aspect can have it's benefits.
Lastly, and this is why I won't buy it., and will be removing it from my system....Something about CFP2 is taxing FCPX's resources, even when it's not in use, and for some reason I can no long scrub clips in the browser smoothly (as soon as I drop them into the timeline they are fine, but in the browser they are unusable for trimming, the playhead sticks and stays stuck till it's good and ready. That on top of re-rendering every time I open the program,(something FCP likes to do with 3rd party plug-ins, so not necessarily CFP2's fault,) but it really has made the workflow unbearable, so I'll be removing it later this weekend.
Thanks for sharing Barry. From everything you describe - I would do the same. Does not sound ready for prime time yet - and my whole thing with staying inside FCPX is for the speed / simplicity.
I'm on a near maxed out non-pro iMac - so would likely experience even worse. I will stick with 1.0 for now - and my occasional dive into Resolve for certain projects.
If I do demo it - at a future time - will let you know how it goes. Cheers.
puredrifting
Major Contributor
Yes, definitely depends on the clients. The majority of our work is for the studios EPK, Home Entertainment and Publicity Departments, most of who only deal with 1080.
We have one client at Fox who likes us to shoot 4K RAW. When I roll 1080 with the Ninja Blade, I am always shooting a UHD XF-AVC as a backup and once in a while, it
gets used. For our own content we are developing, we only shoot 4K RAW.
This is an overexposure test that was motivated by a poorly implemented test published by CVP awhile back. (we believe their test was compromised by improper application of a LUT in post processing). Unfortunately their test has been widely circulated, and appears to show the C500 II as severely hampered by even modest overexposure. I knew this couldn't be true, but figured I'd better test it for myself.
This was shot in XF-AVC CLog2, Neutral Color Matrix, Cinema Gamut and graded using Contrast and Saturation tools in Final Cut Pro (no LUT was used). Whites clip at 5 stops over, and colors begin clipping at 5.5 stops. Skintones are not clipped at 5.5 stops. I would rate the C500II, in XFAVC as having 4-4.5 stops of highlight latitude.
This was shot in XF-AVC CLog2, Neutral Color Matrix, Cinema Gamut and graded using Contrast and Saturation tools in Final Cut Pro (no LUT was used). Whites clip at 5 stops over, and colors begin clipping at 5.5 stops. Skintones are not clipped at 5.5 stops. I would rate the C500II, in XFAVC as having 4-4.5 stops of highlight latitude.
You'll have to rent a camera, and find that out. 
The CVP test chart is a print chart and doesn't have a full range of skin values like the video chart. Based on every chart test like this that I've ever seen, light Caucasian skin tones clip at 1/2 to 1 stop higher exposure than chart white. You should be able to interpolate that. If you can't, or won't. I can't help you. This test isn't for you. It's for me.
While you zero in on whether your pale toe-headed model will look good on this camera, let me remind you that awhile back you said "lets stick to the color and the highlights, Barry". So, whatever your name is... Lets stick to the color and the highlights. One of the big problems with the CVP test was that the C500 color starts fading from the get go...literally at 1/2 stop over and gets progressively worse until the image clips. This is what Liam and I cued in on when we said it was a bad test. You invented a term for it (Ghosting -- which sorry, is already used in video). You can see from my test that when the image is processed properly, the image from the C500 doesn't lose saturation until the image starts to clip. This is a very good result. Now, sit with the Fx9 test at CVP and you'll see that the FX9 colors stay more saturated to the bitter end. this is spectacular, although we need to temper that excitement, as their chart is spectacularly over saturated from the get go. I've tried to match the chart colors as closely as possible in my test, and the camera is doing a very good job at staying very accurate at 4.5 stops over...spilling into 5 stops over. That's something. I'd love to see what the Fx9 looks like with the colors rendered natural to that chart (my guess is the skin tones aren't going to look quite so colorful...but you know..you get what you pay for.)
The CVP test chart is a print chart and doesn't have a full range of skin values like the video chart. Based on every chart test like this that I've ever seen, light Caucasian skin tones clip at 1/2 to 1 stop higher exposure than chart white. You should be able to interpolate that. If you can't, or won't. I can't help you. This test isn't for you. It's for me. While you zero in on whether your pale toe-headed model will look good on this camera, let me remind you that awhile back you said "lets stick to the color and the highlights, Barry". So, whatever your name is... Lets stick to the color and the highlights. One of the big problems with the CVP test was that the C500 color starts fading from the get go...literally at 1/2 stop over and gets progressively worse until the image clips. This is what Liam and I cued in on when we said it was a bad test. You invented a term for it (Ghosting -- which sorry, is already used in video). You can see from my test that when the image is processed properly, the image from the C500 doesn't lose saturation until the image starts to clip. This is a very good result. Now, sit with the Fx9 test at CVP and you'll see that the FX9 colors stay more saturated to the bitter end. this is spectacular, although we need to temper that excitement, as their chart is spectacularly over saturated from the get go. I've tried to match the chart colors as closely as possible in my test, and the camera is doing a very good job at staying very accurate at 4.5 stops over...spilling into 5 stops over. That's something. I'd love to see what the Fx9 looks like with the colors rendered natural to that chart (my guess is the skin tones aren't going to look quite so colorful...but you know..you get what you pay for.)
Last edited:
Again, who I am and my experience doesn't matter in debate- all that matters is evidence, facts, logic, and proofs. My name is John (jcs are initials), and you can see some of my patents in imaging and video here (it appears you don't think I have any experience in this space): https://patents.justia.com/assignee/yoostar-entertainment-group-inc .
Motion blur can be called ghosting and cameras that make people look like Casper are appropriately ghosting skin tones. People disappearing in conversations: ghosting. Many uses of the word when appropriate. Thanks for recognizing the coining of a new use. I think you meant tow-headed? Toe-headed sounds like something from Thumb Wars:
As you may know, cameras know about skintones (and faces) and sometimes perform different color processing in those areas. I've also found that static photos of people don't seem to work as well as a live person for testing.
If the C500 II can do better than CVP in overexposed skin, cool. The debate started when the FX9 was stated as having the best IQ vs. the C500 II: we still need a side by side test ideally to see if the C500 II can match or exceed the FX9 in the same conditions. I'd estimate the FX9 would still win based on sensor tech, however the only way to know is by doing the test.
A friend may be doing a test soon, let them know my thoughts on what should be tested, not sure if I can be there during the test. Perhaps they'll share results on this forum.
Motion blur can be called ghosting and cameras that make people look like Casper are appropriately ghosting skin tones. People disappearing in conversations: ghosting. Many uses of the word when appropriate. Thanks for recognizing the coining of a new use. I think you meant tow-headed? Toe-headed sounds like something from Thumb Wars:
As you may know, cameras know about skintones (and faces) and sometimes perform different color processing in those areas. I've also found that static photos of people don't seem to work as well as a live person for testing.
If the C500 II can do better than CVP in overexposed skin, cool. The debate started when the FX9 was stated as having the best IQ vs. the C500 II: we still need a side by side test ideally to see if the C500 II can match or exceed the FX9 in the same conditions. I'd estimate the FX9 would still win based on sensor tech, however the only way to know is by doing the test.
A friend may be doing a test soon, let them know my thoughts on what should be tested, not sure if I can be there during the test. Perhaps they'll share results on this forum.
Ha...if it doesn't matter, why bring it up? I never asked what you did. It's nice to know your name, John, but see ...you have a little problem acknowledging when you see evidence of something, you seem incapable of grasping the Logic you toss around like a fuzzball, at least when someone else is dishing it up. You've spent a LOT of time crapping on the quality of a camera you've never touched, with nothing but largely "anecdotal" evidence to back you up. If you're an engineer, tell me how often does "anecdotal" get tossed around when you're writing briefs for your patents?
I've shown you that the test you relied on for much of your argument ("the CVP test shows me what I knew all along") was fatally flawed. I've shown you that the C500II has excellent color when overexposed. I've showed you how known skin tone values respond to over exposure. But you literally are incapable of saying..."hmm...yeah..you know...that looks a lot better than what I thought. I wish it was a skin tone test, but yeah...I guess you guys were talking about stuff you understood, that I somehow missed" .
Don't worry, I don't need you to say it...so don't. I know, that in your mind it ALL started with you and what you wanted to see. But that's simply not true. Liam started the thread, and I was in it from the very beginning. We were talking about lots of issues regarding these cameras. Because no one aspect is the one that makes you buy it.
Now if I might make a respectful request: I'd really wish you'd take this fight back to the A vs B thread. This thread is about reporting data from one camera. It might be lonely over here, but I'm just fine with that.
I've shown you that the test you relied on for much of your argument ("the CVP test shows me what I knew all along") was fatally flawed. I've shown you that the C500II has excellent color when overexposed. I've showed you how known skin tone values respond to over exposure. But you literally are incapable of saying..."hmm...yeah..you know...that looks a lot better than what I thought. I wish it was a skin tone test, but yeah...I guess you guys were talking about stuff you understood, that I somehow missed" .
Don't worry, I don't need you to say it...so don't. I know, that in your mind it ALL started with you and what you wanted to see. But that's simply not true. Liam started the thread, and I was in it from the very beginning. We were talking about lots of issues regarding these cameras. Because no one aspect is the one that makes you buy it.
Now if I might make a respectful request: I'd really wish you'd take this fight back to the A vs B thread. This thread is about reporting data from one camera. It might be lonely over here, but I'm just fine with that.
G
Guest
Guest
What is interesting, now that Red brand cameras are actually good, no one wants them anymore. Haha!
But so true about the cheap lens and lens addons, and walking out the door for whatever is offered.
Anyway, i guess all the cameras are good these days. The Canon C500ii is the first Canon C i am interested in.