Canon goes 8K60p and "20+ stops DR" in 4K by NAB

Canon Cinema EOS C700DR
4K Full Frame
4K/240p in Fast Mode, 4K/180p in WDR Mode
BSI Stacked 9.6um 4K WDR Sensor
20+ Stops Dynamic Range in WDR Mode
5ms Readout in 4K WDR Mode
Dual DIGIC DV 8 Image Processor

The C700DR has me very interested. That's larger photosites than the Alexa and 4 times larger than the 8k C500 model. That would give it 2 stops more sensitivity than the C500s. If the C500s is hovering around 14 stops, we might be seeing the first camera with a legit 16 stops of dynamic range.
 
The C700DR has me very interested. That's larger photosites than the Alexa and 4 times larger than the 8k C500 model. That would give it 2 stops more sensitivity than the C500s. If the C500s is hovering around 14 stops, we might be seeing the first camera with a legit 16 stops of dynamic range.

If those specs are real, and they get the formfactor right (Sony Venice, Red Raptor XL or Kinefinity Mavo Edge-esque), it could be one hell of a camera.
 
Very interested in the C700DR compared to an Alexa LF. It has potential considering the large photo sites for sure. Will be excellent in low light / clean shadows as well. 240fps in 4k is not bad for HFR at all either with high DR. For me it will come down to mainly cost and form factor and mount options. I own a RED Gemini / Komodo mainly because I do a lot of gimbal work with smaller crews and need modularity and travel a lot, but rent Alexa when I can on larger projects. I am not fond of the non-swappable RF mount on my Komodo and now V-raptor and it will be interesting to see what adaptors or interchangeable mounts will be available for these new Canon RF cameras. As for resolution, I have been delivering 2k a lot to clients and you still will get enough downsampling in post even with 4k capture. 8K is overrated unless you have clients asking for 4k finals. The cost of delivery (i.e. post processing, drive storage, cards, etc.) goes up with 8K. A lot of people have been asking for a full frame 4k-6k sensor with large photo sites for a while now. My take on RED is that they keep focusing on higher resolution rather than greater picture quality or usability, which is a reason why Gemini / Alexa are still the sweet spot for me over upgrading to something like V-Raptor. If the form factor of the Canon C700 series becomes more like a C200 in size and was more modular for gimbal work and building the camera up or down, it would be hard not to take a closer look.
 
Last edited:
If those specs are real, and they get the formfactor right (Sony Venice, Red Raptor XL or Kinefinity Mavo Edge-esque), it could be one hell of a camera.

I'm asking this in all seriousness: has any camera—other than the Alexa—actually delivered more than 14 stops of independently verified dynamic range?
 
. 240fps in 4k is not bad for HFR at all either with high DR.

Right but it does look like the camera would have maximum DR only at 180fps and below

Also- as far as I know, these rumors have not been updated since May. Usually rumors are more accurate as you get closer to launch date. So, there's a chance that these specs could be wildly off
 
I just hope Canon can prove their 20 stops dynamic range in the court of law if a class action suit is filed. Legally speaking, can companies make arbitrary claims such as with dynamic range if the claims are not true? Is that not fraud?

I'm not seeing much appeal in the rumored C300 specs here. 8K is practically a downside at this point, unless it can still do 4K and 2K perfectly well. Give me an RF mount and a joystick I can push in to make selections and those would be upgrades.

I like a lot of the C300's features so I've been thinking about soldering in one of the ALEV III sensors from my Amiras into the C300 and hoping that I can then get things like a smaller body, auto-focus, better low light, etc.

Actually, I don't know how to solder. But what I could do is get a camera designer to take the ALEV III sensor from an Arri camera, place it in another camera and actually hook it up properly, and then sell that camera. Now, Arri probably has copyright claims on their sensor that would prohibit reselling it in another camera body and design, so this may need to be a black market special.

Yes because DR is not a valid form of measurement dictated by a group. Its sot of like the lux rating cameras used to have. It wasn't a real standard so they could kind of say whatever they wanted to.

With that said nobody is going to make a claim of 20 stops if its not at least in the ballpark. People can fudge the definition of the very bottom and very top stops meaning a error of +/- 2 stops for any given DR claim but its impossible to make a 20 stop claim and not at least get 18 solid stops. No matter what kind of fuzzy math is used. The values are either there or they are not. Its kind of hard to fake it beyond a generous definition of what is clipping in the highs and what is acceptable in that bottom stop in terms of noise and detail.
 
I'm asking this in all seriousness: has any camera—other than the Alexa—actually delivered more than 14 stops of independently verified dynamic range?

You already asked that in #7 :cheesy:

Serious question: has any camera manufacturer actually demonstrated—via independent testing—more than the 14 stops of dynamic range you get out of the Alexa sensor? I know many have *claimed* more DR but I don't recall seeing it ever pan out.
 
Can 20 stops of DR actually sometimes hurt? What I mean by that is when we have that much DR we assume a series of shots will look great but when its time to delivery a rec709 to consumers we are faced with less than desirable details. Sure that can probably be adjusted in grading but for decades we got around dynamic range limitations by using lighting, reflectors, ND sheets on windows and so forth. We used methods to create a better balanced scene without the extremes. Just depending on wide dynamic range means we get good details I the extremes but its harder to dial that down to a better balanced SDR scene.

Wide dynamic range can be a bit of a crutch to toss out high production value methods and in an attempt to get better DR we could end up with a worse final product. Not saying that will happen for everyone. Just that I think its still going to be important to try to balance the environment. 20 stops should not be used as an excuse to shoot as is. It can help gives more control in post but setting up a good environment will still be equally important.

I also think it could have a negative impact on how we perceive what we can deliver. We saw this with HD when we still had to deliver SD or DVD. It was rough editing in HD and then having to watch a SD compressed version. It made SD look much worse and some compensated by trying to make the down conversion sharper which had worse motion artifacts and could introduce aliasing or moire.

I'm not saying we should stick to SDR forever. Just that I think its important to keep in mind the methods and techniques that have been used in productions for over a hundred years. Even when they were shot on film with higher dynamic range these methods were still used.

Just thinking out loud I guess. Obviously HDR when we finally get to a realistic point where we can accurately grade/deliver/watch true HDR changes everything. Right now at best most experience a fraction of HDR.
 
its impossible to make a 20 stop claim and not at least get 18 solid stops.

I have no idea, but RED claimed 17+ for the Monstro VV and I'm not sure anyone tested it at 15. But the whole thing is so loosey goosey, I almost feel like it's better to do what Matt Allard at Newsshooter does and just use it and say, "Yeah, this has a lot of DR. Probably a half-stop more than that other camera I've also used." So, kind of a subjective, comparative analysis.

I want an objective, verifiable standard. We all do. But I think there are so many variables (types of noise, color fidelity at the edges of the exposure range, etc) that it's more meaningful to make a qualitative assessment of how it looks and how it handles varied and challenging circumstances
 
Technically, it's possible with a multiple exposure blend. But I turned down the brightness level on my TV because I don't want it to be blinding me with science.
 
I have no idea, but RED claimed 17+ for the Monstro VV and I'm not sure anyone tested it at 15. But the whole thing is so loosey goosey, I almost feel like it's better to do what Matt Allard at Newsshooter does and just use it and say, "Yeah, this has a lot of DR. Probably a half-stop more than that other camera I've also used." So, kind of a subjective, comparative analysis.

I want an objective, verifiable standard. We all do. But I think there are so many variables (types of noise, color fidelity at the edges of the exposure range, etc) that it's more meaningful to make a qualitative assessment of how it looks and how it handles varied and challenging circumstances

I for the most part agree. when it comes to wide dynamic range I feel like we have two standards. SDR and higher DR. Beyond that the number really starts to matter very little. Be it 12,13 a fake 14 stops or whatever a manufacturer tries to claim. Whats important with dynamic range is we get more than SDR is capable of so we have room to tweak. Its not really meant as a replacement. Shoot against a bright window and 12,13 or 14 stops will all likely need some help with internal lighting to match or filtering on the window to match the interior. How much is needed is not as important as does one need the extra effort or not.

I have also seen a Sony FS5 side by side with a GH5 and the GH5 appeared to have more range. Not sure if the person royally screwed up the Sony but I have seen that a lot actually. Too many reviews mistakingly assume dynamic range is just less clipping. The P4k for example has 13 stops but you only get a ton of boost in the highlights at ISO 1000 which has a ton of noise and very few actually use. Shooting at other ISO modes below ISO 1000 still have 13 stops but if one only looks at clouds they would think it was less.

I also feel like the light bar tests are rather pointless. They tell us how many numbers of bars we can see before we hit the noise floor but they tell us nothing of the actual image resolved. I think this is why sometimes the Gh5 can shoot a real life image that looks like it has better range vs a Sony which should have more. Its like only using a resolution chart to determine if a lens is beautiful when we know some softer vintage lenses look more beautiful. Charts are not everything which is again why measuring DR is kind of pointless and a pseudo science. The number is rather meaningless really if thats all we care about. Charts can also tell us two cameras are exactly 12 stops and yet both look about as night and day different as possible when it comes to range.

Thats why yes I agree with the unit of measurement of "Yeah, this has a lot of DR". At the end of the day. We should only be looking at does it have SDR and some kind of HDR mode either HLG, HDR or Log. Beyond that the rest doesn't matter as much anymore. I'm kind of to the point now where I make sure its more than 10 and then I'm fine. I just bought a Canon R6 which apparently only has 12 stops and so far I think its great. Seems to be lightyears ahead of my GH4 which also claimed 12 stops. For me c.log3 is a way to get beyond SDR and it doesn't bother me that its 12 stops. Obviously if I shoot in a jacked up situation something is going to get clipped but thats true of the P4k, RED and many other cameras I have used. I know in those situations I have to expose for whats important or bounce/add/remove light as needed. Same I would do if I was shooting with film.
 
Just by critically-thinking about it, the most prominent improvement for our humans eyes over the last 30 years has been resolution. With resolution came better color and dynamic range and shallow depth-of-field, but resolution was always #1.

There was a clear difference from SD to HD to 4K and that was the main reason to improve visual technology, photos and video.

Now that 4K seems to be good enough for most and the large sensor sizes are approaching their limit (and colors are usable from everyone), maybe dynamic range is finally next.

__

And sometimes we live our lives and kind of realize one day that technological changes have taken place without really noticing when it happened (like one day you realized pagers/beepers weren't used anymore or standard-definition disappeared).

I think that's what's going to happen with HDR. One day, a lot of the content on YouTube will look like high-quality HDR phone footage with a lot of visible detail.

It will look different than most of the content in 2021, but it will be normal and that's the transition that will have taken place like from SD to HD without really thinking about it.
 
I keep my 4K TV in HDR 10 mode now even though its a 600 nit display. I know I'm not even scratching the surface of the full potential of HDR but I prefer it. I do kick myself that I didn't pay the extra $400 for the higher nit model but oh well. A lot of the more critical colorists I know pretty much poopoo any form of HDR out there right now and told me it all sucks and isn't worth it. Even the $3k TVs don't do it right yet according to them.

Still trying to figure out how much of that is true and how much is snobbery.

Clearly HDR will keep getting better and I think thats a good thing. If this is the worse it will ever be then thats great. The HDR material we create today will keep looking better as the displays get better. Vs a half baked standard today that will be half baked in five years. HDR should be a standard beyond whats capable right now so our content can grow with it. Sort of like shooting 4k when we only had HD monitors. We knew someday that content would be viewed 4k. The side by side split screen SDR and HDR I have watched on my TV are not the blow your socks off kind of demos some have seen at NAB in the past but there is a noticeable difference.
 
I think the RED Monstro is the only camera that matches ARRI's dynamic range. The C500 MK2 and C300 MK3 come close though so I wouldn't be surprised if they made something that surpassed the Alexa's dynamic range. Canon LOG isn't as easy to deal with as Arri LOG though IMO.
 
Their LOG 1 isn't as easy because it's kind of like their WDR (not that flat), but 2 and 3 are great, especially 2.
 
I think the RED Monstro is the only camera that matches ARRI's dynamic range. The C500 MK2 and C300 MK3 come close though so I wouldn't be surprised if they made something that surpassed the Alexa's dynamic range. Canon LOG isn't as easy to deal with as Arri LOG though IMO.

Has the RED been independently verified? Genuinely asking—and if so, I'd love to see a link.
 
Yes because DR is not a valid form of measurement dictated by a group. Its sot of like the lux rating cameras used to have. It wasn't a real standard so they could kind of say whatever they wanted to.

I want an objective, verifiable standard. We all do.

In fact there is an objective, verifiable standard, since 2004, straight from the most famous standards body in the world, the International Organization for Standardization: ISO 21550:2004

It is max well capacity divided by noise floor. This means that the darkest stops will be noisy. But if you measure all cameras by that standard, then you could at least compare them easily. (For example, Alexa may test out at 15, and Canon 16. Then you may, through experience, decide that usually your own acceptable level is the ISO - 2. So for you, the Alexa is 13, and the Canon 14. But now you can get a rough idea just from looking at their specs [and test further if you want to] without having to conduct or observe your own exhaustive tests just to get a ballpark idea of what a certain camera can give you.)
 
Back
Top