C100 with Ninja has 60% more chroma information than a Blackmagic

P

philiplipetz

Veteran
It turns out that the C100 external combo has substantially more chroma information than a Blackmagic Cinema Camera shooting ProRes 1080p. Maximizing chroma information is the key to color grading as well as user experience. Here is a quote from Barry Green in the BMCuser forum. Remember that the C100 sensor is not debayered, like a true 3 chip camera. That is why a C100 plus recorder has more chroma information than Blackmagic ProRes that is debayered, an astonishing 60%.

The debayered ProRes BMCC signal averages RGB information to bring all three RGB values into each point where physically only one of the three RGB values is actually being measured, that reduces chroma information as Barry describes below. In contrast, the C100/ C300 sensor is not debayered so it delivers unaveraged chroma information. The C100/C300 sensor is unique for single sensor cameras in that it takes a 1080p image from a 4k sensor with each color sample coming froma single unaveraged R, G,or B detector so that it acts like 3 chip camera where there are separate unique channels for each RGB point, no averaging to bring information into points that are lacking it.

Since most people plan on shooting BMCC ProRes to save space and grading time a 60% difference (1,036,800 color samples with C100 to an external recorder divided by 648,000 color samples in BMCC ProRes) is an important consideration.

Quote from Barry Green:

Q: Are you saying that the 2.5k bayer pattern of the BMC camera even when loaded into Resolve on a 1080p project have the same chroma latitude and info of a 422 ?

Barry Green:

A: Seems like a simple question, but leads to a complex answer. It depends on what sensor made the 1080p image! But let's assume for the moment that we're talking about a true three-chip 1080p camera, so there's true color info in all of the 4:2:2 decimation, that would give you chroma resolution of 960x1080. Or, per frame, 1,036,800 color samples.

The BMC has 1200x675 chroma samples from its sensor. If you debayered and stored in ProRes 444, you could preserve all that chroma, and would have 810,000 chroma samples. So no, it's not possible to have as much chroma resolution (even with ProRes 444) as a true 4:2:2 camera would deliver.

Now, when shooting ProRes 422, it's a little worse actually, because your 1200x675 will get scaled down to fit into 960x1080, so the net result will be 960x675, for a grand total of 648,000 chroma samples.

The way to preserve the most chroma information is to avoid going to a 4:2:2 codec. Keep it raw, or transcode to a 4:4:4 codec, and you'll have the most chroma the camera can give you. ..."
 
Last edited:
Well, that is not actually that important to color grading as you make it seem. 4:2:2, 4:2:0. It doesn't matter THAT much. Only for keying or doing color mattes where the edges are important. More important for color grading is A) latitude and B) less compression so that you can mess around easier.
 
maarek said:
Well, that is not actually that important to color grading as you make it seem. 4:2:2, 4:2:0. It doesn't matter THAT much. Only for keying or doing color mattes where the edges are important. More important for color grading is A) latitude and B) less compression so that you can mess around easier.
Click to expand...

My colorist shows that samples with less color information break much easier when trying to grade for maximum DR.
 
stefanhuberfilms said:
Hmm, will this be the same on a FS700 with an external recorder?
Click to expand...

I believe that the FS series is debayered, only the C100/C300/C500 sensor is uniquely designed so that is a true bayer imager that acts as if it is a 3 chip sensor array.

This is why the Canon Cinema Sensor delivers a superior color gradients over the debayered single sensor competition.
 
Last edited:
stefanhuberfilms said:
Hmm, will this be the same on a FS700 with an external recorder?
Click to expand...
Probably. It depends on how the FS700 processes its chip for 1080p, but likely, I would expect it could/should deliver full color resolution.

The FS100, no, but the FS700 uses a 4K chip, like the Canon does, like the Red does, etc. Canon processes it differently than others, by avoiding the de-bayer process (and thus sacrificing any prospect of making 4K footage from it, which must seem like a reasonable tradeoff seeing as the C100 & C300 can only record or output 1080p).

Fundamentally the FS700 has the pixels to deliver the full 4:4:4 1080p image like the Canon does, it just depends on whether they process it internally as a 4:4:4 1080p (grouping each 2x2 block of pixels into a single 1080p pixel) or if they de-bayer it, making a 4k image. Even then, a downscaled 4K should preserve enough color information for a true 4:4:4 1080p if done properly.
 
Most threads say no one knows if the the FS700 image is debayered, but some people speculate that it is since there is noticeable decreases in image quality at 120/249 fps where debayering would have to be constrained due to limits in camera processing power. What happens with the external feed is even more of a mystery, as Barry points out. But as Barry also points out the 4k external feed would have to debayered, so would the camera be designed for both bayer and debayered external feeds?
 
Last edited:
On the other hand, does that mean that the future 4K footage one will get from the FS700 will have less color information than the 1080p footage you get wia hd-sdi on an external recorder? Though, it will be 4K RAW isn't it?
 
The 4K external feed will be raw. Whether you choose to de-bayer it (to create a "4k" image, which will be around 3.2k of detail and roughly 4:2:0 color) or you choose to ... well, what do we call it, block-sample it? to 1920x1080 by using each 2x2 block as an individual pixel, well, on a 4K raw image that choice is probably something you can make in post, assuming there's interpretation software that allows you to read the 4k bayer pattern as a 2x2 block per pixel.

does that mean that the future 4K footage one will get from the FS700 will have less color information than the 1080p footage you get wia hd-sdi on an external recorder?
Click to expand...
In relative terms per pixel, yes. In absolute terms of overall color pixels, no, it'll be the same. But in terms of per-detail-pixel, the 4K will be roughly 4:2:0-ish, whereas the 1080p could be up to 4:4:4.
 
This is really interesting. I'm curious to see more from the C100 as I love the idea of the FS700 and BMCC but don't particularly love the footage from the Sony or the handling of the BM. We are in a time where cameras are VERY close to having it all there for a decent price...so close...
 
This is why the Canon Cinema cameras are so much more than other cameras on the market. They employ a brand new sensor technology, which was expensive to research and produce. One of the main reasons I switched from the FS100 to the c300 was that it couldn't distinguish the difference between hot pink and orange. I love having a S35 sensor that acts exactly like a 3 chip camera with the same color reproduction.
 
Using Barry's figures it appears that the C100 plus external recorder, or a C300, has about 28% more chroma information than a Blackmagic shooting RAW. 1037k color samples versus 810k color samples.
 
philiplipetz said:
Using Barry's figures it appears that the C100 plus external recorder, or a C300, has about 28% more chroma information than a Blackmagic shooting RAW. 1037k color samples versus 810k color samples.
Click to expand...

But the bit-depth of those colour samples is substantially higher with the BMC (12-bit Raw or 10-bit ProRes), so the gradations between them should hold more information than the C100's 8-bit, shouldn't it?
 
Grug said:
But the bit-depth of those colour samples is substantially higher with the BMC (12-bit Raw or 10-bit ProRes), so the gradations between them should hold more information than the C100's 8-bit, shouldn't it?
Click to expand...

Yes, highly valid observation. 12-bit with fewer color samples would probably result in substantially better imagery than 8-bit with more color samples, especially in terms of gradation.
 
Barry_Green said:
The 4K external feed will be raw. Whether you choose to de-bayer it (to create a "4k" image, which will be around 3.2k of detail and roughly 4:2:0 color) or you choose to ... well, what do we call it, block-sample it? to 1920x1080 by using each 2x2 block as an individual pixel, well, on a 4K raw image that choice is probably something you can make in post, assuming there's interpretation software that allows you to read the 4k bayer pattern as a 2x2 block per pixel.


In relative terms per pixel, yes. In absolute terms of overall color pixels, no, it'll be the same. But in terms of per-detail-pixel, the 4K will be roughly 4:2:0-ish, whereas the 1080p could be up to 4:4:4.
Click to expand...

Turning a block region (like 2x2 or 4x4 or whatever) of pixels into 1 pixel is called pixel binning.
 
Binning is combining multiple pixels into one, yes, but in a monochrome way. In turning a Bayer pattern RGBG block into a single RGB pixel, presumably there's a different term for that? Or is it simply demosaic'ing the 2x2 block?
 
philiplipetz said:
The C100/C300 sensor is unique for single sensor cameras in that it takes a 1080p image from a 4k sensor with each color sample coming froma single unaveraged R, G,or B detector so that it acts like 3 chip camera where there are separate unique channels for each RGB point, no averaging to bring information into points that are lacking it.
"
Click to expand...

Untrue. There are other cameras that do this. But some aren't cheap. And some aren't for sale.
 
OK, the Canon EOS sensor is the only affordable large single sensor that acts like a thrEe chip array maximizing color information.
 
Last edited:
You must log in or register to reply here.
Back
Top