Welcome to our community

Be a part of something great, join today!

Log in
  • Hey all, just changed over the backend after 15 years I figured time to give it a bit of an update, its probably gonna be a bit weird for most of you and i am sure there is a few bugs to work out but it should kinda work the same as before... hopefully :)

Color space...

K. cromwell

K. cromwell

Well-known member
Joined
Dec 9, 2009
Messages
155
Reaction score
0
Points
0
Location
NY
Website
www.karlonc.com
Why did I read on reduser that Scarlet's color space is 4:2:0

Is this true? To me it just doesn't make sense.
 
4:2:0 is not a description of a colour space anyway - it's a description of Y'CbCr (often referred to as YUV) chroma sub-sampling. I wrote this article: http://www.lafcpug.org/Tutorials/basic_chroma_sample.html a long time back but it does give a good overview of the subject.

Graeme
 
Where did you read that? Of course, you could transcode Scarlet RAW footage to a 4:2:0 sampling codec, but apart from that it makes no sense.
 
Somebody who doesn't understand chroma subsampling is probably claiming that because of how a bayer pattern works... But it's wrong and not applicable to cameras that shoot RAW.
 
Graeme Nattress said:
4:2:0 is not a description of a colour space anyway - it's a description of Y'CbCr (often referred to as YUV) chroma sub-sampling. I wrote this article: http://www.lafcpug.org/Tutorials/basic_chroma_sample.html a long time back but it does give a good overview of the subject.

Graeme
Click to expand...

It was YUV I meant... but to me for Red to do that wouldn't make sense especially if they are competing with 2/3 party, it just turned me off from spending 8k to 28k. I do believe in progress (moving from HVX - RED) but what is the YUV on the scarlets anyway?
 
We don't do YUV. Everything is RGB RAW.

Graeme
 
I might be wrong but since it's RAW you can apply the color sampling you want (for exemple RGB 444).
 
I couldn't find Graeme Nattress's original post, but over on ScarletUser, Dances With Cameras said: "Graeme Nattress says that, after deBayering, the RED cameras give ~78 % real luma resolution and ~54 % real chroma [color] resolution. So, from Scarlet's 3K CMOS Bayer sensor, you'll get a 2.35K real luma resolution and ~1.7K color resolution. Which gives you pristine 2K [+ additional .35Kfor re-framing, stabilizing, etc.]. I understand that 422, 444, etc are video terms. If we have to talk in these terms, at 2K you get an image which is slightly better than '4:2:2'." Given the continual improvements Graeme is making to the debayer algorithms, I expect Scarlet will be somewhat better than these figures. But just because it's encoded as 4:4:4 that doesn't mean all the information is there.
 
1) 4:2:0, 4:2:2, 4:4:4 are not colour spaces - they're examples of Chroma sub-sampling notation for Y'CbCr video
2) A Colour Space is something like REC709, sRGB, Adobe1998, REDColor. They define how RGB colours look
3) We record RAW. RAW neither has a colour space nor chroma subsampling.
4) We do use a Bayer pattern sensor which some think of as chroma subsampling, because of the relative densities of the red, green and blue filtered photosites. However, it's a very different process in reality.
5) 4:4:4 etc. does not say anything about the measured resolution of the luma or chroma, just how it's stored.
6) Sensible thing to do is shoot a target and measure.

Graeme
 
Graeme Nattress said:
1) 4:2:0, 4:2:2, 4:4:4 are not colour spaces - they're examples of Chroma sub-sampling notation for Y'CbCr video
2) A Colour Space is something like REC709, sRGB, Adobe1998, REDColor. They define how RGB colours look
3) We record RAW. RAW neither has a colour space nor chroma subsampling.
4) We do use a Bayer pattern sensor which some think of as chroma subsampling, because of the relative densities of the red, green and blue filtered photosites. However, it's a very different process in reality.
5) 4:4:4 etc. does not say anything about the measured resolution of the luma or chroma, just how it's stored.
6) Sensible thing to do is shoot a target and measure.

Graeme
Click to expand...


In my world, which is much simpler and very much less informed than Graeme's, I am going to take that to mean I can get all the color I need! :thumbup1:
 
I think so Terry!

Graeme
 
Geeky video speak language!

Graeme
 
Graeme Nattress said:
Geeky video speak language!

Graeme
Click to expand...

Well thank you now when people ask about color space, or YUV measurements I will say RED doesn't shoot in that arena, and that it is only defined when you create your final output file.

Thanks Graeme
 
K. cromwell said:
Well thank you now when people ask about color space, or YUV measurements I will say RED doesn't shoot in that arena, and that it is only defined when you create your final output file.

Thanks Graeme
Click to expand...

Hey K.
I don't think your proposed answer "when people ask" is exactly accurate. You are welcome to embrace your inner geek and dive into these topics much deeper so that you can be absolutely sure about exactly what is happening from the time photons hit the sensor until you sit down in the theater. But there are so many topics from sensor to projection that you will essentially be taking a post production course. All that knowledge is great but I think Terry said it best. You get all the color you need.

I think you would find out what you need to know by looking at some footage from the RED MX sensor. Off Hollywoods "brooklyn walk", The MX Reel, Any RED reel will let you know what you need to know.
 
I think this thread would have been a lot shorter if the response to the 4:2:0 questions would have been, that 4:2:0 describes color subsampling, a fitting of data really. RED doesn't subsample, it record the raw data. All the pixels, all the values.
 
You must log in or register to reply here.
Back
Top