F5: When less is more (an ode to SGamut3.cine)

cuervo said:
In the case where the gamut is undefined, or where it's specifically stated "it's an emulation of the original SGamut", I would agree with you Marv. However, Sony specifically calls out Sgamut3 for both the F5 and the F55. Unless they were perpetrating a fraud, they wouldn't define things so clearly. You're wrong about the F5 color space. I can't speak for the FS7. There's numerous white papers defining the Sgamut3 and Sgamut3.cine color spaces, as well as why they exist and how they were derived. None of these papers support your assumption.

Having said that, I don't doubt that the F55 debayer algorithmns are different that the F5. The F55 has a different color filter than the F5. While the F55 sensor has a larger acquisition space, sgamut3 is defined by specification. I would refer you here:
http://www.dvinfo.net/article/acquisition/sonyxdcam/sony_sgamut_vs_sgamut3.html
Click to expand...

F5 and F55/F65 have different CFA, the colour space is definitely different. This was backed up by Alister Chapman, a Sony ICE (Independently Certified Expert)

The F5's colour gamut in S-Gamut far exceeds Rec-709. The filters are designed to match most other typical broadcast cameras, so when shooting in 709, just the 709 range is used and the look will be similar to most other Sony cameras, but the colour filters used allow the camera to see well beyond 709. Not as far as, or with as much separation as the F55 and F65, but still much greater than 709, hence the need to use S-Gamut to capture the cameras full range.
Click to expand...

http://www.dvinfo.net/forum/sony-pmwf5-pmwf55-cinealta/519598-xavc-rec-2020-a.html Post #9
 
I've read Alister's description of color spaces over at DVINFO before. His comments and mine are not in conflict. S-gamut3 is S-gamut3, by specification, whether it's an F5 or an F55. The fact that the F55 color sensor and CFA are different only says that the F55 sensor sees a wider color space than the F5. But, it's throttled back(by something like an in-camera LUT), to Sgamut3 or sgamut3.cine or S-gamut. The F55 sensor probably does see REC2020, while the F5 sensor doesn't see REC2020 . But S-gamut3 is a specification that both cameras see. S-gamut3 is a subset(i.e. it's smaller) than REC2020. As Alister points out, color space of the sensor is not the same as the recorded specification (ie sgamut3) color space.

But, in the end, since Sony has not defined all this very well, if at all, the best any of us, including Alister Chapman, can do is make educated guesses. It's all irrelevant as long as the only viewing color space any of us have is REC709. There are no display monitors that see REC2020. So, until that happens, anything that records in a space bigger than REC709 is academic. The open question is how the larger color gamuts, like sgamut3, are remapped to be displayed by a REC709 display system; and, that is a function of a LUT. Does the LUT clip the "out of REC709 bounds" colors, or does it remap them? And what does a REC709 display monitor do to a larger color gamut if no LUT has been applied?
 
Last edited:
cuervo said:
I've read Alister's description of color spaces over at DVINFO before. His comments and mine are not in conflict. S-gamut3 is S-gamut3, whether it's an F5 or an F55. The fact that the F55 color sensor and CFA are different only says that the F55 sensor sees a wider color space than the F5. But, it's throttled back, whether it's to Sgamut3 or sgamut3.cine. The F5 doesn't see REC2020, and the F55 probably does see REC2020. But S-gamut3 is a specification that both cameras see. S-gamut3 is a subset(i.e. it's smaller) than REC2020.

But, in the end, since Sony has not defined all this very well, if at all, the best any of us, including Alister Chapman, can do is make educated guesses. It's all irrelevant as long as the only viewing color space any of us have is REC709. There are no display monitors that see REC2020. So, until that happens, anything that records in a space bigger than REC709 is academic. The open question is how the larger color gamuts, like sgamut3, are remapped by a REC709 display system.
Click to expand...

First of all, SGamut3 is bigger than REC.2020

http://community.sony.com/sony/atta...ry_for_S-Gamut3Cine_S-Gamut3_S-Log3_V1_00.pdf Page 4

Also, below are a few sources that confirm F5's colour space is visibly different than F55 (smaller)

http://community.sony.com/t5/F5-F55/S-gamut-on-F5/td-p/135691

http://www.cinematography.net/edited-pages/F5 v F55.html
F5 has a CFA which is significantly wider gamut than REC709, but the F55's CFA is MUCH wider gamut. Wider even than film. Wider than anything is currently able to reproduce.
Click to expand...

http://www.dvxuser.com/V6/showthread.php?316999-F55-S-Log-and-S-Gamut
Go shoot SGamut and point it at some stage/rock&roll lights that are LED color mixers outputting a really saturated shade. The F5 will clip one of the color channels, and a 55 likely will not.
Click to expand...
 
cuervo said:
Thanx for that reference. I'd read it but misread the REC2020 vs. Sgamut3 space. You're right. I need to rethink this. I also need to read the other references you quoted as I haven't seen those. Thanx again. Still trying to wrap my head around this.
Click to expand...

No worries, these are all Sony's sophisticated marketing.
 
Spent this weekend just camera testing the FS7 and I am LOVING SGamut3Cine + SLog3. It just looks really nice IMO. Take very minimal grading to get it to shine. Certainly like it more than the SLog2 + FS700 I was using. SGamut3Cine is just nice, nice, nice! It just looks sophisticated, subtle and not over the top.

I just wish Sony would make CineEL a full blown profile with custom white balance and adjustable ISO. Then it would be perfect. I don't think Custom mode is the solution to that because it just looks different than CineEL.

I noticed that the color space for Custom is Matrix instead of SGamut3Cine. Is this why the two profiles look different, even though they are suppose to be similar (or so I thought)?

Edit: I messed around with the other modes like Custom and some hypergammas and those just seem to have way more of the Sony look. CineEl just has it's own thing going. It feels really good compared to the others. I think it's the only mode I'm going to use.
 
Last edited:
cuervo said:
Back to my original question...why shoot the smaller gamut of sgamut3.cine, if one can shoot the larger sgamut3 and LUT it to sgamut3.cine in post? How does the conversion LUT work? Does it remap colors outside of the sgamut3.cine space or does it clip those values to get to sgamut3.cine?
Click to expand...

My reasoning would be that if SGamut3.cine get me the colours I want now - why would I bother with the additional hassle to wrangle the wider gamut into what I actually want? I think the likelihood of revisiting footage to regrade it when you can actually harness a wider gamut is incredibly small - especially when you can already get the colours as nice as you'd want with SGamut3.cine.
 
Grug,
Perfectly good reason. My only comment is that there is some value to recording with the greatest color space possible, then reducing that gamut to a smaller gamut in post. That's the argument for shooting sgamut3.cine, despite delivering to REC709. There's just more bits of data there to contribute to a better final image. However, to your point, the improvement may not be worth the trouble. Applying the Sony conversion gamut is only one step, which my be trivial when stacked with other LUTs and CC's. I'm not even sure that converting sgamut3 to sgamut3.cine would yield the same result as recording straight to sgamut3.cine. That's why I was asking.
 
Last edited:
You must log in or register to reply here.
Back
Top