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 :)

Shooting Redlogfilm vs gamma 3

N

Nick Levanti

Member
Joined
Mar 2, 2012
Messages
18
Reaction score
0
Points
0
Hey guys, question about maximizing the Scarlett's performance.

We recently did a short that we shot on the scarlet using red color 2 and gamma 2 on set.

When it became time to color grade, we brought the footage into rcx and switched to Redlogfilm. This resulted in a lot of noise in the blacks. Everything was shot at a 320 iso. If we plan on grading in Redlogfilm, should we shoot on it? Does it make a difference which profile you actually shoot on considering it is raw? Thanks in advance.
 
since its raw it does not matter.
 
If you are going to "expose" correctly with your profile, then yes it makes a difference because you are burning in different aperture to compensate for that ISO / profile. I would say if you are going to grade in Redlogfilm, then expose the shots under that profile - while making sure that your RAW data is not clipping.

I generally do not touch ISO, leave it to native rating of the sensor. Same with color temp, if possible.
 
Is the noise still there when you apply a grade (basic s-curve)? If there appears to be no noise in the RG2/RC2 files, than the RLF files have just a little (or just as much) noise as the RG2/RC2 stuff.

The thing about RLF is that it's super flat, so your blacks appear to be raised up fairly high compared to when viewing the same clips with RedGamma applied. If they are grainy, it's because you were underexposed on set (which is a possibility if you didn't watch your goal-posts... in this case your shadow/black goal post.)
 
What you "shoot in" is irrelevant as it's the raw data that gets recorded. RLF is designed to meet the cineon spec and hence be gradable just as log film scans are, and invertible to linear light with the standard equations etc.

Visually and mathematically the transform of raw->redgamma2or3 is identical to raw->rlf->redgamma2or3 - in other words you are not inducing noise by grading from RLF, but you're seeing the data in a way that exaggerates dark areas of the image and makes them more visible. When you apply the appropriate viewing lut or grade to rebalance the image it will look nice with a pleasing contrast and low noise.

Graeme
 
Nick - I have had the same situation on two shoots. I am used to LogC on the Alexa, but RedLog functions much definitely for me due to the different color and gamma spaces I can choose. We had viewed on set in RedLog and the histogram appeared to be exposed properly. When we got into post everything below 25% gray had noise in it when viewed.

We had to put an extreme curve just to remove noise. Until I have ample time to test RedLog, I have received advice to shoot/ view and color in in Red Color 3 and Red Gamma 3.

Just my two cents. I know there are people out there who have complete understanding of the difference between LogC and RedLog and how to properly expose. If you experts could give us a bit of knowledge that would be huge!
 
Unless you're viewing RLF through a proper view LUT you're going to get bizarre results if you try to expose from it. That's not what it's designed for.

Graeme
 
Just to be clear, REDlog and REDlogFilm are two different things (different gamma spaces)... I didn't think anyone used straight REDlog any more as RLF is considered to be much closer to the standard.
 
noise doesn't usually show up on the 5" screen like it does on the bigger ones. if you want to be safe, do a proper blackshade at running temp, use lights, expose for something between 320 and 800, check it out on a bigger screen, watch your goal posts - and if all else fails, apply some noise reduction...
 
Hey Graeme,

Is there any chance that you write all of that down e.g., in a little ebook? I'm happy to pay a fee for it, no question.

Or a one day a "RED-Gamma & RED-Log Intensive" in Irvine?

So far I have learned a lot from your posts, but I have already some kind of "fear" to miss one of your replies here ;o) hehe.

Thanks for all your input, it is highly appreciated.

Sassi
 
Graeme's little RED book eh? :-)

Graeme
 
... Graeme's little REaD book perhaps? Oh yes! :o)

(I love to dig deep, and know the stuff. I even contacted Charles Poynton, and have asked him if he can expand his book towards film -- Digital Video and HD 2nd)




Bob, I hope you're right, not little. But just to have all his developments and perspectives/ideas so far in a book would be great.
 
we were talking on another thread about starting a wiki...

graeme - would you be down to edit some of the finer points that we'd undoubtedly misspeak?
 
Wiki -- if active -- is a great tool.
It is a lot of work to keep the content on a high level. If that would be possible: very nice.

Quite frankly, pure source material from Graeme -- I would prefer that. :o)
 
RLF will give you a lot of headaches if you expose to it in camera. I would recommend looking at your image as close to a final grade as possible in-camera. Which for me means RG3. You could build a custom curve on top of RLF if you want to look at the RAW data differently on set. I like to adjust the black point slightly so that I don't get tempted to think that those noisy blacks are actually good data. If I can expose for black at 5% instead of 0% I can always safely crush the noise instead of assuming on set that it was usable based on the 5" screen.

There is another dimension and that is hue. A lot of times an image will look great to the eye but the keyer will see significantly more noise since subtle hue noise can get changed pretty aggressively in the comp. So if you're shooting a really dark greenscreen plate I would not only not use RLF I would shoot for Rec709 or some other pretty aggressively crushed gamma--just to really make sure you're lighting for the least noisy range of capture.
 
Can someone from RED clear this up please :

Blackshading does nothing to effect random temporal sensor noise ( call it grain ) , right ?
I keep seeing people talk about blackshades for noise . Other than possibly sinking the blacks to darken and make noise less visible, it is still the same noise remaining after a black shade.
Black shade is for correcting field unevenness and for fixing hot pixels. Yes ?

Graeme ?


paulherrin said:
noise doesn't usually show up on the 5" screen like it does on the bigger ones. if you want to be safe, do a proper blackshade at running temp, use lights, expose for something between 320 and 800, check it out on a bigger screen, watch your goal posts - and if all else fails, apply some noise reduction...
Click to expand...
 
You must log in or register to reply here.
Back
Top