Welcome to our community

Be a part of something great, join today!

  • 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 :)

Dragon Color science emulate film?

Matthew Glover

Active member
Joined
Feb 5, 2013
Messages
33
Reaction score
0
Points
0
How would you guys feel about this? I think it would be bad ass idea. Of course whatever RED comes up with will be good none the less and the film look can always be made by grading. But how would you guys feel if they went this route?
 
A MX -> DRAGON LUT would be definitely be something interesting to look at, if built properly/controlled environment so it could be of actual use and not just a novelty.
 
What Phil said.
 
How about we start with a basic Log->709 LUT first?

Inexplicably, with all the testing and color science happening at HQ, Red doesn't create LUTs. We're just told we can use Cineon to 709 and move along.

If there were an endorsed RedLogFilm->Rec 709 instead of 'close enough,' there would probably be more than a few happy post folks. I feel dirty every time I put an Arri LUT on R3Ds. May as well put a Canon lens cap on my Scarlet because same difference, right?
 
How would you guys feel about this? I think it would be bad ass idea. Of course whatever RED comes up with will be good none the less and the film look can always be made by grading. But how would you guys feel if they went this route?

There are different films and developing methods to emulate.
Choosing one or the middle ground as a starting point would limit the performance of the sensor.
 
If there were an endorsed RedLogFilm->Rec 709 instead of 'close enough,' there would probably be more than a few happy post folks. I feel dirty every time I put an Arri LUT on R3Ds. May as well put a Canon lens cap on my Scarlet because same difference, right?
It doesn't matter as long as the end result looks good on screen. This is a case where the end justifies the means: it doesn't so much matter how you get there as long as you get there.

There are a million different LUTs, and as Mr. Simic says above, it merely gives you a different starting point for color. It's not a be-all end-all. A good colorist can help you make the right decisions with the fewest possible delays and the fewest compromises in terms of picture quality. No LUT can replace a good colorist.
 
If you try Resolve 10, it now has drop down 3D Film LUTs, both Fuji and Kodak, that work exceptionally well with Redlogfilm.
 
A LUT is what you're looking for but no one LUT will simply make your RED look like its shooting film. There a numerous types of film and even more LUT's to choose from - you can even design your own. The film "look" starts on set by knowing how film will respond to light and preparing the scene so the EPIC captures the scene in the same way.

While post could benefit greatly from RED releasing a sanctioned (and thus standardized) set of LUTs, you may find they won't even fit your definition of the film "look".
 
I look at it this way.

RED's color science should do everything to achieve the very best color rendition across a broad spectrum, be low noise and high Dynamic range, and that is all it has to accomplish.

Saying it should be like any type of film assumes that film has already accomplished this (and it has not).

So to me RED Color Science should "scientifically" reproduce and render the best color most faithful color and then let the artists make the creative choices once they have tested how the sensor reacts to light.

They should never strive to look like anything other than RED.

Battistella



The
 
so, what are the LUTs you like the most today? make your top 3
No LUT can take the place of a good colorist. And a good colorist can't help material that isn't shot well by a good DP who knows what he or she is doing. LUTs are not the panacea you think they are.
 
No LUT can take the place of a good colorist. And a good colorist can't help material that isn't shot well by a good DP who knows what he or she is doing. LUTs are not the panacea you think they are.

ok, assuming LUTs can't take place of a good colorist, can't help bad shot material and are not the panacea I think they are...

what are the LUTs you like the most today?
 
I caution folks on LUT's. They are an excellent way to simulate Film stocks.

But the garbage in > garbage out rule does apply

Just a small word of caution that unless you know what you are doing, LUT's are not going to magically solve everything.

Battistella
 
valeu, paulo :)

and about LUTs, sure they have to be manipulated carefully. IMO, we're a little experienced about presets (let's put LUT as a 'preset' in a broad sense) since we all, at least once, tested after effects and final cut pro preset looks (magic bullet, for example) and found it's not magic.

thing is, LUTs, at least for me, not a colorist at all - though sometimes I have to play as one if a project can't afford a pro - is a good starting point for color grading.

In these shots (scarlet) http://farm4.staticflickr.com/3820/12054291684_8734c1d2a8_c.jpg, I used Tom Majerski's LUT http://www.tommajerski.com/. At first, it was a total mess. Than I fixed things like hue, saturation, contrast bla bla and it became better than a no-LUT color grading would be.

I'm not a pro, everything takes sooo much time when I'm on resolve and results are not consistent, but I also know how to recognize when things are going too bad and when it's time to stop, go back or reset everything. Or even call a pro. So, imo, the problem is not the LUT, it's the operator most of the time.
 
Last edited:
How about we start with a basic Log->709 LUT first?Inexplicably, with all the testing and color science happening at HQ, Red doesn't create LUTs. We're just told we can use Cineon to 709 and move along.
I feel like I'm beating a dead horse, but then again this horse seems to be a Zombie so I'll do it again:Cineon Log is a standardized gamma curve. Rec 709 is also a standardized curve. RED doesn't need to create a custom LUT because it's not a Custom colorspace. Anytime you see a "Cineon to 709" curve it's correct. You shouldn't be using Canon or Arri since they're using non-standard gamma profiles for log. Asking for a Cineon to Rec709 is like asking for a linear to sRGB, the reason they don't make it is because it's not proprietary.

http://www.mediafire.com/download/lzl53zzk93fcecm/LogTo709.zip
 
So to me RED Color Science should "scientifically" reproduce and render the best color most faithful color and then let the artists make the creative choices once they have tested how the sensor reacts to light.

They should never strive to look like anything other than RED.

To me and I think a lot of artists Kodak may not have "Scientifically" reproduced reality the best but most of what we shoot shouldn't look like reality it should look better than reality. I still think RED should buy the rights to Vision 3 for Dragon and offer REDColor4 and Vision #(tm). People choose Alexa in a large part for its color, imagine how many DPs they could bring over if they perfectly emulated the color of Kodak stock. "Here, this is called the Dragon, just set it to "Vision" and you'll get the exact same image as if you shot on film."

I don't know what you mean by RED should strive to look like RED. They already have 6 different color spaces. What's adding a 7th that's highly sought after? If Dragon's color gammut and IR response is similar to film it should (like FilmConvert) be relatively easy.
 
To me it would be great if RCXP could import 3dLUT files. Then I can make a grade in any program and just export it and have it with me on shoots and such and get the transcodes done with it applied on top of whats goes on in REDCXP. Basically a final log2lin user prefered lut that lets me use redlogfilm or any other setting underneath the lut. Also ofcourse Red cine should be able to create 3Dluts.
 
Back
Top