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 :)

REDgamma2 and other examples...

@ Rudi.
There are many reasons the curves (both the luma and the RGBS) are very usefull in the raw developing stage in a way t hat cannot be as good replicated down the signalpath.

RLF RC2 has improved this insanely. I still find the curves very handy for a lot of pregrade work.

Judging from the ab results we get, I see no reason to rethink that..

But that just my 2.5c
 
@ Tim.
What gives you the result you want, is the right workflow for your projects...

But it is a good plan to occationally test different signalpaths, especially after huge improvements like these.

One thing I have learned along the ride is that what was "best practice" at one time, is not neccesarily even adviceable at a later stage when the processing has changed.

Magic is magic and new sauce makes old obsolete.

It is good to have many alternative approaches to an image, depending on where you want to go with it. It's fortunately not like one size fits all, but that is also what makes the redcode so flexible for a myriad of look types...
 
Thanks for the replies Graeme and Gunleik. I'll test a bit more and see what I prefer.
Are there any must see websites/forums/books about colorgrading? I especially would like to know more about the technical aspects.
 
Both technical aspects and the aesthetics are covered pretty well in Steve Hullfish's "The Art and Technique of Digital Color Grading". And then there's the new book by Alexis van Hurkman, which Amazon hasn't delivered here in Germany yet, but judging from his articles (Edit Well or his blog) should be very helpful.
 
I've spoke to Alexis about his new book and it sounds like it should be a good, in-depth read. Amazon hasn't delivered to me yet either, but it should be soon.

Graeme
 
Brook Willard said:
I can't wait until the "RAW view" mode on the RED ONE [and Epic, naturally] let me view REDlogFilm instead. For so many DPs, gaffers and DITs who think in log space from years of dealing with other cameras, being able to check a lovely low-contrast log mode would be fantastic.
Click to expand...

YES!
 
Graeme Nattress said:
Tim, it's about grading preference - how do you like your image to begin? Nothing done to it, a log curve, a more contrasty curve, somewhere inbetween? There's a lot of personal preference involved.
Click to expand...

For a colorist, it's not so much personal preference as it is targeted delivery format as well as viewing environment. For instance, when working in a DI theater, you're usually set up to either a film target or a digital cinema target. If you're targeting film, you are going to be grading through a film print preview LUT, in which case you want the input image to be something relatively close to a gamma of .6 and a Cineon log curve. That is largely what RedlogFilm provides (and also what Arri's LogC provides, although Arri also makes available film targeted color matrices that Red to this point does not). If you're targeting digital cinema, you're probably not using a viewing LUT, and your projection environment is set up to DCI specs, which means a gamma of 2.6 and a white point around D63. Redgamma2 should be pretty good for this, but one might want to use Redlog (or just start with linear light) and create their own gamma curve. If you're targeting HD video, you're going into a Rec709 display world, where the gamma is a "tweaked" 2.4 or so, and the color space is more restricted. For that, Redgamma2 is rather ideal, largely because it largely corrects everything that was not quite right with the original Redgamma (compressed lower midrange, "pushed" whites, harsh faces, etc.) and saves me the time and trouble of manually creating proper contrast from a lower contrast image that doesn't make as much use of the available levels. So for my television work, my first stop would likely now be Redgamma2 (get this into the SDK quickly, please....) instead of the old Redgamma or Rec709 curves, and for DI it would be either Redgamma2 or Redlog for a digital cinema target, but RedlogFilm for a film target. This is, of course, based on the assumption that what's being created is a final delivery element. For archival elements, the log based curves might be more appropriate because in general, they hold a bit more of the information at the extreme ends of the scale.

You might consider this all personal preference, but I consider it proper choices given the particular project's conditions.
 
I can see how RedGamma2 is pretty epic. It reveals a ton of image info. However, I am not trained in color correction and I am having to learn to deal with taking a much flatter image and making it look cinematic. Is there any kind of class or workshop or something I can attend in LA to learn to color correct from RedGamma2?
 
M Most said:
If you're targeting HD video, you're going into a Rec709 display world, where the gamma is a "tweaked" 2.4 or so, and the color space is more restricted. For that, Redgamma2 is rather ideal...
Click to expand...
So there is no additional gain if you use REDLog or REDlogFilm here? Only for DI and digital cinema/film? It's not very likely that I'll be grading for cinema delivery.

Tom Lowe said:
I can see how RedGamma2 is pretty epic. It reveals a ton of image info. However, I am not trained in color correction and I am having to learn to deal with taking a much flatter image and making it look cinematic. Is there any kind of class or workshop or something I can attend in LA to learn to color correct from RedGamma2?
Click to expand...
I would like to know this as well, preferably something online because I don't live in the USA (maybe somebody here on REDuser with the proper knowledge can make a tutorial or something).
I've started reading Alexis's book, it's a great read so far, really useful stuff explaining how and why things are the way they are.
 
TimWitte said:
So there is no additional gain if you use REDLog or REDlogFilm here? Only for DI and digital cinema/film? It's not very likely that I'll be grading for cinema delivery.
Click to expand...

I am not in disagreement with M Most -:)

I would like to suggest, though - that you made some different samples for diferent lighting and exposure cases in your footage and made Redgamma2/RedlogFilm/RedLog developments of those and tested that through your signal chain and software to see what gives you the result you want and what is the best for you to treat and what gives you the most flexibility.
 
Tom Lowe said:
I can see how RedGamma2 is pretty epic. It reveals a ton of image info. However, I am not trained in color correction and I am having to learn to deal with taking a much flatter image and making it look cinematic. Is there any kind of class or workshop or something I can attend in LA to learn to color correct from RedGamma2?
Click to expand...

Color grading is color grading. It doesn't matter if your starting point is a low contrast Redlog image or a more representative RedGamma2 image. The art of grading is to take an image, enhance it, bring out the original photographic intent but with more consideration for story and continuity, and in some cases mold it into what is desired. There is no one way to do this and there is no Magic Bullet (sorry for the trademark infringement, Stu.....'-)) ) that will let this happen with the same approach every time.

Having said that, I'm still a firm believer that the only real way to learn the art of color grading is to spend time sitting in sessions with an experienced colorist, watching them work and asking questions along the way. The system they're using is completely irrelevant. The art of what they're doing, and the approach they take, is very relevant.
 
TimWitte said:
So there is no additional gain if you use REDLog or REDlogFilm here? Only for DI and digital cinema/film? It's not very likely that I'll be grading for cinema delivery.
Click to expand...

The falloff at the extreme ends of the scale is a bit more gentle (film-like, if you will...) with the log curves, but for most cinematography, with relatively sensible exposures, the only thing those curves give you is a flatter starting point. One of the first things you do with a flat image as a colorist is add contrast and reduce the midrange values, in effect, creating a new starting point that is usually very close to what RedGamma2 is giving you to begin with. The only real usefulness of the flatter image is for either visual effects or archival purposes, where the intended final image values are not always desired for technical reasons. But for a colorist, there is little advantage to be gained, and in a "real" grading situation, there is more time lost finding the starting point that a curve like Redgamma2 already gives you. Efficiency is pretty important when you're doing "real" television work. Now, having said that, I would add that I do change the curve, occasionally on a shot by shot basis, when I'm grading a television episode, primarily when it's required to recover hot highlights, or for "kinder" treatment of faces (one of the big advantages of using a grading system that can directly access R3d files - in my case, Baselight). So take what I've said as a general set of guidelines, but not hard and fast rules, because all of this is inevitably dependent on the specific material being graded.

I've started reading Alexis's book, it's a great read so far, really useful stuff explaining how and why things are the way they are.
Click to expand...

I have no connection with nor have I read any part of Alexis' book. I did participate in some interviews and sample sessions with Steve Hullfish and can heartily recommend his book as a good introduction to the subject of color grading.
 
Mike you are a stellar member here, and I have learned a lot from reading your posts.
 
You must log in or register to reply here.
Back
Top