Tip for Making the Most out of RED's Dynamic Range Outdoors

[joshua csehak]

I think we may be misunderstanding each other. What you quoted, and what I stated was a simplistic explanation of why a trick like this is helpful. The lift / gamma / gain controls are great, and very helpful in grading, but they cannot do everything- which is why secondaries are needed. The lift / gamma / gain controls while mostly effecting their section, also bleed into other areas. If you push the lift control enough you can effect the gain. Pulling a key like Phil has suggested here helps to selectively isolate an area without having to track anything- which is the idea / concept that I had not thought of before, so I appreciate the heads up- another tool to add to my tool box.

I think I understand you. Maybe I'm not explaining myself well. I guess what I'm trying to say is, you can pull a key like that in Color, and tweak it based on value, saturation and whatnot, and then make further adjustments within that key with lift, gain, etc. And you have just as much flexibility in tweaking the mask in Color as you do in AE, unless I'm missing something. So I don't understand the benefit to doing it in AE.

 

[joshua csehak]

(BTW, holy cow, gorgeous images!)

 

[Ryan E. Walters]

I think I understand you. Maybe I'm not explaining myself well. I guess what I'm trying to say is, you can pull a key like that in Color, and tweak it based on value, saturation and whatnot, and then make further adjustments within that key with lift, gain, etc. And you have just as much flexibility in tweaking the mask in Color as you do in AE, unless I'm missing something. So I don't understand the benefit to doing it in AE.

Ah, now I'm following you. Yes, you are correct- AE is not the only program that does this, and any grading app worths it's salt can do that. (Which I am thankful for as Color is the only app I have daily access to.) For me what is new about this process is thinking about pulling a luminance key- that is the new piece of info that I'll be working in to my skill set.

 

[Radoslav Karapetkov]

Very useful techique. Thanks. :thumbsup:

 

[Kelley Muro]

Makes sense, real interesting. thanks for sharing.

 

[joshua csehak]

Ah, now I'm following you. Yes, you are correct- AE is not the only program that does this, and any grading app worths it's salt can do that. (Which I am thankful for as Color is the only app I have daily access to.) For me what is new about this process is thinking about pulling a luminance key- that is the new piece of info that I'll be working in to my skill set.

Cool. Good to be on the same page. There is one situation where this technique would be useful, though, and that's if for some reason you weren't able to grade the raw file directly. Then you could make two different versions with different exposures.

That, and of course, if you had precise motion control and a static subject and two takes that were precisely the same, but how often does that happen?

That said, the OP is still very helpful. It's always good to talk about this stuff and explore different techniques. Much more productive than blabbering on about, say, Lost.

(I'm still mad at Lost, BTW. Stupid jerkfaces.)

 

[Phil Bates]

I'll take this tip one step further:

In the original footage exposed for the foreground, you can see some CA (chromatic aberration) on the edge where the foreground meets the sky (the lens was the RED 18-85). Here is a section enlarged to 200%:

As you can see, there is a highly saturated cyan artifact running along the edge. It tends to look worse, or more "out of place" when the sky is fixed:

I eliminated the CA by using a mask and de-saturating the cyan.

First, I grabbed the mask I used earlier:

Then used the Find Edges effect:

To turn this into a usable mask, you need to invert the image and widen the edge. I use a Blur and Levels to accomplish this:

The result is:

I used this mask as a Luma Matte and de-saturated the cyan color of the image placed inside the mask. Here is the Hue Sat dialog:

The resulting image looks like this:

Here is the fix side-by-side with the original:

Like the previous exercise, this tip works on moving images. However, you may need to use garbage mattes to control the effect.

I expect there are many out there who know this trick or others like it. There may even be software that can accomplish this quicker. This is how I do it in AE because I haven't found anything better in that app.

Phil

 

[Tom Lowe]

Phil, is it possible to bring R3D files through Adobe Camera RAW in AE? I really like their post RAW Grad ND, for example.

 

[Terry VerHaar]

Very impressive results! Thanks for the lesson. :thumbsup:

 

[Phil Bates]

Phil, is it possible to bring R3D files through Adobe Camera RAW in AE? I really like their post RAW Grad ND, for example.

Tom, my understanding is that Adobe Camera Raw will not read R3D files. Although we haven't tried it lately.

Phil

 

[Marc Wielage]

Yes, good point, Phil. You do have to blur the edges of the matte in order for this to work. We have the same exact problem with high-end post color-correction tools. If anything, there are aspects of After Effects that are simpler in a lot of ways. But the pro tools are considerably faster.

Blown-out skies and windows are among my pet peeves, and I always try to talk clients out of them unless they're going for a really intense, over-the-top look. I'd always rather see detail in those hotspot areas, and to me, it's distracting when they totally sizzle out, especially if I know there's detail there.

 

[joshua csehak]

Now that's something you can't do in Color! Using Find Edges to kill CA -- pretty freakin clever

 

[Kevin Marshall]

Pretty sure you could manage it in the Color FX room - use edge detection and create a matte from it for Alpha Blend.

Not quite as simple as this AE job, though.

 

[Kim Frank]

I know this technik for desaturating CA but it's nothing more than just desaturating.

Wikipedia:
In optics, chromatic aberration (also called achromatism or chromatic distortion) is a type of distortion in which there is a failure of a lens to focus all colors to the same convergence point. It occurs because lenses have a different refractive index for different wavelengths of light (the dispersion of the lens). The refractive index decreases with increasing wavelength.

Post-processing to remove chromatic aberration usually involves scaling the fringed color channel, or subtracting some of a scaled version of the fringed channel.

http://en.wikipedia.org/wiki/Chromatic_aberration


So the only way to really fix Chromatic Aberration would be to refocus/rescale the RGB Channel:

Fix Red/Cyan Fringe: Adjust the size of the red channel relative to the green channel. This compensates for red/cyan color fringing.

Fix Blue/Yellow Fringe: Adjust the size of the blue channel relative to the green channel. This compensates for blue/yellow color fringing.

I never used this technik, because I try to avoid CA while shooting and it wasn't recognizeable for simple PAL finishing (havn't worked for the big screen yet), but wanted to put this into the discussion.

Once there was a, I think, After Effects Toutorial on this Technik (not the creativecow one! that's also about desaturation) but I can't happen to find it anymore.

 

[joshua csehak]

Pretty sure you could manage it in the Color FX room - use edge detection and create a matte from it for Alpha Blend.

Not quite as simple as this AE job, though.

Oh, I hadn't thought of that! Yeah, probably could. And probably simpler, given you're in Color anyway, and taking it into AE is a whole extra step in the process.

 

[Tim Duran]

Phil, I never stop being amazed at how some Redusers like yourself go to so much trouble to help all of us.

What I don't understand is why a haze or UV filter would not have resolved your washed out sky at minimal effort. What am I missing?

 

[Phil Bates]

Phil, I never stop being amazed at how some Redusers like yourself go to so much trouble to help all of us.

What I don't understand is why a haze or UV filter would not have resolved your washed out sky at minimal effort. What am I missing?

Actually we were using a polarizer, which at the right angle should darken the sky significantly. If the angle wasn't optimal, I am not sure if a UV filter might have been better. Perhaps some competent DP types could answer that question.

Regardless, the RED image has enough latitude to adjust the sky perfectly. The main issue is that the sky and the foreground luminance ranges are so different, you can't make them both look good at the same time (except with a trick like the one I showed). So, when you try and optimize the image for the foreground the sky blows out. If you adjust for the sky the foreground gets too dark. It's like you "can't have your cake and eat it too" kind of a thing. I don't think there is a filter that can compensate for that much of a difference, but I wish there was!

Hope that helps.

Phil

 

[Cail Young]

I don't think there is a filter that can compensate for that much of a difference, but I wish there was!

Graduated NDs would, but it depends on your shot, they're pretty useless if you're reframing.

 

[Nick Burridge]

Graduated NDs would, but it depends on your shot, they're pretty useless if you're reframing.

I think he's talking about digital filters? Specifically ones that can compress the detail/dynamic range captured by the R1 into something 'flatter' and more visible to the human eye. Along that line, aren't Curves a simple gamma adjustment of specific luma levels? Adjusting points along the top end of the curve just pulls down the sky, etc? Not quite the complex 'pick the cleanest channel' approach here - they probably give quite different results - but surely would be fairly close? Since what you're really doing is compressing the high end of values down closer to the low end.

I suppose you wouldn't get the benefit of FLUT as you would exporting multiple passes here, and there's probably other things I haven't thought of.

Also, if you were talking about optical filters, I apologise. :blushing: If so, yeah, grads are one way of dealing with it.

 

[Dan Hudgins]

CA comp

CA comp

I never used this technik, because I try to avoid CA while shooting and it wasn't recognizeable for simple PAL finishing (havn't worked for the big screen yet), but wanted to put this into the discussion.

CA comp will be a bigger issue for Scarlet using C mount lenses.

I have CA comp in the de-Bayer program I am working on, and it helps since C mount lenses can have quite a bit of CA.

But it is not linear over the frame, so just using one scale value each for the red, green, and blue can only align the images in a circle, not all over the frame, to do that you need to distort a grid of values and interpolate the right three scale values for each pixel, something I have not done yet.

Another problem is that small lenses like C mount, are not centered right, so you get issues with the CA being asymmetric, and its not always easy to figure out how much it is off center one way or the other.