The new programmable, stackable and playable tonemapping is here with Corona Renderer 8 RC1!
Link: Current - Google Drive
Apply the operators, tweak the settings, move them up and down to change the order (applied from top to bottom), try the new presets and also create and share yours!
Scene from Evermotion
Hi all, I’ve downloaded the last Daily Build RC4 and I’ve noticed the changes on the ToneMapper.
The new ACES OT seems great!!! It seems that there is no need to use the Advanced Filmic Mapping / Filmic processes anymore or am I wrong?
What should be the process and order for the modifiers now? LUT? ACES OT? Ecc.
Thanks in advance and you did a great job guys!
Dionysios -
We’re glad that you are enjoying the new features!
As to your questions:
Advanced Filmic Mapping / Filmic processes / other operators - I don’t think there is ever “need” to use any of them. 
They are all different, and you can use them if you wish, or if you find them useful in getting the desired look in your images.
Order of operators - here we are also giving the users complete freedom. One thing to remember is that the operators are applied in the order from top to bottom. I would personally strongly encourage experimenting, because even if something is not “technically correct”, it can still produce good-looking results. It’s similar as with using the Corona Physical Material with the Complex IOR option vs using simple controls with just the base color and edge color. Or using the Corona Physical Material vs using the Legacy Material to produce some un-physical results.
I am sure there will be some users with more puristic approach, or who are used to following strict workflows to produce technically correct results in terms of color management, etc. This is fine too, and as far as I know, in such cases it makes sense to only use the exposure operator, and then apply the ACES OT operator.
By the way, this thread is awfully empty! Here are some more examples of the new ACES OT operator in action:
No operator at all vs ACES OT - the results speak for themselves 
- no operators vs ACES | Corona Renderer
ACES OT vs Reinhard Highlight Compression - while these two operators have some things in common, note especially how the highly saturated materials on the spheres react to strong highlights - ACES vs Reinhard Highlight Compression | Corona Renderer
ACES OT vs no operator at all - again, check out the highlights compression and desaturation - ACES vs no operators | Corona Renderer
An experiment: the same scene, using the ACES OT operator in both cases - left is using the ACEScg color space, right is using the Wide RGB color space (the default in Corona) - as you can see the differences are negligible. The mostly visible difference here is in the green refractive sphere and in the overall saturation of the SSS object. There are also some minor differences because of the UHD Cache recalculation - ACES internal color space VS the default Wide RGB internal color space | Corona Renderer
ACES OT vs no operator at all - ACES vs no operator | Corona Renderer
ACES OT vs Reinhard highlight compression - ACES vs Reinhard highlight compression | Corona Renderer
Both images are using the ACES OT operator. One is rendered with the Wide RGB color space, the other one in ACEScg color space - take a look at the blue car - WIDE RGB color space with ACES operator vs ACES color space with ACES operator | Corona Renderer
ACES OT vs no operator at all - ACES vs no operators | Corona Renderer
Both images are using the ACES OT operator. One is rendered with the Wide RGB color space, the other one in ACEScg color space - take a look at the blue car - ACES color space with ACES operator vs WIDE RGB color space with ACES operator | Corona Renderer
ACES OT vs Reinhard highlight compression - ACES vs Reinhard Highlight Compression | Corona Renderer
ACES OT vs no operator at all - no operators vs ACES | Corona Renderer
*By the way, the internal color space can be changed in the Development / Experimental Stuff rollout, and you need to re-render to make it work - https://support.corona-renderer.com/hc/en-us/articles/4402367677713
*Note that both the ACES OT operator and switching between different internal color spaces will make the biggest differences in cases where some kind of color blending occurs (especially in case of highly saturated colors) such as multiplication or division. This happens, for example, when: mixing lights (two spotlights with different colors mix), using refractive materials with colored absorption, using SSS, using the Corona Physical Mtl with colored Clearcoat.
Sorry for the spam!
While it currently seems that ACES OT is the “make it look nice” tool, I assume some people will still want to create their own custom tone mapping stacks to differentiate their images from everyone else using the default setting
Why is aces almost always so contrasty? I didn’t have a chance to test Corona’s implementation yet, but judging from examples here and also what i see in other apps, it’s always the same crushed shadows and overly compressed highlights. Maybe it’s my uncalibrated monitor, i don’t know, but usually when i look at nice images in the gallery, or artwork in general, i don’t see much issues with colours or contrast, so i guess my monitor is not very far off and it’s something with aces itself that i don’t like it. I guess i’m just not a fan of that overly punchy look.
I think this is very, very scene-dependent. My examples are highly contrasty even with no ACES OT operator. When it’s added, they become even more contrasty. In a different scene (maybe exterior? something with soft, diffuse light?) it might look very different.
I understand that, but if it’s one button solution, then it means that in 50% of cases it might be unusable.
Some more experiments. Light with negative color values in different color spaces.
Bonus:
ACES OT vs no operators - ACES OT vs no operators | Corona Renderer
ACES OT Wide RGB vs ACES OT ACEScg - hair color changes - ACES OT Wide vs ACES OT ACEScg | Corona Renderer
Unusable? I think that’s a bit too harsh. If you don’t like the result of applying the ACES OT operator, you can lower the contrast using other operators, like contrast, or Reinhard highlight compression (before or after the ACES OT operator).
Sorry, i probably should at least try it before criticising. Didn’t want to sound too negative.
I would say that if the the rendering does not look at least “quite good” (as far as tone mapping is concerned obviously) after aces display output transform you can safely assume something is wrong with your scene (if “realistic” is what you are after). they nailed the look really well.
ps: sort your display - LCD monitor test images
Thanks for your explanation Maru! As for the Filmic operators I understood immediately the issue but I needed a confirmation from you guys as well.
I love the new ACES OT, I personally find the results natural, highlights and contrast while I play with the exposure the behave like a real Photo Camera in my opinion.
Sometimes but not always I see the darks too darks but with a reflex the effect is the same at the end. I find using the Fuji Luts very helpful with ACES OT. I love the results, really!
As for the ACEScg color space, I see in your toy car example a big difference on the blue cars, we do suppose to use ACEScg color space instead of Wide RGB or is just a matter of final color taste at the end.
I personally do all my final correction process in Lightroom which it uses Wide RGB as well.
Great job, I am extremely happy!
Thanks!
I am playing tonight with the new Tone Mapper and I’ve noticed that once applied the Curve Data modifier and then open the Curve Editor CIE is blocked while the Curve Editor is all white and I see no curves.
Thanks,
Dionysios -
Hello guys,
I haven’t had the chance to test out the new tonemapping but am curious about knowing how it stacks up to the Fstorm tonemapping in terms of Realism and Colours control.
P.s
There used to be a Tonemapping forum were the comparison between the Fstorm, Vray and Corona Tonemapping output is it still active or has the new corona tonemapping been used to confirm the improvement in regards to all the Complaints about the previous implementation?
Just a fast note:
The block CIE issue is only present in CIE standalone, in 3ds Max everything is ok.
I can’t understand why in standalone mode the CIE is soooo laggy and slow.
Thanks -
Hi,
What resolution is the file you are working on in the CIE?
Rowan
Hi Rowan,
the res is 6.100px large.
Thanks!
Hi,
We have some optimisations for the CIE in the pipeline. It should be available in the next build.
I’ll let you know when it gets released.
Rowan
