I was browsing the gallery and came upon an old post by Spike Spigel ( think his name is Sergey ) [https://forum.corona-renderer.com/Small house in Finland ](https://forum.corona-renderer.com/Small house in Finland )
I was just blown away by the realism. He states its nothing more that Corona sun and sky and I read in the comments that he uses a color correction on his corona sky. I tried my best to recreate this but always end in a dirty blue sky. I have also contacted him but I suppose he is not so active on the forums anymore. Can anybody please assist me in explaining to me how I can recreate this rich blue sky
I think that the green colours of the grasses and especially the trees and the dark facades of the house gives the sky that nice blueish vibe. If you make testrenders with a white scenery than it’s just what it is.
My suggestion would be to add more vibrance to your scene and not worry that much about the sky intensity first.
You will never be able to get rich deep blue sky unless your enviroment is underexposed a little bit. Instead of trying to colour correct enviroment and potentially messing with scene lighting, i’d suggest you to use direct override or even replace sky in post processing stage.
You’re big on having the UI as tidy as possible. If correcting a background to be more appealing is nonsense then in what scenario do you see Direct visibility override as valuable? I’m genuinely interested because back in the day I remember comping skies to Vray renders where, IIRC, I used a black environment color in the render so the tree etc edge pixels would be correctly fixed and ready for comping, otherwise there would be artifacts. I’ve forgotten the right technical wording. Is this still something to consider in Corona?
Don´t know what the model assumes, but when I compare photos (clear sky, unmodded) with standard corona sky renders, it´s allways too bright compared to the rest and thus too washed out, even when the effect depends drastically from day/time.
But maybe the rest of my scenes is just too dark Cannot substantiate this any further, sorry. And since 1.5 I additionally have the feeling that the standard exposure for a cSun+cSky setup went from -3 to -3.5 for some reason.
I would setup many scenes by lighting with cSky + cSun but have HDRI in visibility override and reflection override. But as long as (Renderer for architectural visualization - Corona | Chaos) it´s just not possible to do this reasonable.
Curious at which sun angles you get away with EV -3 -3.5? I have toolbar button that creates sun at ~55 angle, adds sky to enviroment and sets EV to -4. If i set EV to -3, then for clear unobstructed outdoor shots, my sun angle should go to 70-75 degrees, which is basically sunset-sunrise already.
And i didn’t notice change in sky brigthness from 1.4 to 1.5. Might it be due to changes in materials (new PBR)?
I can´t speak in angles because the first customer question after “when”, “how much”, and “do I get 300dpi” is usually “Is the sun lighting correct?”. So our default maxstart scene comes with a daylight system and I never cared about it. But -3,5 covers a wide range, obviously you have a low highlight compression value. I use -3,3,3 (EV/HC/Contrast) to start.
Well I agree with romullus. In a summer outdoor scene my EV will typically be around -4,2 with he and contrast both set to 1. And with these settings the sky doesn’t look a bit overexposed.
I´t the contrast setting which acts differently (forgot to check further before installing 1.5.2 on Max2014, reversed to 1.4 again now again). So it´s not noticable when using C1.
Sorry to dig up a 2 and a half year old thread, but this is something I’m currently looking at at the minute. Now that PBR materials/textures are much more commonplace than they were back in 2016 when this thread was alive. I’m curious as to whether you’re still finding -4.2 as a good EV value for exterior summer shots in bright sunlight?
Cannot substantiate this any further, sorry. And since 1.5 I additionally have the feeling that the standard exposure for a cSun+cSky setup went from -3 to -3.5 for some reason.