Difference in adjusting exposure in vfb and in camera

Hi, what is the difference in adjusting the exposure from max/camera exposure controls or within the vray frame buffer? The result can be the same as an image, but do i gain or lose something with each method? thanks

Yes, there is a difference between the two and it’s pretty simple to understand. The principle is the same as in photography. If you’ve ever shot pictures with a real camera, you already know that it’s best to shoot your image with the correct exposure in the first place. This is because if you shoot your image underexposed, when you go and try to fix the exposure later on, you’ll get lots of noise. The same is true with “taking” your images in Vray and Max. Always try to properly expose the image from the camera and don’t try to do very big adjustments of exposure in VFB. You can certainly play with exposure in VFB but just to do small to medium adjustments without adverse effects. If you try to achieve a vastly different exposure through VFB you’ll either notice a lot of noise or renders that take very long to finish.

Thanks for the reply. Well ofc i always have proper exposure from within camera. Just happened we collab with a freelancer that send us some heavily corrected vfb image and it made me thinking.

Though I dont think that what u are saying is the case here. Vray doesnt work exactly as real photography eg. take a look at iso. Doesnt introduce noise. Its just another exposure tool. As u can also have completely fake dof.
Plus the opposite stands: if i am overexposed from camera, render takes a lot to finish cause of the overbright (i think importance sampling that is?). On the other hand changing my exposure in vfb makes 0 difference in rendertime.

Can u give me an example of an underexposed vraycam that u expose it proper in vfb and save it and an example of a vraycam with proper exposure with 0 vfb settings, that shows this problem ur mentioning in post? Cause i have the impression that i save what i see from vfb in 32 bit no matter how exposure is applied. Maybe a dev can enlighten us here :slight_smile:
cheers

I think you didn’t read my post carefully. I’ve never said anything about ISO and that it introduces noise. What I said is do whatever you like with the camera settings as long as the image in VFB looks roughly what you expect when the Exposure slider is on “0”.

That’s exactly what I said in my post. Read again carefully.

You can do this test yourself, don’t be lazy. :slight_smile:

What does image bit depth have to do here? If you render with wrong exposure you’ll either get a lot of noise in post-production or very long render times. Different image bit depth won’t help you with fixing this. Higher image bit depth only makes you able to more easily bring the exposure up or down. It won’t fix any rendertime mistakes such as noisy or long renders.

I never said that u said that haha.
I said vray doesnt work like real camera and that iso was an example. U can be underexposed with lower threshold and compensate for however u wanna render in vray. I m not lazy i just cant reproduce what u are saying here.

And i quote u “If you try to achieve a vastly different exposure through VFB you’ll either notice a lot of noise or renders that take very long to finish.” I read ur post carefully and unless i misunderstand something, the opposite of this stands: from max camera u get longer rendertimes, from vfb nope. Thanks for the answers anyways :slight_smile:

Plus my conclusion is:
If i save, what i see in VFB (I see lets say an X amount of stops or brightness) why does it matter how i set up this exposure?
I ofc understand that VFB exposures on top of my original exposure like post does, but if vray saves this information intact, doesnt matter. So i think a dev can only answer that with certainty.

It clearly has to do with what i am asking here.
I have 2 ways to set my exposure and make it exactly the same. Camera and VFB.
So logically if whatever i see in VFB is being saved intact as 32 bit, why does it matter in post? I still can go up and down a few stops without introducing noise.
I ve done this thats why i post on the first place.

Can you please clearly state what your question is again and what are you trying to do? It’s a bit hard to understand what your problem is. We went from the difference in exposing through cameras and VFB to talking about saving images and bit depths. What exactly are you trying to do?

Because you’ll prevent potentially long render times and noisy renders if you setup your exposure correctly in the camera before hitting the render button. If you like the results in VFB and you’re fine with the level of noise and the render time the image took, then don’t bother adjusting anything in the camera.

I don’t understand this part. Can you please rephrase your question/problem again more clearly? If you’re asking if saving in 32 bits floating point preserves the whole exposure range you see and can adjust in VFB then the answer is yes, 32 bit floating point images preserve the whole range of exposure you see and can adjust in VFB. So for example an image with exposure value of -5 in VFB should look exactly the same as the same image saved to a 32 bit floating point format and exposure value decreased to -5 in post.

Yes sorry if i confused u. Ok lets forget rendertimes. Lets agree, if we are overexposed vray takes way longer to render.
Lets also agree that under or overexposing doesnt matter in vray cause it does not work like a real camera. If u are under u got shadow noise (with 3.3 onwards not as much as in the past btw) u can raise noise thresh, if u are over u got antialising issues but fixable with tonemaping.

A guy send me an image, completely overexposed and he fixed it in VFB. I had no issue in post on whatsoever. I never do this, but made me thinking that however i setup my exposure, doesnt matter in post.

So is that true? Cause besides rendertimes, i see no other difference with setting my exposure in vfb or camera.

Yes as far as I know from my experience, besides rendertimes and noise, there’s no difference in adjusting the exposure in VFB or camera. Just make sure you don’t have “Clamp output” enabled in “Color Mapping” because then it does matter. If you have it enabled, make absolutely sure that your exposure is perfect from camera otherwise you won’t be able to make big adjustments in exposure without ruining the image later on in post. In this case you also don’t need to save to 16 bit or 32 bit because clamped images contain only 8 bits of data. Colors in clamped images have values only between 0 and 1, unclamped images contain colors that span beyond values of 0 and 1. With unclamped images you can have colors that have values of -10 or +10 in your rendered image and in post you can bring them closer to 0 with tone mapping so you can see them on your monitor (monitors cannot show pixels with values below 0 and above 1 without tone mapping).

Thanks, so how is noise related to that? That i do not get.

Well, the “difference” would be mostly in that when setting camera exposure you have more options as you are using the physical camera paradigm. You can play with shutter speed, f-stop, ISO, etc., and get your exposure that way. The VFB exposure, on the other hand, is more like basic EV stops, or a one-parameter tool, but used after the fact of course, which I suppose is the biggest difference between the two. Other than that, it seems to be pretty much the same thing.

These days I adjust to get a good camera exposure (with VFB exposure on 0) as I build up lighting in a project and then enjoy the ability to easily dial-in the perfect look in the VFB instead of a post-processing program, which of course is extremely convenient!

-Alan

Thanks Alan. I do not even touch the vfb btw i do everything in post.
That leads me to another question: clamping highlights from within VFB, reduces my exposure range in post? cheers

Understood. As a one-man shop most of the time, I try to do things as simply and quickly as possible, and the VFB has made that a much easier job of late.

That leads me to another question: clamping highlights from within VFB, reduces my exposure range in post? cheers

Hmm…seems likely but I would defer that question to Vlado and the crew here. If I was doing all post-processing exclusively in another program, I think I would most likely skip VFB adjustments altogether and bring the raw 32-bit float rendered image directly to post and adjust there.

Cheers to you,

-Alan

Thank u very much. Yes, i believe the same and that is the reason that i do not use VFB, but someone more appropriate on this matter can enlighten us :slight_smile: Collaborating with other freelancers, i ve seen all sort of other workflows and because we do have our own, i wanna know what restrictions apply in all these small details cause we heavily rely on post.

Some people send me files with heavy VFB adjustments as i already mentioned, others heavily tonemapped images and I try to respect how each individual artist operates in order not to confuse/restrict him.
But if stuff like this limits our post ability, i must be more demanding on how I want the vray settings.

I just tested “Force color clamping” on a saved 32 bit EXR and it looks that it’s purely for display purposes in VFB, it didn’t have any adverse effect when I loaded and played with the image in After Effects. But let’s hear what the Vray guys have to say too.

I would ask them to send unclamped 16 or 32 bit floating point images with the color correction disabled in the VFB along with a saved “Global color corrections file” (.vccglb) and/or a .cube LUT file depending on where you want to reapply back their color corrections. if you want to reapply them in VFB, the .vccglb file will be best option as you’ll have more freedom than using the .cube file. If you want to apply their color corrections in another software such as After Effects, Nuke etc. then the exported .cube LUT file is the thing you’ll need. You can save these preset files from the button “Globals…” above the Exposure slider. Just click it and then click “Save”. This way you can have the original image without color corrections and you can also reconstruct it with their color corrections for further adjustments on your side.

As a long-timer (old-timer!) on the project management side, I would (and do) insist that anyone working for/with me in 3D needs to have an advanced enough understanding of cameras and lighting to come reasonably close to an acceptable exposure without radical post (VFB or otherwise) adjustments. When this happens, it seems that very often folks are “fighting” the renderer instead of making it work for them.

In the past I have found that if someone was not clear on such things, I could fairly easily straighten them out with some basic instruction in workflow and lighting/camera/material set-up. It’s not all that difficult to understand when you look at it logically and carefully with the will to learn and understand.

Not everyone has a background in photography as I (and many folks here) did, and very often just going over the basics and how they relate to what we do is enough to get them to a point where they are understanding the whole thing on a much deeper, more workable level.

Also, getting into 3D back in the 80s allowed me to understand basic rendering concepts right from where they were relatively quite simple, allowing me to grow in my knowledge as the industry grew, making it relatively easy to show folks how we work this “magic” called 3D.

Best,

-Alan

Balances are delicate in my country! People with strong personalities hehe. Especially when i have to say to a guy that has twice my years in experience how to setup things. And its even harder if the guy against u has the mentality of working without supervision for many years, or even worse, consider himself a great artist :smile:
We try to leave everyone do his/her own things as long as we are ok with the result in order to avoid conflicts and further delays. Ideally when we hire someone we expect to have the damn thing delivered, but many times this is not the case. Often i find myself correcting other people’s files without even telling em. Project management is hard, so i try have as much technical knowledge as i can, so i can debate about stuff with proper arguments :slight_smile:

I hear you and feel your pain. We all have our challenges…

-Alan

Reading this I realize that there is no such thing as a guy who has “twice my years in experience” in 3D! I just knew there was some good reason for getting old!

HA! Thanks for making an old guy feel good,

-Alan