This might have been covered elsewhere in the past (RDNA, SM Poser forums) but I can't remember reading about it, so if it's already been explained/explored then forgive me (and please post the answer!).

Here's a very basic scene:

This is a cube primitive set to white, a plane/floor set to RGB 127,127,127, and a sphere primitive whose material is simply a Superfly Emission node set to white and Strength 500. (I know that's quite strong but it's necessary to show the issue.) Those three elements will remain the same throughout.

Now I add bagginsbill's envsphere prop.

In the Material Room for the envsphere I've simply added a Superfly Emission node to a CyclesSurface, attached the image map (in this case, "shady_patch_4k.hdr" downloaded from HDRI Haven), and set the Emission Strength to 1.

The issue might already be visible. It is, of course, how much dimmer the light from the sphere primitive has become. It might be argued that the light from the HDRI is flooding the scene and making the sphere primitive's light appear dimmer. Which is a fair argument, I suppose.

To test this, I tried lowering the Strength of the Emission node on the envsphere. I lowered it to 0.1.

At one tenth of the original light, I'm still surprised that the primitive sphere's light is (potentially) being overwhelmed so I try reducing the envsphere even further. This time to 0.01

Okay, I'm now starting to think this is getting silly, so I try one more test. This time, I've set the Emission Strength to 0.00001 meaning there should be almost no light from the HDRI in the scene except for any extremes.

At this point, the light from the sphere primitive ought to be illuminating the scene in the same way as the original image, with only the tiniest contribution from the HDRI. But clearly it's not. For some reason, when the HDRI is contributing even the smallest amount of light, the other emissive material is barely doing anything. (Notice the grain in the final image compared to the first image; a sure symptom of low light conditions.)

If I reduce the Emission node's Strength to 0 on the envsphere, the result is an image indistinguishable from the very first image (i.e., the primitive sphere's light is back to its original strength), so it's not merely the presence of the envsphere that's causing this issue (it's still there just not contributing any light).

I've tried this twice now just in case I'd done something to produce this effect the first time around. The second set of renders was done from a fresh restart of Poser.

Why does this happen? Is it only me (I could easily be doing something wrong here and totally not seeing it!)? Is it something particular to bagginsbill's envsphere? (My first guess is "No, why would it be?" but I'm open to persuasion.) Or if it's a Superfly/Cycles thing then why is it happening? (And, more importantly, how can it be fixed?)

Interesting. I've never considered this before!

Another example, but more extreme.

This is the same scene, but I've enclosed the two primitives (the cube and the sphere) with pure white walls and a ceiling surrounding them. (I moved the camera and changed focal length, too, as part of my playing about with the scene but I doubt that this is a factor here.) I set the envsphere's Strength to 0 and rendered.

The image is totally blown out by the primitive sphere's 500 Emission Strength. Just as expected.

Then I set the envsphere's Emission Strength to 1 and rendered again, leaving everything else the same.

So, the envsphere doesn't have to be contributing to the rendered scene; it just has to be emitting light, no matter if that light can be seen by the camera or not.

change the scale of the envrionment sphere to the size of the other emmiter sphere and look what happens, from what I have experienced the bigger the emmiter object, the less the emmision strength has to be for a stronger light.

bwldrd: I know what you mean about the larger the emitter then the more light it will produce (even with the same Emitter Strength).

But it's not the light of the environment emitter I'm really questioning here; it's the light from the smaller emitter (the one you can see in the original 5 images) that I'm wondering about.

Look at the first image. There's no environment sphere in the scene. The little emitter is emitting lots of light, as we'd expect.

Look at the fifth image. The environment sphere is in the scene now but is very dark. Compare how much light is in the first image compared to the last image.

Where has that little emitter's light gone? It's set to 500 Emission Strength in the Material Room in both images.

There are two different issues going on in your scenes.

1. For the ’darkening’ effect it should help to increase the clamp settings in the render prefs or disabling clamp by setting it to zero.

2. The additional noise as seen in your last pic has to do with MIS (multiple important sampling). In Poser MIS is on by default for all light emitters. As the env sphere covers the whole area it does ’kill’ the MIS effect for the small ball. Unfortunately MIS can not be turned off individually in Poser Superfly. Only work around is to not use an env sphere and using the background shader instead (thats the best choice for env lightning anyway). There is a thread on how to set it up somewhere here at Rendo.

Nagra_00_ posted at 4:00PM Thu, 20 February 2020 - #4381155

1. For the ’darkening’ effect it should help to increase the clamp settings in the render prefs or disabling clamp by setting it to zero.

I set the direct and indirect clamp values to 0 and...

This is exactly what I had originally expected (and wanted) to happen, i.e. the small emitter's light is unchanged. THANK YOU! (The noise in the image is something else; it's the light levels I was more concerned about.)

Btw, those clamp settings are 10.0 as default. What is that actually doing? What thing or process is having the value 10 applied to it?

The default light sampling clamp (set to 10) limits the number of lighting samples. Setting them to zero does not mean zero lighting samples, rather it means the lighting samples are unclamped.

When using mesh lighting, I generally set mesh light samples to 20 or so, and unclamp lighting samples by setting both (direct and indirect) lighting sample clamps to zero.