How one could play with the high dynamic range of HDRIs to use with the Environment Sphere?
HDR images can be tone down or up (depending on how the image has been done), how could i change that with the environment sphere?

In Poser 8, I don't think you can. But if you look at the node structure you'll see that there's one for GC (Gamma Correction) - have a play with that, it's quite useful.

I have to try this, merci Anthanasius :)
But how would you do to tone it up?

I have to try with the GC node too :)

Anthanasius has shown a mechanism for changing the brightness of an image using simple multiplication, but in a more complicated way than necessary. If all you want to do is multiply the image with a factor, simply plug it into Ambient_Color and adjust Ambient_Value as desired. That is what all the XYZ_Color/XYZ_Value pairs do. The input node is multiplied both with the color and the value.

For Poser Pro, this is usually all you would need to do. For other Poser's without gamma correction, you should use the shader I supplied with the Environment Sphere, the one called EnvPanoramic. You can use it in Pro, as well, but you don't need to if you simply want to multiply.

In my EnvPanoramic shader, set the Gamma In = 1 because the HDR image is already linear. (Or should be, anyway.)

Set the Gamma Out = 2.2 if not using Poser Pro GC.

Then set the brightness in the HSV node, using the Value parameter. In this example, I loaded a very dark night HDR and then increased the luminance by a factor of 3. This is like using a camera taking the same picture, but with an exposure 3 times longer.

By the way, you used the word "tone" (tone down or up). In discussion of HDR, the subject of "tone mapping" is different from simply multiplying with a factor.

The idea of tone mapping is to transform the image so that you compress the range back into an LDR image, so that nothing is brighter than 1.0. In other words, the range 0 to infinity is mapped to the range 0 to 1.

There are many different math formulas that accomplish this. One of them is the exponential formula used in Poser 8. Most can be made in HSV versions where an attempt is made only to alter the luminance, while preserving the saturation and value. In other words, instead of doing the mapping in RGB color space, we do it in HSV color space. That is the difference between HSV Exponential and simple Exponential tone mapping.

As I said, the exponential type is supported in P8, but we can do it with nodes in any version of Poser that supports HDR (P7 and up).

Another tone mapping formula that I came up with is x/(x+1). (And other similar variants using different constants.) It produces different looking results from the exponential formula. There are good and bad things about both.

If you desire, I can show how to do tone mappings.

Sure i've not really understanding the question

Yes show us !!!

 bookmark!!!!!

OK I'll come back to this. I already promised a window glass shader in another thread that I haven't got round to posting a tutorial on. I also promised AerySoul a week ago to do a "heavy metal" tutorial in the Node Cult and I haven't done that yet either. And I'm preparing a new version of matmatic for release. Whew - I'm way behind. Too bad for you guys I have to do real work for a living.

Ah, man!

I'm curious. Nobody has ever mentioned or shown any renders using the Panoramic effects shader I included. The one that can turn day to night, sunny to stormy, and boring sunsets into crazy over-the-top red and gold eye candy.

Did everybody miss it? I thought it would be really popular.

http://sites.google.com/site/bagginsbill/free-stuff/environment-sphere/environment-sphere-effects

Good question !

I use it very little, i thnil i need to use it more to understand how it work !

Wasn't aware of those effects shaders - I'll have to take a look.

I've just been playing with the Env Sphere again briefly, returning to outdoor scenes. I keep running into problems with the Equirectangular Maps not quite being Hi Res enough for my final output. 

Other issues I have managed to solve when using it with IDL include one where the background (i.e. the EnvSphere image) was rendering differently each time I made tweaks to the sphere's HSV node. I'm now using a dummy Env Sphere scaled 1% smaller than the one actually interacting with IDL and made non shadow casting & invisible to RT  - the dummy Sphere renders a consistent background even if I make changes to the one seen by IDL.
Also I found when using the ground plane as a shadow catcher with the Env Sphere that a lot of light seemed to never be making it to my figure. It turns out that even though the ground plane was just a shadow catcher it was still acting as a barrier to the light cast by the sphere from below and by making it non shadow casting & invisible to RT suddenly everything brightened up again.

Carodan, you've achieve quite a mastery of intergrating the ground into your scene using the Env Sphere.  I'm very impressed.  Maybe I should go back and give it a thorough work out sometime soon.

you realize bagginsbill has a custom shadow catcher, right?  i've never used it (i've never needed one), and i don't have P8 so i couldn't test it there, but it might be helpful:
http://sites.google.com/site/bagginsbill/free-stuff/shadow-catcher

I had difficulties getting bb's shadow catcher to play well with IDL which is why I went back to the ground plane - it's not an ideal solution. Hopefully I'm just being dippy and there is a way to have the full functionality of bb's solution.

good render carodan. what hair did you use?

Quote - good render carodan. what hair did you use?

You mean in the one I just posted in the gallery? It was one of the P7 Kate strand (dynamic) hairs.

I stupidly ended up dumbing down some of the detail on the figure in that image to try and marry her better with the Env Sphere background.

Anyone know of a good source for high quality Equirectangular HDRI's or maps usable on the Env Sphere (preferably free)?
I've found a few of the ones bb suggested ages ago on flickr but could always do with finding more.

btw - for that image I used the technique bb suggested above with the HSV node on the Env Sphere material to tweak the intensity and saturation of the map. It's a good control system.

Carodan:
Try these free at Flickr

http://www.flickr.com/groups/463055@N24/pool/

Quote - Carodan:
Try these free at Flickr

http://www.flickr.com/groups/463055@N24/pool/

Warning! These are not HDR images, despite what the group title says. They were made using HDR techniques, but they are LDR images and do not have high dynamic range. Also, they are gamma corrected to look right on a monitor (most of them anyway). Treat these as LDR in your workflow.

Thanks for the link and warning noted.

Many many thanks for the HSV trick!!! :D Exactly what i was looking for!:)
Now looking forward to see what you do with tone mapping :)

Carodan, on my site you will find several useful links to great quality HDR maps (and some are lower quality). Some of them are equirectangular, some are probes (you'll have to make them equirectangular by yourself):
http://www.vincebagna.com/index.php?option=com_neoreferences&Itemid=157&catid=4

Thanks for the link Vince.

Cool, thanks Vince (nice site btw).

I read on Bagginsbill site that the environment sphere uses spherical panorama images. I also see that in Vue, you can create those sort of images. I tried this but the image on the sphere came out looking blurred. Is there any other setting I need to change to get it clear?

edited to say I also tried using the wide angle lens and that didn't help much.

What size did you render the panorama to?

And what was the focal length on your camera when you tried to use it in Poser?

I have found that higher resolution images give you better clarity.  Now, I'm assuming you are referring to the final render rather than the preview window?

Last week in the Daz forum's, Gregorius (aka IuvenisScriptor here at Rendo) wrote that he couldn't figure out why out why the background always seems to come out a bit blurry in all his EnvSphere renders.

I will copy my answer here.

Because you're zoomed in on a tiny fraction of the EnvSphere image.

If you plan to actually see the EnvSphere image as a background, you have to think about how many pixels you're seeing of that image in your render.

Consider - the EnvSphere image, left to right, covers the entire 360 degrees of a circle. Suppose you're using a "50 mm" focal length camera simulating a 35mm film. Horizontally, that covers roughly 40 degrees of that circle. So you are displaying a fraction of the full image that is 40/360 of the entire image, or about 11%.

Now if the environment image is, for example, 5000 pixels across, you are asking it to cover your render using 40/360ths of that 5000 pixels, or about 550 pixels. If your render is 800 pixels across, you're stretching those 550 pixels to 800 - and you get blur. With a 100 mm camera, you're only using 5.7% of the circle, or just around 283 pixels.

So if I'm going to feature my environment image as my background, I usually use a wider camera (lower focal length). For example, a 25 mm lens gives just under 20% of the circle. With a 5K image, that's about 1000 pixels - usually enough to get a good background.

An alternative is to not use the environment image as the background. Keep it for IDL and reflection purposes, but place another photo (a normal one, not an equirectangular one) directly behind the figure on a one-sided square.

I always assumed that the Poser camera focal length was meant to correspond with the field of view produced by that focal length on a 35 mm film camera. Turns out after experimenting, that isn't so. There is a 1.4x crop factor. I can't imagine where they got that. I might have expected the 1.5x crop factor found on Nikon DX DSLR cameras or the 1.6x crop factor found on Canon DSLRs. Where did 1.4x come from?!?

Given a desired field of view, in degrees, and you want to know what focal length to set the Poser camera to, this equation gives pretty good answers:

12.75 / tan(x * pi / 360)

where x is the desired angle.

Did you know that Google accepts math formulas and tells you the answer? It's pretty cool. For example, to find the focal length to produce a 50 degree field of view, type this into Google and hit search:

12.75 / tan(50 * pi / 360)

Google says 27.34 is the answer.

You can also use this formula to calculate pixel ratios. For example, if your environment image is 5000 pixels and you want to see 800 of them across your render, use:

12.75 / tan(800 * pi / 5000)

The answer is 23.19 mm.

Here are a some more (rounded down) based on viewing angle:

10 degrees = 145 mm
20 degrees = 72 mm
30 degrees = 47 mm
40 degrees = 35 mm
45 degrees = 30 mm
50 degrees = 27 mm
55 degrees = 24 mm
60 degrees = 22 mm
65 degrees = 20 mm
70 degrees = 18 mm

@ Bagginsbill - I did a test Vue render of 1534 X 767. What's the minimum I should aim for? I tried focal lengths from 10 to 30.

@ Hborre - Yes I'm referring to the final render as opposed to the preview.

By the way, since then I think I've guessed where the 1.4x crop factor came from. Whoever coded that may have misunderstood the definitions of field of view for 35mm film, and they may have used the diagonal measurement of a square image instead of the horizontal measurement. The diagonal of a square is the square root of 2 longer than its side. The square root of 2 is approximately 1.414. Thus - a 1.4x crop factor is applied by accident.

My mind just went blank with all the maths!

Quote - @ Bagginsbill - I did a test Vue render of 1534 X 767. What's the minimum I should aim for? I tried focal lengths from 10 to 30.

@ Hborre - Yes I'm referring to the final render as opposed to the preview.

Well you have to use the math above to answer that. It depends on the size of the final render and the focal length of the camera you use.

What size render are you planning, and at what focal length (according to Poser's parameter). Then I can answer what you should be using to get a 1-to-1 pixel rendering from the environment sphere.

I just used the default render window which is roughly 900 X 1000 but I usually like to render out to a height of at least 2400.

Still didn't answer my question. Height doesn't matter. It's width that matters.

How big do you want your final render to be in width, and at what focal length?

For example, if you say you want to render in Poser an image that is 1000 pixels wide with a 50 mm camera, with no blur at all, then your panoramic image size should be 12583 pixels wide and 6291 pixels tall.

Width would be 1500 px at 60mm with no blur.

PS. Thank you for taking the time to answer.

You're not going to like this.

The panorama should be 22506 wide and 11253 tall.

You're right...I don't  :)

I'm guessing it doesn't matter what size the sphere is...

If you drop to 25 mm, still 1500 pixel render, then you need a panorama that is 9992 wide and 4996 tall, or roughly 10000 by 5000. This is the size I look for when I hunt panoramas on the web - 10K by 5K. And I render them with 25 mm cameras to 35 mm cameras (I usually render less than 1500).

Sometimes I find 8K by 4K and they are OK for smaller renders or also OK if I want the background a little blurry.

Quote - You're right...I don't  :)

I'm guessing it doesn't matter what size the sphere is...

Correct. When you make the sphere bigger, the pixels get bigger physically in the world, but they move farther away. The two effects cancel out exactly.

Similarly, if you shrink the sphere, the pixels get closer together physically, but they also move closer to the camera which enlarges them again exactly in proportion.

How many panorama pixels you see in the render is almost completely unaffected by the size of the sphere. All that matters is the focal length. And then if you render bigger, you simply grow those pixels.

Thanks for your help again. I guess I'll stick with the one sided square idea for now!

Even fancier way of computing: http://www.wolframalpha.com/input/?i=12.75+%2F+tan%2850+*+pi+%2F+360%29

you can use a normal 360 image in poser and cheat like its HDRI. you only need to make some control mattes to tell poser what is brighter then 1 or 2 or 3.

if you want to know how bright a light is in a room or sun is then you can use a real HDRI image and use it as reference.

Bagginsbill:

Quote -
Warning! These are not HDR images, despite what the group title says. They were made using HDR techniques, but they are LDR images and do not have high dynamic range.

Oh, I was not aware of it. I'm really an amateur when it comes to photography and technique.

vincebagna:

Quote -
Carodan, on my site you will find several useful links to great quality HDR maps (and some are lower quality). Some of them are equirectangular, some are probes (you'll have to make them equirectangular by yourself):
http://www.vincebagna.com/index.php?option=com_neoreferences&Itemid=157&catid=4

Very helpful, already used two of them in my renders thanks to the links on your page.