Can I achieve the same brightness using an HDRI sky as using a light 
source? Or are there limits on how bright a scene can be?


Mike

Post a reply to this message

Le 18-02-22 à 18:51, Mike Horvath a écrit :
> Can I achieve the same brightness using an HDRI sky as using a light 
> source? Or are there limits on how bright a scene can be?
> 
> 
> Mike

Yes. In fact, an HDRI sky can be TO bright, leading in an image that is 
over exposed like.

For that to work, you need to use radiosity.
Next, you can apply the HDRI to a sky_sphere or a very large sphere 
enclosing the scene.
In the case of the world sphere, it need a finish similar to :
finish{
  diffuse 0
  emission 1
}

If the scene is to dark, you increase the emission value, and reduce it 
if it's to bright.


Alain

Post a reply to this message

Mike Horvath <mik### [at] gmailcom> wrote:
> Can I achieve the same brightness using an HDRI sky as using a light
> source? Or are there limits on how bright a scene can be?
>

I haven't yet used an HDRI image/image_map in a scene, but I would assume that
the maximum rendered brightness on-screen would be clipped to 1.0 (typical
maximum white)-- or rather, as bright as your monitor can reproduce 1.0, just
like with any scene.

The saved file is another matter, if it's saved as HDRI. (POV-Ray can do that,
right? I haven't read the documentation enough to know.)

Post a reply to this message

Am 23.02.2018 um 00:51 schrieb Mike Horvath:
> Can I achieve the same brightness using an HDRI sky as using a light
> source? Or are there limits on how bright a scene can be?

Technically, that depends on the HDRI file format used.

In POV-Ray, light source brightness is stored internally using so-called
"single precision" floating point values; these can range as high as ca.
3.4*10^38.

The Radiance HDR file format uses a proprietary floating point-ish
format with values ranging as high as ca. 1.7*10^38, so not much
difference there.

OpenEXR files are usually stored using "half precision" floating point
values; these can range no higher than 65504 (ca. 6*10^4). However,
OpenEXR can also use "single precision" floating point values as an
alternative. (It even allows for the use of "double precision" floating
point values, which can range as high as ca. 1.8*10^308, but POV-Ray
would limit them internally to single precision.)


In practice, I don't expect the limitations of standard (half precision)
OpenEXR to be of any significance except in fringe cases, such as trying
to model a very bright point-like light source.

Post a reply to this message

Le 18-02-23 à 13:18, Kenneth a écrit :
> Mike Horvath <mik### [at] gmailcom> wrote:
>> Can I achieve the same brightness using an HDRI sky as using a light
>> source? Or are there limits on how bright a scene can be?
>>
> 
> I haven't yet used an HDRI image/image_map in a scene, but I would assume that
> the maximum rendered brightness on-screen would be clipped to 1.0 (typical
> maximum white)-- or rather, as bright as your monitor can reproduce 1.0, just
> like with any scene.
> 
> The saved file is another matter, if it's saved as HDRI. (POV-Ray can do that,
> right? I haven't read the documentation enough to know.)
> 
> 
> 
> 
> 
> 
Yes, you can save as high dyamic range image. Just need to add +fe 
(*.exr) or +fh (*.hdr) to the command line. Then, there is no clipping.

Post a reply to this message

On 2/23/2018 9:50 PM, Alain wrote:
> Le 18-02-23 à 13:18, Kenneth a écrit :
>> Mike Horvath <mik### [at] gmailcom> wrote:
>>> Can I achieve the same brightness using an HDRI sky as using a light
>>> source? Or are there limits on how bright a scene can be?
>>>
>>
>> I haven't yet used an HDRI image/image_map in a scene, but I would 
>> assume that
>> the maximum rendered brightness on-screen would be clipped to 1.0 
>> (typical
>> maximum white)-- or rather, as bright as your monitor can reproduce 
>> 1.0, just
>> like with any scene.
>>
>> The saved file is another matter, if it's saved as HDRI. (POV-Ray can 
>> do that,
>> right? I haven't read the documentation enough to know.)
>>
>>
>>
>>
>>
>>
> Yes, you can save as high dyamic range image. Just need to add +fe 
> (*.exr) or +fh (*.hdr) to the command line. Then, there is no clipping.

I wonder if it would be easier to render my sky and fluffy clouds once, 
and then include it in subsequent renders as HDRI.

Mike

Post a reply to this message

On 24-2-2018 12:09, Mike Horvath wrote:
> On 2/23/2018 9:50 PM, Alain wrote:
>> Le 18-02-23 à 13:18, Kenneth a écrit :
>>> Mike Horvath <mik### [at] gmailcom> wrote:
>>>> Can I achieve the same brightness using an HDRI sky as using a light
>>>> source? Or are there limits on how bright a scene can be?
>>>>
>>>
>>> I haven't yet used an HDRI image/image_map in a scene, but I would 
>>> assume that
>>> the maximum rendered brightness on-screen would be clipped to 1.0 
>>> (typical
>>> maximum white)-- or rather, as bright as your monitor can reproduce 
>>> 1.0, just
>>> like with any scene.
>>>
>>> The saved file is another matter, if it's saved as HDRI. (POV-Ray can 
>>> do that,
>>> right? I haven't read the documentation enough to know.)
>>>
>>>
>>>
>>>
>>>
>>>
>> Yes, you can save as high dyamic range image. Just need to add +fe 
>> (*.exr) or +fh (*.hdr) to the command line. Then, there is no clipping.
> 
> I wonder if it would be easier to render my sky and fluffy clouds once, 
> and then include it in subsequent renders as HDRI.
> 

I would not do that if I were you. The fine details of the sky/cloud 
structures would be lost. The resolution of the sky dome is not that 
good and those images are not intended for crisp backgrounds but for 
(global) illumination.

-- 
Thomas

Post a reply to this message

On 24-2-2018 13:01, Thomas de Groot wrote:
> On 24-2-2018 12:09, Mike Horvath wrote:
>> On 2/23/2018 9:50 PM, Alain wrote:
>>> Le 18-02-23 à 13:18, Kenneth a écrit :
>>>> Mike Horvath <mik### [at] gmailcom> wrote:
>>>>> Can I achieve the same brightness using an HDRI sky as using a light
>>>>> source? Or are there limits on how bright a scene can be?
>>>>>
>>>>
>>>> I haven't yet used an HDRI image/image_map in a scene, but I would 
>>>> assume that
>>>> the maximum rendered brightness on-screen would be clipped to 1.0 
>>>> (typical
>>>> maximum white)-- or rather, as bright as your monitor can reproduce 
>>>> 1.0, just
>>>> like with any scene.
>>>>
>>>> The saved file is another matter, if it's saved as HDRI. (POV-Ray 
>>>> can do that,
>>>> right? I haven't read the documentation enough to know.)
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>> Yes, you can save as high dyamic range image. Just need to add +fe 
>>> (*.exr) or +fh (*.hdr) to the command line. Then, there is no clipping.
>>
>> I wonder if it would be easier to render my sky and fluffy clouds 
>> once, and then include it in subsequent renders as HDRI.
>>
> 
> I would not do that if I were you. The fine details of the sky/cloud 
> structures would be lost. The resolution of the sky dome is not that 
> good and those images are not intended for crisp backgrounds but for 
> (global) illumination.
> 

er.... I might contradict myself here because I may have misunderstood 
what you intend to do :-)

If you intend to build a spherical sky probe to be used as hdri/exr sky 
spheres in later scenes, then yes by all means, do that with a spherical 
camera. However, be aware of my comment above at the same time. I attach 
an example scene I made years ago to show what that means.

-- 
Thomas

Post a reply to this message

Attachments:
Download 'hdri_test1_mp_mt7.jpg' (35 KB)

Preview of image 'hdri_test1_mp_mt7.jpg'

On 2/24/2018 7:15 AM, Thomas de Groot wrote:
> On 24-2-2018 13:01, Thomas de Groot wrote:
>> On 24-2-2018 12:09, Mike Horvath wrote:
>>> I wonder if it would be easier to render my sky and fluffy clouds 
>>> once, and then include it in subsequent renders as HDRI.
>>>
>>
>> I would not do that if I were you. The fine details of the sky/cloud 
>> structures would be lost. The resolution of the sky dome is not that 
>> good and those images are not intended for crisp backgrounds but for 
>> (global) illumination.
>>
> 
> er.... I might contradict myself here because I may have misunderstood 
> what you intend to do :-)
> 
> If you intend to build a spherical sky probe to be used as hdri/exr sky 
> spheres in later scenes, then yes by all means, do that with a spherical 
> camera. However, be aware of my comment above at the same time. I attach 
> an example scene I made years ago to show what that means.
> 

I think that in addition to spherical HDRI formats, there are also cube 
mapped ones. Does POV-Ray support these as well? (Cube mapping tends to 
look nicer IMO.)


Mike

Post a reply to this message

Le 18-02-24 à 06:09, Mike Horvath a écrit :
> On 2/23/2018 9:50 PM, Alain wrote:
>> Le 18-02-23 à 13:18, Kenneth a écrit :
>>> Mike Horvath <mik### [at] gmailcom> wrote:
>>>> Can I achieve the same brightness using an HDRI sky as using a light
>>>> source? Or are there limits on how bright a scene can be?
>>>>
>>>
>>> I haven't yet used an HDRI image/image_map in a scene, but I would 
>>> assume that
>>> the maximum rendered brightness on-screen would be clipped to 1.0 
>>> (typical
>>> maximum white)-- or rather, as bright as your monitor can reproduce 
>>> 1.0, just
>>> like with any scene.
>>>
>>> The saved file is another matter, if it's saved as HDRI. (POV-Ray can 
>>> do that,
>>> right? I haven't read the documentation enough to know.)
>>>
>>>
>>>
>>>
>>>
>>>
>> Yes, you can save as high dyamic range image. Just need to add +fe 
>> (*.exr) or +fh (*.hdr) to the command line. Then, there is no clipping.
> 
> I wonder if it would be easier to render my sky and fluffy clouds once, 
> and then include it in subsequent renders as HDRI.
> 
> Mike

It can be done, but be aware that the clouds won't cast any shadow at all.
Depending on the specific of the rendered scene, this may be a problem.

Post a reply to this message