On 06/06/2014 06:46, Sean Day wrote:

>
> Thanks Stephen, after my first attempt at modelling the ant I thought
> maybe I should revisit Sculptris as it seemed like something I should be
> doing with meshes. However, I like the challenge of just using SDL to
> create images.
>

As someone cannot get blobs to work the way I want them to work. I can 
only gasp and stare. With admiration.


-- 

Regards
     Stephen

Post a reply to this message

clipka wrote:
> Am 06.06.2014 07:52, schrieb Sean Day:
>> clipka wrote:
>
>
> Have you tried a more blurred version of the light probe?
>
> I wouldn't be surprised though if you were in the process of discovering
> a bug in the integration of radiosity and SSLT; so if you don't mind
> digging a bit more, you'd do me a great favor.
>

Hi Clipka,

Going to show my ignorance here but how do I blur the light probe?

Happy to help in any way possible, I tried a different light probe and 
had a similar but more uniform problem (attached). I am happy to upload 
the source if that helps or run test renders myself. The light probes I 
have tried so far have both been from the same source so I will try one 
from a different source.

Both the probes were from http://www.hdrlabs.com/sibl/archive.html the 
backyard and harbour3 ones towards the bottom of the page.

Sean

Post a reply to this message

Attachments:
Download 'ant - harbour.png' (690 KB)

Preview of image 'ant - harbour.png'

Stephen wrote:
> On 06/06/2014 06:46, Sean Day wrote:
>
> As someone cannot get blobs to work the way I want them to work. I can
> only gasp and stare. With admiration.
>

They can be tricky that is for sure, the legs were actually a lot easier 
than I expected (just translate the components along a spline and adjust 
the scale at certain points) it was (and still is) the head that is 
giving me the most trouble, once the model is complete I will upload it 
to the objects or source files thread.

Post a reply to this message

On 06/06/2014 20:35, Sean Day wrote:
> Stephen wrote:
>> On 06/06/2014 06:46, Sean Day wrote:
>>
>> As someone cannot get blobs to work the way I want them to work. I can
>> only gasp and stare. With admiration.
>>
>
> They can be tricky that is for sure,

It is probably something that just comes to you. Too bad I was behind 
the door. When that knack was given out. ;-)

> the legs were actually a lot easier
> than I expected (just translate the components along a spline and adjust
> the scale at certain points) it was (and still is) the head that is
> giving me the most trouble, once the model is complete I will upload it
> to the objects or source files thread.
>


That would be worthwhile. :-D


-- 

Regards
     Stephen

Post a reply to this message

Am 06.06.2014 21:31, schrieb Sean Day:

>> I wouldn't be surprised though if you were in the process of discovering
>> a bug in the integration of radiosity and SSLT; so if you don't mind
>> digging a bit more, you'd do me a great favor.
>>
> Going to show my ignorance here but how do I blur the light probe?

HDRshop is probably the best tool for the job.

> Happy to help in any way possible, I tried a different light probe and
> had a similar but more uniform problem (attached). I am happy to upload
> the source if that helps or run test renders myself. The light probes I
> have tried so far have both been from the same source so I will try one
> from a different source.

Uh, that definitely looks wrong. Can you do me another favor and use 
that same light probe but with (a) a simple sphere, and (b) a 
single-sphere blob?

Post a reply to this message

Am 07.06.2014 00:44, schrieb clipka:
> Am 06.06.2014 21:31, schrieb Sean Day:
>
>>> I wouldn't be surprised though if you were in the process of discovering
>>> a bug in the integration of radiosity and SSLT; so if you don't mind
>>> digging a bit more, you'd do me a great favor.
>>>
>> Going to show my ignorance here but how do I blur the light probe?
>
> HDRshop is probably the best tool for the job.

Dang, forget about that one. It used to be free some years ago, but 
apparently they've made it payware.

Post a reply to this message

clipka wrote:
> Am 07.06.2014 00:44, schrieb clipka:
>> Am 06.06.2014 21:31, schrieb Sean Day:
>>
>> HDRshop is probably the best tool for the job.
>
> Dang, forget about that one. It used to be free some years ago, but
> apparently they've made it payware.
>

That's OK as I have a copy of that, the only time I used it was with 
megapov to make a HDR probe out of my PovComp entry. Fortunately I still 
have a copy on an old backup drive. I will try blurring the probe as a 
further test with the ant.

The non-blurred probe does not cause a problem with a sphere (attached). 
Just about to start a blob render.

Also, I did a render with the building_probe (not from the same source) 
and the ant looks better I think there is still a slight glow but not much.

Will upload the blob version later.

Sean

Post a reply to this message

Attachments:
Download 'sphere.png' (135 KB) Download 'ant building_probe.png' (672 KB)

Preview of image 'sphere.png'

Preview of image 'ant building_probe.png'

clipka wrote:
> Am 07.06.2014 00:44, schrieb clipka:
>> Am 06.06.2014 21:31, schrieb Sean Day:
>>
>>>> I wouldn't be surprised though if you were in the process of
>>>> discovering
>>>> a bug in the integration of radiosity and SSLT; so if you don't mind
>>>> digging a bit more, you'd do me a great favor.
>>>>
>>> Going to show my ignorance here but how do I blur the light probe?
>>
>> HDRshop is probably the best tool for the job.
>
> Dang, forget about that one. It used to be free some years ago, but
> apparently they've made it payware.
>

The single sphere blob also looks OK, when I opened the probe in HDRShop 
I noticed some very bright areas and wonder if they are the cause (or 
maybe if not the cause the reason it is more extreme with this probe, 
also the Harbour3 probe looks OK in HDRShop and has a similar effect).

The bright areas I am referring to are in the attached probe.png.

Also, the single sphere blob (adjusted the size of the sphere component 
to make it closer to the standard sphere) also looks OK, maybe something 
to do with the thickness of the ant/blob or having multiple components 
is the cause.

I have a few more ideas of things to test (obviously the blurred probe 
with the ant being one) but maybe switching off the radiosity on the 
subsurface parameters would be a good one?

Also, when I first noticed this glow problem I stacked 3 spheres and 
rendered them and got a similar issue (hence my original I don't think 
it is an issue with the blob statement). Will see if I can reproduce that.

Sean

Post a reply to this message

Attachments:
Download 'probe.png' (200 KB) Download 'blob.png' (133 KB)

Preview of image 'probe.png'

Preview of image 'blob.png'

clipka wrote:
> Am 07.06.2014 00:44, schrieb clipka:
>> Am 06.06.2014 21:31, schrieb Sean Day:
>>
>>>> I wouldn't be surprised though if you were in the process of
>>>> discovering
>>>> a bug in the integration of radiosity and SSLT; so if you don't mind
>>>> digging a bit more, you'd do me a great favor.
>>>>
>>> Going to show my ignorance here but how do I blur the light probe?
>>
>> HDRshop is probably the best tool for the job.
>
> Dang, forget about that one. It used to be free some years ago, but
> apparently they've made it payware.
>

Here is the problem reproduced by stacking 3 spheres.

Post a reply to this message

Attachments:
Download 'spherestack.png' (92 KB)

Preview of image 'spherestack.png'

clipka wrote:
> Am 07.06.2014 00:44, schrieb clipka:
>> Am 06.06.2014 21:31, schrieb Sean Day:
>>
>>>> I wouldn't be surprised though if you were in the process of
>>>> discovering
>>>> a bug in the integration of radiosity and SSLT; so if you don't mind
>>>> digging a bit more, you'd do me a great favor.
>>>>
>>> Going to show my ignorance here but how do I blur the light probe?
>>
>> HDRshop is probably the best tool for the job.
>
> Dang, forget about that one. It used to be free some years ago, but
> apparently they've made it payware.
>

Forgot to add that the blurred probe did not make any difference..

Sean

Post a reply to this message

Attachments:
Download 'ant - gaussian blurred probe.png' (80 KB)

Preview of image 'ant - gaussian blurred probe.png'