Here's a little scene to demonstrate using media to simulate translucency
(diffuse transmission). Dont' tell me the wallpaper, table etc. suck. I know
and right now I don't care. Just concentrate your gaze upon the lamp shade
and tell me what you think.

I had to use very dense media to get only diffuse light transmission (rgb
1500). This posed a bit of a problem, understandable when you read my post
about media_interaction (in .general). So there are actually 2 lamp shades:
one, the media container, with a clear filter and no_shadow to avoid the
shadow of the media (which would be nearly opaque in this scene); and
another, slightly thicker, using a yellowish filter to create the shadow you
see.

In the process, I encountered other quirks. Using a area light changes the
appearance of the media radically (much brighter and horribly grainy). Using
transmit also makes the media very grainy. I'm still exploring these
problems.

Margus

Post a reply to this message

Attachments:
Download 'lampshade.jpg' (98 KB)

Preview of image 'lampshade.jpg'

Margus Ramst wrote:
> 
> Here's a little scene to demonstrate using media to simulate translucency
> (diffuse transmission). Dont' tell me the wallpaper, table etc. suck. I know
> and right now I don't care. Just concentrate your gaze upon the lamp shade
> and tell me what you think.

The table mearly sucks. The wall paper sucks immaculate. I will let it go
since it is a proof of concept image.
 
> I had to use very dense media to get only diffuse light transmission (rgb
> 1500). This posed a bit of a problem, understandable when you read my post
> about media_interaction (in .general). So there are actually 2 lamp shades:
> one, the media container, with a clear filter and no_shadow to avoid the
> shadow of the media (which would be nearly opaque in this scene); and
> another, slightly thicker, using a yellowish filter to create the shadow you
> see.
> 
> In the process, I encountered other quirks. Using a area light changes the
> appearance of the media radically (much brighter and horribly grainy). Using
> transmit also makes the media very grainy. I'm still exploring these
> problems.
> 
> Margus


Hi Margus,

  I think you are on the right track if media is the choice you
prefer to model translucency with. I still think there are other
alternatives but I am not going to say what until I get a chance
to test my own theories. Your rgb values sound a bit whacko and
I don't understand why you would have to go so high with them.
In this business whatever works I suppose.


Ken's tip of the week:

   An interesting thing happens to media used inside a container if you
add an ior value to the interior statement that also contains the density
statement. I don't understand exactly what happens but it has a tendency
to polarize(?) the grain structure of the media providing a somewhat
smoother appearance. It might be worth trying with the media in this model
to see how different it is and if it behaves better or worse. I tried this
with a color list patterns study I did about a month ago and there was a
noticeable difference in how the density appeared. In the case of where I
used it the difference was better as it both smoothed the density of the
media as well as giving it better visibility in the lighting conditions
that were used.
  Your mileage may differ and some restrictions do apply. See your local
representative for details or answers to other questions you may have
concerning these restrictions and ways you may remedy them. Void where
prohibited or restricted by law. All rights reserved. No other conditions
apply whether implied or otherwise.


 Have a nice day :)

-- 
Ken Tyler

mailto://tylereng@pacbell.net

Post a reply to this message

Ken wrote in message <37112AB4.43CC4E3C@pacbell.net>...
>
>Hi Margus,
>
>  I think you are on the right track if media is the choice you
>prefer to model translucency with. I still think there are other
>alternatives but I am not going to say what until I get a chance
>to test my own theories.

I am appropriately intriqued, yet I respect your reserve and aspire to
contain my curiosity. Quite frankly, I can't imagine what alternatives you
have in mind (except bezier patches)

>Your rgb values sound a bit whacko and
>I don't understand why you would have to go so high with them.
>In this business whatever works I suppose.
>

The scale of this scene is rather small and the walls of the shade are <0.01
units thick. So this value is by no means whacko. Anything less makes the
shade transparent (non-diffusively).

As for your tip: you are indeed correct. Adding ior to the media container
makes it much smoother. I already had ior in the outer (yellow) shade, but
this had no effect... Unfortunately, this also greatly changes the lighting
of the media, to something that is, in my opinion, incorrect. More
specifically, the top of the shade becomes extremely bright - possibly
because the light_source is closest to that part.
Yet another media mystery. I would VERY much like to know the explanation.

>  Your mileage may differ and some restrictions do apply. See your local
>representative for details or answers to other questions you may have
>concerning these restrictions and ways you may remedy them. Void where
>prohibited or restricted by law. All rights reserved. No other conditions
>apply whether implied or otherwise.
>


I will contact my solicitor for a detailed analysis of this statement and
any obligations, financial or otherwise, it may impose upon my person.

>
> Have a nice day :)
>

And a good night to you, kind sir!

Margus

Post a reply to this message

While this is OK, I imagine that the rendering time is quite large... what I
was proposing is for someone to write a specific function for it... not only
would this be faster and easier to work with, it would also have no sampling
problems commonly associated with media...

Anyway, I think you're on the right track though and that's looking quite
good!

Now see if you can do this, a lit candle... :)

--
Lance.


---
For the latest 3D Studio MAX plug-ins, images and much more, go to:
The Zone - http://come.to/the.zone

Post a reply to this message

Lance Birch wrote:
> 
> While this is OK, I imagine that the rendering time is quite large... what I
> was proposing is for someone to write a specific function for it... not only
> would this be faster and easier to work with, it would also have no sampling
> problems commonly associated with media...

  We understood your proposal but there is always going to people willing
to explore the possibilities of using the current abilities of the program
to model the effect that is needed. It may or may not be something extremely
easy is attacked the right way. I have seen things done with Pov that I
would never have dreamed was possible witht the available features but
there it was for all to see.
  The one big problem I have with Pov's current model for media is that
the monte carlo sampling method employed it too limited in scope to be
as useful as it needs to be to cover all of it's intended applications.
Some added options and better sampling features would extend it to become
a very powerful medium.

 
> Now see if you can do this, a lit candle... :)
> Lance.

Been there done that long ago. Piece of cake. See below.

-- 
Ken Tyler

mailto://tylereng@pacbell.net

Post a reply to this message

Attachments:
Download 'kcandle.jpg' (3 KB)

Preview of image 'kcandle.jpg'

he he, that wasn't exactly what I mean by the candle... But it's nice
candle!!!  :)))  I just meant, with translucency... :)  But I'm sure you
already knew that :)

--
Lance.

P.S.  I've uploaded a new website!!!  See below, Colorblind...  (also, it's
best viewed with IE or Neoplanet, because Netscape messes up the frames and
JavaScript, I haven't found a way around that yet...)

---
For the latest 3D Studio MAX plug-ins, images and much more, go to:
The Zone - http://come.to/the.zone
For a totally different experience, visit my Chroma Key Website:
Colorblind - http://www.fortunecity.com/skyscraper/parallax/359/colorblind

Post a reply to this message

Lance Birch wrote:
> 
> he he, that wasn't exactly what I mean by the candle... But it's nice
> candle!!!  :)))  I just meant, with translucency... :)  But I'm sure you
> already knew that :)
> 
> --
> Lance.

  I don't know. Can you find anything uncandle like about the
appeareance of the candle image I posted. I thought not. Ok so
maybe a closer examiniation would reveal the lack of the material
properties you describe, then again it might not. I think I will
keep you guessing for the time being.

-- 
Ken Tyler

mailto://tylereng@pacbell.net

Post a reply to this message

Well, the main thing I was looking at was the light fading in the candle...
(downwards)

You know how when you light a candle the top part is a lot lighter than the
base (and I don't mean the flame there either ;-) he he he

Also, the top part of the candle should be shinnier than the bottom (because
the wax is heated and takes on a shiny appearance).

Just a thought...

--
Lance.


---
For the latest 3D Studio MAX plug-ins, images and much more, go to:
The Zone - http://come.to/the.zone
For a totally different experience, visit my Chroma Key Website:
Colorblind - http://www.fortunecity.com/skyscraper/parallax/359/colorblind

Post a reply to this message

I would like to see an example as created in that high priced, super max,
 do everything better software, made for the "real artist", 3d modelling
 stuff, and those other things too program, you keep mentioning that you
 use. You should be able to do something like this in that max thing in a
 couple of minutes to hear you talk. After I see how it is "supposed" to
 look I will then know what the realism threshold is that I must compete
 with. This will alow me to plan my next move accordingly.

 I'm waiting patiently and please have a nice day.

-- 
Ken Tyler

mailto://tylereng@pacbell.net

Post a reply to this message

On Mon, 12 Apr 1999 13:14:37 +1000, "Lance Birch"
<lan### [at] usanet> wrote:

>Well, the main thing I was looking at was the light fading in the candle...
>(downwards)
>
>You know how when you light a candle the top part is a lot lighter than the
>base (and I don't mean the flame there either ;-) he he he
>
>Also, the top part of the candle should be shinnier than the bottom (because
>the wax is heated and takes on a shiny appearance).
>
>Just a thought...

Do you like this implementation, Lance? This is part of my May-June
1997 entry, re-rendered in 3.1 (no halo). Check the IRTC archive, I've
posted source too (indented and commented :) ). HTH.

---------
Peter Popov
ICQ: 15002700

Post a reply to this message

Attachments:
Download 'CANDLE.jpg' (21 KB)

Preview of image 'CANDLE.jpg'