[LSC] Transparent stickers and color 16

[William Howard]

I've been following this and I've been swinging from the "that's not a 
correct usage of colour 16" camp to the "that's a workable solution" camp 
depending on who posted last, which implies to me that neither approach is 
an ideal solution.

Some thoughts.

Colour 16 is the "main colour", which is not the same as "sticker 
transparent".  Trans-white (clear) is also not "sticker transparent".
Should we actually be defining a new colour?  When the new bluish grays were 
added to the palette the numbers were choosen in such a way that "old 
software" would render them as the old grays.  I never checked to see if 
this worked, as the numbers were allocated by Steve B, I just assumed they 
would.  So we should be able to find a number for the new colour which 
behaves as trans-white in old software.  This would be my preferred 
solution.

Alternatively, do the trans parts actually need to be modelled?  In the real 
world they do, but in a virtual environment any trans part could be modelled 
by leaving a hole - even if this hole is the outside region and creates a 
dis-jointed part.  Purists will throw their hands up in horror at this, but 
other purists will do the same if we use colour 16 in a non-standard way, so 
it may be worth considering.  Another option would be to model the sticker 
as all the trans areas in the main part and all the coloured details in a 
sub-part, that way the user could use the main part if their renderer 
supported the new colour, or just the sub-part if it didn't.

William (still undecided)


----- Original Message ----- 
From: <philohome at free.fr>
To: "LDraw Standards Committee" <lsc at ldraw.org>
Sent: Sunday, January 13, 2008 2:08 PM
Subject: Re: [LSC] Transparent stickers and color 16



> Selon Travis Cobbs <tcobbs at gmail.com>:

>

>> On Jan 12, 2008 12:51 PM, Orion Pobursky <orion at ldraw.org> wrote:

>>

>> > This sounds like a render/implementation problem and not a problem with

>> > the spec.  If transparent material isn't displaying correctly then

>> > either we need to define a better transparent color or the renderers

>> > need to be corrected to better display trans-white.  Unless you can 

>> > give

>> > me concrete and visual examples that this issue is a problem with the

>> > spec and not with the renderer, I feel that the use of color 16 in this

>> > way is definitely non-standard, confusing, and should be avoided.

>> >

>>

>> Due to the physics of the situation, the transparent part of stickers 

>> comes

>> out looking (almost) like it's not there at all when you stick it to a

>> surface.  There are two main reasons for this.  First of all, the 

>> stickers

>> are really thin.  Secondly, they're stuck firmly to the brick, with no 

>> air

>> gap behind them.  If you take a piece of clear LEGO, and put it right 

>> next

>> to another brick and look through it to the other brick, I don't think 

>> the

>> transparent LEGO will disappear in the same way that the transparent part 

>> of

>> a sticker disappears.

>>

>> Because of this, I don't feel that it's a renderer issue.  The two 

>> materials

>> really do have fundamentally different properties, and we don't currently

>> have a way to model a clear sticker's properties in LDraw.

>>

>> There's also a question of pragmatism.  If stickers look good using this

>> method, and can't be made to look good in any existing renderers using 

>> any

>> other method, the users of the LDraw library are going to want us to use

>> this method.

>>

> I completely agree with Travis physical analysis. Maybe the issue could be

> solved also with the definition of a new very transparent (high alpha) 

> color,

> but imho it would be even more confusing and for sure would cause many

> compatibility problems with older LDraw software packages.

>

> Philo

>

> _______________________________________________

> LSC mailing list

> LSC at ldraw.org

> http://five.pairlist.net/mailman/listinfo/lsc

>