[soarchitect]

I agree with Ernesto, Real Scale does not work and it needs to be addressed.

I do not understand how to write software, so explaining the technical difficulties will not help or matter. All I know is there is a major hiccup in the workflow- from 3rd party modeler (such as sketchup) to Maxwell Studio. For Architectural Viz, Real Scale is not an option, it is mandatory.

Since Real Scale is required, I must perform my textures in the 3rd party software. It is too difficult to figure out how to texture in Studio by using a math formula- you can't measure anything in Studio, and what about complex shapes??

The unfortunate result is that Studio pretty much useless. Next Limit created this elegant software, a real engineering achievement that has no equal when it comes to light quality, and I can't use it texture with any degree of accuracy.

The only thing I can suggest is that Next Limit first spend time trying to understand the workflow of Architects (or anyone that needs dimension specific materials), and then see what can be done with the software. Architectural Viz may not be a big market niche for Next Limit, but this industry will never embrace the software if this problem is not addressed. Do you think anyone would use Autocad if you could not use it draw accurately?

That said, I think you guys are great and do impressive work, but it is time you acknowledge there is a big hole that need to get plugged.

[Polyxo]

Real Scale as a concept certainly sounds elegant:
When the renderer works based on physics, it was only consequent if material scale worked like in the real world too.

In fact however the concept of realscale mapping is limited to a box-mapped projection. Such a material on any
complex shape will show the realistic material scale but still look terribly off. Real Scale applied to anything which isn't
a box per se defeats the concept of realistic or even accurate material representation.

Real scale will probably somehow do in quite a few architectural contexts, if one works in this field.
But in order to show a realistic flow of surface structure one will not come around using other projection types too
and likely - as most powerful - UV-unwrapping scene objects.

I therefore honestly don't know what the fuzz is about. I find realistic scaling of surface structure one of the simplest
things to do in rendering, there's no measuring or mathematical formulas required. Of course it still gets done incorrecly
often... in case of doubt one should work with reference images or with material samples laying on the desk.

If I had to say anything I clearly would remove the Real Scale from the program. This option gives users a promise it can
not fulfill and will never be able to fulfill.

[feynman]

I think real-scale materials in Maxwell Studio work really well for many object typologies industrial designers are involved in. Very many two-dimensional materials like sheet metals, engineered wood, sheet plastics, etc. are coming to the fore, with view to digital fabrication, quite similar to the field of architecture. This goes along with the resurgence of more basic forms and morphologies that no longer relates so very much to soap-bar car-styling as used to be the case many years ago. I have not seen an industrial design or product design colleague not getting good results using real-scale materials in MS.

And, sure, if real-scale materials do not suffice to cover a quirky shape, a bit of UV-unwrapping does the trick - although this functionality is not present in Catia, Creo, Alias, Rhino or Solidworks; and usually, in a studio context, you don't really want to fiddle around with too many different softwares (unless the clients pay for it).

[Polyxo]

I may be that many ID's use Real Scale but I deny that one can get good results from objects with discernible texture when just using Box Mapping.
Why using a render engine which doesn't take shortcuts but simply slap textures onto objects even of own Design regardless of their fit? The fact that no
Nurbs application has serious texturing tools built in is an incredible pain, no doubt, but shortcomings in software should not justify using crappy methods.

[JDHill]

Real Scale as a concept certainly sounds elegant: When the renderer works based on physics, it was only consequent if material scale worked like in the real world too.

A nice thought, but far too abstract, I think, since in the real world, somebody has to decide where things begin, are centered, are rotated, and so on. In other words, nothing works automatically, and for a thing to end up with the "material" it has, some human being has either removed all that wasn't the thing (sculpted), or has carefully decorated the thing (painted, etc). In neither case is uniformity an inherent property (and let's not get into the question of what "is" a material -- software is essentially a collection of analogies, and in this case, necessarily imperfect ones, there being no separable "material" concept in the real world, where all is really only molecular structure, and the electrochemical phenomena that arise as a result of that structure), it was either carefully built into the substrate, or was carefully created at the surface of the object, intentionally. Never has it come about simply by answering a single yes/no question.

For Architectural Viz, Real Scale is not an option, it is mandatory.

And it is built into SketchUp, where it belongs.

Regarding the statement that it "does not work," please refer back to my response to you in a similar discussion, here. The fact is that real scale certainly does work, unless you prevent it from doing so by the use of a workflow that defeats it.

Let's illustrate this using the example of SketchUp, which directly asks you to specify what size you'd like your textures to appear on the model. This is great, but it is also exactly what sets up the problem, which is: different applications, when they're aware of a real scale concept at all, go about implementing it in fundamentally different ways. If you tell SketchUp to ensure your textures are of a given size, it accomplishes this by UV mapping faces at that size. But if you then tell Maxwell to interpret texture tile values in its real scale way, you set up an unresolvable conflict, since real scale will not work with the UVs you've asked SketchUp to generate. So you have to choose one or the other: either tell SketchUp that all your textures are 1m square, and let Maxwell do real scale by means of texture scale, or tell SketchUp that they're of a given size, and tell Maxwell not to scale textures at all. This is not something we can automatically resolve after the fact, and the second option is obviously preferable, since you don't want to be looking at 1m tiles in SketchUp.

In the case that your UV-mapping application is Maxwell Studio, and you want to use real scale, all you need to do is use a 1m cubic projection in Objects > UV Sets. Since UVs refer relatively to the mesh, if you later scale the mesh, you would need to click the "Normalize" button for the UV Set, to once again resize the mapping cube appropriately. That's it, though -- it's not necessary to do any math. As far as using Maxwell Studio to map complex shapes, there are really two factors: first, it does not intend to be a replacement for suitable UV-unwrapping software, and second, the question of real scale doesn't even make sense here, since there exists no mutually-agreeable answer to the question of what it might mean to real scale map a model of, say, an hourglass.

[Polyxo]

Real Scale as a concept certainly sounds elegant: When the renderer works based on physics, it was only consequent if material scale worked like in the real world too.

A nice thought, but far too abstract, I think, since in the real world, somebody has to decide where things begin, are centered, are rotated, and so on. In other words, nothing works automatically, and for a thing to end up with the "material" it has, some human being has either removed all that wasn't the thing (sculpted), or has carefully decorated the thing (painted, etc). In neither case is uniformity an inherent property (and let's not get into the question of what "is" a material -- software is essentially a collection of analogies, and in this case, necessarily imperfect ones, there being no separable "material" concept in the real world, where all is really only molecular structure, and the electrochemical phenomena that arise as a result of that structure), it was either carefully built into the substrate, or was carefully created at the surface of the object, intentionally. Never has it come about simply by answering a single yes/no question.

I agree.