[Forester]

OK, just a simple one from me. An Ancient Persia garden in the early morning light. .... Just to keep up some traffic on the site for those who worry that we are all moribund.

Models were made early this Spring - about 250 models dealing with an ancient Persian river city. The reason for this pic is that I have acquired some new HDRI's and wanted to try them out. Also, a couple of the buildings for this city are supposed to be "hanging garden" buildings encrusted with flowering plants. So, I am trying to teach myself to make climbing roses and such in Substance Designer so that I can get rid of the thousands of polygons my climbing rose models create in this garden.

I must say that it is not going well, as yet. I made the roses and the date palms in this picture. (The other plants came from CGAxis.) Making real modesl of climbing roses is massively easier for me than learning to do approximately the same in Substance Designer. But Substance Designer is very cool - so, far, I have some decent climbing rose plants (stems and leaves in 3D) but lack the blossoms and buds. And only a few polys for the wall arches and columns are needed. But the rose flowers and rose flower buds are another matter and I've only been partly successful with those. Just got to keeping working on these until I figure it out.

[Mark Bell]

...I'll take an evening coffee.

Interesting - I don't have Substance Painter so am unable to see whether it is of benefit or not in reducing the work time. I generally get models off the 3DWarehouse site (free) which includes a lot of low poly objects upwards but you'll need to sort through a lot of "useless stuff" to get something of value.

In the previous render I posted, the lighting was also from an IBL which was off my Nokia taken during a site visit for the project. In Studio, I moved the IBL below the horizon line as I didn't want it visible in the render as I'd selected to use Maxwell Sky, but did want the photo used for the lighting. I've found 'real' photos have more 'light data' in their image than a photo that has been through some editing software, and IBL's seem to give more realistic and softer lighting. The other thing I'm learning is to use real life settings for the camera whereas I was often tempted to keep playing with the settings as it renders and push the lighting or camera settings up too high. In the previous image all I did was hit render, let it clean up to about SL18 then adjust the sliders. The other setting which made a big difference to the liveliness of the image was to change the Response to a different type of film. In that example I used Ektachrome 100 (Color Rev) and then adjusted the temperature to bring out the deep blue sky. I'm sure in your image you would get something similar as it already looks clean so you don't need to do any more rendering, just adjust some of the settings to what mood of lighting you want for an outdoor scene. I'm no expert in Maxwell but as I get to use it more the simpler approach seems to give the better renders.

You'll need to check, but I recall Maxwell 5.? when it first came out including a new colour mixer? Nasok did a Youtube video with a bowl of lollies and demonstrated how you could quickly change the random colours of objects. This may be a useful approach to making the landscape more natural and random?

Thanks for posting the render...there is still some life in this Forum...

[Forester]

Thanks for all that information about your photo(s) and how you used them, Mark. Those are valuable and interesting pieces of information. And, ... I'll go play with the Maxwell settings a bit, as you suggested.

[Forester]

A quick clarification so that I don't lead you or anyone else astray...

Substance Painter is a program for painting on 3d objects, as the title suggests. (Not used in the image above.)

Substance Designer is a program for creating materials and shaders. There are a lot of reasons Maxwell users should be learning it, but my primary reason here is because materials made with it can have user-adjustable roughness, normal and height maps that provide amounts of detail no modeller in their right mind would attempt to build into a 3d model. (Because the polygon count would be excessive.) Here are three examples. In the first picture, the vase is just a dead simple cylinder with a very low number of polygons. In the second image, the old gold metal roses bookcover is just a single plane with one polygon. In the third image, thetwo plants I'm playing with just laid on top of a single polygon plane.

Almost as importantly, once I have these materials, they are "user-adjustable" as I stated. That is, I can make the "icicles" in the glass vase be as few or as many as I want - maybe ten different permutations for ten different cylinders as a collection of vases. I can make the gold metal for the book cover be shiny new - or tarnished old or anything in between. I can lay the roses out in any configuration I want, as well as mess around with the border edges. I can raise or lower the embossing to any degree I want for the render. For the plants in the last image, I have a single material whose parameters I can adjust to make a different version of the plant every time I click a button. The plants can have more or less leaves, be taller or shorter, have more or less branches , greater or lesser density, more or less leaf color variation and so forth. So, I can build myself a single material. Apply it to five flat polygons spaced along the wall, and make it look like five different plants, all with just the click of one or two buttons. These plants are not going to look "flat" - they'll have some depth to them. They'll catch light and shadows just fine. And, I can do this all within the rendering engine - "live" - so I can tweak these things until I get exactly the look I want. (Would that I was a better "artist" though, instead of just a simple model-builder! Sigh!)





These are the reasons I've come to believe that almost all of us Maxwell Render folks should be acquiring a basic knowledge of Substance Designer.

[Forester]

OK ..... coffee at dusk ....

[Forester]

Or mid-day coffee...

[Andreas Hopf]

These are the reasons I've come to believe that almost all of us Maxwell Render folks should be acquiring a basic knowledge of Substance Designer.

Well, the glass vase or pewter floor can be easily done with a real-scale material and a cylindrical UV projector for the vase and a planar UV projector for the floor. For more realism, use a displacement instead of a normal map. No UV-unwrapping or other software needed. Industrial, product or packaging designers either buy the necessary textures, or simply draw them in Adobe Illustrator, if the textures are also part of the design task. For architecture visualisation materials, things might be quite different of course. I am still waiting for Arroway or another texture vendor cracking the so far insurmountable solid wood texture problem, where also Keyshot's procedural wood fails ; )

Happy rendering in 2022 to all 12 users

[Forester]

Ah, my error in not making clear my objective for the initial image......

The problem I'd posed myself is to create a procedural material that would allow me to make climbing roses, and some other types of "hanging plants" that would allow me to cover a building with them, without having to build models for the plants.

To start the explanation, I'd early built a collection of some 250 models (buildings, lamps, tables and stools, coffee pots, characters and their clothing, weapons, dhows, etc.) representing the first of the ancient Persian civilizations.



Among the buildings is this simple one, that should be a general-purpose prop for a hanging garden.



After making this one, I experimented with covering it with some plants, but quickly found the polygon count to be excessive. In fact, even the very ugly, sparse covering shown here exceeds anything anyone would want to play with.



So, I put aside the problem and just published the collection as is.

But lately, I've acquired more knowledge of and facility with Substance Designer, so I thought to take up the problem again. Among other virtues, Substance Designer makes materials through "procedural" methods, meaning that it should be possible to make a material that "generates" thousands of leaves, twigs, blossoms and branches, all with variable coloring, placement, scale, and so forth. If I had a "substance" like this, it could be packaged with this particular building, and the end user could create as many different coverages of the building as he or she desired.

So, triplanar projection, and displacement , and "painting" are not issues germane to my self-imposed task.

The images of the coffee pot setting that I posted here are just "target" images for my endeavor. I'd earlier modeled out the various climbing rose vines. I am trying to make a material that approaches the general quality of the rose plants in these images.

At this point in time, I've concluded that I really can't get the depth in the rose blossoms needed for close-up images like these, although I can create sufficient depth and variety for the leaves, branches and twigs. So, for an image like the above, I still must provide the fully modelled climbing rose plants - "hero plants" as these are called. But, since the building that is the subject of the exercise is intended for middle ground and background, it is well worth the effort to create one or more "substances" for climbing roses, wisteria and plants of this sort.

In technical terms, there are some issues in that the procedures for creating a branching plant structure in Substance Designer (or any other program) requires some interesting recursion functions. Having messed around with "trees" since the beginning of computer graphics, these are familiar issues to me - just tricky to recreate in Substance Designer. But two people much smarter than I have given me appropriate "nodes" to work with, and I'm well on the way to achieving what I desire. When I have something worth looking at, I'll post images of the final plants here.

[Mark Bell]

A quick clarification so that I don't lead you or anyone else astray...

Substance Painter is a program for painting on 3d objects, as the title suggests. (Not used in the image above.)

Substance Designer is a program for creating materials and shaders. There are a lot of reasons Maxwell users should be learning it, but my primary reason here is because materials made with it can have user-adjustable roughness, normal and height maps that provide amounts of detail no modeller in their right mind would attempt to build into a 3d model. (Because the polygon count would be excessive.) Here are three examples. In the first picture, the vase is just a dead simple cylinder with a very low number of polygons. In the second image, the old gold metal roses bookcover is just a single plane with one polygon. In the third image, thetwo plants I'm playing with just laid on top of a single polygon plane.

Almost as importantly, once I have these materials, they are "user-adjustable" as I stated. That is, I can make the "icicles" in the glass vase be as few or as many as I want - maybe ten different permutations for ten different cylinders as a collection of vases. I can make the gold metal for the book cover be shiny new - or tarnished old or anything in between. I can lay the roses out in any configuration I want, as well as mess around with the border edges. I can raise or lower the embossing to any degree I want for the render. For the plants in the last image, I have a single material whose parameters I can adjust to make a different version of the plant every time I click a button. The plants can have more or less leaves, be taller or shorter, have more or less branches , greater or lesser density, more or less leaf color variation and so forth. So, I can build myself a single material. Apply it to five flat polygons spaced along the wall, and make it look like five different plants, all with just the click of one or two buttons. These plants are not going to look "flat" - they'll have some depth to them. They'll catch light and shadows just fine. And, I can do this all within the rendering engine - "live" - so I can tweak these things until I get exactly the look I want. (Would that I was a better "artist" though, instead of just a simple model-builder! Sigh!)





These are the reasons I've come to believe that almost all of us Maxwell Render folks should be acquiring a basic knowledge of Substance Designer.

Wow - these look fantastic for a 'painted' surface. I don't see why you couldn't replicate this level of detail on plant life?

[Mark Bell]

These are the reasons I've come to believe that almost all of us Maxwell Render folks should be acquiring a basic knowledge of Substance Designer.

Well, the glass vase or pewter floor can be easily done with a real-scale material and a cylindrical UV projector for the vase and a planar UV projector for the floor. For more realism, use a displacement instead of a normal map. No UV-unwrapping or other software needed. Industrial, product or packaging designers either buy the necessary textures, or simply draw them in Adobe Illustrator, if the textures are also part of the design task. For architecture visualisation materials, things might be quite different of course. I am still waiting for Arroway or another texture vendor cracking the so far insurmountable solid wood texture problem, where also Keyshot's procedural wood fails ; )

Happy rendering in 2022 to all 12 users

This suits me - keep the number of softwares to a minimum and learn how to use what you have better. I think with product and industrial design when the scene is close up on the object there's a greater need for the real life detail - joins, fillets etc. to produce a realistic image which isn't as necessary when doing exterior buildings.

[Forester]

Heartily agree with two of your points.

Close-up views of an object require precise model-building!

Also agree with the point that it is better to try to master a few versatile softwares than to chase every newly arrived application.

But, ... also it is prudent to recognize when technological advances take place that allow for greater control, greater photorealism, ... if that is the objective, or greater ease of use for the product visualization specialist.

For example, in Andreas' image of the floor cart above, we should not yet be satisfied with the image if it is meant to be a commercial product render.

How is it that there are medium-scale scratches on the metal parts, but no edge wear on any of the metal, plastic or wooden objects?

Not even very light edge wear on the edges of the wheel brackets, and the bolt heads are 100% shiny new. The rubber wheels are nicely done. But, also no edge wear or variability in the plywood piece and no variability or edge wear in the white plastic wheel mounts between the metal brackets and the plywood. Sort of unnaturally clean.

If this is meant to be a product promotional image, perhaps some different, more subtle surfacing should be given to the metal pieces than the medium-scale scratches. Otherwise we have a logical (visual) contradiction - scratches on the flat surfaces of the metal - but no scratching where you would most expect it. So, a different, more subtle surface imperfection would improve the image, as would natural variability in the plywood edges and in the white, machined plastic plates.

Adding these touches can be done with traditional tools, such as Photoshop, maybe in post-processing of the render. A good master of these Photoshop surely would be able to add these subtle imperfections to make the image less "CG clean" and more realistic. I certainly agree with this idea.

I think the whole reason that new applications come into being, from time to time, is to make our jobs easier and our work better. So, every once in a while, we have to break down and invest in learning a new application. Knowing when the investment has to be made is the tricky part, of course. Here is a softwear that allows you to have extremely precise control over surface imperfections so that you can create imperfections so subtle that you almost can't see them.

[Forester]

Personally, I don't really care if people want to avoid Substance Designer. I'm simply grateful, on this New Year, for the warm and helpful community of fellow Maxwell Render folks, and for the continual work of the devs on the Maxwell Render application. Easily my favorite rendering engine.

....../
But, I did spend three months preparing tutorials for people in the product visualization field who will need to learn Substance Designer sooner or later.
Actually, the truth is that I spent three months because I chose to work on "glass" which is the most problematic of all the possible "substances".

But, anyhow, if you guys, Mark and Andreas, want to explore this further, here are private links for you to download all the materials in the commercial product tutorial set that I made a couple of months ago. Any Maxwell Render forum fellow is welcome to these. This tutorial set and its accompanying tools was written for a much wider audience, so a bunch of this won't be particularly relevant to Maxwell Render users. You won't need the material on Autodesk Max or Maya (Arnold, Corona, Redshift, Octane rendering engines) or Thea Renderer, for example. This package deals with glass, but everything in the tutorials applies to any kind of material - metals, plastics, stone, etc.

These tutorials are aimed at absolute beginners - people who've never worked with Substance Designer. The first thing that you should take a look at, however, is the first of the "background papers." Skimming over the basics of glass as a material, you should maybe start with the small section on "why you should care about Substance Designer."

http://www.expandingwave.com/clientdown ... ncepts.zip

Two papers in the set above. One is an overview and the second deals with getting from Substance Designer into the rendering engine. As claimed above, you should start with the first paper, and then just flip through the tutorials. Don't come back to the rendering engines paper until you've completed the tutorials. (As a side note, for the rendering engines paper, I only briefly covered the limitations that Maxwell has in handling Substance Designer. I'll probably write a longer, more explicit "letter" to the Maxwell devs describing the issues here. Most of the limitations can be rather easily corrected with relatively simple code changes. But, if the current limitations seem difficult to any of you, I've already prepared a step-by-step paper on how to handle these issues. Just haven't found anything good to do with that last paper. If anyone needs a step-by-step on how to get glass into Maxwell Render, I'll set up a download to this last little paper.)

http://www.expandingwave.com/clientdown ... orials.zip

There is an overview paper describing the contents of the six tutorials - trying to give a reader some idea of what to expect. But, I stress-tested the tutorials on a couple of volunteers - they seem to get most people comfortable with Substance Designer in just a few days.

All illustrations for the tutorials and everything else were made in Maxwell Render.

http://www.expandingwave.com/clientdown ... Ibrary.zip

This is a collection of "tools" for Subtance Designer, and the key item that made this a commercial product, rather than a free tutorial set. Not needed if you are not going to actually work with Substance Designer or glass materials.

http://www.expandingwave.com/clientdown ... ibrary.zip

Completed examples of everything made in the tutorials, and then a few extras thrown in.

http://www.expandingwave.com/clientdown ... isions.zip

A bunch of people who bought the commercial product asked for information about how to convert existing glass materials that had been published by Allegorethmic/Adobe. into glasses they could actually use. So, this is just a bunch of examples of how to modify some of those materials.

So, enjoy! Any comments on the contents are very much welcome.

[Andreas Hopf]

How is it that there are medium-scale scratches on the metal parts, but no edge wear on any of the metal, plastic or wooden objects?

Because the client didn't want any

Industrial, product, transportation and packaging design is not like architectural visualisation, product visualisation, or painting textures for game assets. Very differnt sectors of work. Plus, in design, you are talking about billings in 15 minute increments and deadlines like "please get it done for yesterday". Take something like the Adidas NMD_R1 SCHUH, which upon release came in 26 flavours. No edge wear or ultra subtle things needed that the consumers can't see on their smartphone screens. In design, you need two things: good non-Keyshot cheesyness and above all as much speed as you can possibly muster. And Friday, some hip-hopper comes with yet another shoe idea, in holographic transparent gold and oily teal blue mock-crocodile leather

[Forester]

Fair enough!

Was not making a criticism of the utility of this particular render. Knowing you, this would have been made in this way for very good reason. You always do good work, and it is always exactly suited to the task. Not questioning the quality of your work at all. (And you've made an excellent and laugh-making answer to my rhetorical question! )

Just using borrowing your render as an example of what might be still be needed in a final product visualization for geenral use.

[Andreas Hopf]

Oh, no, critique is always welcome and necessary. What I'm trying to get at is that the usage context (purpose of the image, budget/time available) for renderings matters a lot, as "archviz" and "prodviz" seem to me quite different from what designers in today's high-pressure environments need, and often, the very different usage contexts of certain rendering workflows are not mentioned, and that can confuse particularly the novice, who sees "awesome" but has no clue how long someone actually worked on a render, often as in "archviz" and "prodviz" likely using more than just CAD plus one renderer. Bloom from glare, for example, is much easier done with a few clicks in under a minute in Photoshop, compared with trying to maxwell or keyshot it.