3D-Pangel

I would beg to differ. Let's assume that he uses polygon lights and GI to put a glow into his windows or port holes. Polygon lights can light up surrounding geometry but if there is a glow to be applied that is usually a post-render affect. If all the polygon lights were painted with a texture that had the same luminance value, then all the glows would be the same regardless of their orientation to the camera. As long as the material is being rendered, it gets a glow. But the glows on his windows vary across the model. So that would imply he is using many different textures with different luminance values and is applying them individually to each window in random patterns. Again, a lot of work or pre-built into his kitbashing library. Plus the glows are well done -- not just a bloom of the base color. The colored lights actually have a white center that transitions into the color at the edges. This looks like and individual light to me and not a material with a luminance value. Dave

Well...very deep and analytical. Split flow could be a function of edge-flow in polygonal modeling where the quads flow around the perimeter of a object's face and therefore better define its edges should they be sub-divided. But I do recognize "fuzzy flow" in his work. A trick to add visual interest maybe? One other area of worthy analysis is the proper ratio of inner extrude to extrude of the selected faces. What looks appropriate for the scales you want to achieve while still being able to be picked up at render time. For example, starting with a simple cube and running the divider script on its six faces: This is a standard 78.74 inch cube (or 200 cm). Now selecting all faces and with preserve groups set to off, you inner extrude by 0.1 inch and an extrude of 0.1 inch you get: So when the groves height is 1/2 its width, you barely see it. A few more tests show the following: My overall take-away is that the deeper the grove relative to its width (or when the physical ratio of the grove's heights (F) to width (2I) or PGR is greater than 1, the more defined the overall panel becomes in the finished render even when textures are applied. Thoughts? Disagreements? There are many more ratios to consider and at many more scales to deal with in your models, but this seems to be pointing me in the right direction. Dave

BTW: A HUGE THANK YOU to ICM for diving deep into the work of Ansel Hsiao. With that said, if anyone who is reading this thread knows Ansel or contributes to the many other forums that he frequents, please let him know that we are doing a deep dive into his artistic method. Not for copying but rather out of sheer respect and admiration and to answer the simple question: How does he do this? We are, quite simply, amazed and therefore trying to learn from the master. Similar to how aspiring artists study framing, composition, and lighting from the works of Davinci, Rembrandt, Ansel Adams, etc. I do believe there is no other thread like this in the 3D world simply because the artists are mostly still alive. You want to know; you go ask them directly and hope that they answer. So, if anyone knows him....maybe we should invite him into the discussion? Point me the forum and I will ask! Dave

Great analysis into the lights. The only way I know to get a glow to one light and NOT all lights is to actually place a physical light at that point. If anyone knows how to get a glow to a luminance light using a render engine that is calculated at render time so that it only affects those textured faces pointing directly at camera (and not all polygons with that luminance texture), please let us know. AFAIK, polygon lights are only good for lighting up adjacent polygons, but all glows are a post-process effect which affect all polygon lights. Therefore, I would approach this by building a kit of pre-made light rows that include both the geometry, the luminance textures and the physical lights. These would them applied where I need them. This could explain the proportions you have uncovered (no more than 3 rows, 5% red, etc.). I can't imagine that after all that modeling, Ansel then goes back to place individual lights in hundreds of locations. Maybe he does because his models are extremely detailed and crafted with the same discipline used by artisans hundreds of years ago building giant mosaic frescos out of tiny ceramic tiles. Dave

The overall problem with images as textured is twofold: To avoid repetition, they either need to be very large in size (4K or more) so that ONLY one texture can cover the entire model or individually painted to various sections of the model with different U/V scales for each section. Ideally, they would be procedurally generated as a shader so that their patterns are naturally non-repeating. but for sci-fi this presents a problem because you want to be creating "random" n-sided geometric patterns to represent that the hull was assembled with various plates. Unfortunately, there is no geometric "tile" shader that does this as the ones built into C4D (Mod noise or cell noise) give you perfect checkerboards. Visually, you can see perfectly aligned tiles along the U and V axis. What you need to do is pass multiple cell or mod noise shaders through the fusion shader with varying scales to have any hope that there is a disruption to this pattern: As a color shader, this is not the best either. But if you add it to the reflectance and bump channel while putting a soft blistered turbulence in the color channel to take the curse off a flat single color across the entire surface, you can start to get some interesting results that might work: Here is the shader (using only C4D native shaders) for anyone who wants to expand it and make it better: Untitled 1.c4d Dave

My understanding is that her contract has a pretty expensive severance clause that Disney does not want to spend. That severance drops as the time until her contract expires approaches. So, the "word on the street" is she has until the beginning of 2023. She has been stripped of all creative control for all but the projects she started. Kenobi was one of them. There are others but they keep getting their start dates pushed out. Gee...I wonder why. Dave

I agree that Star Wars brand has taken a nosedive since being purchased by Disney. The last movie in the sequel trilogy was simply unwatchable (it took me four attempts to power through it and if not for the superb FX work, I would have given up). There has been a lot written about all the mistakes taken with the sequel trilogy from a story perspective (here is a pretty good one). Just overall bad storytelling based on poor reasoning that the path to success was to copy the original trilogy with new characters in a new setting without understanding why the original trilogy worked in the first place. So, the filmmakers really didn't understand the source material. George Lucas wrote the first 6 movies as one story and based on some hefty research into what forms our myths and legends over time. He tapped into that lore when he created the story. There was also the disastrous decision to film a trilogy without sticking to a single overall story that tied all 3 films together. Honestly, you don't pay $4 Billion for something with the intent to trash the IP in the first 3 movies. Stupidity brought them to that point....not greed. So, you have to ask, what was the Lucasfilm President, Kathleen Kennedy, thinking when she allowed all that money to be spent on a trilogy but still allow the middle movie to completely ignore the stories direction established in the first movie? Why? When that didn't work out, she brought back the first movies director to rat-con everything done in the second movie while trying to link everything together. What you get is a mess. So why did she allow the second movie to go so far off course when so much money was at stake? If you want to learn more, simply search on "Kathleen Kennedy" in YouTube -- the results will surprise you. But I do have a lot of faith in Jon Favreau and Dave Filoni (the people behind Mandalorian). They get the material, and they are master story tellers. They really excel at character development and taking a character through an arc in a very deliberate manner. Sometimes a bit too slowly for fans because it creates a bit of confusion as to where the story is going (which may explain some negative reviews for the Book of Bobba Fett) but once you see the whole story, you understand. They also understand the art of "fan service" and use it well (the ending of Mandalorian Season 2 probably saved the entire franchise). Side note: The latest Disney+ series, Obi-Wan Kenobi, was pretty bad. Even the new CEO (Bob Chapek) apologized for rushing it to the screen. To set the record straight, Dave Filoni and Jon Favreau had nothing to do with it. It was the same person behind the sequel trilogy, and Solo (the first Star Wars movie to lose money): Kathleen Kennedy. Anyone see a pattern here? Dave

Kent, Are you the main developer for JetFluids? If so, bravo! Really remarkable how far (and fast) JetFluids is evolving. Couple of questions: Have you used the externally exported temperature channel to render fire in Redshift? Any plans to have both density and temperature to be internally cached as well? Also, for the particle motions, is there a sub-step setting? Apart from the first video where you see the fluid clipping the trough (which can be solved by decreasing the sub-step size between frames). it is hard to tell how fluids are reacting to their containers as they are all invisible. I would imagine that one advantage of the grid solver approach is that the particles only exist within the domain (no errant particles flying past the domain borders). One downside is resolution of the fluid solution against a slanted collider object placed within the domain....but things look pretty smooth especially with the viscous chocolate test (in which the rudeness would have been completely missed by me had you not mentioned it. I guess twitter folk have over-sensitized "rudeness radars"....which probably explains why I avoid twitter). So, am I overstating the downside of grid-based solvers? Overall, great work! Dave

Amazing video and captures perfectly my biggest issue: understanding the design aesthetic. EC Henry understand that aesthetic better than anyone (well....maybe not Doug Chiang....but you get my point). You need to think about function and the overall world that model lives in when coming up with the detailing. He made an excellent point in that because SW was set "a long time ago", to get that point across the ship designs actually drew inspiration from old military aircrafts in our own world. Very subtle and subliminal. Bravo! And this is where some procedural modeling techniques fail (IMHO), particularly if they try to be a full service solution with limited control other than population density. Now, RandomFlow is more of a tool than a procedural system and that is why I think it is probably the best plugin for this type of work. Still would like to get a sense of what its export to C4D looks like (hint..hint). Now, I do like the Divider plugin for C4D and have started to play around with it for surfacing the outside of the Death Star. But....I am not there yet. More work to do. Here is a small test for a bump out section of the DS equatorial trench: And how it will fit into the overall model. While in the background, it is still wrong. Unfortunately, no matter how much reference material I have, there is a bit of trial and error going on because I have yet to internalize the design aesthetic - so there is trial and error. Plugins will help, but only when you know EXACTLY what you are trying to achieve. Unfortunately, I only know what doesn't look right. Dave

Honestly, I just loved the "Dutch" or maybe it was a "Hilversum" way of life. Simple things would always put a smile on my face like seeing a senior women (maybe 60+ in age judging by the greyness of her hair) riding her bike with a bag of groceries in the front basket. Or just how many people rode bikes in general. High schools were filled with bikes whereas in America kids seem to stop using bikes for general transportation as soon as they leave grade school (up to the 6th grade). Maybe it is all our real hills (steep ones too) that keep us from riding bikes. Not sure. As an engineer I loved the lock system to prevent storm surges in case of hurricanes. Just a massive engineering accomplishment. Then you hear about some of the long term planning of the Dutch. For example, my Dutch colleagues took me sailing on what I thought was a natural fresh water reservoir. Actually it was reclaimed ocean. The built a dike across a bay opening in the 1950's to both provide strategic road access to the other side of the bay as well as seal it off from the ocean. They then seeded the bottom of the bay with a plant that actually absorbed the salt. Years of harvesting that plant followed by reseeding slowly turned that bay into a fresh water reservoir. Brilliant. And it is true that Hilversum people do think of their town as hilly. While walking somewhere, I got a little turned around and asked for directions. The person said "take a left at that hill". What hill? "That hill up ahead". You mean that slight rise in the road? "Yes". Hey....I loved the topography -- flat works for me. Now, for as long as the summer days are in HIlversum, you do get the inverse in the winter time. Or what is called the "long dark days before Christmas". I have been to The Netherlands at that time of year as well where you get 4 maybe 5 hours of sunlight a day. I had to remember to step outside at lunch time to soak up the sun otherwise my whole day would be spent in darkness. But that is unavoidable and it makes this time of year such a joy! Dave

I spent a full 2 months in Hilversum for work some time ago during the May and June months. Loved every minute of it. Beautiful little town but very expensive to live in. I stayed in a local hotel but everyone else commuted. Trains were the predominant mode of transportation for commuters. There were two main trains that left at 5:30 PM and 8:00 PM. So everyone wanted to be on the 5:30 PM train which meant that work promptly ended at 5:00 PM. What made this really special is that because it was in the late spring/early summer time-frame, the sun did not set until very late and was not fully dark until 11:00 PM. So as my work day ended at 5:00 PM....I had a whole 6 hours of sunlight to enjoy. Usually, I would go for a long run....which is real easy to do in Hilversum as there are no hills. Now I am from New Hampshire so my perspective of hills is a bit different. What a Hilversum local would call a hill is no more than an enhanced speed bump to me. But there was gorgeous county, farms, dikes and scenery to explore and I would literally just keep running without being fully aware of how far I had gone or how to get back....which made things interesting and fun. Afterwards was a shower, meeting with colleagues in the hotel bar for few beers, then a walk to nice restaurant or hop on the later trains for some sightseeing. I would get back around 11 PM and fall asleep while watching the sun set. It was awesome. Beautiful country. Great people. Great food (lots of restaurants from all over the world).....and great beer! This was in the 90's so hopefully HIlversum is still the quaint, beautiful town I remember. Dave

Welcome Christoph. Glad to know I am not the only one who feels that way but rest assured that feeling only last a little while because someone always has the answer on this forum. Dave

Polligon has plugins to help load textures? Really? And in Redshift? Really? Really? I am familiar with the quality of those textures (really good stuff) and shamelessly have helped myself to the free textures in the past. But I never knew you had these tools. Plus they are for free! The tutorials are also very good. I did sign up for email notifications some years ago but never heard anything about the plugins (that I remember). Not sure how I missed this - but hopefully I am a unique case. Also, is this the same Poliigon that Andrew Price (Blender Guru) is associated with as CEO? If so, I found the following quote interesting: Nevertheless, thank you for letting me know about the C4D to Redshift texture plugin. Hopefully some future development ideas in consideration for this new role: Rather than load Poliigon textures directly from your file directory, have the plugin build a separate preview section off-line and load from that or make use of the new Asset Browser Preview capability. Anything that helps with previews prior to loading helps your workflow. Ability to preview/purchase/load directly from the Poliigon site for those people who have signed up for monthly/annual plans. Any plans to expand the texture loader capabilities to importing Poliigon models into C4D and Reshift? Very attractive set of rocks, stumps, etc. at the site. Great news and thanks for letting us (at least me) know. Dave

I actually did notice that but was just trying to make the point that simply splattering panels all over a surface is not enough. You need some randomness to create visual interest but still have an underlying logic driving it as well. But thank you for expounding on Ansel's logic beyond my quick example. He really is the master. I would still love to see one his meshes. To date, I have yet to see him post any on the forums he visits. Is his topology as beautiful as Vector's or is it all triangles? Not to say that triangles are bad for hard surface modeling but as good as Ansel is, you do have to give the nod to Vector as his topology is just amazingly good. Dave

Ditto. I mean just mind boggling impressive and gorgeous to behold. I have always wondered if there is a market amongst CG enthusiasts for a coffee table book of exceptional models and meshes. Something that you could turn to for inspiration or education as I feel there is always something to learn when studying topology of this caliber. Or you could just love the eye-candy. You start with full color page of the finished render in all its realistic glory followed by two or three pages showing the mesh in detail. Also include a call-out box on number of polygons and textures, hours spent to create it, software used, PC used, total render time, and a short bio of the artist. You see this in small snippets with magazines such as 3D World, but it would be kind of need to go deeper with a full-on hardcover coffee table book. I swear it would sell 10, maybe 15, if not 20 copies!!!! 😆 Actually, I jest. This could be along the same vein as "Sculpting a Galaxy" by Lorne Peterson (Head of the ILM model shop) but for the digital age. People value the work put into physical models, but in so doing they completely discredit the mind-numbing endurance and incredible artistry required to create digital models. "It was done on a computer! The computer did all the work". There could be a market for the first book that opens people's eyes to the work that is required. Just a thought. Dave

Greebling does not need to be limited to just mid-size objects to be successful. Again, look at the work of Ansel Hsiao. This is where excessive greebling, especially procedural greebling as produced from PolyGreeble, can get you into trouble. The majority of the detail to be added on large objects should be limited to paneling - again non uniform but following the flow of the general shape of the model or to convey a sense of how the structure was built. To that paneling, you think about specific areas for adding the right details that create a sense of scale. Scale can be achieved with port-hole windows, pipes, vents, gun placements, external ladders, tanks, etc. Again, I refer you to the master, Ansel Hsiao:

Based on the time stamps of both messages, you realize that Cerbera cooked up that design in 9 minutes....and all quads. Just another example of stream of consciousness modeling: the phenomenon where you are so in-tune with the application that you really are not aware of the steps you are taking; they just occur out of muscle memory. Similar to how a good typist is not even aware of the keyboard as the novel is being written, that is how Cerbera works with C4D's modeling tools. He just wills the polygons into existence and the object is formed. I would imagine that there are days after he has modeled the complexities of a 9-iron for a client where he wonders to himself "Hmmm....got that done quicker than I thought. OMG! Its 6 PM already!!!! Where did the day go! I completely missed lunch" ...yes...I am jealous. 😀 Dave

Wow...you nailed it perfectly!!!!! Amazing!!! 😆

The best greebling starts with looking at industrial designs in today's world. Look at the structure of oil tanks, the supports for large hangers, how piping flows in oil refineries. Probably the reason why kit-bashing works in sci-fi model building is because they are using real world shapes from model kits based on real world objects. Thus, when you see this collection of details, your mind accepts that visual language because it is based in a world you unconsciously understand. As an example, consider one of the most famous sci-fi models in the world today: You see these shapes, accept them as real but you don't immediately go "hey -- those are engine components!". So start with basic hull plating techniques but break up that area of paneling with objects modelled from real world industrial components. Also, remember that negative space (no detail at all) also works to add visual interest because it helps direct the eye. Sometimes procedural greebling across every surface creates just too much visual information and can be tiring to look at --- and thus subconsciously read as completely fake. It really is an art and has its own "form language" that actually requires a bit of study to fully understand. I am sure there are rules behind that language. Hopefully that video Iceman posted talks about some of those rules. Dave

Thought I would share this link to an Adam Savage Tested video as a means of inspiration. This video captures the classic visual language for surface paneling that we have all grown up to accept as "real" in sci-fi since 2001 A Space Odyssey. This is exactly what we want to do in CG....and hopefully in less time than it took Adam Savage to do physically (rather long at 40 minuts....really only watch the last 30 seconds to see the finished result but I will watch it tonight) Dave

I am definitely thinking about using Blender -- in fact, do a search on "modeling greebles" on YouTube and the majority of the results are Blender Tutorials! Honestly, it does appear that Blender is the tool of choice for the sci-fi community now that Max is slowly imploding on itself. Thus my interest in seeing how RandomFlow FBX output works in C4D. A simple test to get an idea. As RandomFlow is not that expensive, I could just buy it and give it a try, but maybe the whole Merk experience has left me a bit gun shy. Not sure. But thank you for the test files! I do appreciate it (expecially the Tie fighter). Dave

Ice! Thank you for this. Personally, I think RandomFlow comes closest to the look I am after. PolyGreeble is just a bit too heavy handed and (per my understanding based on the tutorials I have seen) you can't restrict its application to certain selections. Also not sure how well PolyGreeble converts to editable meshes or even if its results can be edited. I would still love to see an FBX export of RandomFlow. Dave P.S. While trying to be fair in my criticisms with PolyGreeble, in the interest of full disclosure, it should be mentioned that I have a beef with Merk. I purchased PolyGreeble via pre-release pricing offer while it was in development many years ago. All seemed good until he got into his motorcycle accident, but I patiently waited for over a year as I had sympathy for his situation. When it was finally released, I sent an email that had the original pre-release confirmation attached and asked how do I get my copy. Complete crickets. I heard nothing back. Not sure if Merk's contact information changed in that time so I tried via Patreon with similar results. So with that type of integrity and customer service, let the buyer beware not only with Merk but with anything sold via Patreon.

Wow!!! Great idea. Hopefully it has more than just two participants (I need all the help I can get). Dave

+1 for this suggestion. Also run a DxDiag command on the PC and check everything is being identified. Maybe a memory stick is not fully seated on the PC or not identified (eg. corrupt). As memory likes to exist in pairs, if one is not seated then that could throttle the data flow across the motherboard. Just spit balling here because the hardware specs of the PC exceed those of the laptop. This is a real mystery and one I would like solved as I am looking at pairing a Ryzen processor with an Nvidia card as well. I would also assume that all drivers (and I mean all and not just those of the GPU) are fully updated. Dave

Well...as a follow up to my Death Star Landing Bay Interior WIP, I am modeling the outside equatorial trench as well. Right now it is just a texture but I would like to add some geometry to it. And the model extents are huge as ships will need to added so everything is in real world scale. The bays themselves are baked down versions of the full scale WIP I did. Literally just boxes with textures rendered with GI from the larger model. Everything is instanced. So I am preparing for a high polygon count once you add the exterior service details (tower guns, hull plating, etc.). Also, the exterior surfaces have a slight curve to them that actually matches the estimated diameter of the Death Star itself (120 Km). A little too obsessive with accuracy? Not sure. So my approach is to copy groups of faces to another file. Then use RandomFlow on those faces to create the hull plating and some interesting details. There is also a good C4D plugin called Divider which may work just as well if the goal is ONLY to create asymmetrical hull plating. Separate those details from the original faces and copy them back into the original model on a separate layer so that they overlay the original outer hull. I have to be really careful with balancing the amount of detail so that you get the just the right impression of detail but not going too far such that you end up with nothing but visual noise and so many pinpoint highlights that the renders look bad GI. In short, it will take an artistic eye that I hope I have to strike that balance. Some people have a "more is better" approach when using procedurally generated greebles or hull plates. Not so. That is why I study Ansel Hsiao's work as he can strike that balance. Dave