Leaderboard

I have a lot of thoughts about this, most of which I have already shared with Maxon. The few bugs that remain aside I am mostly happy with the current SDS modelling toolset, but there are lots of things they could add to it, and few I hope they will improve about it. 1. Smooth edges needs an 'equal spacing' option, so we can turn that bit off when we don't need it, as we often don't ! 2. The brush tools and their smoothing / relax modes need a flood button that applies a set level of brush function to just the selected polys, effectively replicating what HB modelling Bundles's HB_relax used to do. We need to remove the human, brush-based element from these operations and apply these things exactly evenly sometimes. It is noted that Iron tool does similar things, but lacks curvature awareness, so this is mainly needed for the relax mode of the brush. 3. We also need an equivalent of HB_Round edge that works in the way that did, which is not the same as the way smooth edges works. We need both, and both integrated ! 4. Last one from HBMB - the whole start retopo setup and project mesh buttons. With some limitations these tools do still work in latest Cinema, but they really should be native by now. 5. My oldest ongoing fix request - make UV mapping work with OpenSubDiv. It still doesn't since its initial integration back in R17 or so - we get weird SDS distortion that is unsolvable. 6. I want Mirror back in the r-click menu ! 7. Normal Move / Scale / Rotate work on all components now but only appear in menu for polys modelling ops. 8. Q needs to only operate SDS and not any other generator. 9. One sided Bridge patch tool needs improvement so it can patch huge gaps around corners without bulging out / ruining its usabiility. 10. As always, I'd like a boole tool that performs more reliably in more difficult (animated particularly) circumstances, and with a workflow that doesn't involve having to combine operands under nulls or connects at any point... 11. Select similar. 12. Straighten Edges also needs a make planar button set. (Flatten does this, but would be nice to have within SE tool too) 12b. I sometimes miss a dedicated 'Make Planar' button for all component selections, outside the Flatten tool. Whilst I like that tool, nothing quite beats Max's system where we make a selection and can hit a single button to instantly make it planar to any axis and in any coordinate system. We can do it with zero scaling of course, so it's more a small QoL request. 13. Axis in Normal Mode needs to orient more helpfully to single edge selections and retain changes in Axis controls needs to be more predictable and robust. 14. Slide tool needs to not fight polygon snap, as it can do sometimes. 15. Line Cut needs to work more reliably, not insert the extra points it sometimes does, and actually make cuts on any visible line it touches. Currently it misses far too many that have to be repeated. 16. Although its algorithm (particularly in advanced mode) works very well, and much better than anything we had before) I remain unconvinced that Thicken works in the best way it could. With the current system of a mere percentage between inner and outer thickness it is impossible to type specific values in each, or to get specified differing thicknesses inner and outer, which is sometimes required when working with arch-vis blueprints for example... 17. Highlight a)ngons and b) tris buttons in the Mesh Checker so we don't have to piss about with Structure manager to find them ! 18. Slide tool needs negative extension beyond borders. 19. Terminate loop - function that takes a poly selection and terminates a loop within it. CBR

not necessarily in regards to the actual generator, but more in terms of the workflow. just tools that can help us design things more easily, but also allows us to still focus on making good topology easier so it can be subdivided. I've seen some things in 3dsmax that I like. for example, in 3dsmax, there is a unique type of selection known as element that allows you to select irregular cuts with quads in them: see 2:10 in this video: also another cool feature in shown in that same video at 6:10 is the ability to 'hinge from edge' allowing some type of slanted extrude. makes for interesting design using polygonal modeling tools. honestly, I was mostly inspired by the above video by arrimus because the whole premise of the video was trying to show how you can still make great designs while focusing on subdivision surface workflow... and it lead me to think that maybe, we can focus on developing tools that enhance the design process with polygonal modeling and topology in mind. I'm brainstorming my own ideas, but I made this thread to hear some other peoples thoughts and suggestions 🙂 Some of my own ideas revolve around faster ways of cutting shapes that automatically project onto the mesh as your cutting, but not doing too much crazy stuff like beveling and so you can focus more on sharpening and smoothing edges using the subdivision generator.

Better to reference reality rather than other animations. Fortunately, YouTube is a good resource for this as found here: So as you can see, this has more to do with volume builders and solid shapes than with pyro blobs. You could start with two cylinders or two spheres, add a little randomness to their basic shapes so that they are not too perfect and then place them under a volume building. Now align both shapes exactly over each other and slowly animate them moving away from each other while at the same time keying down the connection radius in the volume builder to simulate the cells reconstituting as they pull apart (keyframes are set on the overall "Objects" menu under the volume builder). Just a thought. I hope this helps. Dave

Bit of a face palm moment if you ask me. A friend sent me this earlier, seems the original listing had been taken off their site due to backlash but it’s still searchable through googgle. Full ad https://g.co/kgs/vqWR8aq

This is cool, although it’s pretty overwhelming how quickly this type of technology is progressing. It I’ll be interesting to see how long it takes for text based prompts to 3d to become useable with out considerable clean up by a human. with the ability to iterate on designs almost instantly, with clean results and minimal human involvement it I’ll be a massive turning point going forward.

Maybe it won't need clean up. (I assume you refer to polygon/uv cleanup) I think people will just tolerate the utter mess produced by AI. After all we have enough hardware power. If it's about something static and not reusable like a flyer, poster, web ad, video thumbnail, billboard ad... It won't have to be optimized... just take what the 3D AI spits and refine it with the 2D AI... whatever it takes to produce the best possible result in the less possible time... If it has to be a video animation then toss in the 3D AI model, make a quick rough environment, render the animation and use AI again to take the video and re-imagine it using a prompt best describing your desires ...

Ideally it could most likely just atomically be put through a remesher and be good enough, automatically unwrapped and textured etc. although you’re probably right in that it might not even be needed. which ever way you spin in, iterating time and changes on the fly will be implemented way faster than any human being could do, along with the fact that it can be running all day long 24/7 , 365 with out breaks, with out the need for it to be paid etc. It’s actually kind of astonishing how quickly we’re accelerating towards this. You can almost feel the corporates rubbing their hands together in anticipation. I know there’s a lot of nay sayers down playing the significance of Ai displacing peoples jobs and such, dumping it in the same boat as all the other conspiracy theorists. But I think it’s pretty clear autonomy will change things one way or another, i honestly can’t see it stopping at just being a tool to aid people.

As I see it the most benefited people will be "content creators". Already 50% of my YT channels feed use AI to create their thumbnails. It won't be long till any kind of "influencer" will eventually use a low cost/free online service to create images, videos or add elements to their recordings (face/voice/background replacement or even full-body replacement(anime characters in Japanese-friendly audience, or cartoon characters for shows intended for children)). Avatar usage is already a thing but it requires expensive non-AI-related software/hardware or service provider. In other words, any totally clueless/amateur wannabe-famous brat, granny or looser will be able to boost his channel in YT, X, Vimeo, Twitch, TikTok, OnlyFans by 500% to compete better with the already successful content creators even if their content is totally a clickbait and BS. Adobe had a substantial user percentage occupied by content creators that found the courage to make the leap of faith and learn Photoshop/AfterEffects/Premier to help them with their content instead of hiring someone to do it for them (but even hiring someone for a job like this is still considered part of that user percentage). I think Adobe saw that and was alarmed with the AI wave, and decided to fight fire with fire in order to keep part of that client percentage that felt more comfortable in Adobe's user interface than search and learn to use (again) a new app that is also still experimental/short-lived or soon-to-charge-ehen-out-of-beta. AI should not be available in the hands of the naive, unspecialized and lazy crowd. Too much freedom has been given to the average person that it has already be shown to be dangerous to the integrity of the society (hate-speech, fake news, disinformation, toxic mindsetting, brain washing and manipulation). This is actually true for all kind of technology not just AI, but with AI things will spin out of control even faster affecting more people exposed to it either young or old. We will soon see books titled "Be an incredible artist in 24 Hours !", 50 pages tops, first 4 pages explaining what online AI you can use and how to make an account, the rest just a list of words you can randomly type in the prompt to produce your thing... People will be so much monkeys that the expert edition of the book will have 100 pages with the words separated into chapters like "Decoration", "Poses", "Animals", "Plants".... because people will be so useless that even language will be hard for them to actually form a comprehensive sentence of their foggy thoughts... Oh, wait... we are already there ridiculousness in the highest order https://www.udemy.com/course/prompt-engineering-ai-course/ https://www.udemy.com/course/best-certified-chatgpt-prompt-engineering-and-ai-course/ The land of Content Creators https://img-9gag-fun.9cache.com/photo/a9q0RW6_460svav1.mp4

Anybody catch this - Turbosquid (owned by Shutterstock) is working on this tech: https://www.turbosquid.com/ai-3d-generator Its all marketing right now for the public, but I bet there is some money behind this.

Same. Been finding it overwhelming for the past few months but also frustrating not really knowing what is worth learning now that won't be made defunct down the line. I'm largely a 2D guy that has in the past used 3D sparingly but the past few months I've been going harder into 3D but every day I see something AI related that makes me wonder if it is a waste of time.

As long as they don't release any tech demo I don't have any high hopes. I've seen enough two-minute papers videos with the latest NVIDIA ML-related 3D model generation to know it's not gonna look great. That dragon on the website... totally misleading

i wouldn't say it's a waste of time, at least as a set of skills you can use for yourself. That being said i find it interesting with the hypocrisy surrounding ai "art", i've seen on several social media outlets authors and writers using ai to create cover art, story boards etc with absolutely no regard for traditional artists, whether 2d/digital or 3d having no problem with it, then throwing their toys out the pram when people use ai to write books, scripts and so on. i find it extremely ironic as someone who's workload has been affected by the writers strikes, as many of us have, which in part was due to concerns over studios using ai to displace certain roles as well as pay and other issues, that they can't see the wood for the trees when it comes to this topic as a whole. As if ai will stop at images and not progress to other areas. The naysayers downplaying the significance of things like this will change their tune when ai starts impeding on their given industry/hobby/past time etc. just today on reddit someone made a post asking about ai tools that can create 3d models from prompts in the 3d modelling subreddit, as if it's not a massive slap in the face to every member of that sub who's spent years learning said 3d modelling skills. i think what initially started off as somewhat of a gimmick which non of us really took seriously, in reality could change everything as we know it. it's not going away. People have legitimate concerns, and there's weight to those. It can be challenging as it is to get modelling work these days, a lot of it is farmed out to india, or countries in eastern europe. where labor is far cheaper. i have absolutely no doubt that if/when good enough results are achievable from things like this, it will certainly displace roles, Or alter them to more of a supervisory role, where you're tasked with any minor clean up that may be required. initially i thought it may have certain uses as a tool in some applications, but it's hard to invision where this would fit in without stepping on someone's toes down the line.

Jesus christ. This just shouts cash grab. Don't want to waste years of your life learning to 3d model? neither do we! Become a 3d modeller today! for the low low price of $59.95, you too can create professional looking 3d models. For a small additional fee of 50%, sell all your new 3d models on our marketplace and get rich, TODAY. TURBO SQUID AI, creating 3d models for you, so you don't have to.

i wouldn't be surprised if they're sinking an unreasonable amount of cash into this. The first company that pulls this off to any sort of usable degree is gonna have a licence to print money. Not necessarily purley from the 3d community, but also from people who would like to make 3d stuff but can't / won't take the time to learn how to. capitalizing on people's lazyness/and or inability.

Sometimes it all feels like all this a.i. is trageted at game developers. So that game dev can create infinite worlds, with infinite stories, with infinite characters, interactions etc.. I guess it's time to get into game dev.. 😄

Interesting quote: "AI is not a tool for the artist but a replacement of the artist" Well....I agree...partially. AI is a threat to the artist....but only to the artist being paid to create art for someone else's benefit. AI is absolutely NO threat the artist who creates art for themselves. I do foresee a threat to pre-visualizations artists in the future. If you watch the behind-the-scenes videos enough, what you hear is the Director telling the pre-viz artist what changes to make to their initial concepts. Just as easy to type those directions into the AI program that created the original image. So, concept artists, pre-viz artists, etc may be impacted. But they are talented individuals and there is still a market for original artwork created by an original artist. Of course, the first few AI images that were half-decent have fetched a high price but as more and more of that work floods the market over time, the prices will drop. Pretty soon AI art will be a commodity. So here is my prediction for the near future (or within the next 5 years): AI art (be it 2D or 3D in nature) begins to change how artists are employed by media firms. Artists who are being paid to create art for someone else will find themselves losing their positions. If a business can increase its profits by replacing 10 concept artists with one AI program with no discernable loss of quality, output or threat to delivery, they will do it. A good portion of those displaced artists (those who are in the bottom to middle tiers based on their skills) will find other jobs and leave the industry. They will now create art for themselves in the way that brings them the most pleasure. In short, they will become hobbyists. AI art continues to flood the market. Soon, the thrill of owning something created by AI will diminish. AI will become a commodity and it will only be used by producers and directors in all forms of media to help them hone their vision. There will be very little "human" created art or very little that people can definitely say "this was hand crafted". Hooray for NFT's! The top tier artists who embrace NFT and can therefore prove that their work was created by human hands will become the new rarity! Their work is what people are now clamoring to own: They key being that they can prove via an NFT that their art is original and created 100% by their hand. In short, top tier talent will always find a way to be in demand and make a living. Dave