R21 in one month...we'll know


#21

That sounds doable, but it would not be a huge benefit as the C4D node system would only act as very basic framework to connect A to B - not having significant functionality by itself. Given that the C4D node system is currently separate and different from the XPresso node system, it would not differ much from using a proprietary plugin/external render node system in the first place.

I guess real benefits can only be achieved (long-term) if all nodal systems including XPresso are unified, and all nodes become interoperable on a functional level - a C4D noise node could feed a Redshift texture node, a Cycles color node could be patched into an XPresso vector math node, etc. That way, external renderers really would have to implement only nodes specific to their own engine, and reuse all C4D material and Xpresso nodes that already exist. But the abstraction framework needed to implement this would cause a huge overhead for any practical calculation, so I’m not sure it would be practical. (Not to mention that a render engine may decide to compile its materials internally into some quick-access form, which would be impossible if it had to deal with nodes not its own and go through the framework for any calculatory access.)

What’s more, even if the framework would be completely integrated, we’d still have engine-specific nodes, which means that the great allure of a unified material system fails to materialize. My dream would be that materials become interchangeable (or at least automatically transferable) between render engines, so we would not have to build a material all over if we decide to employ a different engine, and we could take materials from any library to cover our current needs. (I would never expect the final look of the engines to be the same, so switching the render engine within one project / one animation / one distributed render will never be feasible - but a reuse of materials between engines at least doesn’t seem impossible given the fact that most engines follow the same mathematical/physical principles nowadays.)


#22

My dream would be that materials become interchangeable (or at least automatically transferable) between render engines,

That’s the whole purpose of MDL (Material definition language), which Maxon licensed to Nvidia in 2016.

It’s a layer of abstraction that allows the same material definition to work and look the same in different solutions, whether OpenGL or offline renderer, like Substance designer/painter, Vray, IRay and more.

I always assumed the idea was to make one material in C4D work the same in all its renderers (standard, physical, Prorender etc…). If Redshift supports it in the future, then it could be a step closer to what you’re dreaming of.

However, in the meantime,ILM’s MaterialX has been released as opensource and more or less seems to have the same goal in mind… Maybe this is the future…


#23

I agree it would be pretty cool if all these developers truly standardized on a set of nomenclature and functions defining all material types that might be used across apps but it’s a long-shot and in any case won’t happen anytime soon. Somewhat analogous to the web browser situation we have. Over time more and more browsers adopt more and more similar W3C standards but each of them always has a few tags they don’t support and/or proprietary tags other browsers don’t support (mostly MS browsers in the latter case, historically). So the best designers can do is use only tags supported by all browsers.

We might get a scenario like that over the next few years if the app developers all adopt something like MDL (or MaterialX), as Eric noted.


#24
  • You technically could bake it out that way, but if youre using a 3rd party renderer, you’d need to create a separate C4D material, load in the shader, bring in the bake tag, cache an image sequence, and then load the image sequence back into your other material. A vertex map cache could simply be a tab on the Vertex Map that could do everything with one button click, and have it all embedded in the project file or as a separate file. Significantly easier.

  • A fields shader could be used on procedural objects - which is a huge part of the standard C4D workflow. Plus, vertex maps are reliant on the density of the mesh - if a shading effect requires a vertex map, you often need to dramatically increase the resolution to make the shading smooth enough. A fields shader would eliminate that. Also, vertex map based shaders limit you to one material per object; things get really wonky when you try to use that same material on a completely different mesh at the same time - even if it has an identically named vertex map.


#25

Yeah you’re right - those two features would help a lot.

Someone else has to have another stab as to what else will be released! I am ashamed to admit that I like these threads and this one can’t die already!


#26

One can only wonder about improved modeling tools. The volume modeling is new and awesome, but they need to do a bit more with standard modeling, esp. with symmetry and something more modern and sophisticated than soft selections.

I’d love for them to give us an option for various modeling tools to display overlay tooltips with measurements and visual alignment guides.

Fields are so rich already…but they beg to go deeper.

More thoughts soon…


#27

Cogwheel spline 2.0


#28

10+ Things I’d like…that haven’t been talked about:

-Expand limit of polygon count for plane (nice for displacement)

-Move some of the Maxon Labs tools into the main program:
especially: glTF export, Solidify, PyParametric Tools

-Add a double bend deformer (or add option within existing bend deformer)
AEScripts has a tool to do this in c4d right now.

-Add axis options for deformers that currently lack it/need it (like Bend, Twist, etc) Hate manually rotating to get the right orientation.

-Make high-res displacement more performant (better multi-threading or GPU acceleration)

-Add multi-strip option to Polygon Pen tool

-Kill or consolidate current explosion deformers

-Provide a way to create python with node-code- chunks. Include useful node chunk templates for drag and drop use in xPresso

-Allow keyboard shortcuts to be mapped both globally and per layout. For instance you could over-ride a shortcut on a certain layout.

-Biggie: Split code where appropriate for Mac and PC. Tune independently for Metal on Mac and Vulkan or DirectX for PC. This is only way to get superior performance. OpenGL’s future is dubious.


#29

My gut on what gets updated:

  • bodypaint (finally).
  • Uv tools

What I would like to see but probably won’t:

  • Viewport enhancements. It’s great now, but keep pushing it for those small, short projects that are perfect for it. More post FX, volumetric lights, glows, etc. Blender Eevee has me drooling.

#30

I’m pretty sure Bodypaint is being worked on, but there are better tools available right now for UVing and painting. Kind of think that boat has sailed for a lot of people.


#31

They just have to do something that is easy to use/learn, with that most will not bother changing applications.

Will be great if they can go whole span from tracking to scene reconstruction , texture extraction from footage/camera all coupled with easy texturing/UV . I mean is a whole pipeline in one go.


#32

1 click into Rizom…click save and I’m right back in c4d. No importing/exporting or extra steps. Load time and return time are negligble. In fact it might be faster than switching layouts back and forth.

But yes, Maxon needs to finally get UV crap straightened out for the community.

As DM said above…for me and others…ship has sailed. Happy with c4d/rizzom combo.


#33

1.Facial rigging features
2. Xpresso 2.0 “more like the node mat look”
3. Better animation tools

I really am waiting for these


#34

So let’s discuss.
-xPresso…r.u. mainly wanting upgraded UI?
-Better animation tools… can you elaborate? What’s your wish?


#35

Yeah, unless they can build a UV editor better than Rizom, and 3D painting better than Substance, it’ll just be another chunk of the app I pay for but never use.

I’d rather they focused on things like the volume builder with native VDB rendering, better modelling tools, a faster viewport (of course) and maybe revisiting some the old modules that haven’t seen any love for a decade, like Hair, Bullet physics and Pyrocluster.


#36

I wouldn’t say the viewport is “great” until the long-standing object management issue has been resolved, but they’ve improved certain aspects of it for sure.

Blender 2.8 in general is looking pretty fantastic and Eevee is a huge part of that. Overall usability of that app is going to improve quite a bit though. While it won’t be C4D-level intuitive, certainly could be argued though that it will be easier to use than Maya or other apps known for painful learning curves. And at the price of free… something all of Blender’s competitors should take seriously when considering their own roadmaps.

I’m cautiously optimistic each MAXON release R21-R23ish will have a “big deal” aspect to it, due to the new core, new materials system, new render tech they’re working with, etc. Even if the big stuff is overhauling existing workflow components, it will still be a big deal to many users.

R21 will be an interesting release year for anyone who gutted out the “new Mac Pro” wait, depending on how “metal-centric” some of the new features or workflows are. Apple has now finally provided the right hardware path (or at a minimum a much better one than relying on a “Thunderbolt cluster”), and at the same time put the price completely out of reach for anyone who isn’t a big studio boss or who is OK financing a computer like it’s a car. Because $5999 isn’t the base, not really. The base is going to start at “what’s it cost to replace the joke RX 580 card with one of the infinity fabric cards” and the answer is probably going to be around $3000. So base is going to be closer to $8999. Start stacking in more RAM, SSD capacity, etc, call it $10-12K for a typical config people will want, probably more like $20K for a really tricked out machine.

What are people supposed to do with that? That’s literally a car. lol So the PC thing is still in play. For those who waited, we can’t be sure of the best course until you see how deeply integrated Metal becomes into our favorite app. So the waiting game continues potentially. Maybe the thing will be to buy a relatively cheap PC now (still better than the rest of Apple’s offerings) and wait to see if Apple gets hammered on its pricing. If few people buy the 2019 Mac Pro besides studio bosses they may well drop the base $1000 and drop the BTO upgrade costs $x00 per item in a year… but I doubt it.


#37

Yep, this upgrade is certainly going cost me, but I haven’t spent any money on Apple desktops in over nine years, so I can stomach it.

Now Metal seems to be the only game in town, I hope devs (and Apple) really get stuck in and devote time to accelerating apps and using the power of the AMD GPUs. With more cores/threads, faster chips, better, multiple GPUs and a faster overall acrhtecture, I’m looking forward to moving on from my trusty (but increasingly crusty) old Mac Pro. I hope by the end of the year I’ll even be rendering with Octane again, or maybe Redshift.

The Mac Pro has served me well but really can’t keep up with what the software is throwing at it, with things like X-Particles, RealFlow, Forester, huge cityscape models and terrains from World Creator – not to mention just plain ol’ rendering. I’m now really excited about the next six months!


#38

My last car did not cost as much.

If 20K is the word, that’s a quite nice PC workstation plus a five node renderfarm in the basement. Or if you go for GPU rendering with multiple graphic cards, it’s three tricked-out PCs.

That’s some price tag for the Mac feeling.


#39

C4D animation tools are quite good as long as you are animating

motion graphics or any other non bipedal
in animate object.(vehicles etc)

Not wishing for myself , as I have long ago built my CA pipeline with

external tools.

Thus My hypothetical wish ,for those patient blokes who will maintain

their MSA’s, would be,at a minimum: ,

A human IK control rig to retarget motion to native C4D rigs
Like Maya and Iclone.

A motion clip system that can align the root of disparate sourced mocap

without having to bother with some bloody manual pivot object.
Like Iclone, motionbuilder, and the FREE Daz studio’s nonlinear

motion system

And with all due respect, buying an emulator from Insyduim to run the
old Character tools literally salvaged from a Deceased mans
“hackintosh” (Cactus Dan) , not a viable solution for many.

A dynamic cloth solver that works for clothing on Characters
Like Iclone Maya ,Daz studio , even the vestigial Poser.

Some manner of native facial mocap or audio based lipsync option that

DOES NOT require purchasing a $900 USD IPhone
( exceeding one years MSA cost),
Like Maya,Iclone ,Daz studio even the vestigial Poser.

With C4Ds’ Python API ,it would not be too hard to cobble a script to
parse The .dat files from the free open source Papagayo lipsynch utility
Like Iclone,Blender

Even Daz studio has Papagayo as one of four different audio based
options.

(Sample)
https://drive.google.com/file/d/1ME6qMhLjfVCnGcuHIZPrrlQwKNP5WvQ7/view.

I know many will say: “it is better to handkey your lipsync anyway.
those audio lipsyncs are all horrible”

Indeed solutions Like Faceware would be ideal

However 3 “good seconds” of hand keyed lipsync animation per
week ,for one man operations with impatient clients ,is not viable,
Thus an automated solution ,as a foundation upon which to build, is a

must in Mid 2019.


#40

I’ve barely dabbled in character animation, so I can’t speak to your points 3DD. As more of a mograph animation guy…I would like temporal F-curves like we have in After Effects.

I’ve found myself increasingly using Signal for this purpose, which works OK.

The timeline is pretty good but needs some modernizing IMO.