Jump to content

Cinema R25 Release


Guest Igor

Recommended Posts

It's been very interesting following this thread for the last few weeks and reading all the different and passionate points of view, so I thought I'd throw my two cents in.

 

Speaking frankly I'm pretty confused by the current direction of C4D. Working procedurally can definitely be a cool and efficient way to work, but Maxon's effort here seems to be mostly reactionary to get on board with what they perceive to be the new fad. IMO a procedural workflow really shines for dynamics and effects, and since C4D has no fluid system and not a much of a dynamics or particle system to speak of, I find little value in the procedural workflow.

The idea of trying to compete with Houdini doesn't make much sense to me, since it will likely be many many years before C4D can even be considered a contender, and in the meantime C4D is trailing it's ACTUAL competitors in so many areas.

Maya is already the artist friendly answer to effects with fantastic capabilities, while Blender is a great FREE alternative with an impressive host of features to boast. Neither software is perfect, but each one has areas where they excel and take no prisoners, such as Maya's amazing uv capabilities, or it's native renderer Arnold, or other fantastic features like Bifrost, Xgen, and Mash etc.

Blender has really taken leaps and bounds to compete with Zbrush, and though Zbrush is the undefeated champion in terms of sheer polygon power, Blender has developed a rich sculpting program, and should be considered a viable alternative. Blender even beat Zbrush to the punch with their cloth brushes, although Pixologic quickly introduced this feature with highly impressive results, just like any smart developer would do to keep it's rep. Blender also has a realtime renderer, which is great for lookdev and maybe even final renders.

 

This realization really hit home for me with a recent project; I finally gave up on C4D's neglected hair system and reintroduced myself to Xgen with the intention of using Arnold's proprietary USD system to render out of C4DtoA, when I had the sudden epiphany that Maya SHIPS with Arnold allowing me to forgo all that export hassle and render out of Arnold directly using Maya. Despite it's clunky and disorganized interface, I find myself moving to Maya more and more, which is ironic because C4D's interface is far superior to Maya's and that's one of the few 'features' Maxon chose to mess with.

 

I could ramble on and on, but the basic point is that C4D has fast become a master at none, and really doesn't have a specialty area anymore. In my recent projects  I have found myself using C4D less and less, opting to use other packages throughout the pipeline that are much more efficient and capable than C4D.

I am confused by Maxon's decisions and wonder if they themselves know what their long-term goals are, but I guess time will tell.

 

Thanks for reading 😀.

Link to comment
4 hours ago, Matches said:

Blender has developed a rich sculpting program, and should be considered a viable alternative. Blender even beat Zbrush to the punch with their cloth brushes

 

I think the blender community is going to really miss Pablo Dobarro (he developed the sculpting and painting tools). And by the sounds of it, it was the community itself that drove him to quit.

Link to comment
2 hours ago, Matches said:

since C4D has no fluid system and not a much of a dynamics or particle system to speak of, I find little value in the procedural workflow.

 

The unspoken assumption is that they're working to eventually put all those into the app - redo dynamics, update the particles, maybe even do fluids if they're feeling generous. But none of those were possible in the way that they would have liked with the old core, so they decided to do a new one. It's hard to see the new core and a decade worth of efforts towards it being designed to present nicer caps and bevels or more cloner spheres in future GSG tutorials. Ditto that object tracking, motion tracking from several years back.

 

Maxon stuck Pyrocluster into the app way back in the day (before many University age students were born in fact) so someone eons ago must have once liked the idea of effects in C4D. And Stu Maschwitz, now running a bunch of stuff at Maxon, is also an FX guy. But I don't see those cool things happening till they've done all their core stuff. Which is maybe 2023 or 2024? Who knows.

 

Chad Ashley's long sigh near the end of his R25 recap maybe reflects your thoughts. But whatever. They'll get there eventually. 

 

Blender's 'everyone gets their say/everyone wants to be heard/users speak directly to the devs/devs have to justify everything they do each fortnight in public' system seems a double edged sword if it leads to stuff like Pablo packing it in. But Blender is cool.

 

 

Link to comment
4 hours ago, kbar said:

 

I think the blender community is going to really miss Pablo Dobarro (he developed the sculpting and painting tools). And by the sounds of it, it was the community itself that drove him to quit.

 

I prefer to quote Pablo himself on a fundamental issue in the "expectations management process" (whole thread)

 

 

Also, "the community" appreciates Pablo's efforts; it seems to be the usual vocal minority that criticizes Pablo for implementing certain tools instead of fundamental speedups. (While I appreciate speedups, I doubt that this was ever meant to be the focus of Pablo's work, and I also doubt that the loudmouthed demanders have any clue about how difficult and far-reaching changes in the critical foundation of a program actually are.)

 

Now let's continue criticizing Maxon for the same thing 💥🦊☠️😇

Link to comment

The issue with Pablo was more deep. He had no enough knowledge of coding to manage the whole sculpting. I think it was more a problem of Blender Foundation resources than from the users. Pablo was not a coder or developer, he is an artist that went to code to improve the tools he works with.

 

Link to comment

If the foundation of the Blender sculpting needs to be changed because it can't handle "enough" polys or is not "fast" enough*, then this is clearly a much more difficult issue than adding tools on the UI level. Changes in data storage / handling speed / parallelity / GPU usage need to be addressed by specialized software architects who know their craft, and it will definitely have repercussions on all Blender areas (unless it's far more modular than I believe it is).

 

I don't know the actual Blender code (no, I'm not going to review a few million code lines now...) but it seems more like C4D's core problem, or LightWave's core problem. Change the deepest fundament, and it takes quite a while until all levels of the program have caught up with it...

 

...not to mention that someone needs to start the work on that ground level in the first place.

 

* always taking into account who wants to do what with it...

Link to comment
19 minutes ago, Pinin said:

The issue with Pablo was more deep. He had no enough knowledge of coding to manage the whole sculpting. I think it was more a problem of Blender Foundation resources than from the users. Pablo was not a coder or developer, he is an artist that went to code to improve the tools he works with.

 

We have quite a few coders at Maxon that come with an artist background and they are treasured. You need both, the hardcore experts on computing structures, efficiency and algorithms and those that include an artists view. Prime example might be one of the developers that laid the foundations of MoGraph.

Link to comment
10 minutes ago, Pinin said:

The issue with Pablo was more deep. He had no enough knowledge of coding to manage the whole sculpting. I think it was more a problem of Blender Foundation resources than from the users. Pablo was not a coder or developer, he is an artist that went to code to improve the tools he works with.

 

The most powerful combination in software development is when a user is the developer.  Only with that combination will you get the tools that are both needed and work the way you need them to work.

 

As companies grow, that pairing of abilities can be lost.  If you read the job descriptions, you find more of a focus on a candidates proficiency in the core language (C++, python, etc) and who capable of transforming technical papers into code.  Rarely do you see the requirement of being a 3D artist in the job description.  Some companies like Insydium, overcome that gap by hiring a "resident artist" such as Mario Tran Phuc who pushes the platform to ever greater and greater capabilities. That works!

 

Not sure if Maxon has a "resident artist" on the payroll.  The closest I have seen that happen was probably in R24 as most of its new features were tailored the needs of artists like Beeple (eg, the ability to grab stuff from a library pretty efficiently and scatter them all over your work in kitbash fashion).  But I am not sure if Beeple is on the Maxon payroll (nor does he need to be).

 

Going back to Matches post, he has hit the nose square on the head:  How is Maxon competing?  Is it even competing?  I loved this example he provided on Blenders sculpting tools vs. Z-Brushes:

 

Quote

Blender even beat Zbrush to the punch with their cloth brushes, although Pixologic quickly introduced this feature with highly impressive results, just like any smart developer would do to keep it's rep. 

 

So when was the last time you saw Maxon quickly compete with a competitor on a new feature?  Maybe with Maya on the addition of dynamic MoGraph modifier in R22 when Maya was adding fluids to their motion graphics suite of tools.  Unfortunately Maxon was only able to implement particles effects while Maya was doing fluids.  So points to Maya on this round.

 

I know we all think the new "core" will bring a vast treasure trove of riches to C4D once "fully" implemented...but that transition is now looking like a 10 year journey and that is just too long in this world.    Sorry, but better bevel capability should not be the only benefit we should be seeing from this new core at this point.  I know there are some viewport improvements, but it seems that you need to be using Scene Nodes to appreciate that benefit.  If Scene nodes is the ONLY way we will see the full manifestation of the new core, then that is a HUGE problem because it totally changes the one thing that keep users tied to C4D: the way we interact with the program.  Right, wrong or indifferent, Scene nodes changes that interaction.  Capsules restore a good deal of that "ease of use", but I am not seeing the increase in viewport performance -- especially with the greeble modifier.  Increase the polygon count on the source object and watch C4D slow to a crawl.  That is not how I thought the new core would behave.

 

Honestly, the fastest growing development at Maxon is the license server (IMHO).  Certainly more changes there than on C4D's ability to catch-up on features to the competition.  Not including Redshift as the default renderer for C4D but cancelling its perpetual licensing is a one example of what is a priority to Maxon: growing  revenue via subscriptions with their existing user base rather than growing the user base with new features to the software.

 

This priority does indicate that Maxon is not even trying to compete on features.   This should not be surprise to anyone as it has been the #1 complaint from the user community.  Don't even try to defend it because the weight of evidence is against you (and this thread is long enough).

 

Dave

 

 

23 minutes ago, srek said:

We have quite a few coders at Maxon that come with an artist background and they are treasured. You need both, the hardcore experts on computing structures, efficiency and algorithms and those that include an artists view. Prime example might be one of the developers that laid the foundations of MoGraph.

Wasn't that Per Anders?  And didn't he leave?

 

Dave

Sorry...but I simply do not have enough faith to be an atheist.

Link to comment
11 hours ago, Cairyn said:

If the foundation of the Blender sculpting needs to be changed because it can't handle "enough" polys or is not "fast" enough*, then this is clearly a much more difficult issue than adding tools on the UI level.

 

The issue with creating a sculpting system in a program that is built around the traditional concept of polygons with points, where all that data needs to live on the graphics card, is that you soon hit limitations on the Graphics card itself. To achieve millions of polygons you need to take a fundamentally different approach to how the data is handled... which is what ZBrush did.  The limitations of doing that is that you no longer have traditional vertex data, so you can no longer do all the other work that you would normally associate with a 3D animation package. Can't animate moving 3D points if the concept of that XYZ point no longer exists. So for most DCC apps  they won't ever reach the "100 million" polygons scale for sculpting unless they effectively break it away from all other systems and has its own separate viewport, tools and workflows. Then to get it back into the DCC app it would need to be baked down to make it usable.

 

The benefit of having a sculpting system in a DCC, even with the limitation of maybe 4 to 6 million polys, is that those same tools can work with everything else in that DCC app as well, even on low polygon models. The tools can be completely integrated because it can use all the exact same data as everything else. This is why the C4D Sculpting tools can work on both high resolution "Sculpt Tag" based model as well as a regular Polygon Object directly in the viewport, or on a Pose Morph.

 

A dedicated sculpting tool like Zbrush will always exist and be the master at sculpting at incredibly high detail. Because they can focus on doing just that and not have to think about any other workflow.

Link to comment
14 hours ago, Pinin said:

 Pablo was not a coder or developer

 

So what was he doing when he was working on the Blender code for a year or two. Gardening?

 

If he can jump into the code of a 3D software app and improve it to pretty much universal acclaim, he's probably more of a coder than most people. That said if he walked because people were bugging him to do more than he was able or wanted to do, that's fair enough. 

 

Maxon could tell all their developers to hold regular public meetings about their work, plans, schedule and productivity, and make them go sit in the naughty chair without breakfast if Fred Random from Reddit was unhappy with what he was seeing, but I dunno if it would help to be honest.

Link to comment
2 hours ago, kbar said:

 

The issue with creating a sculpting system in a program that is built around the traditional concept of polygons with points, where all that data needs to live on the graphics card, is that you soon hit limitations on the Graphics card itself. To achieve millions of polygons you need to take a fundamentally different approach to how the data is handled... which is what ZBrush did.  The limitations of doing that is that you no longer have traditional vertex data, so you can no longer do all the other work that you would normally associate with a 3D animation package. Can't animate moving 3D points if the concept of that XYZ point no longer exists. So for most DCC apps  they won't ever reach the "100 million" polygons scale for sculpting unless they effectively break it away from all other systems and has its own separate viewport, tools and workflows. Then to get it back into the DCC app it would need to be baked down to make it usable.

 

The benefit of having a sculpting system in a DCC, even with the limitation of maybe 4 to 6 million polys, is that those same tools can work with everything else in that DCC app as well, even on low polygon models. The tools can be completely integrated because it can use all the exact same data as everything else. This is why the C4D Sculpting tools can work on both high resolution "Sculpt Tag" based model as well as a regular Polygon Object directly in the viewport, or on a Pose Morph.

 

While doing this you limit the Sculpt tools to basic brush deforming.

 

The benefit you mention is actually not true, because it's in a single direction: you can indeed use the Sculpt tools on every polygon objects, but you cannot use standard cinema 4D tools on "Sculpt tag" objects - like deformers, booleans, generators or modeling tools. It's absolutely not an "integrated" solution. A Cinema 4D "Sculpt tag" object is in fact not that different from a Zbrush model, it must be converted to make it usable.

 

I personally don't care about traditional vertex data. I want to sculpt whatever I want with (almost) infinite details. I don't mind doing it in a separate window and exporting it. It such an effortless automatic process.

 

For me what makes Cinema 4D's Sculpt obsolete is:

 

1. The density limitation. 4-6 millions might seem a lot, but it's insufficient to add small details on a model (like scales on a lizard or wrinkles on skin). By today's standards 5 millions polygons is no longer "high polygon".

 

2. The lack of dynamic tessellation (adaptive topology). This is why I stopped using Cinema 4D's Sculpt. Dynamic tessellation allows me to be truly creative, to start from any shape, to add or remove geometry, to increase details locally, etc... Without it Cinema 4D's Sculpt is just a limited low performance brush deformer. 

Blender, Mudbox or 3D Coat all offer dynamic tessellation.

 

 

 

Link to comment
×
×
  • Create New...

Copyright Core 4D © 2023 Powered by Invision Community