[Dmitriy Berdnichenko]

Here is one more interior test in Maxwell. This one is render out much faster around 2 hours with denoiser.

[feynman]

I love Maxwell, I do, but every time I face stuff like this I keep thinking Arnold might be the better option.

I wish Next Limit were acquired by Autodesk and then given the financial resources to address all pending issues like speed, bug response, etc. to save us in industrial design from the awful "keyshotisation" and trend to quick-fix ugly CGI.

Alias, Catia, Rhino, NX, Creo, SolidWorks - what are industrial designers, a huge global user group, that need top of the line images to do? Is Arnold the way to go? What is just not commercially feasible is to switch and re-learn a new render software every 1-2 years.

[LadleSky10]

Well . . . business decisions are none of my business but, I'm not so sure an acquisition will get your intended result.

Has anyone heard of an app for the iPad called "123D Sculpt+" ? That was a fantastic, simple to use, powerful little sculpting app that I used all the time. It increased my sculpting productivity ten-fold. Imagine the ability to do the rough draft start of a sculpt while reclining on your sofa, sitting at the airport or on a plane (truly mobile sculpting) and then export said sculpt to your desktop sculpting program of choice to finish it up. Don't tell Autodesk, but I'd have paid as much as $200 for that app, probably even $300 --that's how much it increased my productivity.

Well, that app was discontinued and according to the website it was because they were spreading themselves too thin and had to discontinue all their 123D apps. Personally, I think the better choice would have been to include the app (and its continued development) as a freebie for Mudbox users rather than completely getting rid of it.


So, like I said, not so sure an acquisition will get the result you're expecting. (Heard of Mental Ray?)

[feynman]

It none of my business of course, but the alienation of tens of thousands of potential users sounds either foolish or could be a typical sign of cash flow problems at NextLimit. Hard to say since the company is not publicly listed and no balance sheet or cash flow statement are available.

[LadleSky10]

I feel that my attempt at self-deprecating humor might have gone awry. My apologies.

It was merely a recognition that I have no information whatsoever regarding the behind-the-scenes business decisions here.

1. It could be an attempt to reposition for a potential acquisition.
2. It could be a better valuation of the benefits offered by the software.
3. It could be the result of strains from trying to maintain compatibility with an assortment of ever-changing plugins.
4. It could be any or all of the above.

So, the humor was a poor attempt to balance my lack of information with an urge to comment on a phrase that I've seen often in forums: "I wish Autodesk would acquire" . . . and each time the resulting thought in my head is: "are you sure that's what you want?"

[feynman]

Well, it is obvious the company is underfunded, for whatever reason; recently the Rhino plug-in guy left after only three months and dropping SolidWorks support (while after many years also not offering any Alias, Catia, NX and Creo plug-in, but that's another topic), is worrisome. Until the late V3 version, everything was quite ok (apart from the dreadfully slow speed) by using MR Studio and our Rhino and SolidWorks users were happy, too.

It could well be that NextLimit is only concerned with architectural visualisation (and even there, it is in a small niche), and if that is the case, the company should maybe say so, so industrial design studios can change to another renderer, having waited forever for some competitive speed increase, for example. I like the whole concept and quality of MR, but maybe it is warranted to migrate all 50 installations to another program rather soon, and then hoping that that one will prevail for some time.

There is nothing worse than pestering a large user base with heaving to learn a new renderer every two years. In architectural visualisation, maybe that is the norm, I don't know.

[LadleSky10]

Yes, you are right.

5. It could be inadequate funding.
6. It could be pushing a major feature update out of the gate too soon in response to user or investor demands.
7. It could be a long-term vision strategy.

Another company made a major change not too long ago (probably calculated in the loss of some customers) and a large swath of its user base revolted with great hostility. That company made a few minor tweaks, stuck to its long-term vision and today that company is in a much better position. I'm referring to Netflix.

At any rate, I certainly have no clue regarding the business decisions here but I trust they're well thought out.

And with that, I'd like to take the time to thank the developers and all the people who put great effort toward making what is already an excellent product even better.

[Max]

My humble opinion is that Maxwell Render got cut out from rendering market due to its nature.
We live in a world where now rendering means speed, live preview, instant feedback, PBR, interactive. click click click, done.

Maxwell doesnt really suit anymore into this mechanism sadly. Every other render engine out there is extremely fast for every purpose, its very well integrated and have alot of features.
Dont get me wrong, i love maxwell and always will, but the thin line between physically accurate and pleasing image is way too blurred nowadays.

You need a very well trained eye to spot some rendering difference between a maxwell render and any other engine out there...momst of the clients dont care, its just about speed speed speed.
And who gets the job done faster.

I dont think maxwell will ever change because its built in a way that is impossible to compete with these other engines. they would have to rewrite the whole code. Reason why its userbase and visibility is quite shrinking.
I might be wrong but thats just my opinion.

How long has been since 4 release, and honestly the engine is still the same, GPU has more issues that benefits, how many people use it for professional work and reliable workflow in GPU?
Because i dont even bother, i just use cPU and thats it.

Just an example, the render on top of this page, from Dmitriy Berdnichenk, it took 2 hours + denoising.
Just scroll trough evermotion.org and see how many of that quality or even better are rendered in 20 minutes with Corona or Vray or even Redshift, Fstorm, Octane.

I feel like Maxwell Render ended up like Arion.

[choo-chee]

Maxwell doesnt really suit anymore into this mechanism sadly. Every other render engine out there is extremely fast for every purpose, its very well integrated and have alot of features.
Dont get me wrong, i love maxwell and always will, but the thin line between physically accurate and pleasing image is way too blurred nowadays.

well I think MR is kind of "click click click" as it's very easy to understand with the physical workflow. and they have a very good product after all, esp. with the denoiser. they just need to find a way to undo the very very very poor mistake they did with the new licensing system and poor gpu support, and they can be back on track...

[LadleSky10]

There! Well, that's something I'd like to explore. What exactly is it that you find problematic with the new licensing system?

a. Is it the pricing model?
b. Is it the separation of the plugins?
c. Is it the handling of nodes?
d. Is it something else entirely?

You see, I've only been a user since the tail end of version 3 so I'm not as attached to the prior licensing system as most. But, I am curious to know more.

(I should probably mention that I think treating each plugin as a separate piece of software is a good business decision. I also like the stepped pricing model for the various plugins --also a good business decision. The GPU issues, well, I think that's just a result of pushing it out too soon. Patience will likely solve that problem.)

It would be great to hear other opinions and, as well, to have some clarity on the issues.

[choo-chee]

I'm using MR since version 1.0. It was very nice to have the option to use maxwell's plugins as a "homemade FBX" fromat , and collaborate work from all sort of 3d apps easily. You see, veteran customers (such as me) got used to having the option to use maxwell in any way they needed, not limited to one specific 3d app (at least, without paying extra...). So many of use feel that dev is too slow (took lots of time for a decent denoiser to arrive and don't mention the GPU rendering being useless) and that upgrading actually took something away...

[feynman]

Just an example, the render on top of this page, from Dmitriy Berdnichenk, it took 2 hours + denoising. Just scroll trough evermotion.org and see how many of that quality or even better are rendered in 20 minutes with Corona or Vray or even Redshift, Fstorm, Octane

If you had to migrate around 50 MR users (product/industrial design with Alias, Creo, SolidWorks and occasionally Rhino), which of the ones you mention would you consider worthy pursuing?

[LadleSky10]

It sounds to me as if the separation of the plugins is the main issue here.

So, would the following option be a viable solution?

What if you could purchase a special version of Maxwell Studio (i.e. a bit more money than the regular version of Studio) that comes with all of the previous plugins including the discontinued plugins?

The only caveat here would be that these plugins would merely act as "dynamic links" that open up Maxwell Studio allowing you to render your model and, of course, save the Maxwell file. You would have the full version of Studio but there would be no integration of the plugins with your 3d Software whatsoever (you would have to separately purchase the fully integrated version of a given plugin for that).

Because it's a "dynamic link" to your model it will update as you make changes to your model. Any materials that you apply in Studio will, of course, update as well.

Does this seem like a reasonable solution? (Please know that I'm merely a user, a creative solutions type of guy , and I have no affiliation with the company nor do I know anything about its business decisions.)

[Max]

Just an example, the render on top of this page, from Dmitriy Berdnichenk, it took 2 hours + denoising. Just scroll trough evermotion.org and see how many of that quality or even better are rendered in 20 minutes with Corona or Vray or even Redshift, Fstorm, Octane

If you had to migrate around 50 MR users (product/industrial design with Alias, Creo, SolidWorks and occasionally Rhino), which of the ones you mention would you consider worthy pursuing?

I use Redshift and never looked back. Its extremely fast and really good quality, multi gpu support, great integration ( i use softimage but its well integrated in 3ds max, maya houdini c4d). Its also cheap has multi gpu and out of core technology meaning your rendering and scene management is not limited by your GPU ram, which is crazy good when you understand what this means.

[feynman]

Redshift does not work with Alias, Creo, SolidWorks, NX or Rhino (no OBJ import/export workflow according to the sales flyer PDF), so for product and industrial designers, it's unfortunately not an option. That leaves Arnold, Corona and Octane to avoid the dreaded Keyshot/Vray.