[Maximus3D]

I just watched the presentation video, sorry i missed the liveshow because i was busy.

From what i could see it looks like a very nice release, lots of new goodies in there. Procedurals are most welcome, so is the fog and all those external thingy's. I'm missing one examplerender i would like to see and that's one featuring a slit experiment shooting a lightbeam through a prism with the fog activated. Will this produce visible spectrum colourbeams through the atmosphere or not. I know Thea is able to do this and visible laserbeams with the help of it's atmosphere thingy.

Keep up the good work guys!

/ Magnus

[feynman]

Many nice features, but - no GPU rendering or other means of speed improvements for designers doing car, product and packaging stills? That's really a bummer. Just buying faster CPUs all the time is no answer. I hope at least the network render feature is more reliable now.

[Mihai]

Personally I don't see how much slower Maxwell is compared to other renderers in those markets. I mean compared to how long it takes you to make a final, good quality image. From start to finish. Everybody always compares render time only. I see a lot of pointless "tests" on the web, people being proud of how fast x renderer did something and you see the image and can't help but ask, so? Who's going to buy that? How long actually do you have to spend until you make that image final?

Especially now you have instant FIRE feedback when moving objects, custom alphas, exclusive lights, boolean objects...so there are numerous new features to save you time. And considering everything that Maxwell is calculating, and all the features it supports, I would certainly not say it's much slower than what's out there already if you try to reach the same quality level.

I can understand the people doing interiors only, this takes time in Maxwell. Then again, you don't have to worry that your final render will have splotches or other interpolation artifacts, and that "quick" 3 hour render time means re-tweaking, re-rendering for 3 hours. While you could have been working on another project.

[feynman]

Sure. Lots of useful things for non-architects and non-animators here. All very well, no doubt. But clients grow more demanding by the year, breathing down your neck with comments on "slow" software usage : )

Anyway, one can always plonk a faster processor in the network until V3 becomes available to us...

What makes designers happy (who enjoyed these features already way back in 1992 by way of Alias, etc.) is procedural textures, exclusive lights and fast material presets that can then be fine tuned in a more advanced way. Especially the latter feature will be a boon to these hundreds of novice users I meet in workshop who just can't be bothered with too many settings, used to the one-click-culture that prevails today ; )

[Mihai]

Well it was almost obligatory that we designed a simpler to use material creation approach. Even though render guys tell us they have no problem understanding our system compared to other material systems out there - the people who aren't so familiar with rendering tell us, pretty much without exception, that the most difficult part they find with Maxwell is creating materials. So yes I think for them they will find these material types really nice, and much more sophisticated than the current Wizards. Which are also modal...you click, create...want to tweak something, you have to start over.

[feynman]

Yes. Ease of use functions are very critical. After all, hundreds of novices in the field of architecture and design will eventually end up in studios, taking their experience with them, spreading opinions, which is why they're a critical marketing factor. It's very much like who is the influencer. As in cars, where it's by and large the kids influencing parental decision-making or in BYOD where now management forces the IT department to allow tablets at the office and "just make the damn things work" : )

[photomg1]

Mihai ,
I wished I'd said it before but I thought you did a great job on your presentation for v3 . I've also grown to like very much next limit as a company (there are not many I would put in the same bracket) . I've seen some of the grief next limit has had on the forums(ernesto springs to mind,sorry ernesto!) and think all of you handle it very well , much better than I could thats for sure. I also believe you really are trying to put the best product out there that you can.

I'm not trying to add to the grief level, believe me . What concerns me is that there are not many paying clients who will pay to wait for the quality you are talking about. Thats the reality I have experienced firsthand unfortunately . I've seen a twenty strong studio who have tested maxwell , and rejected it not on quality but on speed(they loved the quality). The majority would rather go for a "it's good enough" approach from my experience to save time.

That is the harsh reality,and that is what we are competing against.I just dont see things changing much for maxwell till speed is managed better.

I'm making my own choices and trying to stick with maxwell but when you are staring at a 24-60 hour render for a print res file I do wonder what I'm doing.Can't keep relying on render farms its just not always financially viable. I was really holding out for possibly gpu rendering support on your roadmap as its just a much more financially viable approach for a freelancer than building a mini render farm. I'm still hoping next limit is exploring it even though its not in the new release!

It would have helped some of us greatly if you had offered us the option as Martin Brinks had mentioned to be able to limit certain factors of maxwell for us to gain a speed up if we needed it ( bounces etc) . It would have just made maxwell much more practical in the same way that you are now offering light linking.

Anyway roll on v3!

[Mihai]

60 hours? No, I would not wait that long to render an image for a paying job. What is the usual resolution you render at for print size?

Regarding render farms, I believe the usual large size print render would go for about 200 euros on Ranch or Rebus? I hope most of your clients can accept that as the cost for a large size print and you get paid at least 1000 for such an image, plus that 200 for the render farm. Otherwise I don't see how anyone can make a living doing this. Keeping in mind taxes, software and hardware costs, rent, and yeah food. At least not in the western world.

[Asmithey]

I use the Ranch mainly, sometimes Rebus, for all my renderings. It really is not that expensive. When you do your proposal add a little in for rendering out put. My rendering costs on the ranch, for exteriors, usually run between 25 to 50 us dollars. Interiors usually between 60 to 120 us dollars. Exteriors usually take 15 to 30 mins. Interiors usually take 30 to 50 mins. I do not think I have gone over an hour yet.

My image size is always 2400 x 3600.

It is well worth it I render to 20 or 22 sl every time for interiors.

Farms are they way to go. The Ranch by far is the best.

Aaron

[photomg1]

At least A3 at 300 dpi preferably 6k by 4k. Its not like you will ever just do one version of a render , it will go through iterations that will soon add up. Personally I've never found it easy for example to judge texturing on a smaller version for example.

I did try and pre-empt in my post the fact that renderfarm's was going to get farmed out as a reply .... that as a reply I think is missing the point, even more so speaking about how much should be charged to justify the use of render farm is even more deeply missing the point.

I have never gone over 8 hrs with a biased engine usually its a lot less.That is the point , there are alternatives and plenty of them and a few of them will give " its good enough" result for the majority . Why is vray so prevalent for example.

It doesn't matter how many new features are added. If you don't make strides addressing the speed issue it will always be the number one problem for maxwell render gaining wider adoption.

[feynman]

In industrial, product and packaging design, renders are a necessary by-product, so common that one can not charge for them explicitly, because many are needed for one's own presentation purposes to sell a concept and the cost cannot be absorbed so easily, especially on pitches. Therefore, render farms are usually out of the question for most studios, and I know over a hundred in various countries in Europe. The "please can you do it for yesterday" and "good enough, that'll do" paradigms, whether we like it or not, rule; and so "dumbing down" options like photomg1 suggests would be welcomed for non-architectural and non-animation users of MR.

[Polyxo]

In industrial, product and packaging design, renders are a necessary by-product, so common that one can not charge for them explicitly

Yeah, I also see a huge difference in workflow between Visualization Specialists and Industrial Design work.
Feynman has put that nicely - there's a constant need for "in process" renders which are very important for internal decision taking. Such renderings may be used try out
surface finishes, combination of colours, placement of labels etc. The option to create such imagery on individual workstations is crucial. I agree that using renderfarms
or even network rendering weren't very helful in such contexts.

No Viz-Specialist would send his scene off to a Farm or kick of a Network rendering when still amids tweaking the Scene composition, light and materials - that would not
make sense, right? Using these powerful but static systems only makes sense when all aspects of the Design of the Sujet in question are considered final, as well as all other
components of the composition. Well, renderings in Product development stages often depict something which is equally unfinished...
Rendering in Product Design in contrast to Viz-Contractors work often does not take place in the end of a development cycle: It rather equals snapshoting highly volatile
prototypes. And in this different scenario – again for reasons of work economy – rendering should take place on the machine of individual Designers as these store the full
editing history.

[Fernando Tella]

I vote too for a "dumbing down" option. It could be as simple as a "realism slider". When you lower it, it starts cutting bounces or other options. You would know you were speeding up the render at the cost of quality, but it could come handy in some situations.
Maybe an advanced panel where you could specify exactly what you want to spare would also be useful.

[Fernando Tella]

Could those 2 sided materials be used to make windows with glass as seen from inside and ghost from outside? It would let the ligth come in easier (as if there were no glass) and also have the look and reflectiveness of glass from inside.

[Bubbaloo]

You guys who are complaining about speed are buying a Ferrari and complaining about it getting bad gas mileage.