Jump to content

Maxwell 1.0 Release is pushed back again


Devin Johnston
 Share

Recommended Posts

I think the thing that you guys are missing is that offices, like the one that I work at, won't hesitate to buy a slow rendering engine, with a render farm, if anyone in the office with three days working experience can get good images. We make ALOT of changes/variations and need to express them in our visuals. The best way to do that is to have an incorporated visual workflow: in house.

 

As soon as we incorporate modeling programs like Revit we won't need consultants. I'd say this is five years off, but this is the appeal, and long term intent, of the software. Lower overhead so we can reduce costs for our clients and be more competitive. Have big VR renderings for every client meeting.

 

If anything freelancers should be BEGGING Maxwell for a feature rich buggy release that makes interns cry. Crappy software with a big learning curve makes for good business. Architects pay for your knowledge of how to make it work.

 

-Joe

Link to comment
Share on other sites

  • Replies 464
  • Created
  • Last Reply

Top Posters In This Topic

Thanks Joe, I was actually working on a post that was saying the same thing, but I got a little long-winded and was trying to pair it down. ;-) This is why I see NextLimit being bought out be Autodesk eventually. Once the render times are reasonable (through core optimizations and increased cpu speeds).

 

But that's only going to help you guys in the design process, which for my firm is very little of the work we do. For finished marketing images, you'll still need a trained artist to work from photos, satellite maps, etc. to place the building in its environment and produce a visually pleasing image.

 

Meanwhile, once the CAD jockeys have gotten to that level, we'll be onto realtime visualization hardcore! I don't ever see engines like this taking work away from us. Rather the technology improvements will boost our industry to the next level.

 

Shaun

Link to comment
Share on other sites

Thanks Joe, I was actually working on a post that was saying the same thing, but I got a little long-winded and was trying to pair it down. ;-) This is why I see NextLimit being bought out be Autodesk eventually. Once the render times are reasonable (through core optimizations and increased cpu speeds).

 

I really don't think that will happen. Autodesk will buy MentalRay long before they buy anything Maxwell. They are investing a lot of time and money in the support of MR which is now embeded for free in Max and Maya. They don't want to start over with speculative software like Maxwell. They would buy razil or Vray long before Maxwell, and they are not making any moves to buy any of those by a long shot.

Link to comment
Share on other sites

Autodesk will buy MentalRay long before they buy anything Maxwell...They would buy razil or Vray long before Maxwell, and they are not making any moves to buy any of those by a long shot.

 

Maybe they would want it for the same reason they wanted Lightscape--murder. Why build up their products only to lose business to some upstarts. If Autodesk was really interested in developing accurate-minded light propagation software they could have kept working on Lightscape. For whatever reason they chose to dump it, along with whatever it cost them. They probably more than recovered their loss with added sales of Viz and Max. But they did not kill Revit when they bought that, so there's no telling. When is there ever?

Link to comment
Share on other sites

Reading what juan said in the maxwell forum (that they reached a dead end and created the engine from scratch) make me think.

 

Was "bad" buffos again correct?

When i was shouting that its not normalt o wait 6 months in an almost complete engine to see some progress, i was trying to explicitly get that answer. This was my suspicion from the start.

 

The funny think is that i dont mind, and if i was told, do you want to wait or get a refund, and was explained that they will do this and that, i would have said that i would wait.

Honest deals make good friends.

Instead they closed curtains, they paniced, and tried to save the boat.

I dont know if they will save it in the end.

 

For now, the engine we have is below any expectations. All pictures are badly lit, we dont know if the initial problems (call me dielectics, were solved and so on)

 

If you want my opinion (based only on intuition) i think this approach is going to work. Why?

The answer lies to the material definition system.

They created a unified approach, and the key is the attenuation parameter. It will work nicely with the sampling path system (if one knows how the integration scheme works he will understand)

Infact with the approach they chose they could even create nonlinear optics effects.

 

Another plus is that once ONE material works all will work.

I dont know about speed , but i think that the material system, although not a complicated one, will be a reference for future programs.

I would like to remind everyone , that simple solutions are great solutions.

 

Although their behaviour to me is ****** (almost XX emails and no answer) i believe that scientifically they are on the correct track.

Link to comment
Share on other sites

Ughh, Chris! You got me there. I forget how heavily ADME is sinking itself into Mental Ray -- didn't I hear Autodesk is gonna license it as the renderer for ADT now? In any case, it would be a waste of the investment they've already made in MR, but if Maxwell's ease of use comes to fruition it may be too tempting. Push button rendering is exactly what Autodesk's architectural design software is missing, and Mental Ray is a little complicated for the casual user. Perhaps they'll just write a simpler interface with built-in presets rather than make another costly acquisition.

 

But then again, they do have more money than God.

 

Shaun

Link to comment
Share on other sites

Hey Shaun,

just noticed that you're from Portsmouth. My brother lives in Kittery Maine so on vacations I'm out and about there. Also noticed that you guys did the renderings for those Condos being built downtown. Portsmouth is such a quaint new england coastal town... I don't know how I feel about developers getting a chunk of downtown. All those cool little shops are going to be snatched up by Starbucks pretty soon...

 

-Joe

Link to comment
Share on other sites

Hey Joe!

 

Portsmouth's a great little "city," and as someone who's lived his whole (albeit short) life in northern New England, it's one of the only places I know that still has a vital downtown atmosphere. 95% of the shops are locally owned, or at the very worst a tri-state franchise location. There are two four and five story condo/retail/office/hotel buildings going up now that we've worked on -- one I think is the most massive structure in the city to date. You know how tiny the downtown is, so two buildings are a massive change.

 

The nice thing is the city has mandated that the new buildings have relatively tiny retail spaces and many of them -- it discourages corporate franchises and keeps rents affordable for local shop owners. We're pretty crabby New Englanders, and it can be a bitch getting anything "exciting" built here architecturally, but the community is marvelous.

 

I took these photos out my office window. Really can't beat the view. I drive home through Kittery every night (I live in Kennebunk, ME) -- I'll give a shout to your brother!

 

Shaun

Link to comment
Share on other sites

Maybe they would want it for the same reason they wanted Lightscape--murder. Why build up their products only to lose business to some upstarts. If Autodesk was really interested in developing accurate-minded light propagation software they could have kept working on Lightscape. For whatever reason they chose to dump it, along with whatever it cost them. They probably more than recovered their loss with added sales of Viz and Max. But they did not kill Revit when they bought that, so there's no telling. When is there ever?

 

I didn't mean to say that Autodesk has any intention of anything. What I am saying is that the Maxwell guys are trying to create something to get bought out. That is why they are so stuck on the 'it's new', 'it's different', 'it's photon based' issues. They are trying to create something with capitalization capability. That is why they are trying to do it differently. That is why it doesn't matter how they treat the customers or how badly they slip. The customer is just an infill between the product and the buyout.

 

Of maybe, just maybe, writing a ground up rendering engine is totally insanely difficult!!! Maybe they are just really having a hard time getting it to work!??!!! ;)

Link to comment
Share on other sites

In this era of industry heavyweights like the Microsofts and the Autodesks, I think the best anyone starting out a new piece of software can hope for is an eventual buyout.

 

Anyone have any idea how many guys NextLimit has working on the Maxwell engine? I know nothing about programming, but the VRay team seems incredibly small for the power and popularity of the software. Maybe NL needs to get themselves a Vlado? That man's a superhero!

 

Thanks for the compliments on the photos, guys. ;-) I need to get my act together and revamp my website so I have a place to put all this stuff.

 

Shaun

Link to comment
Share on other sites

Anyone have any idea how many guys NextLimit has working on the Maxwell engine? I know nothing about programming, but the VRay team seems incredibly small for the power and popularity of the software. Maybe NL needs to get themselves a Vlado? That man's a superhero!

 

As far as I know, they are 23 people at Next Limit. Don't know if it's Maxwell dev team only, Maxwell + Realflow dev teams, or the entire company...

 

Concerning Vlado and Vray and bla bla bla, everybody seems to forget (or ignore ?) that the reason Maxwell was 'unbiased' from the begining was for scientific purpose, not image quality. Image quality is just a consequence, not the cause.

The scientifically correct approach of the calculus makes it slow. I don't think they need Vlado or anyone else. No matter how bad they are at PR and project planning, I think they at least deserve a bit more respect on the programming side...

 

Bash them on the lateness, on the atrocious communication, on the GUI bugs, etc...

But please let's stop the poor comparison with Vray. It's never been supposed to be a direct contender.

They could have choose to make a "classical" fast GI renderer, but they did not. And it's not because of incompetency on the programming side :rolleyes:

They sticked to the unbiased approach to keep their "mathematical results" true and scientifically useable. And it's slow.

I've learnt that fact a bit too late, and I would have waited for Vray for Maya if I had knew it from the start. I need a production tool, not a geeky analyser.

But... once again... it's their approach and their goal.

They are not stupid n00bs in need of a Superhero to show them the way.

Link to comment
Share on other sites

As far as I know, they are 23 people at Next Limit. Don't know if it's Maxwell dev team only, Maxwell + Realflow dev teams, or the entire company...

Concerning Vlado and Vray and bla bla bla, everybody seems to forget (or ignore ?) that the reason Maxwell was 'unbiased' from the begining was for scientific purpose, not image quality. Image quality is just a consequence, not the cause.

The scientifically correct approach of the calculus makes it slow. I don't think they need Vlado or anyone else. No matter how bad they are at PR and project planning, I think they at least deserve a bit more respect on the programming side...

Bash them on the lateness, on the atrocious communication, on the GUI bugs, etc...

But please let's stop the poor comparison with Vray. It's never been supposed to be a direct contender.

They could have choose to make a "classical" fast GI renderer, but they did not. And it's not because of incompetency on the programming side :rolleyes:

They sticked to the unbiased approach to keep their "mathematical results" true and scientifically useable. And it's slow.

I've learnt that fact a bit too late, and I would have waited for Vray for Maya if I had knew it from the start. I need a production tool, not a geeky analyser.

But... once again... it's their approach and their goal.

They are not stupid n00bs in need of a Superhero to show them the way.

 

I've never read (or heard) that they designed maxwell as a tool for scientific applications. If it were so, they wouldn't have targeted maxwell to high end 3d applications like max and maya, and they would've included scientific analysis tools since the beginning (like lighting analysis). They have stated that maxwell could be used in the future for scientific lighting analysis through mxi files.

Link to comment
Share on other sites

Concerning Vlado and Vray and bla bla bla, everybody seems to forget (or ignore ?) that the reason Maxwell was 'unbiased' from the begining was for scientific purpose, not image quality. Image quality is just a consequence, not the cause.

I really have trouble swallowing that statement.

 

if maxwell was originally intended for 'scientific purpose', then there is a serious flaw at the core of the entire operation.

 

http://mywebpages.comcast.net/ThomasAnag/Misc/Maxwell/Prism02.jpg

 

that is obviously not correct, and although it may be a 'neat' image, it is obviously of little scientific value. there are other, better, faster renderers out there better suited to scientific application.

Link to comment
Share on other sites

My understanding is that Maxwell was developed using scientific principles of wave energy propagation to more accurately depict the way light truly behaves. It's also true that several Universities assisted Next Limit in the development of Maxwell, however I don't think it was ever intended to scientific use, I'm not even sure what that would entail other than to identify lumens on surfaces. Maxwell is a scientifically correct, or physically correct render engine taking a completely different approach to generating an image, as a result the calculations take a long time to complete. In my opinion it is a direct competitor to Vray, Final Render, Mental Ray and any other high end engines you care to include in that list. If it weren’t why have Vray and Final Render implemented a Physical Sky module, there trying to catch up. I have no doubt that they hope to be bought out at some point but I doubt any company would do that before they are able to complete their first full version.

Link to comment
Share on other sites

Well, on the Nextlimit homepage, at least the old one, it was mentioned that maxwell was intended also for use by defense contractors (eg. military), eg., simulation, etc...

 

You all have valid points but I think NL's intentions lie somewhere in between being a scientific application, or a purely cg one, and one of the problems in this rc release is that one gets the sense that they want it to be both, at the same time, which is pretty difficult to achieve. Assuming that the majority of the forum users are cg artists, some improvised polls on the forums showed that at least 80% of the users had very little interest in the patchwork complexities of the maxwell stand alone, or 'studio.'

 

The new material system (the main culprit at this point for the core hang ups) is another good way of demonstrating how NL is shooting itself in foot try to make a single app targeted for two very different markets (cg/scientific) - the new material system (bsdf+coatings) is quite innovative, and when working, should create some way-out stuff seen very rarely up to date, but on the other hand it can be way too complicated for an artist who wants to concentrate on content, not fight with nanometer attenuation and complex ior measurements such as al2or.ior or tasi2-b.ior. At a certain point, it can get a bit too technical for the non-scientists out there.

 

NL seems to be missing a clear focus on where to go with this thing...

Link to comment
Share on other sites

Mhhh. Don't get me wrong, i like Maxwells quality as well - but at the same time i know modelling the screws in wooden furniture would be more physical correct, but i just don't do it ;)

 

I understand your point, why create an engine that's physically correct when products like Vray can make images of almost the same quality without the speed issues. I suppose it's the idea of being able to create an image that is as accurate as can possibly be and not having to sacrifice quality to do it. It's the glass is half empty or half full argument, which one is correct I don't know, it's up to the artist to decide which tool fits their situation best.

 

it can get a bit too technical for the non-scientists out there.

NL seems to be missing a clear focus on where to go with this thing...

 

I completely agree with you on this one, I think one of the big selling points of Maxwell pre RC was that it was so simple. I'm very concerned that they are replacing the complexity of setting up lighting with materials so now you have to spend hours tweaking materials instead of lights. I'm not sure which one is better but again this Maxwell Studio is not something I asked for or wanted when I purchased Maxwell.

Link to comment
Share on other sites

What I understand, based on tons of discussions in the maxwell forum, is that you will have the option to keep your current workflow, assigning simple materials and rendering from your modelling app, or export the mxs file, and tweak it further in the studio. Also they said that you will be able to apply mxs pre-made complex materials within your main app, through a special material, "mxs material" and render from there.

Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
 Share


×
×
  • Create New...