Maxwell 1.0 Release is pushed back again

[Ernest Burden III]

there's an outside chance they just may be able to actually release a final version before 2006. (I wouldn't count on it, though)

 

It won't happen. Even if they can get what's being released now to be functional and stable, there are still all the host plugs that have been ignored since summer, not to mention all the Mac people to serve. I do not see them hitting all those posts in a few weeks.

[Gaze]

Q. How much $$ did you guys pre-pay? Was it less than the current $495? I noticed their site now states "Last Chance" for $495. Then it goes up to what?

 

Also, (forgive my newbie ignorance) what is it that is so promising about Maxwell that prompted you to pre-pay?

[Devin Johnston]

We purchased when they were selling Maxwell for about $300 per license, the reason we did it was because of the phenomenally easy setup compared to other engines, and the reality.

 

I've been messing with RC2 for a few minutes and I have to say they have done a lot of work in the last few days. This one is much more stable and I'm actually able to do some pretty quick preliminary images without any problems. The interface still needs work but it's looking better than it was this morning. I still don't think they will have it ready before the end of this year.

[Gaze]

So, after these experiences, would you recommend a new user stay away...or do you feel things will work out okay?

[Devin Johnston]

I would say to stay away until the full version has been released; God only knows how much longer we will be messing around with this RC version. Only drawback to waiting is you don't get to play, and you will be paying twice the price, but you don't take the chance of Next Limit going belly up if they can't come through with Maxwell. I believe they will but it could be up to another year before we see anything close to the final version.

[Gaze]

Thanks...I'll stay on the sidelines then. My last question is whether at twice the price, would the added realism be worth the $1K, or is it more of a $300~$500 tool.

 

Goodluck with it

[jackb602]

Thanks...I'll stay on the sidelines then. My last question is whether at twice the price, would the added realism be worth the $1K, or is it more of a $300~$500 tool.

Goodluck with it

 

Well the ENORMOUS if is wether NextLimit is able to someday deliver a product that will reliably allow users to make images of the same quality as those shown in their galleries. IF that happens Maxwell will, in my opinion, produce the most photorealistic renderings I have ever seen, hands down. If you plan to use it professionally, and have the CPU resources to use it, I think it will be well worth $1000.

 

Jack

[otacon]

Well the ENORMOUS if is wether NextLimit is able to someday deliver a product that will reliably allow users to make images of the same quality as those shown in their galleries. IF that happens Maxwell will, in my opinion, produce the most photorealistic renderings I have ever seen, hands down. If you plan to use it professionally, and have the CPU resources to use it, I think it will be well worth $1000.

Jack

 

What do you mean? The images in the maxwell gallery are made from its users.

[jackb602]

What do you mean? The images in the maxwell gallery are made from its users.

 

At the beginning (i.e. last winter), most of them were from NextLimit or people working closely with them. But my larger point is that Maxwell is not even close to being production ready. Some very patient users, generally using the more robust plugins, have produced great images, but I would never use Maxwell in its current state to deliver something to a client. I've had to bite my tongue several times over the past year when I've found myself tempted to mention the great new render engine I hope to be using soon. I already gave NextLimit my $400, but I won't give them another penny until a production ready release of Maxwell exists.

 

Jack

[manta]

Thanks...I'll stay on the sidelines then. My last question is whether at twice the price, would the added realism be worth the $1K, or is it more of a $300~$500 tool.

Goodluck with it

 

Well I think when its ready and fast enough, yes, its actually priced right in between Vray and Brazil...The only thing I'm not sure of is the licencing, whats thier current policy ?

[AJLynn]

Well I think when its ready and fast enough, yes, its actually priced right in between Vray and Brazil...The only thing I'm not sure of is the licencing, whats thier current policy ?

 

1 license per 4 CPUs (dual cores are 2 CPUs, hyperthreading P4 is 1) regardless of number of PCs - e.g. 2x dual core machines or 4x single core. Student licenses are for 2 CPUs.

[Gaze]

Are the licenses transferrable? meaning can an owner sell his license? Looks like some disgruntled early adopters are wanting their money back, maybe they want to sell if possible.

 

Mainly interested in arch viz...do you suggest I'd be better off looking elsewhere if I'm mainly focused on realism/stills at this point?

[chuck]

Gaze-

 

the beta was/is a seriousy fun program to use. the realism is there, but it still had a few issues:

 

1. a lot of noise/long render times

2. glass didnt let direct light through...a big problem for architecture.

3. clip maps (opacity) didnt work quite right.

 

it's easier to set up than vray...but it's quite a bit slower. i'm using beta again and waiting for an "RC" to be released that i can actually use. i'm not going to download every patch they send out...i need to move on for now.

 

RC is way more hassle than it's worth right now, even though i've been using beta for the past few months and know my way around it.

 

chuck

[Devin Johnston]

I have to say in defense of the RC2, it is much improved over RC1, Next Limit has fixed many of the initial bugs that were plaguing RC1. There is still much to be done but I can say it is faster than the beta was.

[Christopher Nichols]

There is still much to be done but I can say it is faster than the beta was.

 

By how much? Even a factor of 2 or 3 would not be enough. In order to be competative with Vray it seems it would need to be by a factor of 10 or 20.

[Devin Johnston]

I'm comparing the speed of RC2 to the alpha and beta, but I agree there is still much more room for improvement. I do have to point out that Vray wasn’t always as fast as it is now, this is still a piece of software in development and from what I've heard the engine still hasn't been optimized to it's fullest extent. It's also important to note that the cooperative rendering feature is going to basically remove the speed issue to those people who have access to even a small render farm.

[Christopher Nichols]

It's also important to note that the cooperative rendering feature is going to basically remove the speed issue to those people who have access to even a small render farm.

 

Really? I thought you had to pay more money for extra procs? Either way, it does not help for animation. When rendering animations, you always need a farm and you would only do it one frame = one cpu. And 12 hours a frame (as opposed to 24 hours frame is not going to cut it. At vid rez, you need to be around 20 min frame (per computer) for an high quality render, maybe 1 hour a frame for an exceptional rendering. I just don't see it happening.

[Devin Johnston]

Really? I thought you had to pay more money for extra procs? Either way, it does not help for animation. When rendering animations, you always need a farm and you would only do it one frame = one cpu. And 12 hours a frame (as opposed to 24 hours frame is not going to cut it. At vid rez, you need to be around 20 min frame (per computer) for an high quality render, maybe 1 hour a frame for an exceptional rendering. I just don't see it happening.

 

With the beta I've rendered out print sized images (5100x3800) over night, these were exterior images but never the less it is possible to use Maxwell in very specific situations now to do over night renderings. I'm sure by the time 1.0 is released they will have increased the speed of Maxwell so that at least with the 4 CPU license you are given it will be possible to render out any print sized image over night.

 

You’re correct that you have to buy more licenses to get more processing power, and to do animations you will need more than one license if the animation is anything complex. However based on what I've seen with RC2 I'm pretty sure you will be able to render out even the most complex scene at 720x480 within a hour or two with little or no noise left over. I'm not saying animating in Maxwell is going to be for everyone, unless you have a render farm and are willing to pay for additional licenses you probably won't be doing much of it.

[William Alexander]

By how much? Even a factor of 2 or 3 would not be enough. In order to be competative with Vray it seems it would need to be by a factor of 10 or 20.

 

Are they really trying to compete?

The nature of the program says to me it will never be as fast as Vray at the same moment in time. As the system capabilities increase, Maxwell times will become more acceptable. Ease of use and approaching one button perfection are the real selling points and right now anyway have trade offs, such as render times. The Photon/FG based render engines were developed to overcome the tediously long render times/engine that Maxwell needs to create the quality it can produce. Is that a reasonable statement?

 

Guess I'm saying this is a tool for the 'Artist's' studio, front end and tail end (conceptual & Marketing) of CGVFX/Film...certianly not production. The iterations and fine tuning for materials and scenes with these render times just don't add up for fast turn around on projects. Yet alone being paid to sit and watch/wait for it to render, cause you'd want all resources committed to getting it done

 

Two more cents

WDA

[Ernest Burden III]

One of the (so far unproven) claims about Maxwell is that it does not get slowed down much by stuff that cripples other engines, like caustics and many, many lights, suggesting that the benefit is in advanced scenes. That would be nice.

 

I was hoping for better interior stills, which have been a real slowdown in the native Cinema engine. But now I've got FinalRender and it has made those much quicker, and better looking.

 

But if you throw more CPUs at Maxwell then of course it will speed up results--but the same number of CPUs/core thrown at other products speed their results, too. There is no point comparing 20 core on MWR to 1 core on vRay. If you have 20 cores, you have 20 cores. FR2 bucket renders up to 10 core per licence. MWR is only 4 per licence, so I get more core from my one FR2 licence than my two MWR licences.

 

And Maxwell is a concept. FinalRender and vRay actually work (at least on a PC--is there a Mac version of vRay)?

 

My head is about to explode from reading the community-in-denial on Maxwell's forum. While I'm not there to bury Ceasar (yet), I don't see the point of standing around praising him either. Naked Emperors usually disappoint.

[Christopher Nichols]

One of the (so far unproven) claims about Maxwell is that it does not get slowed down much by stuff that cripples other engines, like caustics and many, many lights, suggesting that the benefit is in advanced scenes. That would be nice.

 

Well that is the whole point of the PPT in Vray too. The problem is that it is slow to use it... although still faster then Maxwell as I have been told.

 

I think it is misphrased. Rather than say, it is just as fast when you put in caustics, you can say, it is just as slow when you don't have caustics. Therefore, it you are getting no optimization or acceleration.

 

Basically, if you are going to use DOF, caustics, blurry everything, yada yada... you have the option of doing it with PPT, but generally speaking you use other faster techniques for 99.9% of the time.

 

Oh and you can bake lighting in Vray... can you even do that in Maxwell? A GI package without baking lighting for an archviz firm seems insane. Especially for animation. Once I bake the lighting in Vray for a static scene, I can get 4 mins a frame or less at vid res. Baking the lightign for a static 300 frame anim takes not more then one hour on a single CPU. I am just saying.... that is just me.

 

I think people got seduced by the push button approach and the physical sky/sun thing... oh and the Gallery. Their Gallery is always a powerful one. I think that is what seduced them in the first place. Vray's weakest point is their Gallery.

 

I also think that people are still busy defending their choice to buy speculative software. While they may be defending it now, I really doubt they will ever do it again.

[_PopArt]

Funny situation. NL releases the RC1, a total watershed disaster. Looks like it needed at least 6 months to be repaired. Total shock anger by the customers. 2 days later, a patched up RC2 comes out, a bit more usable, but the basic core problems are all still there. But on the forum, it is met with joy, save the few, reasonable sceptics.

 

Classic trick - looks like the RC2 is really the RC.5 (an older build) - clever psycological trick which at first disappoints everyone (in the most extreme way possible), and then pull out an older build which works a bit more, and most people start screaming for joy, forgetting completely the true, substantial issue - that the new core engine is in bad shape, and they still arent able to fix (after a year) - the most basic issues...

 

Clever guys...

[Fran]

Count me in among those sceptcs. It is obvious that they fixed some of the more glaring bugs with the studio and the installer. That means little or nothing to me when the render engine is so messed up. If it is not restored to good working order in the next few days, I'll be very concerned. I'm amazed at how people are oohing and aahing over renders that look like they are beginner POV Ray renders.

[Ernest Burden III]

Rather than say, it is just as fast when you put in caustics, you can say, it is just as slow when you don't have caustics.

 

I sorta figured that was evident, but probably good to articulate.

 

Oh and you can bake lighting in Vray... can you even do that in Maxwell? A GI package without baking lighting for an archviz firm seems insane. Especially for animation.

 

Good question--there has been talk from NextLimit about baking, but as far as I can recall it was not a promise for v1.0

 

There is baking in FinalRender2 though I haven't tried it yet.

 

Chris, what version of vRay was being shown to us when you and I were talking at the Chaos booth at Siggraph? Because whatever it was, it was fast, looked great (very much like Maxwell) and was bucket rendering. Also, didn't they say the guy over on the side was working on a Cinema connection--or could? My memory has gone a bit fuzzy, and since you were there too, I'm hoping you can fill in the blanks because I was impressed with what they were showing us.

[rivoli]

2 days later, a patched up RC2 comes out, a bit more usable, but the basic core problems are all still there. But on the forum, it is met with joy, save the few, reasonable sceptics.

 

actually there are more than a few sceptics over at the maxwell forum, and there's a whole bunch of threads/posts about the engine core's poblems. nearly everyone is aware of those problems. fact is it already happened once, when they switched from alpha to beta (or whatever you like to call them): images were blurry and full of any kind of dot, and they sorted it out in a matter of days (maybe a couple of weeks), bringing the core engine back on track again. at this point i guess people are just confident they will do it again, so they tend to concentrate on the new mat system for example, or they try to figure out how this new "3d environment" works.