User avatar
By Ernesto
#371245
Hello,

I am working with real scale materials (Maxwell)
They are supposed to be defined in Meters as it says in the Material editor.
The volumes to receive those materials should be UV mapped 1 meter by 1 meter to match the previous material.
In my case, as I am working in cm, I decided to set an UV of 100 by 100 cm.

When I use a material which tiled map is only 0,40 meters, the materials show off 6.25 times bigger.
In order to correct this I had to work with an uv of 16 by 16 cm which is not right.

But in case the Tiled Map is bigger than 1 meter, it shows 2,5 times smaller, so I had to use an UV of 250 x 250 cm to correct it.

All this in the same scene, which is totally crazy, since in one case it seems to be working in inches and in the other in feet, but I am working in cm...

Thanks

Ernesto
User avatar
By Mihnea Balta
#371392
Real scale materials don't really work in anything other than Studio. When you use a real scale material, Studio generates UVs automatically, and changes them every time you change the size value (it works like a projection node). In Maya you already have UVs, so all that setting does is set the tiling to the inverse of the "size" value.

We are considering making a similar projection operation in the plug-ins in the future, but for now you just have to edit the UVs manually, like you would for any other object.
User avatar
By Ernesto
#371417
I am not sure if I understood what you said...
It is not that complex.
It seems that we are talking about diferent things...

Look at this:

The way to work with fixed measurements materials in Maya:
All we have to do is set an UV of 100 x 100 cm to all the objects.
Then we use materials previously prepared, which has a fixed measurement. For instance if I have a map, showing 16 tiles 40cm x 40cm each, it would represent a surface of 160 x 160. Then the material definition in Relative mode would be: 0,625 x 0,625.

In case of a Real Scale Material, the only diference would be that it would be set to Meters, and 1,6 x 1,6 and it should work in the same way. There is no need to consider "similar projection operation as studio". Everthing is already there. You only have a Bug with the units. That is all!

I am waiting for these things since the Beta versions. I really though that in the current version it would have been solved, since this is basic and simple compared to all the complex stuff you have done.
Please consider a pluggin fix. It is not that complex, and it produces lot of problems in the user end!

In fact, I did a simple test: I have my scene in Maya, set as detailed above, and exported to studio without any editing. Then assigned the REAL SCALE Material, and it worked perfectly.
If I do the same in the maya scene I get units inconsistencies, as described at the begining of this topic.
Therefore the problem is clearly in the Pluggin.

I really would resign all the new features of Maxwell if I could get these things fixed properly. I have really lost all interest in upcoming new features. They will not be usefull under the current circumstances. Again (and excuse my insistence, but nobody at NL seemed to listen seriously to my concerns) in my opinion, new stuff development should be stopped, to concentrate all the energy of the team in fixing all the current bugs (or problems) that conspire against the user, that are not a few at all. I would love Maxwell to be a tool, rather than an obstacle that everyday offers me new unpleasant surprises, that makes me delay, therefore downgrade my profesional quality in order to reach the deadline. If I would live in a world where i could spend all the time I want, to investigate, and collaborate to solve Maxwell Problems and at the same time make a living I would do it with pleasure, because this is my interest. But I cannot do that, having to pay such a price for it. I do not mean the price of the software, I mean the cost of all the time lost trying to make things work, to realize (a week later) that they were not expected to work, as you explains to me in this case. It is really a pitty, because I think Maxwell is one of the best render engines, if not THE BEST, but implementation is not at the same level. I know that some people take these comments as atacks or insults, but that have nothing to do with my interest. All I write is the hard true, as I sincerely sees it. I could be wrong, and in fact I would love to be in a mistake, that is why I am exposing this comments with the hope that somebody would correct me if I am wrong. But nothing like that happened. The long silence is confirming the arid panorama, therefore hopes for a change is dying.

Yours

Ernesto
User avatar
By Mihnea Balta
#371453
Ernesto wrote: The way to work with fixed measurements materials in Maya:
All we have to do is set an UV of 100 x 100 cm to all the objects.
How do you do that?
Ernesto wrote: Then we use materials previously prepared, which has a fixed measurement. For instance if I have a map, showing 16 tiles 40cm x 40cm each, it would represent a surface of 160 x 160. Then the material definition in Relative mode would be: 0,625 x 0,625.

In case of a Real Scale Material, the only diference would be that it would be set to Meters, and 1,6 x 1,6 and it should work in the same way.
So does it work if you use relative mode and set the tiling to 0.625?
Ernesto wrote: In fact, I did a simple test: I have my scene in Maya, set as detailed above, and exported to studio without any editing. Then assigned the REAL SCALE Material, and it worked perfectly.
Yes, because Studio created new UVs when you did that. It's not using the UVs which you set up in Maya.
Ernesto wrote: Again (and excuse my insistence, but nobody at NL seemed to listen seriously to my concerns) in my opinion, new stuff development should be stopped, to concentrate all the energy of the team in fixing all the current bugs (or problems) that conspire against the user, that are not a few at all.
For some reason, you're the only user I know about that constantly runs into show-stopping problems. Maybe you're just unlucky or good at spotting these things, or maybe you're exaggerating the impact some problems have on your workflow and productivity. Either way, rather than posting a diatribe each time something doesn't work as you expected it, you could instead attach a scene which exhibits the problem. That would actually allow us to fix things. Telling us how you think we should run our development or business is much less constructive.
User avatar
By Ernesto
#371472
Please do not change the focus.

The problem is that the product is not working properly. The amount of problems that were found in a short time is abnormally high.
The nature of those bugs are normal only in initial development stage, but are not normal for a final version. Maxwell 2.7 as you mentioned was corrected and improved for years, but as we can see it was not enough. We have obviously something wrong here, since it seems not to be any diference between a new release and a final version, as for the amount of Bugs found. Development should end, when you reach a professional level, not when you decide arbitrarily, leaving users alone to deal with basic and rough bugs that shouldn´t be there specially at that stage.

Perhaps you didn´t liked me to talk about the way NL is working. But I am pointing it because that is the problem from my point of view! Any work is the same. First you start it, obviously you will have lots of problems because it is something new, then you work to improve it until it is prety finished. But Maxwell is starting new features at any stage, and the obvious consequence is that you will never have an acceptable level of refinement, because when you fix an old bug, you are introducing a new one!

If you need a scene to study the problem, you can ask it to me, and I will send it to you, but please do not complain because that is not my work.

Yours

Ernesto
User avatar
By Mihnea Balta
#371483
Ernesto wrote: The problem is that the product is not working properly. The amount of problems that were found in a short time is abnormally high.
The nature of those bugs are normal only in initial development stage, but are not normal for a final version. Maxwell 2.7 as you mentioned was corrected and improved for years, but as we can see it was not enough.
Well, I wasn't trying to change focus. I was just saying that people do stuff like this short film with it without complaining, so clearly Maxwell is usable (we didn't get a single support question from these guys, as far as I know). There are lots of people who produce great work with it, which invalidates your point. Maybe those people run into bugs occasionally, or have to use a workflow which isn't ideal, but they don't find show stoppers every 5 minutes. Or maybe they don't consider clicking an extra checkbox a show stopper, or they use the right tool for the job they're trying to do, or something.

Anyway, it doesn't matter. You clearly think you know better than us, so there's no point in dragging this further.
Ernesto wrote: If you need a scene to study the problem, you can ask it to me, and I will send it to you, but please do not complain because that is not my work.
Yes, it's not your job to debug our software. However, when you run into a bug, you seem to spend a lot of time and energy philosophising about the morality of the bug, when reporting it properly would be much easier, and infinitely more likely to lead to a fixed build. You don't have to do this, of course, but you can't expect me to divine what doesn't work and why based on a wordy description that I don't really understand (maybe due to the terminology).
User avatar
By Ernesto
#371489
Again, lets focus in the key.

That some other people haven´t found any bug, does not means that the bugs we are finding are inexistent.
Therefore If we accept that they exist, it doesn´t matter who discovered. All what have to be done is fixing them.
Anyway, exposing myself in this way, is not desirable for many others that might share the same concern.
In any case, although this could be unpleasant for you and for me, and I am sorry for that, the goal is to improve the product/service and having everybody happy. In other words I am trying to help. Also I am trying to use Maxwell as a permanent rendering software, since I would prefer not to move to other renderers.

Only you know how many customers have bought Maxwell, and how many have migrated, or haven´t upgraded.
But These problems are surely having an influence, although many people prefer not to talk.

The key (as I see it) is providing SAFE versions. I have wrote about this some time ago. Users must have the choise to get a really new version obviously with some problems due to the recently implemented new features, or a more stable version, that could lack of the new features. What is really bad, is not providing any stable version. Usually the last version before a new upgrade (the current version) should be the most stable version of all, but in maxwell there is no diferentiation and all the users have to deal with the consequences no matter if they can or not.

The main critic, I do, is that you shouldn´t be working in V3, until you finish the necesary refinement in V2. The last version of a software before an upgrade is presented, should be the most stable of all the versions. And this point is the one that is clearly not fair, in your current policy. If I am wrong, tell me why you should prefer the opposite? Up to now, despite all the words and phrases against my person, nobody answered the question, and this silence is also not fair for our point of view (customers)

I am not inventing this. This way of working is the standard for other software developpers, and let the users to chose how much risks to run, deppending on the circumstances. I love to explore new features, and love also to help in development, but when I have a deadline, I need stability. I really hope you could implement something like this, and it will be great for everybody, otherwise, the professional customers will have to leave to search the necesary stability to guarantee to their clients a stable deadline, and a predictable quality.

This is not something I decided it true, and impossible to modify, this is only my opinion. If you think this is not desirable or not possible, all I expect is an explanation of the reasons that support the oposite idea, and I will surely understand. In any case disqualifiyng the messenger, only makes stronger the message.

I am preparing a special scene so that you can see what happens.
In that scene you will find two models, both with the same UVs of 100 x 100 cm, in the first (which name refers to normal materials) everything works as it should, but in the second (which name refers to real scale materials) we have two diferent behaviours depending on the size of the tilable map used in each material, as follows:

Image


If the Map is smaller than 1 meter by 1 meter, the texture gets huge, and if the Map iks bigger than 1 meter by 1 meter, it reduces drastically the material. You will see that it is not anything subtle, but very drastic!

Image


It should be in you inbox now.
I hope it helps.

Ernesto

By the way, talking about not clear termonology, which is your mother language? mine is spanish.
Regarding the images that you use to request, why do not enable image attachments in this forum. It is difficult for me to include images with the current procedure. At least (if you are concern of the available disc space) you could enable this option in a special subforum, where users could explain these things that could help you to fix bugs. I think that it would be good if you facilitate the free work others do to help you.
User avatar
By Mihnea Balta
#371622
Ok, so what I don't get is: how did you set real scale? Because the Maya plug-in doesn't support that. There's no option for enabling the "real scale" flag on textures in Maya. Did you import a real scale material from a MXM file?

We will change the plug-in to invert the tiling values when it imports real scale textures from MXM files. That should fix the problem, but when you modify the tiling later in Maya, you should keep in mind that you have to use inverted values too. For example, if you want to use a tiling factor of 4 meters, compute 1/4, and set the tiling value to 0.25.

Or, if you're using MXM files a lot, you could consider referencing them instead of importing. That way the real scale flag should work correctly.
User avatar
By Ernesto
#371675
Ok, so what I don't get is: how did you set real scale? Because the Maya plug-in doesn't support that. There's no option for enabling the "real scale" flag on textures in Maya. Did you import a real scale material from a MXM file?

Yes

We will change the plug-in to invert the tiling values when it imports real scale textures from MXM files. That should fix the problem.

That sound PERFECT!

but when you modify the tiling later in Maya, you should keep in mind that you have to use inverted values too. For example, if you want to use a tiling factor of 4 meters, compute 1/4, and set the tiling value to 0.25.

That will not be a problem, since all models to be used with real scale materials will have a fixed UV, and real scale materials tiles shouldn´t be changed, since the scale should be fixed too. Real scale materials are predefined in MXM files, because the key of using them is that they are UNIVERSAL, and are intended to be used in many diferent projects, and should NEVER be changed, since in such a case they would loose their real scale.

Or, if you're using MXM files a lot, you could consider referencing them instead of importing. That way the real scale flag should work correctly.

Yes I am using MXM files a lot. Really? I didn´t knew that! i will try.
Some minutes later:
I did tried using referenced MXM and I get the same results as with imported MXM materials.
Perhaps I didn´t undertood your explanation regarding Referenced Materials.
Anyway the proposed fix will be perfect!

Ernesto

after a lot of years doing arch-viz... almost 20 a[…]

render engines and Maxwell

Funny, I think, that when I check CG sites they ar[…]

Hey, I guess maxwell is not going to be updates a[…]

Help with swimming pool water

Hi Choo Chee. Thanks for posting. I have used re[…]