Author Topic: Daily Builds 1.0 - 1.4  (Read 244997 times)

2015-08-07, 17:38:28
Reply #360

pokoy

  • Active Users
  • **
  • Posts: 1850
    • View Profile
From the first solution option one makes more sense and I would say this solution is what I'd like to see... but without checkboxes I know I will end up missing them.

Now the jackpot question: Is it completely impossible to stretch the UI a bit and allow for a two line layout per layer? Like:

[X] Amount [100]....................[X] Amount [100].........
[-------- Material X --------]       [-------- Mask X --------]

Where both amount spinners are independent. So the layer can be 50% opaque but the mask still goes from 0-100% and vice versa. AND where I can disable the whole layer, obviously, but I can also disable the mask if needed.

Is this where all common sense breaks? Why design a solution that inevitably leads to limitations if we could just add a line in the design? It would still be more compact than the Composite Map for example, which isn't something people complain about every day.
« Last Edit: 2015-08-07, 17:51:03 by pokoy »

2015-08-07, 17:47:19
Reply #361

racoonart

  • Active Users
  • **
  • Posts: 1446
    • View Profile
    • racoon-artworks
Ok, currently there are two plausible variants on the table:

First solution - Two spinners
- One spinner for material opacity/amount
- One spinner for mask influence

[...]

For me, it's Solution 1 with option 2. Its easy to understand, flexible, and only needs 2 spinners. Solution 2 is not as flexible but imho an acceptable compromise.
Any sufficiently advanced bug is indistinguishable from a feature.

2015-08-07, 17:58:26
Reply #362

pokoy

  • Active Users
  • **
  • Posts: 1850
    • View Profile
Rereading the options from solution 1 makes me think the explanation is missing what happens when exactly.

From what I understand the main amount spinner should not affect the masks ability to block the material, in case main spinner is at 50%, the mask's black should still remain completely black, white will only allow for the opacity the main spinner is set to.
The main spinner should NOT act as a multiplier for the mask, this is what I understand option 2 will do.

2015-08-07, 18:33:24
Reply #363

Ludvik Koutny

  • VIP
  • Active Users
  • ***
  • Posts: 2557
  • Just another user
    • View Profile
    • My Portfolio
Ok, currently there are two plausible variants on the table:

First solution - Two spinners
- One spinner for material opacity/amount
- One spinner for mask influence

[...]

For me, it's Solution 1 with option 2. Its easy to understand, flexible, and only needs 2 spinners. Solution 2 is not as flexible but imho an acceptable compromise.

Vray uses option 1, so that could bring another wave of complaints :D

2015-08-07, 22:44:22
Reply #364

antanas

  • Active Users
  • **
  • Posts: 269
  • Hmm ...
    • View Profile
 A tremendously long awaited layered material got me a bit overexcited so I made 2 versions of that materials gui mockups which, as I imagine, could be much more useable and fool proof than the one which is currently implemented. There's totally no need to copy vray's crappy and unintuitive VrayBlend material cause, imho, people which are used to v-ray and don't want any improvements should just stay with v-ray and not plague corona with their workflows ))
 To the matter at hand - first mockup is based on modified multi\sub-object material logic and the second one is based on modified composite map's one. Both of those are quite familiar to most users and those for which those aren't a part of the daily work, usually don't need such advanced materials as multilayer anyway)) so I think that even given the differences with originals people will easily understand how to use either of those.
 In both cases I've used material and map preview windows cause I think and I think many will agree that working on complex materials without those is a major pain, especially when one needs to tweak that material later on and by that time totally forgets which map\material was put where and exactly why or imagine some studio work, when different people work one after another on the same scene\models\materials and so on.
 Visual representation in CoronaLayeredMtl material\map slots is a must imho - for instance just try to imagine how tremendously crappy and unwieldy would be afore mentioned multi\sub-object material if it would not contain those previews as it does now.
 Both of those mockups are are of course pretty crude and can be made prettier, more logical, and simply cleaner looking but I think as a reference those will suffice for now, so people please do tell me that you think of those. Personally, I would prefer something similar to v2 as imho it is cleaner, more foolproof and much more understandable from the first glance, but of course that could just be my overexcited imagination ))

2015-08-07, 22:55:13
Reply #365

pokoy

  • Active Users
  • **
  • Posts: 1850
    • View Profile
Now we're talking! Fully agree with your opinion about Vray's layered material, I always chose the normal Blend material because I didn't like their logic of how masks worked.

From your mockups I'd choose v2. You're right about the previews, funny it wasn't mentioned yet but now that I've seen them there's no reason why they shouldn't be there.

Great idea. With the previews being there, the short slots aren't that much of a problem. But they might be a tad too short... Still, I like it!

2015-08-07, 23:12:55
Reply #366

Ludvik Koutny

  • VIP
  • Active Users
  • ***
  • Posts: 2557
  • Just another user
    • View Profile
    • My Portfolio
Wow, I never though something as simple as LayeredMTL could be made to look like so complex beast O_o

2015-08-08, 00:54:38
Reply #367

antanas

  • Active Users
  • **
  • Posts: 269
  • Hmm ...
    • View Profile
Wow, I never though something as simple as LayeredMTL could be made to look like so complex beast O_o
Well for those kind of materials need to be informative and functional or else they are just not needed at all as there already is a standard blend material which Pokoy mentioned above. Still I'm just suggesting something like that and of course it can be made simpler\cleaner and not as cluttered, but imho V2 interface, or rather some sort of derivative from it, is already quite simple and production proven, well not it itself but the composite map after which it was modelled.
And for something really complex )) see something like https://cdn.tutsplus.com/cg/uploads/legacy/000_QuickTips/016_Maxwell_Decals/MXED_Decals_600x400.jpeg or any other multi-layered Maxwell renderer's or Fryrenderer's\Arion's materials, yet still It's the only thing I miss about those engines as one could do pretty much anything using those materials if not for their sad render speed, bad integration to a standard 3dsmax interface and most sadly really poor support of standard and 3rd party plugins and shaders.

2015-08-08, 04:11:20
Reply #368

Christa Noel

  • Active Users
  • **
  • Posts: 911
  • God bless us everyone
    • View Profile
    • dionch.studio
Now we're talking! Fully agree with your opinion about Vray's layered material, I always chose the normal Blend material because I didn't like their logic of how masks worked.

From your mockups I'd choose v2. You're right about the previews, funny it wasn't mentioned yet but now that I've seen them there's no reason why they shouldn't be there.

Great idea. With the previews being there, the short slots aren't that much of a problem. But they might be a tad too short... Still, I like it!
agree with pokoy :D
Ok, currently there are two plausible variants on the table:

First solution - Two spinners
- One spinner for material opacity/amount
- One spinner for mask influence

[...]

For me, it's Solution 1 with option 2. Its easy to understand, flexible, and only needs 2 spinners. Solution 2 is not as flexible but imho an acceptable compromise.

Vray uses option 1, so that could bring another wave of complaints :D
then the team should choose "stay away from another wave of complaints" but keep the user need on them side.

2015-08-08, 10:38:02
Reply #369

juang3d

  • Active Users
  • **
  • Posts: 636
    • View Profile
So far latest DB is working pretty great, no crashes and it even feels a bit faster, I may b wrong, it's an old scene and I did the reset settings thing.

Cheers!

2015-08-08, 23:51:03
Reply #370

PROH

  • Active Users
  • **
  • Posts: 1219
    • View Profile
Hi. Played with the 2015-08-07 daily, and it has been working great - except with the latest file I tried. It's an old file with some water-testing, which is converted/reset so there's no "legacy". With this file I'm having big problems with max freezing and the renderer wont stop calculating when canceled (both IR and normal). I tried to narrow down the problem, and I think it's caused by a strong AO effect in the water-material. If I use another material without AO everything works as expected.

I've attached a part of the scene (Max 2014), so others can try it and see if they get the same problem, or it is caused by something else in my system. To reproduce the problem:

1) Open the file and press "Render"
2) After 1 pass (or more) press "Cancel" (or "Stop" in IR)

Now the problem should be, that the renderer won't stop calculating - you need to shut down Max to stop it.

Thanks in advance.

2015-08-09, 03:22:13
Reply #371

antanas

  • Active Users
  • **
  • Posts: 269
  • Hmm ...
    • View Profile
 Hi PROH, I tried to render your scene using max 2016 and 07 08 daily, indeed it doesn't stop using both IR or normal render, so yeah I can confirm that strange behaviour and probably you should file that as a bug on mantis bugtracker right away.
 Just to be sure I even tried to rebuild that material using new Corona Layered Mtl but ended up with the same sad result and without such when coronaAo is disabled, so probably you're right and it is triggering that.
 It got me curious and I tried to downgrade to the older versions, because I tend to use coronaAo map for many purposes and haven't got such bugs earlier so here are the results : 1.3  07 08 daily - bug is present, 1.2.1 - bug is present, 1.2 no bug, 1.1.1 and 1.1 are bug free too - so it is quite easy to guess that it was introduced in 1.2.1 and further on, I just missed it cause I completely missed 1.2.1 and after using 1.2 jumped right to a 07 08 daily only because I wanted corona layered material pretty badly\madly but did not render much yesterday or today and as a result hadn't noticed that.
 Thanks for finding that bug out as just right now I'm working on some older scene and the deadline is on Monday - that scene is using some coronaAo for the roads\walls dirt etc and it would be me panicking, cursing and plucking my hair off just about tomorrow for now I'm back to 1.2 and thanks again ))

2015-08-09, 13:03:25
Reply #372

PROH

  • Active Users
  • **
  • Posts: 1219
    • View Profile
Hi antanas. Thank you for testing and confirming. I've reported it at Mantis (ID 0001242). Hope it will be fixed very soon, since it makes everything after v1.2 useless for (my) work.

2015-08-09, 16:03:53
Reply #373

cgiout

  • Active Users
  • **
  • Posts: 158
    • View Profile
If it helps i can confirm too that it won't stop.

3dsmax 2015 + daily 08/07

2015-08-09, 17:15:58
Reply #374

PROH

  • Active Users
  • **
  • Posts: 1219
    • View Profile
Thanks :)