Messages

31

[Max] General Discussion / Re: PBR kinda materials

« on: 2016-08-24, 14:01:25 »

Have they given any indication of how quickly they plan to add those 'waves' of scans? The results of the 3rd and 4th waves should be interesting.

Not that I won't enjoy experimenting with the environmental stuff in the meantime. Looks good!

32

[Max] Resolved Feature Requests / Re: CoronaDisplacementMod to Mesh

« on: 2016-07-19, 15:42:50 »

Corona's displacement when set to world size, is no better than max displacement modifier and when set to screen it's useless in baked form. So what's the fuss about?

In screen mode it's not useless to me if you are working to a predetermined camera/resolution. Consider for instance a heavy-poly object which is only partially visible.
You could go manually splitting the object at the edges of the camera but I'd rather use this method. It's only a tip after all, so nothing gets changed.

Also, considering possible benefits in world mode, doesn't Corona's displacement calculate adaptivity based on the detail in the displacement map as well?
I honestly don't know if it works in this way or not (have to go back and look at some of the threads from when adaptive displacement was introduced) but theoretically this would make it more efficient than Max's modifier.

33

[Max] Resolved Feature Requests / Re: CoronaDisplacementMod to Mesh

« on: 2016-07-19, 13:03:29 »

Thanks for that tip. It's very useful!

I am getting the same effect as PROH upon import of the proxy. My object was 10 times too small (Scene & Display Units were in cm). Not a huge problem to remedy temporarily as it's an exact amount.

34

[Max] Resolved Feature Requests / Re: CoronaDisplacementMod to Mesh

« on: 2016-07-18, 17:55:09 »

I would really like to see something introduced like this as well. Just to have the potential to create whatever the adaptive Corona displacement would calculate as a mesh - or at least an option within the current CoronaDisplacementMod to cache a calculated mesh in RAM.

I spend a lot of time, when doing test renders, waiting for the same displacement to be calculated again and again*.
The 3ds Max Displacement modifier is an alternative option in these situations but it lacks the adaptivity that is a big bonus for the Corona version.
I would accept that a 'baked in' calculation using Screen Size would not work very well if you're changing viewpoints but quite often you're at a stage when you have a fixed viewpoint so think this would still be very useful. I actually be pretty interested in seeing what the adaptive displacement had done with the mesh if you could look at this in the viewport after a render.
Resolution and Camera could be drawn directly from the main render settings at the time couldn't they?

* I think I raised something like this before in an earlier thread and it was argued that being able to use Interactive Render made this request somewhat redundant.
I can accept that to some degree but I'm still of the opinion that this would be extremely useful if you prefer to use the normal render to perform tests.

35

Gallery / Re: Visualization of tiles #2

« on: 2016-07-12, 13:38:58 »

Is the bump mapping on the wooden floorboards matched to the diffuse mapping? It looks as though the knots in the grain are in different places in the bump & diffuse.

36

General CG Discussion / Re: Modeling Standards for Archviz Furniture and Turbo/Meshsmoothing Modifiers

« on: 2016-06-27, 13:17:27 »

On the metal straps...erm... If I called them metal termination blocks would that change the degree of realism?

Yes, fair enough. Good luck lifting the bench though :).

Below is a suggestion of how to approach the whole thing in principle - keeping a Turbosmooth workflow rather than using fixed edge chamfers.
There are pro's and con's to each but I've always preferred the former. With the way that the quad chamfering works these days I think it's almost as simple.

1 & 2 - Model basic shape without any supporting edge loops. As Freakaz has posted already, circular elements are more correctly modelled using minimum of 8 sides.
3 - Unwrap the mesh at this stage because the overall shape won't change from here on in. It's easy and clean to unwrap now because there's no additional edge's crowding corners where seams go.
(collapse the UVW Unwrap at this stage. You could keep stacks of modifiers as an alternative. I prefer collapsing).
4 - Select corners that you want to be chamfered in the final result. Apply a quad chamfer with the following settings (with tension set to 0 this doesn't move your original edges and basically has the effect of adding 2 edge loops at the corners).
5 - The result with Turbosmooth modifier applied. As Dubcat has pointed out, with 2 loops at the corners you shouldn't get any issues with stretching. The UVW Unwrap from Step 3 is maintained and should not be adversely effected by the chamfering. So my earlier post regarding the edge loops along the length and width was unnecessary really for the particular aim of stopping the texture stretching (although sometimes it's useful to create a more evenly spaced mesh for other purposes).

I would try and consider any supporting loops at corner edges as temporary 'pins' as much as possible that can be applied (ideally using quad chamfer for speed) and removed temporarily if required to make major changes to the basic shapes. If you are changing the basic shape then you'll have to apply a new UVW Unwrap and relax the shape again but your seams should be intact. There are obviously many alterations that are best performed without removing any edge loops that you've already applied but I've always thought it best to try and keep it so that they're not set in stone.

37

General CG Discussion / Re: Modeling Standards for Archviz Furniture and Turbo/Meshsmoothing Modifiers

« on: 2016-06-25, 14:52:38 »

Found out about the texture stretching as I mapped at no turbosmooth. 

You can control the stretching by adding more loops in the geometry to break up long narrow polygons. For instance, if you added in 10 loops along the length of your metal strap and a few loops along the width (I'm guessing at the moment you've just introduced additional loops at the corners to sharpen them). This is effectively similar to what you'd be achieving by collapsing at 1 level of Turbosmooth but by adding the loops in manually you can keep your original '0-level Turbosmooth' structure at the corners and other important areas which makes the mesh more flexible to edit in the future.
You can UVW Unwrap and collapse before you add these additional edge loops, just for ease.

I think, by the way, if those are metal straps that wrap around the timber legs then it would improve your model to make them realistically as such. At the moment it looks as though the timber legs disappear into a solid block of the painted metal.

38

Gallery / Re: London Apartment

« on: 2016-01-27, 19:31:37 »

I really like the late sunlight you've used, it suits the colours in your scene. The styling is interesting too.
The rug looks fantastic generally. I guess it's displacement. The effect just seems to come apart a bit in places (the back of the rug in the 3rd image doesn't look so convincing).
The velvet material on the sofa appears to be glowing also!
Very interesting images though, I like them!

39

[Max] I need help! / Re: Which format for animation image sequence?

« on: 2016-01-24, 14:03:01 »

I think there are some free options here for compositing that might be of interest:  https://www.blackmagicdesign.com/products/fusion

I've not used them, or looked at the features properly, so sorry if that's a useless suggestion! It just stuck in my mind from somewhere that they're available for free.

40

[Max] I need help! / Re: Corona SSS for dummies ?

« on: 2016-01-22, 17:33:25 »

I think the terminology for absorption is counter-intuitive when it comes to colours because a pure red colour (at R1.0, G0, B0) would be absorbing all of the colours other than red wouldn't it? I know this is the same as many other properties, i.e. diffuse/reflection colour, is the colour that's being returned by something, but as the word absorption is used you expect it to work to positively remove that colour rather than remove it. It still makes way more sense to define the colour that's not being absorbed though. It's just the wording.

Is there any definitive description for how scattering colour operates as well? The description you've pasted there for how the absorption works is really useful, something along those lines. I always think of the scattering colour as an internal multiplier but now I don't think it can be technically as it would be doing something similar to absorption colour if it were.

41

[Max] I need help! / Re: RAM or Processor bottleneck? - Rendering issue

« on: 2016-01-22, 13:36:42 »

Ye, I'm beginning to suspect that I need 64GB of ram based on the responses here. Just want to be sure before I ask my boss to cash out for it :S

I'd have said you could be fairly sure that's the solution. Your symptoms completely match what I've experienced when running out of RAM during the initial stages of a render or operating close to the limit.

42

Work in Progress/Tests / Re: Living Room

« on: 2016-01-20, 14:12:09 »

I'd advise to go through some of the textures as well, in particular the timber ones, and check on the scale and position of the mapping.

The scale of the texture on the wooden floor looks very big for example. Depending on where you got your textures from, some of them have a real world scale noted with them which can be useful for a reference point (Arroway, CG source etc.). You've also got the floor mapped as though it would have a continuous texture across the joints but realistically each plank in that floor is a separate piece of wood so the texture should differ for each. There are a few approaches to this. The Floor Generator and Multitexture Map available at CG Source are free and they automate that process. Also, just details like the vertical seams visible at the texture edges on the timber rafters in the ceiling. They don't belong there in reality so it's best if you can avoid them by adjusting the mapping.

You've got some bad geometry on the wall that faces the camera on the left hand side. I guess there are some big triangles with odd shading running between the corner of that downstand at the top of the wall and the corners of the shelf lower down. I'd model it with quads instead or however you prefer (just not with the big triangles).

43

[Max] General Discussion / Re: Cores x Clock

« on: 2016-01-17, 16:10:31 »

there is no premium anymore AFAIK, only home vs. professional. Home does not have remote desktop, so it is unusable for me anyways. And even if not, the price difference is about 45 euros, which is really nothing in context of buying dual xeon workstation.

Sorry that was a bad choice of words - I meant premium in the general sense of the word as opposed to the official title (forgot that they used that as an actual title for previous versions).
I seemed to manage to pay £99.99 for the privilege of upgrading from Home to Pro but still, I agree the price difference is fractional and wouldn't really sway decisions at that level.
I thought it was useful knowledge nevertheless as it's not immediately obvious from Microsoft's Windows 10 comparison pages that this limitation exists.

44

[Max] General Discussion / Re: Cores x Clock

« on: 2016-01-17, 12:44:27 »

One factor that stays unanswered in that windows thread is that you do need to go above Windows 10 Home in order for Windows to recognise more than 1 CPU.
So, in any pricing comparison between faster single cpu systems and slower dual cpu systems (if that's where the original post is leading), you'd need to account for the price of a premium version of Windows.

45

Gallery / Re: Ingersoll Road

« on: 2016-01-13, 20:56:03 »

Did you extract any of the textures from the original photo's to use for the concrete? I only mention because I was taking a look at the real life project and noticed that the markings on one corner of the column are incredibly similar in both versions - but it's just that one corner! It's very nice work either way. I really like the one with the tiny amount of red in the berries amongst all of the greys.