Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.


Messages - dj_buckley

Pages: 1 2 [3] 4 5 ... 59
31
[Max] I need help! / Re: "Thin Shell"
« on: 2024-02-19, 09:56:48 »
This way the base layer may still have effect, this is how physical material (and probably the physical phenomenon too) works.
If you put the base layer to full white, it should be as if it was hidden. Give it a try.

Do you mean in my Rayswitch example?

32
[Max] I need help! / Re: "Thin Shell"
« on: 2024-02-19, 00:32:50 »
Yup - as I say, it's a super basic shader.  There isn't even a refraction colour in the Physical Material.  As I say it could be working as expected.  I'm just trying to clarify.  I guess the issue with physically reducing the IOR to 1 would be the reflection strength, so assuming I don't want the 'darkening' I can get around it using RaySwitch and just set IOR to 1 in the GI slot version of the material.

33
[Max] I need help! / "Thin Shell"
« on: 2024-02-18, 21:48:02 »
If I've got some window glass setup as 'thin shell' should the glass block light at all?  I was always under the impression it stops the objects being refractive (the equivalent of ior 1) and from casting shadows/blocking light as the light is now passing straight through and not being refracted.  Now I think my understanding is wrong.  I'm just testing an interior scene and when hiding the glass the brightness of the whole interior increases quite a lot.

It's a super basic shader.  0 diffuse, thin shell, 1 refract. 1.52 ior.

A bit more testing and leaving Thin Shell ticked but changing the IOR to 1, I get the same result as if I'd hidden the glass.  So what exactly does 'thin shell' do.  Does it simply stop light rays bending?

34
[Max] General Discussion / Re: Camera Film width setting
« on: 2024-02-16, 10:40:24 »
I'm correct in thinking you also need to change it when camera matching to a backplate shot in portrait mode?

For example my camera sensor D810 - is 35.9mm x 24mm.

So when I'm matching to a landscape shot I pretty much leave it at 36mm, but if I'm matching to a portrait shot then I'd change it to 24mm

36
If a texture has white in it, then I need to darken the whole texture. Right?
But why should I darken the texture that has white in it, together with all the other colors in the texture, and not all other textures in the scene, that don't have such bright colors,  at the same rate?

This is a mine field. 


"But why should I darken the texture that has white in it, together with all the other colors in the texture"

If a texture contains white as a value of 255,255,255 then it's just wrong and all of the other colours in that texture are likely to be the wrong values relative to the white.  But simply put white whites at 255,255,255 is wrong.  So it's easy for these 'guides' to highlight that.  Same with black, although I personally think that black PBR standard of "30-50" is also wrong.  But I'm no expert.  Far from it in fact so who am I to say.  So take the rest of what I say with a pinch of salt.

Check out Martins article here for more info on 'blacks' - https://www.racoon-artworks.de/blog_PBRfromrulestomeasurements.php


"and not all other textures in the scene, that don't have such bright colors"

As far as the 'other textures' go that are full of mid-range values, who knows.  That's for you to decide (or verify).


"At the same rate"

Not all textures are made equal, I'm assuming you grab textures from various different sources and each creator will have their own workflows and processes to capture said textures, who knows if they're all done to the same "standards" or any standards at all.  So you might have two textures from two different sources and they might both be wrong.  But they might not both be wrong to the same degree relative to each other, if that makes sense.


In my opinion, following a 'strict' PBR workflow is only really possible if you're creating all of your own assets from scratch.  Failing that, ideally you should be checking and correcting each and every texture to get all of your textures from all of your different sources as consistent as possible in terms of the 'correct' values.

Wanna know what I do ... make sure there are no white whites and no black blacks, then eyeball the rest.  As already mentioned, if it looks right, it is right.  When I'm creating a material, I tend to have real world references up on screen, of that material under the same or very similar lighting conditions as the lighting in my 3D scene.  I then get the overall shader looking roughly correct in terms of reflection/roughness/bump etc (I often do this with no albedo texture, just a mid desaturated red colour), and then add the albedo to the mix and tweak using CoronaColorCorrect nodes (brightness, saturation, hue) until my scene and the reference look pretty close.  Once they're pretty close then I'm happy.

I do use the Albedo render pass element as Tom suggested to check my scenes just before rendering, and if anything is red, then I consider it 'out of range', and so I tend to fire up IR and drop the diffuse level on that objects material until it's no longer red in the albedo render element and back 'in range'.  I only do this for small things though where you wouldn't really notice if their textures are really the 'correct' values or not.  And for some reason it always seems to be book covers and artwork that need fixing :)

Jumping back to strict PBR stuff - "In Range" and "Correct" are two very different things.

Hope it helps (assuming what I've written isn't absolute nonsense).


37
[Max] I need help! / Re: wxWidget Debug Error
« on: 2024-02-09, 17:29:37 »
The native one in windows called 'Remote Desktop'.

In answer to your second question.  I wouldn't know.  It's happening on one particular node and I only ever connect to that node through Remote Desktop.  It's just sat in the corner of the room with no monitor or peripherals.  I have another node that I also connect to through Remote Desktop and I've not seen the error on that node.

The problematic node is a Threadripper 3990x running Windows 11.  The other node AND my main workstation are both Intel i9's running Windows 10 if that helps.  Although I should stress I've been using this setup for years with no issues, it's only since upgrading to Corona 11 and now Corona 11 Hotfix 1 that I've started encountering the issue.

38
[Max] I need help! / Re: wxWidget Debug Error
« on: 2024-02-09, 15:11:34 »
If you could send us any random scene where this consistently happens to you

That's the problem, it's not consistent - it's random.

I use Remote Desktop yes.

39
[Max] Feature Requests / Re: Tone mapping - Grouping
« on: 2024-02-07, 14:41:12 »
+1 would save a tonne of clicks for testing different setups.

40
You can use Corona converter to quickly convert standard 3ds Max bitmaps to CoronaBitmaps.

Does this now work safely with all image formats?  There were issues previously with tiff files

41
[Max] I need help! / Re: wxWidget Debug Error
« on: 2024-02-07, 12:38:23 »
I've already explained that its completely random and has happened with lots of different scenes, sometimes it happens, sometimes it doesnt.  It's not any scene in particular.  I could open a scene now and it doesn't happen, I could open the same scene again tomorrow and it does happen.  Completely random.  Am I meant to just pick a random scene to send over?

The error message is quite clear no, surely it means something to the Corona team?  Also this doesn't seem to be a new issue.  You have tonnes of reports (3 pages when you search the forums) of this dating back an entire decade.  i notice some of them are marked as 'Solved' without any actual resolution ...

42
[Max] I need help! / Re: wxWidget Debug Error
« on: 2024-02-07, 11:24:30 »
Yes

43
The heavier the scene the better.

Isn't this the other way round with this specific issue? I think complex scenes where background takes relatively small portion in terms of calculation time, are not affected as much as lightweight scenes, particularly single object renderings against monotone background.

I was thinking along the lines of this, a heavy scene, the black background takes up roughly half the image, a scene that already takes a while to render so any shifts in rendertime will be noticeable rather than negligble.

44
Out of interest, why does Corona downsample like that?  Speed purposes?

For things like that, where stuff is happening behind the scenes that we're unaware of and at the cost of quality, it would be better to give the user a choice? "Keep Original Environment Map Resolution" as a tickbox for example?

45
Regarding the black background slowdown - I was able to find a report for Cinema 4D and it's scheduled to be investigated (fixing it based on a C4D report would also fix it for 3ds Max). Of course we can log additional reports with additional files, so if you have an example or simple reproduction steps, please do share them.

By asking us to submit scenes, are you suggesting that you're unable to replicate the slowdown at your end?  Surely you have a tonne of scenes by now, submitted by users that you can test on?  Simply open any scene, render, then render again with environment override directly visible set to black and compare the render times?  The heavier the scene the better.

Pages: 1 2 [3] 4 5 ... 59