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 - James Vella

Pages: [1] 2 3 ... 8
General CG discussion / Re: Rhino to 3DSMax
« on: 2019-10-14, 15:04:13 »
on the triangles issue, have you tried the new rhino7 wip quadmesh plugin? It looks quite interesting.

Also using the built in Mesh from surface/polysurface with some adjusted settings can do a pretty good job, not 100% but in this case I couldnt find a triangle.

This seems to be a different issue than the OP. I think maru provided a solution for you that would suit the situation in your separate thread;topicseen.

What I am proposing above is basically similar to what sprayer mentioned, render the interior at the desired exposure and then use a mask of the glass to lower the exposure of the exterior (in post production). Or if you use a pass management system like vexus/rpm you can just have 2 separate renders at different exposures (which could work in your case too especially if you dont want the trees to interfere with your lighting).

General Discussion / Re: corona vfb v image in photoshop
« on: 2019-09-27, 19:45:44 »
2) in PS set colour profile to Dell's monitor

Just be careful doing this as when making your textures in photoshop you will want your color space to be in RGB as this is what 3dsmax will expect.

General Discussion / Re: corona vfb v image in photoshop
« on: 2019-09-27, 07:50:51 »
Correct me if Im wrong but you wont need the last step (windows color management device profile) as Photoshop is using CAL1 already to display the wider gamut as per juraj comment below.

But this will once again only work with color managed applications, so I suggest only to use hardware calibration (either default sRGB, or custom CAL1 sRGB later), and using generic color profile in Windows (use sRGB IEC61966-2.1, not default "display profile").

In 3dsmax you are seeing the VFB with CAL1, but thats not what the render actually looks like (but it doesnt matter we opted for option 2 in juraj scenario).

So now in photoshop we create a new file and select 'dont color manage' as juraj mentioned.

You can then set Photoshop to either ignore color profiles (it will do the calculations in generic wider gamut but it will still show sRGB on your monitor because the monitor is doing soft-proof automatically with the clamped sRGB profile

Import your render, it will look the same as the VFB. Adjustments from here on will use your monitor gamut so you can adjust colors as needed. So if you want to ensure it looks like it does in Photoshop in the JPG you must Export As - Convert to sRGB.

(A good idea to select 'embed color profile' as well to make sure other software knows its an sRGB profile).

In Photoshop > Proof Colors - Internet RGB should appear the same as well.

**edited for clarity

General Discussion / Re: corona vfb v image in photoshop
« on: 2019-09-26, 16:46:58 »
It doesn't matter what you save from 3dsMax (8/16/32bit, any file format), because the software isn't color managed as a whole and neither is the image exporting process (regardless of renderer).

Well there is a difference between exr and png, png is slightly more desaturated - but just a little bit.

2) WRONG colors, but identical image transfer: If you want the SAME (WRONG) colors like you see in 3dsMax on wide-gamut monitor, you have to attach the same color-profile your monitor is using. This will give you wrong colors (you aren't converting between color-spaces but simply transfering the same values). You can then convert to sRGB when sending it to clients or uploading to website.

+1 this method. It seems to be what is wanted in this situation right, to match the vfb? You can then adjust what you need in photoshop further if you need to, otherwise you can just adjust it if you prefer using the VFB.

General Discussion / Re: corona vfb v image in photoshop
« on: 2019-09-26, 15:30:15 »
have you tried saving your image as 16 bit half float exr and loading that into photoshop? does it look the same as the vfb?

are you photoshop settings set to default - or have you changed the color settings management policies?

are you working in a different color space in photoshop? if you are - are you selecting 'assign profile' or 'convert to profile' (if neither ignore this last question for now).

I need help / Re: Light conversion to watts questions
« on: 2019-09-26, 13:05:01 »
Well its also related to the size of the light.

for example:

20cm corona light sphere @ 900 lumens = 0.841 W/(sr.m^2)

10cm corona light sphere @ 900 lumens = 3.363 W/(sr.m^2)

The sun is quite large, so im sure its been set by the scene units and calculates it via that. Im not sure exactly how large etc but for specifically matching a corona light to a coronalightmtl on a sphere of the same size it should give you the same result. From my testing I find as long as Im sticking within realistic light bulb sizes and matching lumens it looks very close to the reference. For your other questions im afraid I dont know.

Hello James,

Thank you for pointing CProxyLister,  but I didn't want a lister for several reasons. Essentially, too many proxies on the projects I was working on, so I thought a lister wasn't the good approach. The good thing IMO is maybe to use both scripts, depending on users' needs ;)

I see your point, simplicity is always a beautiful thing. Ill do my best to break it :D

Great work, would be nice if you could select which proxy from the list such as this script ive been using for awhile, good to see a comparison for speed maybe.

I need help / Re: Light conversion to watts questions
« on: 2019-09-26, 09:39:43 »
I usually use Lumen (lm) as its easier for the conversion (in my case)

If you scroll down the page I have listed information on the conversion for lights.

lumen 250 = 25watts = 0.015 W/(sr.m^2)

Not entirely sure about the logic using W/(sr.m^2) maybe someone with more technical experience can chime in, however my point is most standard lightbulbs are roughly between 25-150 so the chart does a pretty good job for (lm).

To answer your question regarding coronalightmtl/sun i find the 'intensity' value matches W/(sr.m^2) - so in the above example if you put 0.015 into the lightmaterial intensity it will be 25 watts or 250 lumens. My conversion technique is quite simple really i just create a corona light, put the lumen (lm) value in according to the chart, change the type from lumen to W/(sr.m^2) and it gives you the 'intensity' value.

I need help / Re: How to save background on exr
« on: 2019-09-16, 17:59:42 »
THX! opening on phosothop with the free exr plugin and checking split alpha do the trick!

Great to hear, good luck!

I need help / Re: How to save background on exr
« on: 2019-09-15, 10:26:21 »
once you have placed the image in the override direct visibility that should do the trick. When you load the file into Photoshop did you select Alpha Data Channel: As Alpha Channel or As Transparency?

As Alpha Channel should give you the output you are looking for. If you are using a different software to view the exr file then you should find the option that is enabling the alpha by default (which will make the background either transparent or black).

Feature requests / Re: Adaptive Rendering
« on: 2019-09-08, 09:00:57 »
Maybe a nice feature would be to use more samples on the selected objects? So if you're trying to work on a texture on a particular object, then that object renders faster if selected in viewport, while the rest of the image remains more noisy.
Currently with "render Selected" the interactive rendering starts over. With this new feature the adaptability would be interactive depending on the selection.

It actually does work this way in interactive rendering.

Make sure you disable Clear VFB Between renders in the System tab in Render Setup (Settings).

edit: this seems to be effective with smaller scenes. After testing with a large scene you are right to some degree there does seem to be a recalculation at the start of the interactive but the noisey parts dont update, but it is slower (probably due to calculating reflections of many objects).

Why not to use just exposure on camera?

This too. I actually thought thats what Tom also meant by the compositing method, either mixing 2 hdri or 2 images at different exposures. Similar to how you would do a HDRI merge in photography.

I agree with TomG with the composite method for this particular workflow.

However you could also fake the bounces by putting window lights for secondary bounce thus giving you a built in solution (wont be the same but can be effective).

Pages: [1] 2 3 ... 8