1
$\begingroup$

I'm working on a scene in Blender where I've created a blue glass jar with an aluminum lid. Unfortunately, I can't get the material to look the way I want, matching my reference image. I've already experimented with different settings in the Shader Editor, but the result still doesn't look like the reference.

I've attached a few screenshots of my current material settings. My goal is to make the glass and aluminum look more realistic – I'm especially struggling with the reflections and transparency of the glass, as well as achieving a metallic look for the lid.

Here are some details about my setup:

  • Glass Shader: I'm currently using Glass BSDF.
  • Aluminum Shader: For the lid, I've set up Principled BSDF.

I'd really appreciate any tips on how to improve the realistic look. Maybe I'm also missing something with the lighting or render settings?

Reference picture Reference

Glass shader settings Glass

Aluminum shader settings Lid

My workspace Workspace

Current render result Render

$\endgroup$
4
  • $\begingroup$ That jar doesn't look like you've hollowed out the inside. Have you, and are you sure the normals are facing the correct way? $\endgroup$
    – TheLabCat
    Commented Nov 6 at 22:52
  • $\begingroup$ You know the jar in the reference image is itself not real, right? $\endgroup$
    – TheLabCat
    Commented Nov 6 at 22:53
  • $\begingroup$ @TheLabCat Yes, I’m aware that the reference isn’t a real object, but it does look much more realistic than my render. Thanks for the tip about hollowing out the inside—great suggestion! As for making sure the normals face the correct way, do you have any tips on how to check or adjust that? Thank you! $\endgroup$
    – Alex
    Commented Nov 7 at 9:01
  • 1
    $\begingroup$ For checking the normals go to the Overlays menu in the 3D Viewport, there is a Face Orientation checkbox. It works in all modes, but I guess in Solid mode it's best visible: the outside of objects should show blue, not red. If it shows red, go into Edit Mode, select all with A, then Shift+N to recalculate normals. But it does not look like a normal problem in your screenshots (or, even if there is a problem with the normals, they are not responsible for the general look of the glass). More on this in my answer. $\endgroup$ Commented Nov 7 at 13:56

1 Answer 1

2
$\begingroup$

For the aluminum shader, I would not change a lot. In my opinion it is just a bit too "shiny", I would increase the Roughness more. So I did not bother to model a cap or any real top for the jar in my answer.

As a preliminary, in the following example renders I have these settings under Light Paths > Max Bounces in the Render Properties:

  • Total: 64
  • Glossy: 16
  • Transmission: 64

bounce values

These are the important ones for glass materials, the others are negligible (but of course they should not be 0, not even the Volume bounces if you go for a volumetric shader in the glass, but it's less important there as those bounces are used for Volume Scatter).

So, for the glass material: first of all, I know there are a lot of tutorials teaching to mix a Glass BSDF or a transmissive Principled BSDF with a Transparent BSDF. And in some cases this helps especially when you want sharp shadows inside a glass object etc. But for the most part, a simple Glass BSDF or Principled BSDF with a Transmission > Weight of 1 are completely sufficient.

Here an example, just a colorless (= full white) glass material. On the left with transparency mixed in like your example, on the right just the glass. By the way, I left the Roughness at 0 to match your setup, but personally I would never use it like that - no glass is perfectly smooth, I would rather use a value of 0.03 to 0.05 usually.

glass + transparency vs glass

In my opinion the simple Glass BSDF looks even better, because it seems unrealistic that the jar does not throw a faint shadow on the plane below since glasses are not perfectly invisible to light.

So I'll go with the simple version for the rest of the explanations. Of course a Glass BSDF seems an obvious choice if you want to make a glass material. Another way would be to take a Principled BSDF with Transmission > Weight set to 1. If it looks darker, that's because the Base Color is a light grey with a value of 0.8 by default, not full white. So I change that first. Now comparing the Glass BSDF to the Principled BSDF, there seems to be no real difference between them:

glass vs principled

However, you will notice a very big difference if you want a colored glass like your jar with a dark saturated blue for example. While the Principled BSDF gives the surface a specular layer which reflects without tint, the Glass BSDF has only reflections tinted by the glass color like a metallic surface and these mirror-like reflections are what your jar is missing:

colored glass vs colored principled

To get something similar for the Glass BSDF you have to mix it with a Glossy BSDF, this mix should be angle dependent so a Fresnel node is a good choice here.

glass + glossy

If you want to go with the Principled BSDF but the reflections are too strong for your taste, you are often told that lowering the Specular > IOR Level will reduce/remove the reflections. While this is true for dielectric (= non-metallic) opaque diffuse surfaces, this has no effect on the specularity of metallic and transmissive materials. So in order to reduce the reflections, you can lower the Specular > Tint color from full white to something darker, blueish:

specular white vs specular blue

The Volume Absorption node in your setup is not really needed. When you make the surface such a strong saturated color as if the glass is color-coated on the outside, the volumetric effect is hardly visible.

Yes, it is not just helpful but rather mandatory if you want to create something like solid-colored glass which is colored throughout the material, so that the color is more intense on thicker parts. For that I would use a white surface color or just a light color with low saturation and then a strong color and high density for the Volume Absorption node. The white or light blue surface color would also help with the Glass BSDF having better reflections without the need to mix with a Glossy BSDF.

Here is a comparison between different surface colors in the Glass BSDF and varying Density values in the Volume Absorption node. The surface color affects the reflections and of course the overall color, while the volume density influences how much the material thickness affects the color. Of course this can be done with a Principled BSDF as well.

comparison surface color and volume density

To come to a close, to get more examples how the initially mentioned Max Bounces can be important to get good glass materials, read my answer to this question:

Grey artifacts in glass bottle

If you are not happy with a simple glass material and want to use Light Paths to mix in some transparency, I would not Add the Is Shadow Ray and Is Diffuse Ray together, but rather plug the Is Shadow Ray and the Diffuse Depth into a Math node set to Maximum to get a mix factor. This factor can also be multiplied by values between 0 and 1 to change the amount of transparency. More on this in my answer to this question:

Glass object's surface problem

$\endgroup$
6
  • $\begingroup$ Thank you so much for such a detailed and insightful answer! I really appreciate the time and effort you put into explaining everything so thoroughly. There’s still so much for me to learn, and your answer has given me a lot of valuable information and new approaches to try. I’ll experiment with the Roughness adjustment for the lid and I’m especially interested in testing out the Glass BSDF vs. Principled BSDF approach with the different color and reflection behaviors you mentioned. Thank you again for your patience and for sharing your knowledge. I’ll work on implementing these suggestions! $\endgroup$
    – Alex
    Commented Nov 7 at 19:38
  • $\begingroup$ I've tried out all the suggested variants, and the result is much more realistic now. However, there’s still an issue: in the rendered image, there’s a sort of hazy or matte layer at the widest part of the jar. I’ve also tried adjusting the Max Bounces under Light Paths, especially Glossy, but unfortunately, it didn’t resolve the problem. (I can't post screenshots here in the comment function, other ways to share a screenshot?). Are there any other settings or techniques I could try to reduce this matte effect? Thanks again for all your help! $\endgroup$
    – Alex
    Commented Nov 7 at 21:57
  • $\begingroup$ @Alex You could edit an image as additional information into your question. Hard to tell what a hazy or matte layer is without seeing it. Is it always in the same spot or does it change with the viewing angle? If so, maybe it is just a reflection of something in the environment. You could also upload your file here: https//blend-exchange.com and follow the instructions there to edit it into your question. $\endgroup$ Commented Nov 7 at 22:41
  • $\begingroup$ Thanks so much! I used blend-exchange.com to upload my file and adjusted the roughness as you suggested. Both methods—mixing Glass BSDF with Glossy BSDF and using Principled BSDF—gave good results. Here’s the file: blend-exchange.com/b/BSslqDYK The matte shimmer around the jar only shows up after rendering, and I can’t figure out why. Any advice for fixing this or improving the realism overall would be greatly appreciated. I'm working on a small online shop, so every tip helps a lot. Thank you! $\endgroup$
    – Alex
    Commented Nov 8 at 9:32
  • $\begingroup$ @Alex The whole html code at blend-exchange.com is supposed to be edited into the question to place the file there. Anyay, I see a matte shimmer in the lower half of the jar, I guess that's what you mean (a pic would have helped). It is a reflection of the environment. Of the Shadow Catcher to be precise. If you want to keep the object but do not want it in reflections, go to the catcher's Object Properties and under Visibility > Ray Visibility disable Glossy. Of course it affects Transmission as well, but if you disable that it looks as if it was pitch black behind the jar. $\endgroup$ Commented Nov 8 at 10:28

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .