Some artist's using metallic values in painted metal for guns?

guitarguy00 · Apr 2020

I have been noticing that some high level weapons artists have metalness in their materials in where it would be considered painted metal(which would mean no metalness). I have seen this on multiple occasions on Artstation and also in texture packs. This is very confusing to me as in reality, only the exposed underlying metal surfaces(through edge wear, scuffs etc) would have a metallic value but not the actual painted surface.

I'm not talking about weapons like a Desert Eagle where the body is obviously metal with no paint on top whatsoever. There is also 'gun metal' which is metallic but that is never black or even near the color black. Here is an example of what I am talking about:

https://www.artstation.com/artwork/qAorrL

You can see that only the wooden parts in the metallic map are black(as in non-metallic) but everything else is metallic. The magazine looks painted black to me but has a metallic value in the map all over. It even has edge wear revealing the metal material underneath but yet the whole object is metallic?

Is this done just for aesthetic value(it just looks better?) or is there such thing as a black colored metal that is not painted?

rollin · Apr 2020

Imo this is the result if this "OMG IT'S PBR!!!" credo.

Reality is you can make things look good even if you completely do everything 'wrong'.

Kanni3d · Apr 2020

Yeah, the result of straying away from the rules of PBR looks good here when done carefully, but yes it isnt physically correct. (and would turn up red if you were to run it under a PBR validator filter in substance)

FrankPolygon · Apr 2020

The finishing treatments and protective coatings for metals like iron and steel aren't strictly limited to dielectrics like paint, varnish, wax, epoxy, etc. Two common finishes for firearms of this vintage are bluing and parkerizing. Under harsh lighting, a fresh example with a good surface finish could be mistaken for some other kind of coating.

Bluing is black iron oxide Fe₃O₄ and parkerizing is generally manganese phosphate Mn₃(PO₄)₂, zinc phosphate H₄O₁₂P₂Zn₃ or iron phosphate FePO₄. Both coatings can be derived from iron. Black iron oxide coatings are conductive and phosphate coatings are non-conductive but aren't barriers to galvanic corrosion. Color variations in the coatings can be influenced by the conversion process, materials, contaminates, byproducts, dyes, etc.

They're both used as protective coatings because they resist further oxidation and hold a thin coat of oil, which helps protect the underlying metal. The type, color and saturation of the oil can also have an effect on the appearance of these coatings. Certain oils and oil contaminates also have conductive properties.

The quality of the starting surface finish also has an effect on the appearance of the final surface finish and can influence how a coating will behave. Fine machining and polishing will produce a smoother, richer and uniform surface finish where rougher machining without polishing will produce a rougher, patchier and uneven surface finish. In more contemporary production there's also stipple patterns and micro abrasions from shot blasting and tumbling processes that produce more surface texture.

All of these factors can be manipulated to produce finishes of varying quality. Everything from a deep glossy sheen to a rough matte finish. Add the color variability from different coating processes, material quality, machining quality, etc. and it produces a wide range of real world examples to choose from.

Compare WWII production from early to late and the effects of surface finish and treatment processes becomes more apparent. Early in the war there's more of an emphasis on quality. Late in the war there's more of an emphasis on quantity.

It's relatively common to see late war examples with severe surface finish problems like chatter marks from dull tools, gouges and scratches from improper machine settings and all kinds of other surface imperfections that no one really had the time to care about. Over the course of the war there's also a wide swing in finish color, quality and luster for both bluing and parkerizing.

Towards the end of the war, if the parts passed the go / no go tests then they were good. Almost all of the prewar cosmetic stuff was a waste of time and effort. A good example is Singer: they tried to tool up to produce M1911's but did such a good job with the fit and finish they were told it was a waste of manpower and they should produce something more complex instead.

Reference images are great but they don't always provide these details. Older stuff is usually pretty easy to figure out because it has patina but newer stuff or new old stock with pristine finishes can be a bit trickier to figure out. That said: bluing and parkerizing are still fairly common coatings.

For that particular asset, it looks like the material values are mostly correct, relative to the real world properties of the materials being represented.

If an artist is following the basic guidelines and principles of the PBR system it still leaves a to be decided by personal taste. There's a similar discussion about the importance of PBR and material values here: https://polycount.com/discussion/comment/2711887/#Comment_2711887

gnoop · Apr 2020

I just don't make too much of metalnes variation since in our engine it creates ugly halo effects . So it's either all flat metal or not, and often kind of slightly grayish for outdoor dusty structures.

That typical metalnes kind of PBR is just another pain in the a.. you fight with regularly that had supposed to help artists by making art fool proof but in reality ended up as just limiting your ability to make important subtle touches.

guitarguy00 · Apr 2020

FrankPolygon said:

The finishing treatments and protective coatings for metals like iron and steel aren't strictly limited to dielectrics like paint, varnish, wax, epoxy, etc. Two common finishes for firearms of this vintage are bluing and parkerizing. Under harsh lighting, a fresh example with a good surface finish could be mistaken for some other kind of coating.

Thanks heaps for the detailed reply. I have started to look into parkerized steel and it is used quite frequently. Now I have a question that you most likely already knew I would ask, is parkerized steel considered 'metallic'? The Parkerized Steel in Substance Painter says it is Metallic but not sure if this is accurate.

The slide of this Glock is Parkerized, would I have it as metallic?

Image: https://us.v-cdn.net/5021068/uploads/editor/6s/pr45ozetiab9.png

zachagreg · Apr 2020

@guitarguy00 That is accurate, the "Parkerizing" or Parco-Lubrite finish is still considered metallic due to the chemical breakdown on the surface when introduced to the metal. Still results in ferric phosphate or Iron(III) phosphate. A lot of the finish is micro noise which would show up in your roughness map and some in your metalness.

FrankPolygon · Apr 2020

In the context of PBR texturing (with the metalness workflow) the phosphate coatings aren't pure metals but they can have visual properties similar to metals so they'd fall into the metalloid category. A lot of stuff is still approximate and there can be contextual differences so it's important to experiment with your tools / target engine and focus on the visual accuracy of the results.

To quote the substance PBR guide:

"When creating materials for PBR, it is helpful to think in terms of metal or non-metal. Ask yourself if the surface is metal or not. If it is, you will need to follow one set of guidelines. If it is not, you will need to follow another.

This can be a simplistic approach as some materials may not fall into these categories such as metalloids (a mix of metal and non-metal), but in the overall process of creating materials, distinguishing between metal and non-metal is a good approach and metalloids are an exception. To set up guidelines for materials, we must first understand what we are trying to create."

https://academy.substance3d.com/courses/the-pbr-guide-part-1

https://academy.substance3d.com/courses/the-pbr-guide-part-2

When it comes to personal opinion: I'd have to agree with Zach and gnoop. Using a metallic value helps generate the sheen from off axis lighting and the micro texture of the surface finish needs to be controlled with inputs from roughness, normal, etc. It's normal to want to have hard answers for things that seem like binary options but the real world doesn't always translate 1:1 into a graphics pipeline.

There's also a lot of variation in these phosphate coatings. To replicate one you might need to use a metallness value but to replicate another you might need to use a non-metal value. It's been my experience that best practices, guidelines and real world values are a good starting point for working with PBR textures but they aren't the be all end all.

Whether or not something is "right" can come down to what the asset looks like at render time. Poorly done dirt and wear layers on-top of materials with poor separation and read has the potential to be more problematic than using slightly unrealistic base values for materials. Sometimes it's more helpful to make sample assets and judge the results rather than guessing at what works better in theory.

guitarguy00 · Apr 2020

gnoop said:

I just don't make too much of metalnes variation since in our engine it creates ugly halo effects . So it's either all flat metal or not, and often kind of slightly grayish for outdoor dusty structures.

That typical metalnes kind of PBR is just another pain in the a.. you fight with regularly that had supposed to help artists by making art fool proof but in reality ended up as just limiting your ability to make important subtle touches.

Hmmm interesting. What engine would that be?

guitarguy00 · Apr 2020

zachagreg said:

@guitarguy00 That is accurate, the "Parkerizing" or Parco-Lubrite finish is still considered metallic due to the chemical breakdown on the surface when introduced to the metal. Still results in ferric phosphate or Iron(III) phosphate. A lot of the finish is micro noise which would show up in your roughness map and some in your metalness.

Thanks for that, for the micronoise I could use something like "white noise" or "blue noise" in Substance with a higher scale in the roughness. For the metalness side, would you recommend a very subtle amount such as 0.8(with full metalness being 1) on that layer?

zachagreg · Apr 2020

yea that sounds like it wont throw any strangeness for rendering, i would also experiment with directional noise mixed in on another layer as well as the milling and machining process will usually leave directional patterns along the length of metal pieces.

guitarguy00 · Apr 2020

FrankPolygon said:

In the context of PBR texturing (with the metalness workflow) the phosphate coatings aren't pure metals but they can have visual properties similar to metals so they'd fall into the metalloid category. A lot of stuff is still approximate and there can be contextual differences so it's important to experiment with your tools / target engine and focus on the visual accuracy of the results.

Thanks again. There is definitely alot to digest there. I did read those PBR guides briefly but the first one definitely seemed like theoretical overload for me. The second one was alot more understandable. I will give them another look over. Any other documents, links or videos you recommend for me to help in this area? I really want to improve my texturing.

Also, how do you usually find out if a gun has been parkerized? Can you tell from looking at the metal? I am struggling to know if this standard Uzi is parkerized or if it is just a painted metal/ceramic coating.

Image: https://us.v-cdn.net/5021068/uploads/editor/dd/vrfn59dhk3tw.jpg

FrankPolygon · Apr 2020

The short answer is yes. With some research and experience it is possible to look at a single picture and get a rough idea of what coating was used but it's not going to be perfectly accurate. There's a lot of contextual information missing and in this case a single picture on it's own isn't representative of anything other than what's in that picture.

A good starting point for determining finishes is to look at the time period the item was produced in. Prior to the 1920's it's going to be mostly case hardening and bluing. Prior to the 1970's is mostly going to be bluing and parkerizing with some anodizing for aluminum parts. From the 1980's to the 1990's there's some weird stuff with painted finishes. Then from the early 2000's on it's more exotic stuff like ceramic coatings and specialty finishes.

Another thing to look at is the purpose of the manufacturing run: was it a military run or a civilian run, which country purchased it, etc. There's certain time periods where there's massive production runs and a lot of variability from different processes, materials and design changes. In WW2 there were batches of 1911's that came out this weird olive green because the parkerizing bath at the plant was contaminated. In Vietnam there were a bunch of experimental versions of the M16. This is why context is important.

Looking at the example image above: The barrel and the bolt are probably parkerized because they have a grainy surface texture and dull gray sheen and sheet metal parts are probably coated with something else. It doesn't look like parkerizing because the surface finish is too even and too glossy. It doesn't look like bluing because the highlight is white and washed out but it's not possible to say for sure that it's NOT a very dark bluing that's reading differently because of the lighting in the picture.

This is the problem with trying to use a single picture to cover all of the nuanced details. It's just too limited. Multiple pictures of the same object or video showing different angles and lighting conditions is a much better source.

A quick text search of the production history shows that there three common finish variations: parkerized, parkerized then painted and blued. Comparing it to other photos in a cursory image search and it seems that this one is likely parkerized then painted but only on the sheet metal parts. The internals are likely all parkerized.

But that still doesn't answer the question because the reference image hasn't been qualified. This is a pop culture icon so there's a lot of knock offs and movie props floating around. The assumption is that those don't count. Then there's the issue of different manufactures using different materials and finishing processes. So this brings up the accuracy question...

The originals were manufactured by IMI and this is manufactured by a different company. Different manufactures have different specifications, different materials and different processes so it's likely there's some variation.

If you're trying to replicate a model with specific provenance for a specific time period then this reference image isn't usable much less accurate. Loosening those requirements to "a standard model" is still problematic because who's standard model are you looking for? What condition is acceptable? Do refinishes during official service or post service rebuilds count? Where do you draw the line on what you're looking for in a "standard model"?

In the strictest sense of "standard model" it still doesn't seem usable (as a material reference) because it wasn't produced by IMI. The qualifying standards have to be loosened to include knock offs for it to be an acceptable reference. To get to a "standard model" that will require gathering additional references and qualifying those as well. From there the options are: average out the material's appearance to where it matches most of the finishes in the references or pick a specific subset and replicate that finish.

It all comes down to how accurate something needs to be. Single references can be used to prove that something is possible but the real question is how probable or common was it and does this fit with the theme of the project. This is why it's my opinion that any excessive obsession with technical accuracy over the art and story elements isn't the best use of resources.

Pedantic for sure but once you go beyond eyeballing general references then it becomes a more nuanced discussion. The question that should be answered first is: does this serve the object's story and will the audience even care? If the answer is half no then this level of accuracy is something that's nice to have but isn't essential and if the answer is a full no then it's probably a waste of time. The other side of this coin is the people who do care about these sort of things are VERY pedantic.

In most cases the sensible thing to do seems to be gather a pool of qualified references, start with a base material and adjust it by eye until it's close enough to hit the quality standard of the project. If you have access to samples and can test and scan then that's a different situation. Most artists won't have access to these objects so it's going to be a process of eyeballing it anyway. Which is why gathering and comparing a lot of references is important.

Udjani · Apr 2020

You can find ''dark metal'' pretty often on artstation in props, one thing that i found playing with it is that if you make a dark metal is better to set the roughness usually a bit higher. If the metal is too glossy it can look good in renders on marmoset, but they will likely look pretty bad on a game engine since you don't have 13 lights following the gun at the right camera angle all the time.

zachagreg · Apr 2020

Udjani said:

You can find ''dark metal'' pretty often on artstation in props, one thing that i found playing with it is that if you make a dark metal is better to set the roughness usually a bit higher. If the metal is too glossy it can look good in renders on marmoset, but they will likely look pretty bad on a game engine since you don't have 13 lights following the gun at the right camera angle all the time.

This is a really good point to make, a lot of what you are going to do for a prop is dependent upon what your target output is. If all you want to do is make a render out of marmoset then thats fine. But if your target is a game engine if it doesn't look right in your target then you've lost the plot so to speak. PBR is a way to guide that consistency between target render platforms as physically accurate systems will respond to lighting in a predictable way. It is possible to work outside of those guidelines as well.

rollin · Apr 2020

FrankPolygon said:

(...) adjust it by eye until it's close enough to hit the quality standard of the project. (...)

I love how this is practically the tl;dr

However, thx for your detailed information. I have the feeling everyone should have a FrankPolygon around to ask questions about metal coating!

Some artist's using metallic values in painted metal for guns?

Replies