Learn How to Create a Beretta PX4 Storm with Unique Details

Introduction

Hello, again! I'm Eric, and I'm a Lead/Senior 3D Artist who loves firearms, especially within game art. For those reading who are also weapon artists, if you have joined The Weapon Room Discord, then you would know me over there as "epitaph".

I am an active and helpful contributor for all things technical! Since my last 80 Level article, the industry has continued to fall into a deeper rough spot, and as a result, I've unfortunately faced a layoff.

I kept myself busy by staying active in the weapon art community, mentoring/reviewing work over at Wrong Button (weapon outsourcing studio), and creating projects such as this one to keep my portfolio and skills sharp and active. Thankfully, I've recently found myself a new role!

Beretta PX4 Storm

It's interesting to think that sometimes all it takes to get inspired for a project is coming across an unfamiliar but interesting-looking weapon candidly, especially when the media you discovered it from is not a video game or weapon-oriented content. A cliche moment of when something "speaks to you".

After figuring out what that cool-looking gun I was looking at was called, analyzing and admiring the forms and silhouette alone is what sold me further into initiating the project. Giving it a closer look got some thoughts and gears spinning, and before I knew it, I was already equipped with a strong enough PureRef file to get started on this project.

To be more particular, the unique, largely angular shapes of the slide and soft flowing forms of the frame/trigger guard look very sleek and comfortably ergonomic. This piqued my interest when I realized this would be a great CAD surface modeling project to boot, seeing how this doesn't have that typical aggressive and primitive look, and that there are almost no extremely harsh/sharp shapes on any of the components that you would typically expect on most handguns.

What filled up the last of my inspiration was the fact that I generally like to synergize projects as much as I can: a friend's start-up project required a handgun, Wrong Button needed an extra handgun to add to their marketplace efforts, and I personally thought it was a great-looking project for my own portfolio. Everyone wins. Personally, knowing that it was going to satisfy multiple projects helped a lot with this project's pace and deliverable quality.

Working with References

When it comes to reference, I like to keep things practical, significant, and generalized. I didn't want anything too granular or detail-specific since I wasn't going for an unnecessary 1:1 recreation with this project. Also, with that note, barring scanned data, the closest way to achieve 1:1 accuracy is to have the gun literally right in front of you as a direct reference.

Collecting hundreds of granular references always seemed way too excessive for our needs for game art. Tarkov is actually a fine example of developing weapons as close to 1:1 as possible with game art, and those legendary weapon artists work with deactivated firearms right in front of them as they model to achieve that accuracy. 

World of Guns was also an incredible help in seeing how parts were roughly shaped and fit together, but not necessarily for modeling/texturing reference. Don't get me wrong, gathering and studying plenty of references is still important, but choosing to fill in some of the blanks yourself using your own mental/visual library is rewarding and proves to be a useful skill when dealing with the inevitable lack of reference for certain parts.

There were a couple of driving factors for the many design changes in this project: primarily, it was because this was being developed as a redistributable product, and as weapon artists, we constantly face trademark/copyright issues with our work. Unfortunately, it isn't as simple as changing the name/brand/logo, so some identifying features would have to be tweaked, changed, or omitted.

I was very sensitive to the fact that any design changes had to leave the gun still looking recognizable while still eliciting a "this FEELS like a PX4 Storm, but it isn't" response. Secondarily, it's time-consuming, laborious, and frankly boring to make a 1:1 exact replication of the original.

And for what purpose would that be, anyway? We're artists, it is simply much more liberating, fun, and easier to free-form some design choices yourself instead of having to study reference and fret about every inconsistency.  Below is an example of some of the design changes I had noted/planned for development, keeping the above two points in mind:

Modeling

For modeling, I used 3ds Max to block out, clean up, and UV, Plasticity for modeling the basemesh (using blockout as a rough guide), and ZBrush for creating a high-poly out of the basemesh. Approaching the intricate details is a stage all of its own. It's something that shouldn't be done too early on.

Doing so can not only tunnel-vision your work but also invite a high likelihood of needing to re-work or re-fit things, as newly established large forms/shapes could invalidate the accuracy of those details. Essentially, once the base large/medium forms are modeled and confirmed to fit well with each other is then I start to look at tackling each of the more complex details. 

The rear of the slide, in particular, is an example of something I spent more time than I'd like to admit: it had many compound shapes that affected one another. If I tweaked one, then what I thought looked acceptable somewhere else now looks wonky.

This form was a great example of proving that it is much faster to start from scratch than trying to adjust what doesn't work, a brutal indication that the flow of large -> medium -> small modeling wasn't quite right.

Plasticity

Plasticity has been involved in my workflow for nearly two years now. It caught my eye just as I was getting the hang of parametric CAD modeling in Fusion 360. What I appreciate most about Plasticity is its intuitiveness. It's incredibly easy to pick up after watching a few quick-start videos online, enabling anyone to start modeling right away.

The fact that it's a direct modeler instead of a parametric modeler makes it that much more instinctive to do most of your operations. I also really appreciate the fact that it has an active and open-minded development team that listens to its incredibly active community, releasing anticipated features and fixes (both really refreshing to experience after dealing with other giant software entities).

Plasticity has recently stepped up its OBJ exporter quality. MOI is a better alternative, but I still stand by using PiXYZ for superior CAD to polygon conversion. The observed results after comparing with other exporters are generally much easier to manage and tidy up.

I personally like to work with lower resolution topology and adding fidelity when cleaning up, as I find it is far easier to manage, clean, add, and space loops out evenly than it is to painstakingly remove a dense amount of loops with a denser export. I will say, though, it is a big shame that PiXYZ cannot export as n-gons/export without triangulation. MOI/Plasticity's n-gon export makes sections much easier to visualize and deal with.

Topology is very context-sensitive. Should many questions be answered to justify topology decisions: what is the asset for (actual gameplay use or something set dressed in the environment)? How close is it typically seen on screen (first person vs third person vs item-pickup on the ground)? Are there any other scenarios (360 inspectable customization)? 

Knowing that this weapon is for first-person use and would very likely be inspectable in customization features gives me enough justification for making sure there is plenty of density to go around, nearly universally, for establishing well-defined silhouettes and not cutting corners on areas that are outside of typical FPV (since it is inspectable). 

Frequently checking glancing angles and respective camera view distances helps determine if a particular detail or feature has too little or too much density. Being constantly zoomed in where the detail takes up your entire screen space invites tunnel vision, which often leads to breaking that important context we established earlier.

Disabling wireframe on and off also helps judge if you can get away with a lower amount of topology while not affecting the shading too negatively. Regardless of context, though, areas that are occluded receive much less topology for obvious reasons.

Details that are convex-shaped get a lot of topology reduction as well, since their silhouette practically cannot be appreciated, and also that mesh smoothing takes care of most of the work.

UVing is straightforward: split by sharp edges, relax, straighten edges where possible, align to pixel grid, and pack together with suitable margin and padding values.

However, the actual process is more nuanced than that. UVing becomes a balancing act of placing your sharp edges and managing your mesh shading to grant yourself as few seams as possible (leveraging beveled edges to replace sharp edge placement and forgoing a UV split is a great example).

Conversely, relying on sharpening all angle changes to achieve flat shading can do the job, but not without generating a ton of UV splits/texture seams, which is often not ideal. Controlling your shading with bevels/support loops not only greatly reduces your UV splits/texture seams, but also gives you the bonus of more rounded corners.

Something that I've heard years ago that stuck with me: "If you aren't planning or imagining your UV situation while modeling/cleaning up your LP, you are already behind." I tend to plan my UVs as I'm modeling, well before I even get to UVing.

What that looks like is simply just being mindful and making pre-conceived modeling decisions for your UVs. Some examples of that sort of behavior would be analyzing mesh shading, judging sharp edge placement, and predicting if something would be an issue for bakes/textures down the line.

If there is some pesky shading encountered, my initial reaction is to experiment with adding some edges/control-loops to minimize the shading as much as possible, allowing me to keep UVs contiguous and seam-free. I avoid placing a sharp edge as my first solution, only as my last if experimenting with other methods wasn't ideal.

This analytical "future-proof" thought process helps greatly to make UVing that much more pain-free since a lot of those issues and kinks are worked out long before they could become a surprise.

Texturing

When it comes to materials, I'm looking for something eye-catching and identifiable from a distance, something that shows glossiness details well (stainless steel), all without clashing with another material choice, a careful choice of balance. You don't necessarily have to commit to the same finish that you see in your reference!

Once decided, and some additional specific reference of those materials is collected, I start my texturing process using Substance 3D Painter in a very similar fashion to blocking out a model. Start with large, "close-enough" material values both in color, specularity, and gloss, see how they read from a distance (to judge value contrast and readability) and glancing angles (to catch the light to observe gloss/specular readability).

Medium details can then be laid on top of this solid, established foundation, and finer/small details as finishing touches or characteristic details. Material definition is critical here, we want our materials to not look vague at all, our work should behave and have all of the characteristics of the material in mind: closely analyze, experiment, and nail down those values that define a material (polymer being noisy in height, not as glossy against stainless steel which is bright and glossy as a simple example of defining two materials that are neighboring each other).

I typically never texture one material at a time, I bounce around to encourage balance and to not get too deep into one material, and having to start that entire process again from scratch when it comes to the next material. There could be some context deep into one material that would be great to transfer to the next material, but would not be at all possible if the other material is still a blank white fill layer.

Try to ensure other materials are "caught up" relatively close. Here's a GIF of how I built up my polymer material, for example:

A lot of presentation ideas and scenarios that I had in mind were an accumulation of what I had come across when searching for reference at the start of the project. In fact, I kept a separate section in my PureRef file solely for presentation inspiration/ideas stumbled upon: the way the weapon was laid out on surfaces, the angle/perspective of how the weapon was held in the hands of operators, and the cold, sterile indoor lighting used in many images are all things that caught my eye.

For presentation, I tend to present my project in various lighting setups, as I feel like rendering a dozen shots with one setup can be a little stale and not totally show off how versatile your project can look in various scenarios. Almost all of my renders are simple 3-point lighting set-ups and follow the same ideology however, fewer lights are more.

A main key light that lights the model up well enough without blowing out any values, a fill light or two to subtly light up the more occluded areas that still need attention, and a strong rim light to accentuate silhouettes and forms. For outdoor, more natural scenes, it’s the opposite: more reliance on the HDR and a couple (really not much more than two) directional lights can go a long way in having your work feel natural and grounded in the environment.

For the indoor, studio-lit environments (I call these "sterile" environments, since it's artificial, clean, and often cold lighting), I rely much less on the HDR amount and more on the placement and balance of directional/omni lights. 

Once lighting is set, remember that once you start "presenting" your work by rotating and moving your model around, the lighting could very much change since the angles of the model and direction of the lighting no longer play off of each other. Just be sure to tweak the lighting if you ever nudge/adjust your cameras or model.

Conclusion

This project took about two months for me to complete. I started it right after my last project wrapped up in the summer. Admittedly, it was a leisurely, but disciplined and consistent pace of about 1-4 hours a day of work (depending on the stage) throughout the weeks.

The other endeavors aforementioned kept me busy, but were welcomed as they allowed me to take some much-needed mental and visual breaks from the project. Other than learning something unfamiliar, such as surface modeling/CAD modeling, the main challenge I faced was to make sure my project was impactful in the end.

There is so much phenomenal weapon art frequently produced, compounded with stunning lighting and presentations, that I certainly felt the pressure from the challenge of making something that could maybe hold up to those standards. Typically, the process I enjoy the most is generating a very nicely shaded and optimized final game-res mesh to bake from with very particular UVing.

This phase is quite literally "make or break" for the project. But for this time around, what I really enjoyed was the fulfillment received with surface modeling in Plasticity. Being able to establish forms and shapes with no continuity/tangent issues (after lots of trial and error, of course) was really fun and rewarding. 

A lesson learned here was not to be afraid to iterate: instead of pacing around wondering how to tackle a shape, just give it a stab and iterate upon it to achieve your shape. If it's not working, try another approach entirely instead of fighting with something that isn't working. It's quite difficult (and sometimes not even feasible) to nail down a complex shape right away.

Keep track of everything as you produce your work (if you wish). It's virtually free and easy to track when using tools like PureRef. I like to keep details like feedback notes, render/presentation ideas, specific texture detail from other subject matter that would look cool on this project, design changes/personal notes, previous iterations screenshots, etc., in my PureRef file.

I find that it helps get you started to finish up the next task/stake, and also helps to look back and reflect on past efforts and internalize the entire long process that often flees away from our consciousness. Plus, it is handy to have when demonstrating or sharing information with others. My advice to artists would be:

• Reference is extremely important, but do not be bound by it. Use your creativity to fill in the blanks and treat the project as your own creation.
• Great textures cannot cover up bad modeling and/or poor UVs; take a lot of care and time towards what is integral.
• Tackle things Large -> Medium -> Small regarding practically every area of production.
• Scrutinize from a distance; always make decisions with your project context in mind.
• Develop your materials almost in tandem with each other. Try not to leave other materials too far behind, since they can synergize with each other very well and speed up texturing.
• Re-use "generic" detail fill layers across all of your materials. Not only does this speed things up, but it keeps your materials consistent and grounded with each other.

If you've come this far, a sincere thanks for reading this. I hope you were able to take away something useful and informative that you can apply to your future projects! Please feel free to give this project a closer look at ArtStation.

I would greatly appreciate it if you dropped a like! I also would like to plug my latest product, The Ultimate Bag of 3D Tricks, if you are interested in wielding some more advanced 3d techniques and processes to amplify your next project!

Thank you, 80 Level, for another opportunity to write about my latest work! I enjoyed putting this all together with the hopes that there is some useful information that can help other artists tuning in. One more piece of advice: if you really want to grow as an artist and build out connections, be sure to involve yourself with like-minded communities.

The Weapon Room is a fantastic example of a great community of artists with professional feedback for hard-surface art (though all artists are welcome, of course).

Eric Correia, Senior/Lead 3D Artist

Interview conducted by Amber Rutherford