Part of my job involves transcribing as many parts as I can into CAD form. This is for both my own reference and experimentation, as well as simply something enjoyable for me to perform. I've put this page together to display some of the paintball-related items I've modeled in the past. Hopefully this can help some people learn and/or understand how the markers work (learners by dissection). Even if they don't help anything, they're still fun to stare upon.

These images are example assemblies of multiple solid models of individual components. Each part of the marker is recreated then inserted in position with the surrounding components. The creation of assemblies like this can take anywhere from a few minutes to several hours, depending on the complexity and number of components. Generic parts such as o-rings, screws, barrels, and other common parts can be re-used on multiple assemblies (as long as they're visually compatible).
Takes this... ...into this:
Almost all the images on this page are solid models. This means the files were created in 3d space using the most infinite dimensions and curves. The alternative to solid models are called wireframe models, which are primarily used for graphics and 3d parts that aren't actually something you can hold in your hand. Solid models are used for parts production and manipulating of objects that appear in real space. An example of wireframe can be seen at the bottom of the page.

Impulse with a Tapeworm and most other stock parts.

Stock Ion with a stock exoskeleton (updated March 2006), transparent view.

Stock Ion similar to the view above, except this is a section view cutaway to show the internals. This file was used to create Chiumanfu's Ion super animation.
I'm told the paintball in this model looks like a meatball. The paintball is actually "rust" texture, although I admit it does look pretty meaty.

Smart Parts 360º swiveling QEV valve, made for the Ion (and pictured on an Ion). These pictures help show the capabilities of CAD work in that the QEV is relatively small and complex internally; the pictures show a good magnification of the model which is easy to do with a computer model.

Nerve with factory upgrade LPR extender, body porting, and Humphrey solenoid.

Same Nerve as above, except this is a section view to show the cutaway. I plan to re-make this picture with more accurate parts as well as o-rings, when I get around to it.

Shocker SFT with Parker solenoid, stock bolt, and most other stock parts. The second rendering shows some of the o-rings in their positions. These models aren't as detailed or complete as the other Shocker models I have below; I made these models back in 2003 and hadn't yet bothered to complete them to exacting specifications.

Shocker NXT transparent render. This is my most complete assembly (over 150 different parts, more than 400 connections between components). The marker has pictured an HE bolt and Parker solenoid, since they come on the NXT models. I made the second shot after being bullied into making a battery snap from somebody on shockerowners.

Shocker NXT transverse cutaway, with o-rings.

Shocker NXT perspective renders. This viewpoint mode compensates for the distances between viewpoints and objects, and adjusts their size accordingly.

Studio renderings of a Shocker NXT.

Shocker Sport marker with internals. These were probably the first paintball renderings I made, back in 2003 or so.

Studio renderings for a Shocker Sport, using more updated parts than the previous internal renderings.

Dye DM3 Matrix, not all parts finished but you get the idea.

AGD E-Mag. The body and firing assembly is complete, but I haven't finished the rest of the gun (except for the frame and rail, pictured).

AKALMP Excalibur, no circuit housing or frame yet.

Excalibur again, this time a transverse arrangement to show the internals (bolt and LPR on bottom, hammer and valve on top).

Invert Mini, sans regulator.

Body components for a PGI Mayhem.

Body components for an AirStar Nova 700.

Body components for a Deadlywind Aedes. This isn't an exact model, rather a representation I made on my own, but it's similar to the original.

Transparent Parker K4H03 solenoid from a Shocker/Impulse/Nerve/Bushmaster/Defiant/Intimidator/Tribal.

Same Parker solenoid from above, except with a solid cutaway view (second rendering is from the Shocker NXT model).

Shown here are a couple circuit boards (Morlock original, Predator2 board, and a Mini-Morlock). The Mini-Morlock board is actually a functional model and includes a working circuit. Inventor isn't made to model this type of thing, but that's not a problem for me.

Three generations of Shocker/Nerve upper boards. As with the above boards, these are functionally working in terms of circuit schematic. However, the components aren't to-scale (I made these models without measuring).

These are early models I made during the design phase of me SuperIon. I used this model to judge the required distance between the guns, and assess other design factors before I started doing things in reality (design twice, cut once?).

Demo for a Freak barrel system. Pictured is the aluminum 2002 series Freak back using a 12" Freak front.

Similar to above, except this is "perspective mode" to show the angles in a realistic setting.

Two examples of stress tests on CAD models (these tests were done using different software). The first picture shows a stress test based on air pressure being shunted through the bolt; the second picture shows a stress analysis on the weakest points surrounding the pressurized air chamber.

Renderings of regulators can be found on the How regulators work page.

This page is continuously updated with new assemblies and models as I finish and publish them. Coming soon are some models for a DM4, JBS King Cobra, additional regulators, and others.

Background:
These images and others like them are examples of solid model assemblies created in Autodesk Inventor. I'm an expert in many CAD programs and as a result I spend infrequent time transcribing equipment into CAD form, for my own reference and records.

Many people have asked me what I use as part of my job, or for hobbyist projects. Well, the answer isn't straightforward; for professional use it depends on the format required by the client, though for personal use it depends on the task at hand. I prefer to design parts principally in Autodesk Inventor, and make 2d prints using AutoCAD. When machining parts I write the G-code by hand, or use Mastercam to generate it if the surface is too complex.
Over the years I've used many programs, and am experienced in many environments, including the following: Pro-Engineer, Inventor, SolidWorks, AutoCAD (including 2d prints, 3d wireframe, or even old-school 3d solids), COSMOS FloWorks, COMSOL, ANSYS, Vectorworks, ABAQUS, CATIA v5, Think3, Solidedge, Inventor's ANSYS plugin, Surfcam, Mastercam, Autodesk 3ds max, Cinema4D, Autodesk Maya, Maxwell, and others. I'm not an expert in each and every of those systems, but I do have experience in each of them, and expertise in some.
Many of the above programs are geared for professional use and are expensive as a result. I only use a handful for my own use. There are many additional cheaper and even free CAD programs available (3dCAD, turboCAD, Alibre, etc), though naturally most of them will lack the advanced abilities found in a mainstream package.

A professional operator shouldn't find most paintball marker components to be excessively complex. Typically, markers tend to use a lot of small, simple parts arranged together, which won't take too long to create given a reasonable amount of detail. The complexity of the model will also depend on how the modeler chooses to create it in 3d space. Some of the more complicated components, such as those with complex internals, threads, or freeform surfaces may take more time to create due to the dependencies that the various surfaces hold with one-another. The majority of complicated parts are still relatively simple, unless it's an external part with cosmetic milling. However, those external parts with advanced cosmetic milling or advanced freeform curves will often take longer to model due to the operations that must be performed. Operators with a good spatial-cognitive sense are likely to have an easier time visualizing things in order to create the necessary functions before they're actually made. Unfortunately a side-effect of an extremely well-tuned cognative sense is subconsciously trying to mentally modeling objects when seeing them driving down the road, watching television, etc. I suffer from this and other related quirks quite heavily.