The cutaway is an excellent tool for an automaker to show off the sweet technical guts of its latest car. They’re also great educational tools to help young gearheads and aviators learn how their vehicles and engines work. But how do you think about how these cutaways are made? Sure, some are just photos of an actual vehicle chopped apart, but what about rendered images?
This morning, Jason showed us the glorious cutaway images put out by General Motors in the 1980s. I mean, the GMC S-15 is such a simple truck, yet GM gave it an incredible cutaway image.
Reader Pneumatic Tool gives us stunning details on how these images are made:
I’m (primarily) a graphic artist by trade and was in college during the late 80’s/early 90’s. These kind of drawings always fascinated me, so I paid close attention to the breakdowns of how these things are/were done as I learned various illustration techniques. When I was in school, the first shots of the digital revolution were being fired…these cutaways were created well before that. With the exception of the Toronado (photo composite), these were hand rendered. That doesn’t mean that they were *hand drawn*…it’s likely that the isometrics were drawn in CAD and handed over to the artist who traced them on his board. That’s how it was done…if a drawing looked too good to be true, it was traced, and yes you can transfer a tracing from a photo or print onto an opaque illustration board by hand – I did it many times (just a matter of learning the tricks).
Now the real fun begins. What I’m seeing here is a frenzy of mixed media; gouache (opaque watercolor), markers, technical pens, airbrush, colored pencils, you name it. Orchestrating all of that is where the the artist shines. Using an airbrush for illustration involves putting a sheet of transparent material (frisket film) over a drawing that has been broken down like a paint-by-numbers project (shapes represent different colors or shades), then you take a x-acto knife and veeeery carefully cut out all of those shapes. Using some kind of photo reference (or your mind) you lift off these shapes one by one and color each area individually, adding gradient within each. When you’re done with that area, you carefully replace the frisket, rinse your airbrush, and start the next one. The problem is that by working like this (lift off, replace, lift off another, replace), you have no idea what the overall illustration actually looks like as your working – fun! Basically, you have to imagine what the right thing is and work towards what is in your mind. You only know if you’re right when you lift all the frisket off. If it worked, you’re golden…if not, either start over or try to fix it at the peril of making it worse.
Once you’re done with the airbrushing, you go back in and kick up the areas that need it – adding pure white highlights with whiteout, details with colored pencils, maybe some solid water colors, technical pens, straight watercolor, graphite, etc.
Ultimately, these people did a lot of work to create illustrations like this in the pre digital age. Honestly, there’s no comparison between now and then. It was truly an art unto itself.
Things appear to be going bad for Stellantis and the Ram 1500 is now behind the Toyota Tundra in market share. Reader H4llelujah offers insight from the dealership side:
I hate that I keep harping on this and whining about Stellantis, but as a Ram salesperson, there’s a few factors at work here:
The biggest one is the amount of people trading in Ram leases for another Ram has absolutely CRATERED. People are not going to pay 499 a month for 3 years, and lease the EXACT same truck again for 699 when they can slide over to Ford, GM, or Toyota, and try something new for less money. Stellantis is stubbornly keeping rebates low, and lease rates high, for what reason I don’t know.
2nd, they dumped the hemi, leaving 2 options: an unknown (boosted) engine, and a reliable yet severely underpowered V6. GM and Ford will happily still sell you a V8, again, for less money.
3rd, Quality has just not gotten better. The 14-18 Ram bodystyle was incredibly reliable, and the 18 and newer style has been less so. Heater cores, Manifolds, front end parts, leaking back windows are all recurring issues. You fix them, they go bad again within a year. It’s not enough to make a man swear off the trucks, but it’s just enough to make a man think about trying something else.
It’s such a mess here at Stellantis, I’m actively planning my exit from the car business. I stake my reputation on the cars and trucks I sell, and from what Ive seen the past year, I cannot trust this manufacturer to make anything other than the wrong choices. I’m about fed up, and it looks like others are as well.
Have a great evening, everyone!
The main reason Ram was a constant #3 was because it was more affordable than GM and Ford. It’s no longer the case, it’s customers are less loyal than those two brands as well as Toyota’s customers.
Bring the rebates already and customers will start getting the Ram again.
Engine downsizing didn’t negatively affect Ford nor Toyota, and wouldn’t affect Ram either as there are many 2024 samples with the Hemi laying around on dealer lots yet no one takes them home
Those art nerds at the College for Creative Studies really have done some excellent work for the Automotives. Still the best place if you want to get into car design.
As for the Hurricane I6 you can get..RAM buyers are the toughest market to shift from the “no replacement for displacement” demographic, but if they got the people from the GME-T4 team that I worked on (previously called the “Hurricane”) I do trust their work. Since it’s 2 years into production (the Chrysler build quality tax) I’d give it a good thought.
While I don’t have experience with either, I’d rather have the Ram I6 than the Ford turbo V6. And I can’t believe I’m typing this, but I’d take either over a V8. I’m done with 12mpg vehicles.
But since I can’t buy new cars, my opinion doesn’t matter. I’ll take whichever one is still reliable past 200k miles.
Will not rant about AI and the loss of artistic integrity.
Will not rant about AI and the loss of artistic integrity.
Will not rant about AI and the loss of artistic integrity.
*clutches colored pencil set against chest protectively
waves exacto knife around defensively*
I will not rant about AI and the loss of human integrity!
AI can’t do details like this. Sure you can ask it make a cutaway image of a car, but even if you specify the make and model, it has no idea what the technical inner workings look like. It would just make shit up.
AI can work if the details don’t matter. But the thing is, details do in fact matter for most forms of art that have moved past the basic concepts phase.
Most AI art can’t pass the three second test. If you flash the image quickly, it looks amazing. Stare at it for any amount longer and the errors start jumping out.
Some things AI can do very well like the human face. But basically all faces are laid out the same and the amount of image data out there of just headshots is immense. Plus no two people look alike, so as long as the features are sized sanely, the face is going to look believable.
I’m a guy who went to art school BTW.
I agree. AI is pretty easy to spot.
For now. But the technology will, at some point, grow at a near exponential rate and one day there will be entire generations of humans that can’t tell the difference.
Damnit!
I will not rant about AI.
That’s actually pretty debatable. AI is only as good as the data it’s being fed. It isn’t god. At some point the amount of good data dries up and these models start being fed data created by other AIs. Open AI is already doing this.
Will it increase your data set? Absolutely. Will it make the model respond with more appropriate responses? Maybe, sometimes.
There’s a ton of speculation that the current AI models may actually be reaching their peaks due to how they work. They’ve already been fed most of the available data that exists in the world. And it’s not enough.
I hear voices in my head. They are voices of reason.
I don’t dare believe you, because I so strongly hope that you are correct.
People keep saying this, but I think they’re misunderstanding how the technology works and/or believing the marketing hype. Current AI is not improving at an exponential rate. It’s improving at a pretty marginal rate, and there are likely to be diminishing returns moving forward.
The thing to keep in mind is that machine learning isn’t new. Some of the generative applications of it are, but this isn’t a brand new thing that we’re just figuring out. The example I keep going back to is the IBM Watson AI on Jeopardy. It was impressive how well it worked. It also hallucinated obviously wrong answers to things. Compare that to ChatGPT. It’s impressive how well it works. It also hallucinates obviously wrong answers to things.
So have we made exponential progress in the intervening decade+? Not just no, but I think there’s a compelling argument that the underlying technology hasn’t improved much at all. Maybe it still will, but I don’t think it’s the foregone conclusion many people assume it to be.
I went to art school too (the places you can go with an MFA!)
AI is a really cool tool, but I can’t imagine anything more useless for technical drawings. Sexy rabbit robots in flying saucers fighting with cowboys on space squid? Oh yeah.
A pair of scissors? … Well actually AI scissors are pretty hysterical in a drop acid and make pottery sort of way, and the absolutely have the essence of scissors, but they are very much cow tools.
I think in the arts it’s at the “photography will put painters out of business”, or “synthesizer will put musicians out of business” stage.
The competing with idiots who will work for nothing and clients who don’t know the difference is a thing, but its been that way for the past century.
I know so much about so many lost technologies, like how to strip dye transfers. Do not miss waxers at all.
Yeah I’ve messed around with AI art a ton. I even had a midjourny account for a while. AI is great at making surreal art because you can’t compare it to anything else. What’s to say you sexy rabbit robot is better than my sexy rabbit robot. What does that even look like?
But as soon as you ask it to draw out a character sheet needed to 3D model the sexy rabbit robot it literally just rendered, it shits the bed. That has specific requirements that cannot be fudged and generalized away.
Pretty much my point exactly.
I used to know an artist that had the old Ford dealership in Santa Monica as his studio. In what had been the service area/indoor parking and so on he had hundreds and hundreds of folding academy tables, Each covered about 2 inches deep in publicity photographs, Hollywood studio stills, advertisements, magazine clippings, art postcards, and just about everything from soup to NASA moon photographs. Anyway, he would just walk through all the stuff sifting through it, sorting it out, finding things that sort of related to each other, but not really. He would find stuff that looked interesting, but most of it didn’t look like his art. When he found something that looked like something that he would do, there it was.
It’s like William Burroughs or Philip K Dick. Anybody, well lots of people of my acquaintance anyway, can take the right kind of drugs, and have hallucinations and visions. Organizing it, editing it, figuring out what it means, and making it interesting to other people is entirely different.
AI is like a research librarian on really excellent peyote. Sometimes you get really amazing stuff that you would not have tought to combine but sometimes you just get puke.
I do NOT miss making Matchprints.