The Man Who Invented the Alpha Channel: Alvy Ray Smith on Pixels, Pixar, and the AI Revolution

Episode 100 of the CG Pro Podcast — a conversation with a true pioneer of digital image making.

We hit a milestone with episode 100, and it felt right to mark it with someone who was there at the very beginning of everything we do. Alvy Ray Smith co-founded Pixar, co-invented the alpha channel, created the HSV color model that's still baked into every piece of software you use today, and recently wrote A Biography of the Pixel — a sweeping history of how digital images came to be. It was one of the best conversations we've had on the show, and this post pulls out the best bits for you.

Pixels were never little squares

Ask most people what a pixel is and they'll say something like "a tiny square on a screen." Alvy's been trying to correct this myth for decades, and he devoted a whole book to it:

"It's not a little square, folks. It never has been — ever in the entire history of pixels. That seems to be just a comfortable myth."

What's even more surprising is how old pixels actually are. Alvy traced the history back and found that the first computer — Baby, built in Manchester in 1948 — already had pixels. Computation and pixels were born together.

The moment everything changed: seeing color for the first time

Before Alvy landed at Xerox PARC, computer graphics existed — but it was all black and white. He'd been vaguely aware of it and, by his own admission, kind of yawned over it. Then a colleague named Dick Shoup built a paint program with color, and dragged Alvy in to see it.

That was the moment. Alvy talked his way into Xerox PARC, started painting on the machine every day, and watched people walk through the room where it was set up. Half of them stopped dead, minds blown. The other half just kept walking.

He ended up getting fired from Xerox PARC — they decided not to pursue color, another in a long line of famously bad decisions from a company that also passed on the desktop computer. But by then Alvy knew exactly where he was headed.

How the alpha channel was born — overnight

After Xerox PARC, Alvy ended up at the New York Institute of Technology, working alongside Ed Catmull (who would later co-found Pixar with him). They had something almost nobody else in the world had: enough memory to think big.

Ed Catmull was working on a hidden surface algorithm — the classic computer graphics problem of figuring out what's in front of what — and hit a snag. He needed a way to render his code over an arbitrary image, not just a hard-wired test picture. He mentioned the problem to Alvy.

They kicked it around. The solution came together quickly: add a fourth channel to the image data that carries opacity information. Then composite two images together using the formula — alpha times A, plus one minus alpha times B — at every pixel. Alvy went home, wrote the code overnight, and called it alpha because that was already the variable name in the formula. He named the format RGBA.

He didn't fully grasp what he'd done until years later. The alpha channel didn't just solve Ed's problem — it freed images from being rectangles. With alpha, shapes could float. Soft edges became possible. Digital compositing as an art form was born. Alvy later started an entire second company, Altamira Software, built on that single insight.

The fight to protect that invention wasn't easy either. A British company called Quantel later claimed patents on airbrushing (which Alvy had also built, years earlier) and sued companies across the UK. Alvy testified in London, got called a liar in court, and lost. But when Quantel came after Adobe and Photoshop in the US, Alvy was ready. He worked with Adobe CEO John Warnock to take the case all the way to invalidation of all five patents.

HSV: why you can actually pick colors

If you've ever adjusted hue and saturation in any software, you've used something Alvy created at Xerox PARC. Dick Shoup's original paint program used RGB mixing — which is how video engineers think about color. Alvy, being an artist, wanted to work the way artists work: pick a color off a color wheel, then lighten or darken it.

He asked Dick for the algorithm to convert RGB to HSV (hue, saturation, value). Dick said it didn't exist. So Alvy went home and worked it out. He wrote it up, and it shipped. It's still in use everywhere — not because it's perceptually perfect (Alvy is the first to admit it's "a crude approximation of perception") but because it works.

The Genesis Demo and how Pixar got on George Lucas's radar

After New York Tech, Alvy and Ed moved to Lucasfilm — convinced George Lucas wanted them in his movies. He didn't, exactly. What he wanted was to digitize the filmmaking pipeline: editing, audio, compositing, logistics. That was fine, but it wasn't the big screen moment they were after.

The break came from an unexpected direction. Paramount was making Star Trek II: The Wrath of Khan and wanted some computer graphics in it. ILM (Lucas's special effects house) pointed them at Alvy's team. They wanted a 60-second sequence. Alvy listened to the brief, went home, and designed a shot that would use exactly what his team could do.

The result was the Genesis Demo — a sequence showing a dead moon transforming into a living planet. It took six months to render 60 seconds of footage. But Alvy designed it with one specific goal beyond satisfying Paramount: he knew George Lucas was always acutely aware of the camera in any film he watched. So they built a camera move into the sequence that was physically impossible — no real camera could have done it — but served the narrative perfectly.

The day after the premiere, Lucas stepped into Alvy's office, said "great camera move," and left. That was enough. Word spread to Steven Spielberg. Things started to happen.

(A side note: to keep themselves sane during those six months of computing, Alvy's team — including the late Lauren Carpenter, the first of the original Pixarians to pass away — embedded an Easter egg in the star field at the end of the Genesis Demo. Lauren had used a real star catalog to find a star that might have an Earth-like exoplanet, Epsilon Indi. The final shot shows the Big Dipper from that star's perspective — with one extra star in the handle. That extra star is our sun. Nobody in the audience ever knew.)

Starting Pixar: 45 rejections, a Hail Mary, and Steve Jobs

When George and Marcia Lucas divorced, Alvy and Ed saw the writing on the wall. They walked to a bookshop in Marin County, bought some books on how to start a company, and got to work.

What followed was bruising. Thirty-five venture capital firms said no. Eight corporations said no. General Motors and Philips of the Netherlands finally said yes — then fell apart overnight when H. Ross Perot publicly insulted GM's board while Alvy was in a room in Manhattan finalizing the deal.

In the limo to JFK, Alvy and Ed had one idea left: call Steve Jobs. Jobs had already offered to invest — at half the valuation they'd wanted — and Lucasfilm had passed. Now, with nothing else on the table, they called him back and asked him to match the GM/Philips offer. He did.

What followed was five years of hell. They ran out of money three times. Each time, Jobs would berate them and then write a cheque, taking more equity in exchange. Eventually he owned the whole company outright. The employees had nothing.

But they had their team, and they had Toy Story in the can. Jobs took the company public with nothing in the bank, banking on the film's reviews from New York critics. It worked. He became a billionaire overnight.

Alvy's verdict: "I don't like Steve Jobs at all. But he did give us the money when nobody else would."

What Alvy thinks about AI — and what artists should do

Alvy started his career in AI. He came to Stanford in 1965 specifically to study artificial intelligence, stayed for two years, concluded it wasn't going to happen in his lifetime, and pivoted to making the first computer-animated film instead. Now, at 82, he's watching it all come together in ways he couldn't have predicted.

He's not a doom-and-gloomer about it, and he doesn't think it's the singularity either. His take is more grounded — and more useful for working artists.

He told a story about David DeFrancisco, an artist he'd worked with at Xerox PARC back in the 1970s. David had been making art with pixels for 50 years. Alvy reached out and asked what he made of the AI tools.

David's answer: "Alvy, I'm in love with it. Everybody thinks artists are about rendering pictures. We're not. We're about selection — deciding what's important. Usually, you had to make the image to say this is what you should look at. Not anymore."

David said he used to produce maybe 12 keepers a year — pieces he was proud enough to sign. Now he produces hundreds. The AI generates mountains of junk, and his job is to sift through it and find the things worth pursuing. He added: "It may be the end of humanity as we know it. But I don't care. It's the most creative I've ever been."

Alvy's advice to artists mirrors exactly what John Lasseter once said to animators who were frightened of computer graphics: it's a tool, not a threat. Engage with it. Help shape it. Because if artists don't, the tools will be built without them — and they'll be worse for it.

"Quit fighting it. Engage. Take part in it. Help shape the tool, because if you don't, it's going to be wrong."

On creativity, and what makes it work

Near the end of the conversation, Alvy and Edd got into a riff on creativity itself — what it actually is, and how to keep it alive. Alvy brought up Einstein's definition: combinatorial play. He loved those two words. Creativity isn't inventing new elements from scratch; it's finding new combinations of existing ones. And play is the key — you can't force it, you have to let it happen.

His practical advice for anyone mid-career in a fast-changing industry:

First, figure out what you're actually good at. Not what you wish you were good at. Alvy thought he was going to be an animator. He was wrong. Knowing that — and accepting it — freed him up to do what he was genuinely exceptional at.

Second, go where the action is. He moved across the country multiple times, nearly overnight, to be in the right room at the right time. He acknowledges that's easier without a family, but the principle holds: proximity to the work matters.

And third: engage. Don't just read editorials about what AI means for your industry. Open the tools. Make something. See what happens.

The name Pixar

One final thing — the origin of the name. Alvy chose it, and it came from his New Mexico upbringing surrounded by Spanish speakers. He'd always been fascinated by the word "laser," a noun that looks like a Spanish verb. He wanted a word with the same quality for their new machine.

He suggested PIXR. Lauren Carpenter pointed out that "radar" sounded high-tech, and that "-ar" was also a valid Spanish verb ending. PIXR became Pixar. When it came time to name the company and nobody could agree, Alvy said: just name it after the machine, everyone already knows the machine.

He thinks of it as a fake Spanish verb that means "to make pictures." Or, more completely: "to make fake pictures."