Living in the Future

Just a quick post because this is so cool, and the blog is the easiest way I have of sharing audio. I just bought iZotope RX3 to help with my auditions. All I have is a cheap (yet surprisingly good) Shure USB mic, a room with only so-so acoustics, and a neighborhood and appliances that provide a lot of background noise. I learned about the software because the engineer on my audiobook project uses it to process out the mouth clicks.

This is like science fiction, which is what Layer Paint looked like to me when I first saw it retouching images. Layer Paint? Oh, I'll have to tell that story in a later post. Let's just say that many of its features later showed up in a product called Photoshop.

This really is like Photoshop for sound. A total game changer. My serial number arrived just a few hours ago. I sat down and recorded this quick test, and then processed it a step at a time. What you will hear in this one minute sample is the same 15-second audio four ways:

  1. The raw audio. Note the background noise, mouth clicks, and iPhone chime.
  2. The Denoise feature has pretty much made the room noise vanish.
  3. DeCrackle has killed almost all the mouth clicks.
  4. I literally painted out the sound of the iPhone chime.

The UI displays the audio like this:

The blue is the waveform you're used to seeing. The yellow/red is the audio spectrum, with low frequencies at the bottom and high at the top.

The blue is the waveform you're used to seeing. The yellow/red is the audio spectrum, with low frequencies at the bottom and high at the top.

See that horizontal streak in there? No? Oh, let me dial it over to show just the spectrum.

That iPhone text tone stands out as a horizontal stripe at one frequency.

That iPhone text tone stands out as a horizontal stripe at one frequency.

The software has a time/frequency selection tool that let me draw a box around that horizontal stripe where the offending tone was. Then the Spectral Repair feature has a way to attenuate it. Really slick.

I'm not claiming this is polished audio at all. That's the point. I could have gone from the raw recording to the finished one in about five minutes. Without the offending text tone it would have been good to go in one minute.

Pretty cool. And a lot cheaper than hiring a studio or finding another house.

Sixteen Hours Before The Mic

A certain Neumann U-87 and I have spent a lot of time together at the studio over the last couple of weeks. This next Monday, December 30th, we'll reunite briefly to put the final touches on recording The Seventh Angel. That's kept me busy enough to not update the blog in a while. Expect more pearls of wisdom and tales of derring-do during the holiday break.

I just finished reading The Ship That Would Not Die, the true story of the USS Laffey, which was attacked by more Kamikaze aircraft than any ship in history: Twenty in one day. The uncle of a good friend of mine was on board for that battle, and lived to tell the tale.

Between that and the thriller I just read to my friend the U-87 I must say that I'm content to be peacefully writing and narrating from terra firma rather than enduring combat on a warship. It sounds terrifying.

The Seventh Angel - Coming to an earbud near you

I'm excited to announce that I've just started production on the audiobook of Jeff Edwards' military thriller, The Seventh Angel. Expect a lot of authenticity: Jeff didn't just research the Navy well, he actually served in it, doing ASW. He's been a dream so far, even sending me a pronunciation guide for the foreign words and nautical slang I'll encounter in the book. Back when Pixar made television commercials we could only dream of a client who understood how long it takes to do things.

He joked that I should read the whole thing in the Star Command voice from Toy Story that you can hear on my reel. To date our only disagreement has been about how many minutes it would take for the listeners' eyes to bug out.

I'll be recording the narration with a very experienced engineer at Live Oak Studio in Berkeley, located temptingly near the Gourmet Ghetto. The project has already given me the excuse to get my pastrami and pickles fix at Saul's.

The book will be released on Audible, probably very early next year. Expect updates as the project progresses. My experience at ACX has, so far, been very pleasant. I'll no doubt blog about that more at length after the book ships.

Ships. See what I did there?

Try the veal.

Make It More Religious

ILM, Industrial Light and Magic, was quite the place to be in the mid-eighties. During the decade starting with Star Wars it had provided the visual effects for a vast majority of the top ten grossing films of all time. One of the key players behind all of that, of course, was Dennis Muren. We were fortunate to be able to work with him on Young Sherlock Holmes because working for Dennis is fun, professionally satisfying, and very educational.

Of course Dennis is also a great student. He learned about the promise of computer graphics from the Lucasfilm Computer Division's so-called "Graphics Group", the core team that would eventually form Pixar. With CGI now commonplace it's hard to realize that in 1984 it was still mostly a laboratory curiosity. Only a few CG shots, starting with Ed Catmull's own digitized hand in Futureworld, had been used in features up to that point.

Even Tron only had fifteen or twenty minutes of CG. Don't forget that none of the Star Wars films had any CG at all beyond Larry Cuba's Death Star briefing display in the first film and the "hologram" of the later Death Star that the Computer Division did for Return of the Jedi.

While we were working on YSH Dennis wanted to study the CG in The Last Starfighter, so he got a print and had it loaded on the rock-and-roll projector at ILM. It was an impressive effort. While there were some shots that had obviously been done at lower complexity than others, a few of the hero shots were so good that Dennis said, "I'd completely buy that as real".

One of my biggest lessons came when a scene rolled by that was a master shot of a large hangar bay full of starfighters. Dennis suddenly called, "Wait!" The projectionist paused the film on a frame. Dennis was already using his pointer to trace rapid circles all over the screen, and I was squinting to see what was wrong with the shot. His eye for detail was legendary (though he also knew when to say, "Forget it, nobody will see that").

"Where am I supposed to be looking?"

The lightbulb was going off in my head. Dennis went on to explain that a shot had to be designed in a way that directed the viewer's eye. There was no contrast, just a frame full of busy details, all about the same value. Suddenly I saw that there was nothing specific to see, no one thing to draw my attention. In a later post I'll write about the obsessive degree to which Pixar directs the viewer's eye. That lesson from Dennis was probably where it began.

Years later I flashed back to that day when I saw the first space battle scenes in The Phantom Menace. But I digress.

I think Starfighter gave Dennis some hope that these Computer Division guys were going to be able to deliver a usable Stained Glass Knight. There was more than a little trepidation about going that route since that kind of character, integrated into live action, had never really been done before that. In fact, it almost wasn't done this time.

Our first attempt was completely rejected by director Barry Levinson. The script called for a knight in a stained-glass window to come alive, alight in the church, and frighten the poor priest so badly that he would panic, dash outside, and promptly be trampled by the horse drawn wagon that someone just knew would be passing by at that moment. The perfect murder!

The first design had the window break into colored shards, floating blades of rippled shower-door glass that gathered in the air to form a three-dimensional Roman centurion. It had all been modeled, animated, lit, and roughly composited into the scene, and we were all pretty happy with it. It seemed quite menacing. But it didn't work for Levinson, so very late in the game our keen-edged Roman was scrapped and a whole new design had to be executed.

 

Stained Glass Knight, Take 2

Stained Glass Knight, Take 2

Of course not only the Knight was redone, but that required a whole new window. When John Lasseter painted the texture map for it he included our initials in the ribbon-like scrolls near the bottom. You could barely read them in the theater, so they'll never be seen again unless the movie is rereleased in Blu-ray. On DVD our Easter egg is a smudge. 

The tilt-up from the Knight's hands to his face was the first-ever rack focus CG shot in a feature film, and it remained the only one for many years. Rob Cook figured out a way to render depth of field (see the update, below) and pitched the shot to Dennis. He went for it, and it got coded up, but the computational requirements were insane. I hand-carried two disc packs (a whopping 350 Mb each) to CCI in Orange County where I spent two solid weeks using borrowed time on two of their best computers to render the last 130 frames of the shot.  Meanwhile, up in San Rafael, Bill Reeves was computing the rest of it.

When we showed him the first test Dennis, who usually gives insightful notes very quickly, spent an uncharacteristically long time watching the screen. "Hmm. That's interesting." He said that real lenses zoom slightly when racking focus because the focal length of the lens is actually changing. Our perfect computer lens didn't breathe.  "We can put that in." "No, leave it. That's how lenses should work, but it's not how lenses do work. I kind of like it." Rob later calculated that the field of view change that particular rack focus would have produced a change of perhaps 6 pixels in a 1500-pixel wide image. That's less than half a percent, but Dennis sensed that it wasn't there. Of course, that breathing was carefully added to the lens package on WALL•E many years later. But back to our story. 

All of this should make it clear why we were so nervous when it came time to screen the result for Levinson. We had already started over again from scratch once. There really was no time to make major fixes, and probably no time for even minor ones. We sat on the back row of couches in the old Kerner screening room, listening to the ratchety growl of the projector as the film rocked and rolled, back and forth, back and forth, while Muren, Levinson, and a producer sat silently in the row ahead of us.

Finally, Levinson spoke. 

"I dunno. Can you make it more... religious?" 

The panicked glances on the back row probably looked like a scene from a Three Stooges short, but with a lot more Stooges. We searched each other's faces for an answer. What the Hell did that mean?

And then Dennis showed why he now has a hall closet full of Oscars. Calmly, he waved us a subtle shush with his hand and shot a "don't worry" look over his shoulder.

The very next day the shot was screened for Levinson again. He loved it. Dennis Muren is Dennis Muren because he knew that "religious" translated to telling Optical to put two more steps of diffusion on the glow element pass. The rest, as they say, is history. 

When we convened with him in the lobby of the Cinema in Corte Madera after the crew screening of the finished film, we were trying to come to grips with the feeling that maybe the movie wasn't as good as we all hoped it would be. Dennis had another lesson for us. 

"Sometimes you can get so close to a movie as you're making it that it makes perfect sense to you, but the audience goes 'huh?'" Keeping objective distance is possibly a film director's most important and difficult skill, and it's something John Lasseter excels at. But we wouldn't really reap the benefits of that skill in full for another ten or so years yet -- when we made our own feature film.

 


Update: 

I have corrected the rack focus part of the story based on input from Rob Cook himself, who wrote to me about the experience of realizing what the renderer, Reyes (Renders Everything You Ever Saw) would be capable of doing:

Depth of field was part of Reyes from the very beginning. I was working alone in the office one night when it suddenly dawned on me that the stochastic sampling principle could be used not just for anti-aliasing and motion blur, but also for many of the thorniest problems in rendering: depth of field, soft shadows, glossy reflections, etc. I remember sitting there and being struck first by how incredibly powerful this one simple technique was and then by how odd it felt that I was the only one who knew it at that moment. The next day other people at Lucasfilm would know about it, and the following year everyone in graphics would know, but the rest of that evening had a kind of eerie solitude about it. I know other people who have had such moments, and I'd heard a few of them talk about it, but it was surreal to be experiencing one myself.

I remember how excited Rob was about it when he explained it to me, which was probably within just a few days of his epiphany. Sometimes I think my experience in that Graphics Room on the second floor of the old C Building was akin to being there the day Ook and Og ran into the cave yelling, "Ooh! Look what I invented! The wheel!" 

Lagniappe:

Ralph McQuarrie got the accolades for production design on Star Wars, but Thomas Kinkade did all the heavy lifting.

There is absolutely no subtext implied by linking a gallery of abandoned toy factories to a post about the old ILM building on Kerner. 

Which, in turn, does not mean that's why I feel like listening to a sad clown

It's Not About You

Sharing Time

I learned a lot working at Pixar for over thirty years, often referring to it as the world's best film school. Starting as I did, as a janitor at Lucasfilm, pretty much everything I learned about filmmaking and computers was on the job. When I joined the Computer Division in 1982 I was surrounded by highly intelligent, generous people eager to share what they know. That strikes me a good model to emulate.

This is not to claim that I'm nearly as smart as they. When I got to know Penn Jilette in the early 90's I told him about the caliber of people I was with. He said it seemed familiar, saying that he "went from being the brightest kid in my class to the dumbest guy in my peer group." Exactly. And while I'll never be as generous as Ed Catmull and that amazing team he assembled, I can try.

So, among other things, I plan to use this blog to share some of whatI learned at Pixar about filmmaking, writing, improvisation, collaboration, and other such creative pursuits. This first post is about getting some distance.

You Are Not The Work

One thing that struck me in the old LFL and early Pixar days was how many people came from architectural backgrounds. Architect schools were, and perhaps still are, apparently a mix of design, engineering, and boot camp. I've heard that professors would rudely sweep a student's painstakingly-crafted model to the floor just so they wouldn't get precious about the work. I guess it did toughen them up.

There's a better way.

Learning to separate yourself from the work you are doing is one of the hardest lessons to learn in any artistic field.  Quite understandably the person who has labored lovingly over a screenplay, novel, drawing, sculpture, or building design can easily take criticism of the work personally.

Our creations are our babies. Tell me that my dialogue is clunky, my photograph improperly framed, my building uninhabitable and I'm perfectly justified in feeling that you are attacking me. Right? 

Only if I'm an amateur.

Steely-eyed professionals understand that notes should never be taken personally. They have thick skins, and ice water coursing in their veins. They know better than to get emotionally invested in what they've created.

And if you believe that last paragraph, I have some prime beach-front real estate in Florida for you. It's true that pros make it about the work, but it's still hard. Of course they feel it. But the right culture, atmosphere, and attitude can get people to a healthy place where it really is about the work.

At CalArts  the drawing instructor who taught John Lasseter, Joe Ranft, and many other animation luminaries told his students at the beginning of the year, "Everybody has five thousand bad drawings in them. And we're going to start drawing until we get them all out." Note that he expected the drawings to be bad, and expected the artists to eventually produce good ones.

Get Lots of Eyeballs

Pixar picked up some of its culture and process from Lucasfilm. Working with Dennis Muren on Young Sherlock Holmes was an education on its own. In fact, I should do another post on just that experience later. An important process we picked up there was of how to run dailies. Most film productions screen dailies, or rushes, for the director and editor to pick their favorite takes and quality-control the shoot. At ILM that was taken a step further.

The previous day's shots were put on the "rock-and-roll" projector, which could run forward, backward, fast, slow, or even still. The entire effects crew would gather to see what came back from the lab from the day before. A room full of eyes scoured the shots over and over again, looking for flaws. Dennis would make the determination as to how serious the flaw was. He could call "final", send it back for a fix, or dub it a "CBB" (Could Be Better), meaning that it would go in the final film if they ran out of time to get a better take. To this day Pixar still uses final, fix, and CBB.

Stained Glass Knight from Young Sherlock Holmes.

Stained Glass Knight from Young Sherlock Holmes.

Animation Dailies at Pixar became, and still is, a gathering of the director and all the animators on the show. Artists are encouraged, nay required, to show work in progress. Nobody is allowed to work until "it's done" and then show. Everybody in the room is welcome to make suggestions. They are all about how the shot could be better. Then the director decides which notes the animator should address. As John put it back on Toy Story, "My job is not to have the good ideas. My job is to recognize the good ideas."

This applies to the directors as well. The "Brain Trust" (a name chosen sarcastically, but which stuck) is made up primarily of other directors. When they meet to give notes on a movie the notes can be frank and brutal. But the notes are about making the movie better, not about what anybody did wrong. 

For years now each film has been screened for people in the company, but outside the production, at least three times at different stages. Everybody is encouraged to give notes. Everybody. Security, Cafe staff, PA's, everybody. The producer makes a digest of those notes and they go into the hopper at the Brain Trust meeting. Good ideas can come from anywhere.

Bully for Pixar. What about me? 

Suppose you aren't working with a giant Pixar-sized production. Perhaps you're a writer making a solo pitch with a producer and his or her staff. How might your attitude make or break the deal? At AFM a couple years back a speaker told about a meeting where the producer actually asked the writer if the protagonist of, I think it was a RomCom, could be a monkey instead of a guy.

How would you have taken that note? 

Here's what that very smart writer said: "Interesting idea. Let me think about it and get back to you." 

In the end the protagonist remained a human. The writer impressed the producer so much that since then they have done multiple movies together. 

Making a constant mental effort to separate yourself from your work can pay off in much better work and much more work. It's hard. That script, that scene, that image, might feel like your baby. 

They say in the business that sometimes you have to kill your babies. If they aren't making the movie better, out they go. It's going to sting. The real pro is the one who can identify his own targets among his darlings. As Thomas Mann put it, "A writer is someone for whom writing is more difficult than it is for other people."

 


Laignappe

You know you need a baby duck slide.

It might put you in the mood to meet more friends

This could not possibly be more on target. 

I salute consistency

Risotto alle Fragole

I like food. A lot. So, while this is not a cooking blog, do expect some culinary detours. 

This summer my daughter and I spent a couple of weeks in Florence. Just two nights before we left we discovered an amazing restaurant called Aqua al 2. There were a couple of dishes we had there that I wanted to try to duplicate. I spent a few months and several trials coming up with a blueberry sauce for Fileno al Mirtilio

Next up is Risotto alle Fragole, or Strawberry Risotto. I started with a recipe I found on line, but axed the celery right away. Seriously  — celery? This was not only my first attempt at that recipe, but my first attempt at making real risotto the stand-there-and-stir way.

It was pretty good. I'm going to dial down the butter (believe it or not) and use a bit more cheese on the next run. 

Breaking Good

Since this new web site is going live close to Halloween it only seems fitting to share a little project my daughter and I have been working on. 

Namely, my Halloween costume.

She helped finish the shave, tint the beard, and post-process the photo. I'm pretty happy with how it turned out. This was just a test run. With any luck we'll get an even better result this weekend.

WalterWhite1.jpg

Update: 

This is what I looked like on Halloween:

There's always something cooking in my kitchen.

There's always something cooking in my kitchen.