Got Rock Opera?

I’ve always thought the term Rock Opera described two things that belonged together only in the most twisted of contexts.  Add to that a narcissistic “hero” called White Gold and place him in a whitewashed world called Milquarious,  and apparently you’ve got something the California Milk Producers Board can get behind to propel the “Got Milk?” campaign into the annals of advertising – um, footnotes?  I’m still scratching my head about how my ancient analog animation computer, Scanimate, got sucked into the Vortex.  Literally! Seriously! I’m not making this up!

It seems the look of  “analog light” – that glowy chromic warmth that characterized much of the computer animation seen on TV in the seventies and eighties – is back in demand again.  It would also seem that all the digital wizardry of modern-day imaging software can’t quite manage to get the same “look”. I guess its like the audio afficionadoes and their love of vacuum tube amps.

Thus it was that I ended up spending three days recently shooting hours of  glowy, warping, swirling animation so that when White Gold and his buddy get sucked into “the Vortex”, they fall through a sea of high-definition Scanimation before landing on… wherever it is.

Joe Mullen adjusts Scanimate
Joe Mullen adjusts Scanimate (Click the pic to see more)

The Scanimate was never a high-resolution machine, due largely to the limitations of the NTSC Television standard, and to some sixties-vintage pickup cameras.  Enter the RED Digital Cinema Camera,  the latest whiz-bang technology transforming Hollywood from celluloid to digital moviemaking.  Capable of  shooting at up to 4,000 pixel images (about 12.5 Megapixels in consumer digital photo speak) at up to 120 frames per second,  this expensive little toy makes it possible for my old Scanimate to take on a new life in the digital world. And here in East Tennessee, Frank Weeks at Digital Cinema South just happens to have one!

By way of background,  I worked in Hollywood in the eighties at a company that had two of these strange machines.  The obscure company that built them only produced eight of them, and for about ten years, the airwaves were filled with “electronic animation” they produced.  Everything from the sizzling spiraling patterns that numbers and letters transformed into on “Sesame Street”, to the guy’s expanding stomach (“IndiGESTion”) for Alka-Seltzer, to the Death Star sequence in the first Star Wars movie – all were produced on Scanimate.  As far as I know, I have the only remaining working machine.

Joe Mullen, who does motion graphics for recently flew out from LA to supervise the work for three days. Joe said: “The cool thing about Scanimate is that the CRT monitor gives off this kind of analog light that is very hard to simulate digitally.” (see the video below)

So as I understand it, being only peripherally involved in this whole campaign,  “The Battle for Milquarious” officially launches October 5,  so if you want to see what a Rock Opera about Milk – with vintage analog animation – looks like be sure to check it out!
I’m not making this up! Really!

7 thoughts on “Got Rock Opera?”

  1. Fascinating! Scanmate far predated the (self-claimed) pioneering vector graphics I saw as a student at Univ of Mich (early 70’s), and our own digital animation work of the mid-late 80’s. Not even Disney has this technology (their Pixar-tech based CAPS system fired up near end of the 80’s). Will be fascinating to see how this technology evolves in the digital age!

    offers some warp-morph animation drawn traditionally and then animated by digital warping (would have been fun to see what scanmate could do with the characters!)

    Best wishes!

  2. Uhm, have you seen that “Russian Singer” video which is becomming an internet pheonomenon. If you look at the title, there seems to have been some sort of russian project for doing the same.

    Anyhow I think I might have a copy of some scanimation from the Scanimate shipped to RTL. It’s on an old Video 2000 tape from 1985 I have. If everything goes well, I can get you a “decent” consumer grade copy of it in a few weeks. Essentially I need to install a S-Video output to it as my framegrabber has serious problems with luma croma seperation if the timing is off.

  3. You might want to take a look into BBC Engineering Number 93 from March 1973.

    On page 5 of the PDF file there is an article with the boring sounding title “Colour Film Transmission in BBC Television”. However if you flip through it to page 12 of the PDF-file you will probably be pleasantly surprised. 🙂 It’s point 3.4 Special Facilities I’m talking about.

  4. Thanks for the great information!

    My HP monitor does not have a mesh screen. I know what these look like in Tektronix scopes, and beam is very sharp, with little to no bloom depending on intensity. The monitor has a 17″ CRT, so it’s big for electrostatic deflection. Resolution is excellent.

    My goal at the moment is simply to get a standard raster on the screen, with good linear ramp signals on the deflection. I am pretty good with electronic circuit design and understanding, so I would love to get any suggestions you may have on building these ramp circuits. I have horizontal and vertical drive pulses for sync, if I need them.

    After that’s working I’ll probably play around with summing amps to modulate the deflection first, then move on to multipliers later.

    As for the intensity compensation, the HP 1317A monitor has an internal device called “Phosphor Protect,” which dynamically controls beam intensity based on the deflection signal gain in both the horizontal and vertical aspects, so that should work nicely, in theory.

    As for patching, I have a ton of 1/4″ audio patch bays, giving me 150 or so ports to work with, and I can parallel several sockets to do the same thing as patch cord stacking does with the Scanimate.

    If you have any ramp circuit suggestions, or even copies of the Scanimate schematics on the those, that would be amazing.
    I’ll give you updates as I go.

    -Tom Ostertag

  5. Tom:

    You have taken on a very interesting project!

    While I was at Image West, we began development of a newer system that would allow digital control and memory of effects setups. We used the same HP XY monitors you have but ended up replacing the CRTs (you can see a pic of this at
    The HP CRTs employ a fine mesh screen that acts as a lens to produce a wider angle deflection.
    The downside of that is that the spot size blooms by a factor of almost 10.
    HP used electrostatic deflection which has better frequency response, while Scanimate uses electromagnetic deflection, again, keeping the spot size small but limiting the frequency response.

    Scanimate’s “CPU” has horizontal and vertical ramp generators, each of which was followed by a multiplier allowing for size modulation. There was also a pair of multipliers after that with a single control voltage for overall size, called “depth”. Later, they added 4 multipliers in a 2×2 matrix with sine/cosine control voltages for rotation. (We actually extended that to a 3×3 to give full 3D rotation in VersEFX). Even today, high bandwidth low noise analog multipliers are rare. Everything went digital and people stopped developing that kind of stuff.

    Later animation oscillators were developed, and they can lock to vertical top or bottom or even other signals. Making a good linear sawtooth and a distortion-free sine/cosine are critical. I could give you some circuit suggestions if you are up for building something but its not simple.

    The scan lines are counted and there are comparators to trigger “segments” so that you have 5 separately controllable pieces of raster you can manipulate. This is good for individual words.

    The hardest problem is intensity compensation. When the raster shrinks to a dot, the intensity has to be reduced to avoid burning a hole in the CRT. Also if you have grayscale video, any size modulation has to be dynamically compensated for. Increasing the size of the raster requires boosting the intensity, smaller requires reducing it. A lot of calculus went into defining and designing that function.

    We wanted to use electrostatic deflection so that we could deflect even at video speeds.
    That requires wideband multipliers that come with – wideband noise.
    We had custom CRTs designed for smaller spot size to increase over all resolution, but we found that then when the raster was blown up, the scan-lines were so sharp that there was less “fill” from the beam’s spot-size. So in effect, all our high-bandwidth sharper image efforts produced an image inferior to the old Scanimate.

    The Scanimate colorizer is also a work of art. It can smoothly transition from making one shade of gray into a color to making the next higher one a different color without any of the usual frying edges associated with switcher keyers. The circuitry to do that is pure genius and should have been patented!

    Lately I have been shooting the CRT with a 4K RED camera and the imagery on the CRT holds up really well ( see )
    But there is no digital equivalent of the magnificent work the colorizer does. I have considered making a HD colorizer but the analog bandwidth required would make that a difficult if not impossible project. Figuring out how to do it all in digital as a plugin for after-effects would be the way to go, but if you don’t have the kind of glowey source images that come from rescanning a CRT, it would be of little use.

    Finally, Scanimate is “programmed” with patch cords that are stackable. Vector electronics used to manufacture them but of course there is nothing that uses them and they are no longer made.
    I only have a handful of them and the pins are an odd diameter (0.09″) so that is being very frustrating when it comes to making more complex animations.

    I’m glad there is such interest in the kind of images Scanimate produces. Oddly enough, none of the digital technology available today can produce that look.

    Please do keep me informed about your progress and let me know if you need any help.

  6. Dave Sieg,

    My name is Tom Ostertag, and I am an avid collector of television production equipment, and your site about the Scanimate system is of tremendous interest to me.

    I am currently in the process of building a simple analog video manipulation system based on the Scanimate design, and I would love some help with a few things.

    Currently the system consists of a Sony industrial black and white Vidicon camera for the graphic source, an HP 1317A XY monitor with electrostatic deflection to display the manipulated image, a Panasonic single CCD camera to scan that image, a Panasonic WJ-5600 switcher on the input end, and various oscillators/tone generators for deflection distortion.

    My goal is just to create a system that can perform simple horizontal and vertical modulation distortion to the image. I am not going to attempt to implement triggered perspective oscillator circuits just yet. The video source is standard 525 line interlaced video, so I am not dealing with progressive scan either.

    I am having trouble generating accurate, linear sawtooth signals for the initial horizontal and vertical scan. Do you have any suggestions on how to build something to generate these signals? My switcher has an internal sync generator, with horizontal and vertical drive outputs, so they can be used as oscillator lock sources if need be. Any copies of Scanimate documentation on this subject would be a huge help, if that would be at all convenient for you.

    Any help on the subject would be greatly appreciated. I am completely amazed by the design of the Scanimate machines, and your work on keeping the system running is wonderful.

    Thank you,
    -Tom Ostertag

Leave a Reply