Computer control of Analysis Projector

[carllooper]

Here is a computer interface I've built for a NAC 16mm Analysis projector:

http://youtu.be/MJCzh63jQx0

It's basically three solenoids mounted on the projector's original remote control. I was going to tap into the remote control's inner circuit, but after opening the remote and looking at the circuitry it wasn't hard to convince myself it would take forever to decipher it. So put the cover back and on and came up with this alternative: mounting solenoids. The bracket was printed on a 3D printer (designed in Lightwave 3D). The solenoids were a $1 each. The solenoid control board (out of shot) is a 4 channel relay board from Phidgets http://www.phidgets.com/products.php?ca ... _id=1014_2 - a bit pricy ($60) but comes with a USB interface and API, the rest being a relatively simple programming job.

The three buttons under computer control are:

1. Run/Still
2. Forward/Reverse
3. Frame Advance

The frame advance is the main button used. It is syncronised to a Canon EOS 60D, also under computer control (using the SDK provided by Canon). The Canon takes 5K 16bit RAWs of each film frame. The turn around time for each frame is about 5 seconds - there is a bit of overhead in transferring each frame, via USB, to the computer, and doing the RAW to TIFF conversion. I'm doing the conversion on the "fly" (if one can call it that) but could always do the conversion afterwards instead of on each capture. Nothing is saved to the camera's cards as such. Goes straight to hard disk.

C

[Nicholas Kovats]

Great effort, Carl!

I wonder if you could speed up capture with USB 3.0 or Thunderbolt (10 Gb) connectors? I take it that TIFF is a container w/ compressed codec? Is it better than jpg frame compression?

NK

[hirudin]

if it could be converted RAW -> 32bit EXR, now we're talking turkey, but that's gonna need alot of space.

[carllooper]

Great effort, Carl!

I wonder if you could speed up capture with USB 3.0 or Thunderbolt (10 Gb) connectors? I take it that TIFF is a container w/ compressed codec? Is it better than jpg frame compression?

NK

The Canon's interface is USB. It's not bad. I'm not concerned too much about the time it takes. The fastest rate would be saving to the camera card, at about 5 frames per second, which is the fastest the camera does RAW capture, but you would still have to eventually transfer the data from the camera via it's USB, to the computer at some point.

TIFF files are very basic files - uncompressed pixel data, but also a very old file format, so well supported by graphics programs: Photoshop etc. The Canon RAW files are readable by current versions of Photoshop, but can throw a spanner or question mark in other programs. I save a JPEG as well (doesn't take up much extra bandwidth). The problem with JPEG is that it's only an 8 bit format (so a lot of dynamic range is lost in translation) and JPEG introduces artifacts (through it's compression methods) that become more pronounced the moment you try to do any grading.

At 5K, roughly speaking, each pixel of the captured data represents a 2x2um square (0.002x0.002 mm square) of the original 16mm film. So each pixel is the same size as the largest possible silver bromide crystals in a typical film. Until I get an electron scanning microscope I'll have to live with this limit

C

[carllooper]

if it could be converted RAW -> 32bit EXR, now we're talking turkey, but that's gonna need alot of space.

Oh yes, any format can be done. I save the RAW files so can always export to any other format such as EXR. The conversion to TIFF is just for my workflow. The RAW files are about 25 MB per frame. The TIFF files are 105 MB per frame.

C

[S8 Booster]

although on the "wild side" of the contents of your original post it reminds me of a project i experimented on in the early ´90s.
I had/have this wonderful MAC Quadra 840AV which with the equally wonderful adobe Premiere 4.0 (YEah I know Cristoph hate me for saying this ) - but - the Quadra could control (most) (higher) (end) VCRs frame by frame via a Sony wiring set with a few electronics components. it worked like stop motion - AND - the Quadra could capture images up to 4096 pixels wide in any proportion id like. Now, that were the days...

of course there was no way the standard MAC could handle a decent bit flow with that image size (without some extra EXTRAS) but it worked out of the box. any file format one could think of was available but for a somewhat normal frame rate both frame size and compression rate would have to be set to something like, eh, VHS - BUT - films could me made in stupidously heavy sizes and format but the processing was S - L - O - W.

The AP 4.0 was the only NLE which has ever made me feel like working with film - ever even though it was not as good as the Adobe Audition bundle version with Photoshop before that. the film strip could be exported between the two and and photoshop could do automated image correction to the un-imaginable frame by frame.

where did it all go?

anyway, some day ill re-find the Sony documentation for the frame by frame capture because i have it somewhere. i can not find any information about it on the net for the time being.

Shoot - el-filmo.

[carllooper]

I had/have this wonderful MAC Quadra 840AV which with the equally wonderful adobe Premiere 4.0 ...

Yes those were the days. At the time they were quite a significant leap forward in computing power. We can easily forget that. They were slow but nowhere near as slow as their immediate competition. Indeed they seemed like miracle machines. In those days one more readily allowed the machine to do it's thing for hours or days on end ( I was doing computer animation at the time), the result being the reward for one's patience and faith. Today everything seems as if it needs to be done quickly or not at all. As if the faster the machine the less time it should take. Which in principle should be true, but then the more efficient a machine the more you ask of the machine. I still like projects that require time and patience. Of course, the Quadra's would be stupidly inefficient by today's standards, as will today's machines be stupidly inefficient by tomorrow's standards. Our grandchildren will no doubt have a good giggle at what we consider high tech today ...

http://www.youtube.com/watch?v=pHc7-37i ... _m&index=2

C