HOME
the Gifts
In early 21st century USA, a landscape of dissonance and dysfunction, it is comforting that there exists at least one sector of human activity where things are getting better... 

A sector where, every year, costs go down and functionality goes up. Where improvement happens so fast, that a decade doesn't just mean "much better", but rather the opening up of whole new frontiers of possibility.

This, of course, is the microelectronics sector, and its offspring: sensing, storage, displays, computation and communication.

Carlovision, which involves the electronic storage and processing of thousands of high resolution photos, is enabled by the amazing gifts of microelectronics.
1991, 20MB hard drive, $1000
Holds 20 photos
S T O R A G E
It doesn't seem so long ago...  In the oddly western-themed Frys Electronics in Palo Alto, I reluctantly signed $500 onto the credit card and took home my first hard drive.  It was big, fast and awesome: a 20 MB Seagate (left). 

It was 1991, the Jurassic period of personal computing, and I had, up until then, used 1.4MB floppy drives for "mass storage".

After the requisite 2 day ordeal of getting it installed, I was in the big leagues: booting off hard disk, and reading and writing huge files in the blink (or 3) of the eye.  The drive made cool sounds, and it seemed unlikely I could ever fill it up. 

Things have changedToday, 20MB is a laughable amount of storage, barely enough to hold 20 1MB photos.  The little SD card at right can hold 32 thousand such photos, and the common 1TB hard drive can hold 1 million

The plunge in "cost to store 1 photo" is stunning: with the drive at left  (in 2012 dollars) costing 50 dollars per photo, and the drive at right, .0087 cents per photo.  That's a 574,000 fold cost reduction.  
 
2012,  32GB,  $44
32 thousand photos
2012,  1TB,  $87
1 million photos
C O M P U T A T A T I O N
When computers "just work right", as they typically do, it's hard to appreciate just how much work they are doing.  Tap a thumbnail of a photo and it instantly fills the screen.  Zoom and pan: no problem, it's responsive and smooth. But under the hood, billions of computations and transfers are taking place, and only recently have processors become fast enough to handle this sort of task. 

Just the act of opening a photo to the display is a monumental feat of organized action:  the compressed jpeg file must be transferred to RAM, then systematically deciphered into a grid of millions of red, blue and green pixel brightness values.  And since the picture's native grid size is likely different from that of the display, it must be scaled - another computation-intensive operation.  Finally, the processor must tell every one of the display's 6 million subpixels (closeup, upper left) how bright to be, and voila - a photo on the display.  Elapsed time, 1/10 second.

And, according to the plot below, next year that 1/10 second will become 1/20 second, and 1/40 second the year after that. This increasing amount of computation per dollar enables any and all of the following: higher photo resolutions & better zoomability, faster frame rates, higher display resolutions.  In short, the potential of computer mediated photoset viewing is good and getting better, fast.  
Closeup of a display pixel,
composed of 3 subpixels. Each subpixel's brightness is independently controlled by a processor. 
Closeup of 22 x 18 grid of display pixels. (396 pixels). With this many pixels, simple images start to become recognizable. A typical display contains a grid of 1920 x 1080 pixels (over 2 million).
Seems like it's not slowing down... 
Will all this processing power bring a Utopia of long healthy lives free of drudgery?  Or will it enable a hyper-efficient neo- Stalinism?

So far, the signs are not encouraging.  But on the bright side, the prospects for beautiful computer mediated 3D panoramic time lapse photography are excellent.
D I S P L A Y
In a photography system like Carlovision, all links in the chain from the photographed scene to the user eyes are important.  The display, being a direct link to the human visual system, is especially important, and is still the weakest link.  As far as applicability to 3D photograph viewing, electronic displays have recently passed into the realm of "fairly good".  

The problem is that the human visual system is so good that it's very difficult to fully feed.

We see with very high resolution in the center of our visual field, the fovea, and low resolution over a huge, almost hemispherical field.  And we see in stereo.   

At this juncture, feeding the fovea in stereo is the best we will typically get. The situation awareness that comes from peripheral vision will have to wait for some unimagined technology, or be provided by large & expensive wrap-around arrays of 3D displays.

Having said all this, the advent of inexpensive HD (1920 x 1080) displays (left) has been a major leap forward in photo viewing.  Inexpensive passive 3D HD displays are also available, and for the first time have brought good quality computer mediated 3D photo viewing to the mainstream.  

And the near future looks promising since Apple restarted the stalled "march of the pixels" with their so-called Retina displays.
Bad case of the jaggies: Early personal computer graphics.  128 x 48 pixels, 2 colors (on and off)
kindof a photo ...
CGA: 320 x 200 pixels with 16 colors
Photos look fairly good in 2D or 3D...
HD: 1920 x 1080 pixels, 16 million colors
the march of the pixels . . .
works of  Carl C Pisaturo