h a l f b a k e r yA hive of inactivity
add, search, annotate, link, view, overview, recent, by name, random
news, help, about, links, report a problem
browse anonymously,
or get an account
and write.
register,
|
|
|
First, for those unaware, a little info on how Jpeg compression works. The image is cut into 8x8 blocks, which are then transformed to give their spacial frequency components. Then, depending on the quality desired, a certain amount of unwanted data is thrown away. Finally the data undergoes a number
of lossless compression techniques.
The Physical Jpeg is giant electromechanical display. Instead of pixels, it uses tiles corresponding to 8x8 blocks.
On the first day of the installation, members of the public are enticed to send in pictures via social networks and mobile gizmos. The Physical Jpeg gladly springs to life and the carriage at the top zooms back and forth, dropping tiles into the grid like a game of Connect-4. The end result, however, is terrible, because the magazine contains only 8 different types of tile.
On the second day, the magazine is extended. More tiles are added, each printed with an 8x8 section of image. The choice of which tiles to print is dictated by the 'quality' parameter of the Jpeg algorithm, which has been artificially extended to well below the normal range.
By the third day, images are almost identifiable on the display. There is no colour info (we are not totally mad) and the quantization is evident. As we move from having a huge number of tiles in a small number of bins to a large number of bins with relatively few tiles, we can optimize by overprinting: almost all tiles will be a linear combination of the few basis tiles we started out with.
Upon finishing each picture, the tiles are released and fall down to a tray where they are collected and meticulously sorted by the machine. For some people, this may be the most exciting part of the exhibit.
65,536 types of tile are present by the end of the week - a number dictated by the sorting mechanism more than anything. With individual bins impractical, the carriage now has to root through buckets of tiles before finding one that it needs. The end result is still a greyscale picture of embarrassing bit-depth.
The Physical Jpeg is efficient at a microscale yet grossly inefficient at its overall goal. It is a triumph of construction and a tribute to the algorithms it is trying to describe. It is a Rube Goldberg machine that, paradoxically, is much simpler than the real method of decoding a Jpeg. Principally it is a poignant reminder of the futility of mechanical engineering in the 21st century.
Can get a bit tricky description wise..
photography_20rewor...20avoid_20verbosity
[not_morrison_rm, Jun 28 2014]
Illustration
http://mitxela.com/other/pjpeg/anim.htm
[mitxela, Jun 28 2014]
How a Linotype works
http://www.youtube....watch?v=6wHiddZOfa8
When I feel down, this film never fails to cheer me up. [mitxela, Jun 28 2014]
Please log in.
If you're not logged in,
you can see what this page
looks like, but you will
not be able to add anything.
Destination URL.
E.g., https://www.coffee.com/
Description (displayed with the short name and URL.)
|
| |
//futility of mechanical engineering in the 21st
century// |
|
| |
Which aspects of the futility of engineering were
you thinking of? The Internet? The ISS or Hubble
Space Telescope? Artificial hearts? The channel
tunnel? |
|
| |
The Honda Prius, obviously. |
|
| |
// we are not totally mad // |
|
| |
You DO have to be mad to work here? |
|
| |
<shuffles toward door> 'Um... doctor? am I cured yet or can I stay?' |
|
| |
//The Internet? The ISS or Hubble Space Telescope? Artificial hearts? The channel tunnel?//// |
|
| |
Those are all terrific examples of engineering feats. Even so, I would like to draw your attention to the parts which say "mechanical" and "21st century." |
|
| |
Because it's a Saturday and I have literally no objective in life, I decided to spend the afternoon animating what this might look like. [Link] |
|
| |
//I would like to draw your attention to...// |
|
| |
Could you perhaps provide some examples? |
|
| |
Oh, and [+] for the illustration. |
|
| |
Yes, what MB said so eloquently. |
|
| |
I can't think of any. But that's the point. In my eyes, all of the great engineering achievements of the last fifteen years have been electrical, civil, medical, computational... but not mechanical. So what I meant is that trying to solve today's problems through mechanical means is futile. |
|
| |
Ah, OK, so the final bit of your idea should be read
as "the futility of trying to solve 21st century
problems with mechanical engineering", rather
than "the futility of 21st century mechanical
engineering". |
|
| |
I would still disagree, however. There are some
mechanisms (some developed in the 21st century)
which just make you want to weep with joy at
their simple ingenuity - often far superior to
electronic gizmullery. |
|
| |
And there are many problems which cannot be
solved by the application of electronics and
software. A bridge, for example, is usually best
implemented in hardware. |
|
| |
So, I am still unsure of the point you are trying to
make. |
|
| |
Respond carefully - your bun hangs by a thread. |
|
| |
Marvellous mechanical engineering still exists, yes, but not on the scale that it used to. |
|
| |
Have you seen that 1960s film about how a Linotype machine works? (I'll find a link - really worth watching if you haven't seen it.) The level of complexity borders on the absurd, yet there was an era when the Linotype was not only practical, but by far the most economical method of typesetting. It's certainly possible that greater, more ingenious, more elegant, or more complex mechanical achievements will take place in the near-future, but the economic incentive has gone. |
|
| |
There are instances of elegant mechanical solutions in the 21st century, but vastly more problems are solved by removing moving parts entirely. And I admit this is quite often a better, cheaper, more reliable solution. But it has lost its visual appeal. |
|
| |
(Software solutions can be beautiful, but usually an inherent ugliness is induced because the resources are so immeasurably cheap that there is no incentive to optimize. Again, it's the economics that I'm complaining about. There are a number of examples (e.g. demoscene) of just how elegant programming can be. In a way, although we are oblivious to it most of the time, the ugliness of code is a far greater problem to me than the disappearance of mechanical engineering. It is not uncommon for the efficiency of equivalent solutions to differ by many orders of magnitude, but the sole reason to choose one over the other would be 'elegance'. ) |
|
| |
What point am I trying to make? None, I suppose. To be brutally honest, I had written an "art" idea and felt a kind of social pressure to suggest a meaning. A real artist might have left the interpretation open. The only reason I want to build this is to gawk at the moving parts. |
|
| |
//Software solutions can be beautiful// but are an acquired taste: the elegance of a Dykstra algorithm, the zen-ness of a Jackson structured batch program, the stark wonder of a hand-optimized assembly routine... <sigh> mostly in contrast to the "Lookit ! a big turd made out of smaller turds, made out of even smaller turds; isn't it marvellous" stuff that's called "programming" only by the people selling it and the lawyers that are there to make sure it stays sold. |
|
| |
In other news, I have no clue what the Idea is about, but will probably figure it out in a bit. [edit: or watch your link, which makes it pretty obvious] |
|
| |
//the futility of mechanical engineering//
//some examples//<b>
3D printing would be the best new example of great mech eng. But pretty much anything you use, see, touch, or even eat, has involved a mechanical engineer somewhere in it's lifecycle.
Oh, and kudos on the animation. It's quite impressive. |
|
| |
Nanotech gears, materials optimization in component design, hybrid vehicle design and vehicle integration, manufacturing requirements, these are all largely mechanical problems. |
|
| |
Having seen the illustration, I doff a fuzzy image of my hat to you, [mitxela]. [+] |
|
| |
+ I'm impressed by the illustration, if not bewildered
by the actual description. JpegBun |
|
| |
I like the way that we are presented with a digital rendering of a physical analogue of a digital rendering of a physical object. |
|
| |
Yea, bun for the amazing illustration. |
|
| |