?

Log in

No account? Create an account
Benjamin's Journal
 
[Most Recent Entries] [Calendar View] [Friends]

Below are the 20 most recent journal entries recorded in Benjamin's LiveJournal:

[ << Previous 20 ]
Friday, November 14th, 2008
7:07 pm
different ways to get colour
So results on this parallel branch of research,
i.e. inkjet-printed colours with white backlight,
have been good: at least for certain kinds of
art, this will be cool. I still see a need for
blended-LED lighting or other techniques, to
get the super-saturated colours that so many
of my ideas need (if they are going to have colour
at all -- always going to the extremes, that's me).
But this being able to design colour schemes
on the screen, and then see (sort of) the same
thing in foil-dots, this is very satisfying.

The colours don't look even vaguely like they
do on the screen, so obviously one initial need
will be to print out a "swatch sheet" with examples
of what certain significant colour mixes will look
like. So it's not very "WYSIWYG", really; though
with some effort I suppose I could make the colour
that appears on the screen be numerically different
from what will be printed, adjusted to look the same
visually; i.e., I could calibrate the colour space.

Oh, so but anyway: of course a big part of adjusting
the colour space is knowing what the backlight will
be. And I wasn't joking, I did buy my very first
compact fluorescent bulb: see, I'll even risk mercury
poisoning for my art! Heh heh. My real risk is lead from
all that solder over the years, actually...
OK, so I tried it, but the quality of light was even worse than
I was expecting -- and I'm sure you know I bitch about
fluorescent lights and their poor spectral quality,
pretty much nonstop. When you actually try to use
one of these C.F. bulbs to illuminate specific colours,
wow. There just ain't much colour there, everything
looks greyish and awful. Well, maybe it's the washed-out
colours of the inkjet printer, can't do much about that...
But no: when I tried backlighting with a good old
incandescent filament, ahh! Colours! What a treat for
the eyes! Yes, it made a big difference. Hmmmm.
All the issues of reliability, heat, etc.... But can't deny,
the colours look a lot better with a nice 25W black-body
filament behind them. So that's an interesting result,
albeit not exactly welcome. I imagine white LEDs will
be a lot closer to the C.F. bulb, than to the incandescent.
That's another of my "holy grails", the black-body LED.
Friday, October 31st, 2008
12:43 pm
talked myself back into it again...
More about light and colour.
Houston, we have R3.

OK, so each of the three cone types in the retina, L, M, and S,
have something along the lines of a bell-curve shaped response.
L and M overlap heavily, S is pretty far out there by itself.
(564 nm, 534 nm, and 420 nm.)
The whole visual range is about 700 nm .. 400 nm.
L and M each have some degree of response across almost
the full width; S is narrower, covering only about the top half.

So here's why it is *not* necessary to have a variable-frequency
narrowband source or sources, to hit every R3 colour. Or why I
am currently leaning towards being self-convinced (in direct
opposition to what I've read from the Experts) that it's not
necessary.

For each of the three cone types, there is some wavelength,
which is hard to notate here but I've been calling Lambda-ms
for max-separation; this wavelength maximizes the ratio of
the response of the cone in question, over the *sum* of the
responses of its two neighbors. Sum, because of linearity
and superposition, ya see. Correct me if I'm way off track
here, someone!

E.g., Lambda-ms-L maximizes L / (M + S).

These Lambda-ms wavelengths are *not* necessarily the same
as the peak wavelengths of the three cones. But they must
exist. Or else I'm uck-fayed in the head. (Or both, I suppose...)

Can you see where I'm going with this?
Any given single-frequency stimulus (sorry I keep mixing freq
and wavelength, I'm an engineer, not a Scientist!) will in
general excite all three cones; but it *cannot* excite any
particular cone more "asymmetrically" than the Lambda-ms
signal would. It can be *equal* to the separation of the
Lambda-ms (if and only if it *is* Lambda-ms), or it can have
*less* separation, but it can't have more, by definition.

So if you have three narrowband, fixed-frequency sources,
tuned to Lambda-ms-L, -M, and -S, then you can brew up
any possible combination that a single narrowband stimulus
at any other frequency would produce.

Treating that single stimulus as a "Dirac Delta" and applying
linearity/superposition again (along with a liberal dose of
hand-waving, if you hadn't noticed), we thus hit all of R3
(any R3 signal can be decomposed into a sum of single-frequency,
i.e., spectral, signals).

Unless I'm missing something.
Because it can't be that simple, can it???
Sunday, October 26th, 2008
9:04 pm
more about colour
The plot thickens.

So I think my "theorem" last time is *not* true.
It is *not* possible to create any R3 colour using
(any) set of three fixed wavelengths.
Or so I am reading. I just don't know what's
true or false anymore!

*If* what I read is correct, then the "boundary"
colours, i.e., the fully-saturated or monochromatic
colours, can only be produced by themselves.
An infinite continuum of wavelengths is necessary,
to truly reproduce what the eye can see.

I'm still writing code to test this out for myself
(taking as given what I also read, that the eye response
is linear and thus I can apply superposition).

But if it's true, then at least I can be smug and
"I told you so" about it: pseudocolour displays really
do suck, they really can't do the full job that they should.
But this is less satisfying than actually having a plan for
how to improve the situation!

Cause here's the thing, just because triplets based
on fixed wavelengths and just varying the coefficients
can't do it (I'm still not convinced, but what if), still I
believe that three degrees of freedom are all that is necessary,
almost has to be, and of course they are called H, S, and V.
We already have this measurement system, it's just that no
physical device properly implements the variable "H", so
instead we map this space into something with fixed wavelengths,
where -- it seems, I continue to protest -- there is a necessary
loss of "gamut", places you can't get to in the original colour space.

But a device that really implements variable frequency and
bandwidth, should be able to produce every R3 colour.
For example, picture a prism, with two independently-movable
shutters that can chop down the spectrum from either end,
leaving a variable-width band in the middle. That, with a
dimmable source bright enough to punch through the full
defined brightness, through the narrowest slit we choose to attempt,
and you have it: *one* ideal pixel.
Then just Moore's Law it a little bit...

But if three degrees of freedom are all that is needed, how come
there's not a lossless transformation into a space with fixed frequencies?
Isn't there a theorem in there somewhere?
I don't know enough math, to realize how futile it probably is to
hope for this, so what the hell, I'll go ahead and hope. And tinker.
8:08 pm
colour and human vision
OK, here's what I have so far.

Perception is determined entirely by the linear
combination of responses of the three types of
cones (except for those special girls with four).
This space is called R3colour (picture the 3
being superscript).

As much as it bothers me to think of faking out
the eye with doctored wavelengths, as a practical
matter it should be possible to
reproduce any R3 combination, using the right
combination of three narrow-band sources at
any of a broad range of chosen frequencies.
I think this is a theorem.

But in practice, colour reproduction systems are
unable to hit a large outer band of the R3 colours --
and some of the most important ones if you ask me.

Is it because their sources are not narrow-band enough?
Is it because they have chosen a poor triplet of
frequencies?
Both of these factors are nowadays easy to overcome,
if they were ever a problem.
Therefore, my theorem above "must" be wrong.

Or else, there is a (big, IMO) improvement to be
had in display technology, and we are all looking
at the world through washed-out, colourless goggles,
a good percentage of the time.

Darn, I spend all my time figuring out how to
improve the realism of electronic sounds, then
I wander over to the light department, where I
figured they had this all in the bag long ago,
and you turn over a rock, and wow, what a mess.

What have you rocket scientists been doing with your time?
Sunday, October 12th, 2008
8:10 pm
my foil punch
I updated this google site I made before,

http://sites.google.com/site/cncfoilpunch/
.
It now has an attached jpeg image, showing some
"finished work". Actually, it's still not the "real deal",
because it's lit by plain white light, not by the red/blue
LED mix. But white light is better for "proofing" the
foils -- and I might like it for some images, as the
final thing, too. But this one in particular, I think,
will be getting the red/blue. Anyway.

Don't worry, I am working on other images, I know
I seem to be (and am) obsessed with this one.

It's really a multi-day (but pipelineable) process
to move an image through this technique, from
line drawing to foil behind glass.
Wednesday, September 3rd, 2008
7:15 pm
still not dead
(But good things come to those who wait!)

Hello again, oh LJ-Abyss, I gaze into thee again.

In the interim, one big project has been -- did I
tell you about it? -- this foil punch thing.
I've built the machine itself, now; further work is
centered in the software arena.
I also just got an old document scanner working,
for the same project.

The overall plan is to punch tiny holes in aluminum
foil (optionally painted black), and shine light
through: result is a little reminiscent of the Lite
Brite [tm], if you remember those.

Well anyway, it'll be a while yet before I have
actual artistic results to show, but I made a google
web site with some pics: so far, just pics of the
machine.

https://sites.google.com/site/cncfoilpunch/

OK, see ya in another two years! Or sooner,
who knows.
Saturday, January 26th, 2008
9:34 pm
anyone know about FIR filters?
I drew some pictures to show what I don't understand.

fir_confusion.gif
Friday, January 25th, 2008
10:59 pm
Tux Paint
Ah yes, some great open source software, for kids.
I almost had to write something like this myself, but
here it is and it has (roughly) all the right features.
Nice work, whoever!

www.tuxpaint.org

Unfortunately, it has a limit on image size, 400x300
or something, otherwise I'm going hmmm, does
everything "xpaint" does and more (not hard), saves into
a good format to not lose information (PNG), nice
small cross-platform program (unlike "The Gimp",
Photoshop, etc.): could be my new "vi for images"
(with the goofy sound effects turned off, of course;
actually, come to think of it, "vi" has some annoying
sound effects too). Oh, well.

This is me, not Kayleigh, in case it's not obvious:
20080125215709.png
20080125215709.jpeg
(Two formats of same image, Firefox on this Mac knows what to do
with .png but maybe some browsers written by guys named Bill might not.)
Sunday, January 13th, 2008
10:42 pm
Wednesday, December 26th, 2007
11:38 pm
avr_synth v0.0.15
BTW, our house Linux box seems to be
available on the Net again. I have put a
recent version of the avr_synth code up
on the webserver. I will also be making
this available on sourceforge.net,
being one of the first software projects I've
done which might have interest to anyone
beyond myself. Very exciting!

It's not quite to the "ready for prime time"
stage yet, but here's the URL anyway.

http://216.162.198.117/~guest/ben/src/avr_synth/
10:44 pm
piano
Also in the Kayleigh news.

http://216.162.198.117/~guest/ben/src/audio/abc/kayleigh/song_idea.Dec_27_2007.gif

I think I notated that right. Look at all those
black keys!
I can hear all kinds of fugues and canons and stuff,
based on this...
10:36 pm
Best Guess Spelling
WOSUP.ONA'TI'M.
A'FLAWR WOCT
SOF.R'AND HAN2'FLAWRS'
WOCT.V.SAM SPED
VAN A'BRD CAM
OAAAAAA!!!



Once upon a time,
a flower walked
so far and then 2 flowers
walked the same speed
then a bird came
Oaaaaaa!!!
Tuesday, December 25th, 2007
4:03 pm
my new girlfriend
My plant is still alive, I am happy to report.
I'm glad she was able to survive my early
experiments and learning curve, in terms
of what she needs to thrive. Kinda like the
feeling, once we made it through my daughter's
first night after her birth.
I learn so much from this plant. A friend
asked me if I've named her yet. No: I have
a feeling, she already has a name, I just don't
know what it is yet.
I'll tell you one thing, she loves water. I've
never seen a plant consume water so voraciously.
Right through the leaves. Her leaves are incredibly
agile; she moves them around and changes their shape
all the time, throughout the day, taking in light,
adjusting them into little concave cup shapes to catch
the water better, moving one leaf to not shadow
another, etc.; and I have seen her move a big mature
leaf right up against the edge of a fresh new baby
leaf, protecting it and also channeling water to it.
I sprinkle water directly on her leaves, many times
a day. You can see her little water-pores stand up,
like goose-bumps. Her leaves get all dark green
and luxuriant-looking, with the veins practically
glowing a bright yellow-green.
Within a minute or two, the surfaces of the
leaves are dry again. A lot of her leaves have long
slender points, which she curves around underneath
like little hooks. A drop of water tends to collect
right in the hook: and after a few minutes, it'll be
gone. Not dripped off, but slurped up.
The top two mature leaves, she uses like her "hands",
she moves them around a lot more than the other
leaves. Once the latest pair of baby leaves grows big
enough, she seems to suddenly transfer the "hands"
responsibility to the new pair. It's always one pair, never
split between two pairs or whatever. She raises her
"hands" high up to welcome the water in the mornings,
then adjusts them to line up in a plane with the next pair
down, forming a cross, to receive the light. And at
night, when she's ready to go to sleep, she droops her
little hand-leaves down until the tips just brush the tops
of the larger leaves below. This happens before I turn
off her light, not in response: I take it as the sign to
turn off the light. It's definitely not just a general change
in turgor-pressure, like we learned about in high school
biology. The other leaves stay just about the same, sticking
straight out from her square-section stem, and just the
one pair lifts up and down dramatically. But she certainly
can move the other leaves independently as well. One
time I wasn't there to turn on her light until late in the
afternoon. She had cranked her two largest leaves around
at a crazy angle, in order to face the dim light filtering
in through the open doorway. Again, not the other leaves,
not some simplistic "vegetative" reaction to light. She picked
the pair of leaves with the most surface area, and moved
only those.
Tuesday, November 20th, 2007
4:28 pm
Finally, an actual WMD threat. Hello?
(cue music) "Osama's got a nuke"...

Yay, it's another US-supported dictatorship in
the Middle East, crumbling in the face of rising
Islamic militancy. Remember the Shah of Iran?
No? Well, the Iranians certainly do. They don't
forget things quickly over there, none of these
pesky four-year election cycles (or more like, they
have the elections to placate the UN and keep the
money and weapons flowing in, but nobody
notices because nothing changes; ah yes, we were
just speaking of Pakistan, weren't we).

Who's next? Let's see... other prominent dictatorships
(oh, excuse me, "democracies" that just happen to
have military-backed career rulers and surprisingly few
rights for women), whose regimes are backed by the US,
detested by their own people, in danger of collapse in
the face of rising religious fundamentalism...
Islamic Arabia?

Iran might well *want* a nuke -- though I don't accept
the commonly-held neocon "wisdom" that there is no
other explanation for their behaviour; remember how
Saddam kept us all guessing about *his* imaginary
WMD program?

But as of now, Iran does not *have* the nuke. And
they won't, for at least a good decade. We're trying
as hard as we can to give them the maximum
incentive to proceed with a WMD program as fast as
possible, true; but things still take time, especially
in the land of opium and corruption.

Iraq doesn't have the nuke, and never did; they barely
have electricity and running water, ever since we "liberated"
them back to the stone age.

Pakistan has the nuke.

So quick, let's get a war started with Iran, so that our
military can be doubly stretched -- umm, make that
triply -- at just the time when we might actually
need our military to protect our own country from actual
dangerous and technologically-sophisticated aggression
.

"Thank you for calling the United States Military. All of our
representatives are currently serving other customers;
however, your call is very important to us. Please stay on the
line, and we will --" BOOOOM.

Nice work, neocons. Funny how the guys who crow the
most about "fiscal responsibility" are the ones who have
driven the economy straight into the ground, in a surprisingly
short time, after it was doing pretty well. And funny how the
guys who crow the most about "spreading democracy" are
the ones who are propping up the dictatorships and opposing
democratic changes of leadership, all over the globe, while
they simultaneously and systematically dismantle the
democracy in our own country. And funny how the guys who
crow the most about having a strong military and
"supporting the troops" are the ones abusing those troops
in ever-extending deployments into a quagmire with no
real plan for success, wasting our military power -- to say
nothing of the credibility of our "American values" -- on stupid
and poorly-conceived projects in countries that have never posed
a serious threat to us (or even Israel), while the real threats to
our homeland grow without restraint or opposition, fueled (heh heh)
by the flames of hatred that our own imperialistic policies are
feeding.

OK, it's not actually funny at all.
Friday, October 19th, 2007
9:21 am
money talks too much, but once in a while it says something funny
The one part of Paul Allen's little corporate Disneyland
in South Lake Union that I am in favour of, is the trolley.
I guess everyone else already knows about this, but I had
not noticed before what this spells.
It's the South Lake Union Trolley.
Lucy in the Sky with Diamonds, man.
There's a related website, ridetheslut.com.
Tuesday, October 16th, 2007
2:03 pm
musical chairs
Speaking of exploding population...

Remember the game of musical chairs? I always hated it,
but then, I hated most organized games for kids, regular
stick in the mud me, so that's not so surprising. But man,
to trip out on the implications of the game. What are we
teaching kids, with that?
Actually, I could see it being a little more innocuous in,
as I imagine it, the Victorian parlor game for adults that
it originated as (my own mental fabrication, based on
zero known facts). A chance to accidentally bump up
against the opposite sex, a staider form of "Twister".
Yeah, baby. Oh, pardon me, Miss Murple, I do believe our
legs may have inadvertently come into contact during that
last round; therefore it is my duty as a gentleman to beg
your forgiveness for my unintended transgression of
your purity.

Everybody dances in a circle; this part I would have liked.
But then, the music cuts off abruptly, fun's over, everybody
dive for a chair. And there aren't enough chairs. Slowest
one loses. Ha ha. Let's go salt some slugs now.
There's scarcity in the universe, so you'd better grab fast.

Me, in these situations, I hate 'em, and it always just makes
me want to stand back, apart from the rush, and say like what
are we all doing here? Let's go get some more chairs, who
wants to help? At the airport... in traffic... so many places we
live out the musical chairs game. Checkout aisle. Oh, go ahead,
I have two items and you have a cartload and a half in one cart,
plus a phone call in progress. I don't give a fuck, I'll read all about
Britney and the space aliens, I'm sure your whole life is a
bigger hurry than I'd ever agree to put up with.
It's all about moving at the right speed.
Sometimes it's 20 in a 30 zone. And yes, officer, sometimes it's 110.
Ah, they never understand...

And stop cutting off that fuckin music, I was diggin it.
11:19 am
Mona Mona
Our sweet border collie, Mona, is ill.
I am exceedingly concerned.
I could tell she was sick by the energy lines,
but I can't do anything about it.
Thursday, October 11th, 2007
4:32 pm
Support The Troops: someone's got to, since the army doesn't.
Oh my god. What incredible assholes.

730 vs 729 days.

A bunch of National Guard troops (who shouldn't even
be in Iraq, but that's a different rant) just returned, after
serving for two years straight. Which is, absent leap years,
730 days.

But there was a teeny little administrative clerical error. Some
of the soldiers had "730" entered as their time served, but
a bunch of others (from the same company) had "729".
Innocent little error, hey, what difference does one day make?

Interestingly, though, it makes all the difference: 730 days is
the threshold to receive full educational benefits. 729 or fewer,
means you get only partial benefits. I forget the amounts, but
it makes a big financial difference.

Yeah. So these god-damned war pigs who keep cowing the
spineless Democrats into submission with this "support the
troops" mantra, are fucking over the troops to save a few bucks.
It's the same mentality as hiring two part-time workers to cover
one full-time job, so you don't have to pay benefits; except it's
even more blatant.

Now everyone is scrambling around, trying to manage the PR
problem, trying to pass legislation to fix this, etc.: how come
some general can't, with one stroke of a pen, put this to rest?
One stroke of a pen (or a mouse click, more likely) is all it takes
to send them into the jaws of death. Can't we send them to
college as well?

I don't know how many people are affected, but suppose it's 100.
If I were a rank-and-file soldier, I'd offer to donate 100 of my days,
one for each person who got screwed. I hope someone thinks of
this approach, it'd shame those bloodthirsty assholes in Washington.

Politicians hide themselves away,
they only started the war.
Why should they go out to fight, they
leave that up to the poor.
Wednesday, October 10th, 2007
5:41 pm
LSD
LSD, which chemist Albert Hofmann called his
"problem child", was first synthesized on Nov 16, 1938.

63 years later, on the same day, my child was born.
But she's not a problem -- most of the time.
Tuesday, October 9th, 2007
10:08 am
Wow, I never knew vodka was available in plastic bottles.
I was recently speaking of my compulsion to fix
or help complicated basket cases. "Case" in point.
I have this friend. He's great and everything,
except for this minor little quirk where he is
trying to destroy himself with alcohol. As Ozzy
says, even with earnest effort, this takes a while.

I've seen plenty of light-duty alcoholics; arguably,
I've been one myself, though I certainly never sank
into the addiction to the extent that I did with
nicotine, way back when. Phew, glad to be out of
that miserable quagmire! Actually, I am thankful for
my prior addiction to cigarettes, it showed me what
the whole world of substance dependency is all about,
and it probably helped me not get addicted to (or even
try) the many far-worse substances I've been offered
over the years.

That's right, kids, try cigarettes: they helped me.

OK, so maybe that should help me to understand my
friend. But damn, I've heard about, but never before seen
in action, a true-blue alcoholic of this class. It's not
just a social crutch or a little overindulgence with him.
We're talking, the big question first thing in the morning
is how to get some alcohol. Goes great with coffee.
You have to really plan ahead to manage an addiction
like that -- a task which alcohol doesn't help with!
But where there's a will, there's a way.

So here's the thing. I have a weak spot for basket
cases, I have an irrepressible compulsion to Help People,
I guess it's all a challenge to me or a way to sharpen
my "chops" as a healer. But even I can tell that this
guy is too much for me, unwise to take on. Or not even
that it's too much, it's just that there is no traction:
I can't help those who don't want to be helped. So I
feel like it's a waiting game: how far do you have to sink
before your survival instincts kick in? How much do I
have to watch, and resist the impulse to do something?

They have support groups to help with this: Al-Anon is the
main one I guess, for spouses and friends of alcoholics.
But they seem to go by the same 12-step credo as AA.
I don't feel like I can fully sign on to this. There's
too much emphasis on surrendering control. I'm sure it's
a great idea for most people and helps them feel so much
better about their place in the world, but I can't quite
accept the idea myself. Well also, I think these groups
(though they make a point of saying "friends" too), mostly
cater to spouses, or others who (in reality or in their
feelings) have no escape, who are stuck with the alcoholic.
In such a case, I'm sure it's good to get in touch with how
much you can't change the other person. In my case, I do
have a compulsion to "fix" this person, but I don't believe
myself to actually be stuck with him in any sense. So maybe
I'm just not close enough to him to warrant Al-Anon. But
what I'm really looking for are, indeed, ways that I *can*
influence the situation, not ways I can get to like having
no influence.
[ << Previous 20 ]
My Website   About LiveJournal.com