What Are You Working On? - December 2014

[ECrownofFire]

[QUOTE=ZenX2;46618102]I just realized that complex adaptive systems are analogous to perlin noise in that they turn the garbage/noise of extreme complexity into something meaningful by being self-similar
Basically perlin noise is why the universe works[/QUOTE]

There's also the chance that they never reach an equilibrium and the universe remains in chaos forever.

Kind of like Dwarf Fortress or Minecraft.

[chaz13]

[QUOTE=Pangogie;46619589]I have no idea what I'm looking at, but it looks science-y and complicated!
Geez, dude.[/QUOTE]

I'm not entirely sure what I'm doing either :v:

But in those pictures, it starts with SR-CT (synchotron-radiation computer-tomography, basically a special 3d xray machine) image you can see (it's got 1022 slices so it's 3-dimensional). It then needs to be filtered so the black dots you can see (called lacunae, cavities in which the ostecyte bone cells reside) are white and everything else is black.

There's numerous filters which could be applied, pictured here the maximum entropy filter is used but I'm going to implement more.

After that, a sort of 3D flood fill based algorithm is used to track the positions of each lacunae in 3D space and determine their center of mass, area, etc and the center points are plotted in blue on the scatter graph.

This is just the data-processing part, I haven't started anything to do with the actual project aim yet :v: Next I need to fit ellipsoids to each lacunae somehow - not sure how to do this just yet.

[cartman300]

[img]http://i.imgur.com/GZLKs4m.png[/img]

Operations for rationals.

[proboardslol]

Pretending to know what I'm doing with Fourier transformations. As part of a test phase, I made a 0.5hz wave in audacity and saved it as a .wav file. I did an octal dump (-d flag) into a text file, and then parsed that data into a simple graph. What I found odd was this:

[img]http://i.imgur.com/9JLQ8LW.png[/img]

The data, when graphed, looks instead like a cosecant function rather than a sin function.

Is there anything I can make of this?

[Berkin]

[QUOTE=proboardslol;46624543]Pretending to know what I'm doing with Fourier transformations. As part of a test phase, I made a 0.5hz wave in audacity and saved it as a .wav file. I did an octal dump (-d flag) into a text file, and then parsed that data into a simple graph. What I found odd was this:

[img]http://i.imgur.com/9JLQ8LW.png[/img]

The data, when graphed, looks instead like a cosecant function rather than a sin function.

Is there anything I can make of this?[/QUOTE]

I could be wrong, but this looks to me like a case of arithmetic overflow/underflow where the value wraps around when it goes out of range.

[proboardslol]

[QUOTE=Berkin;46624600]I could be wrong, but this looks to me like a case of arithmetic overflow/underflow where the value wraps around when it goes out of range.[/QUOTE]

Do you mean with the graphical processing of it? I don't think so. The window was resized to that position.

this is the version with a 50px vertical offset and a 400x300 window

[img]http://i.imgur.com/4SpoLOK.png[/img]

[editline]2nd December 2014[/editline]

Ah wait I see: the data is represented by audacity centered from the middle, but it's represented in octal as switching. I'm wondering what kind of math I can do to it to put it back to normal...

[Berkin]

[QUOTE=proboardslol;46624653]Do you mean with the graphical processing of it? I don't think so. The window was resized to that position.

this is the version with a 50px vertical offset and a 400x300 window

[img]http://i.imgur.com/4SpoLOK.png[/img][/QUOTE]

If you exported a WAV file from Audacity, it's most likely signed 16-bit uncompressed PCM. This means you'll have negative values. Depending on how you projected your sample data, you may have ended up with a different value range. You should check your samples to make sure they're still in the correct range; converting them back to a WAV file is a good way to verify this.

It's also possible that the origin point of your waveform renderer is not in the center (thus the wrapping).

[proboardslol]

OH FUCKING DUH

it rolls over at negatives. I haven't interpretted the data as signed so it just assumes I'm talking about positive numbers.

[img]http://imgur.com/dG87kiw.jpg[/img]

see, it switches between zero and the size of an unsigned small int.

[editline]2nd December 2014[/editline]

fixed it

[IMG]http://i.imgur.com/GPOlzzF.png[/IMG]

[editline]2nd December 2014[/editline]

fuck I love being smart

[sidneys1]

Refactoring is important, I've been told.
[img]http://imgur.com/fqd0ODV.png[/img]
Edit:
[img]http://imgur.com/ugiUe4j.png[/img]

[sloppy_joes]

made a basic raytracer from scratch for my graphics class:

[img]http://i.imgur.com/68zpq6v.png[/img] [img]http://i.imgur.com/iTorENm.png[/img] [img]http://i.imgur.com/mhCTSNW.png[/img]

might try extending it so it does photon mapping, so i can get those sweet sweet caustics.

[editline]3rd December 2014[/editline]

i dont think the refraction works perfectly, maybe i should fix that first.

[Cold]

Well since you've made the ray-tracing part you pretty much have already implemented half the photon mapping stuff.

[Foxtrot200]

So, a while ago, I wrote a [url="https://github.com/VerbalExpressions/"]Verbal Expressions[/url] implementation in Lua. I hadn't really used it for anything and thought I'd fix it up and [url="https://github.com/Apprehentice/vex"]put it on GitHub[/url] for others to use.

Again, I haven't really used it, so it's probably buggy, but it's documented, so there's that. If you find a way to improve it, please do. I would love you long time.

[Darwin226]

How many samples per pixel do you usually need to make diffuse surfaces look good in a ray tracer?

[peep]

woo first highlight
cool

still chugging along
slowly

[Fourier]

[QUOTE=Darwin226;46627058]How many samples per pixel do you usually need to make diffuse surfaces look good in a ray tracer?[/QUOTE]

I think 4.. :)

You can also calculate needed samples on the fly, depends on, if the points of the rays that hit the surface are really close to each other.., if you know what I mean.

Imagine you fire 4 samples at sphere, but one ray does not hit the sphere, but the wall behind the sphere. This means points of the rays are not close to each other, and in this case you increase number of rays for this pixel :)

[chaz13]

[QUOTE=Fourier;46628760]I think 4.. :)

You can also calculate needed samples on the fly, depends on, if the points of the rays that hit the surface are really close to each other.., if you know what I mean.

Imagine you fire 4 samples at sphere, but one ray does not hit the sphere, but the wall behind the sphere. This means points of the rays are not close to each other, and in this case you increase number of rays for this pixel :)[/QUOTE]

I use about 100 :v:

[Darwin226]

[QUOTE=Fourier;46628760]I think 4.. :)

You can also calculate needed samples on the fly, depends on, if the points of the rays that hit the surface are really close to each other.., if you know what I mean.

Imagine you fire 4 samples at sphere, but one ray does not hit the sphere, but the wall behind the sphere. This means points of the rays are not close to each other, and in this case you increase number of rays for this pixel :)[/QUOTE]

Not sure that matters. You pretty much only care about screen space distance. 4 is also kind of a small number. If you have a 100% light source and everything else dark except for one reflective surface then you only have 2^4 possible values for any pixel. 16 colors don't make a smooth gradient.

[DarKSunrise]

[thumb]http://files.1337upload.net/messagetest0000-31aa02.jpg[/thumb]

generating random bluescreens with rant

[esalaka]

I guess it's technically not programming, but I can't see any other suitable place to complain about TeX.

It produces lovely output but goddamn is it awful to write.

[editline]4th December 2014[/editline]

I'm writing something for school and as it involves formulae and graphs and MikTeX is pretty much the only suitable software package I had installed, I'm using LaTeX.

[alien_guy]

[QUOTE=esalaka;46631462]I guess it's technically not programming, but I can't see any other suitable place to complain about TeX.

It produces lovely output but goddamn is it awful to write.

[editline]4th December 2014[/editline]

I'm writing something for school and as it involves formulae and graphs and MikTeX is pretty much the only suitable software package I had installed, I'm using LaTeX.[/QUOTE]

I think its pretty awesome. infinitely superior to using word.

The biggest issue I've had was getting citations to work properly.

[esalaka]

[QUOTE=alien_guy;46631480]I think its pretty awesome. infinitely superior to using word.[/QUOTE]

Precisely. It's just the sheer amount of macros that go in simple formulae that's kind of... disturbing.

\begin{align}
	P_{DC}&= {10\log_{10}(\frac{\SI{1}{\milli\watt}}{\SI{1}{\milli\watt}})}~\si{\decibel{m}} = {10\log_{10}(1)}~\si{\decibel{m}} = \SI{0}{\decibel{m}}\nonumber\\
	P_{AC}&= {10\log_{10}(\frac{\SI{0.0049}{\milli\watt}}{\SI{1}{\milli\watt}})}~\si{\decibel{m}} = {10\log{10}(0.0049)}~\si{\decibel{m}} \approx \SI{-23.10}{\decibel{m}}\nonumber\\
	\nonumber\\
	P&= {10\log_{10}(\frac{\SI{1.0049}{\milli\watt}}{\SI{1}{\milli\watt}})}~\si{\decibel{m}} = {10\log_{10}(1.0049)}~\si{\decibel{m}} \approx \SI{0.021}{\decibel{m}}\nonumber
\end{align}

[img]http://i.imgur.com/Ofj5uBZ.png[/img]

[editline]4th December 2014[/editline]

Of course, it's not [I]necessary[/I] to use the \SI macro but it's the simplest way I found to write units in non-italics.

[editline]4th December 2014[/editline]

It's kind of ironic in that dBm isn't an SI unit

[editline]4th December 2014[/editline]

Oh, I just spotted a mistake there (10\log{10} should be 10\log_{10})

[alien_guy]

[QUOTE=esalaka;46631509]Precisely. It's just the sheer amount of macros that go in simple formulae that's kind of... disturbing.

\begin{align}
    P_{DC}&= {10\log_{10}(\frac{\SI{1}{\milli\watt}}{\SI{1}{\milli\watt}})}~\si{\decibel{m}} = {10\log_{10}(1)}~\si{\decibel{m}} = \SI{0}{\decibel{m}}\nonumber\\
    P_{AC}&= {10\log_{10}(\frac{\SI{0.0049}{\milli\watt}}{\SI{1}{\milli\watt}})}~\si{\decibel{m}} = {10\log{10}(0.0049)}~\si{\decibel{m}} \approx \SI{-23.10}{\decibel{m}}\nonumber\\
    \nonumber\\
    P&= {10\log_{10}(\frac{\SI{1.0049}{\milli\watt}}{\SI{1}{\milli\watt}})}~\si{\decibel{m}} = {10\log_{10}(1.0049)}~\si{\decibel{m}} \approx \SI{0.021}{\decibel{m}}\nonumber
\end{align}

[t]http://i.imgur.com/Ofj5uBZ.png[/t]

[editline]4th December 2014[/editline]

Of course, it's not [I]necessary[/I] to use the \SI macro but it's the simplest way I found to write units in non-italics.

[editline]4th December 2014[/editline]

It's kind of ironic in that dBm isn't an SI unit

[editline]4th December 2014[/editline]

Oh, I just spotted a mistake there (10\log{10} should be 10\log_{10})[/QUOTE]

Why can't you just write it like this if the output is essentially the same?

\[ P= \mathrm{10\log_{10}(\frac{1.0049mW}{1mW}) dBm = 10\log_{10}(1.0049) dBm \approx 0.021 dBm} \]

[img]http://i.imgur.com/Xuz9m92.png[/img]

[Berkin]

[QUOTE=DarKSunrise;46631429][thumb]http://files.1337upload.net/messagetest0000-31aa02.jpg[/thumb]

generating random bluescreens with rant[/QUOTE]

I've been having a really shitty week and you just made me laugh. Thanks for sharing this.

[Pappschachtel]

My toaster wanted to boycott my bachelor degree xD
Created a nice light show and a shortcut...
Instead of kitchen's fuse only, the master switch got triggered... no electricity in whole house xD
luckily android studio saves your stuff!

[mcormack1994]

Hey guys, check out my Leap Motion 3D Jam entry!
Rating is now open so please do rate and leave feedback!

[url]http://itch.io/jam/leapmotion3djam/rate/14886[/url]

Sadly you need a DK2 and a Leap to run it but here's a video of me trying to make breakfast.
[video=youtube;hkyegFnFl9w]http://www.youtube.com/watch?v=hkyegFnFl9w[/video]

[esalaka]

[QUOTE=alien_guy;46631648]Why can't you just write it like this if the output is essentially the same?

\[ P= \mathrm{10\log_{10}(\frac{1.0049mW}{1mW}) dBm = 10\log_{10}(1.0049) dBm \approx 0.021 dBm} \]

[img]http://i.imgur.com/Xuz9m92.png[/img][/QUOTE]

That would be essentially the same if you bound units with ~ instead of a space, yes.

For one, because I didn't know of \mathrm - and second, because I also use units inline, in which case \mathrm would actually make it longer.

[Darwin226]

[QUOTE=Pappschachtel;46632534]My toaster wanted to boycott my bachelor degree xD
Created a nice light show and a shortcut...
Instead of kitchen's fuse only, the master switch got triggered... no electricity in whole house xD
luckily android studio saves your stuff![/QUOTE]

Are you saying you don't vigorously mash Ctrl-S every few characters you type?

[Berkin]

I finally figured out how to add syntax highlighting to RantBox.

[img]http://i.imgur.com/O1hN4uz.png[/img]

It's amazing what such a small thing does for readability. I should really make a standalone Rant editor, though.

[Darwin226]

Lately (and for a while now) I haven't been sharing what I do because it's not visual stuff and I think it would be boring to read about.
Is there anyone here that would find things like [URL="http://lukahorvat.github.io/programming/2014/12/03/type-level/"]this[/URL] or [URL="http://lukahorvat.github.io/programming/2014/11/18/haskell-memoization/"]this[/URL] interesting?
I'd like to contribute some content as long as it isn't just filler everyone scrolls past.

[mastersrp]

[QUOTE=Darwin226;46635082]Lately (and for a while now) I haven't been sharing what I do because it's not visual stuff and I think it would be boring to read about.
Is there anyone here that would find things like [URL="http://lukahorvat.github.io/programming/2014/12/03/type-level/"]this[/URL] or [URL="http://lukahorvat.github.io/programming/2014/11/18/haskell-memoization/"]this[/URL] interesting?
I'd like to contribute some content as long as it isn't just filler everyone scrolls past.[/QUOTE]

Seems pretty interresting, although I don't have the time to read it all at the moment.

Anyway, a thing that might be useful is stating which language you're specifically writing about. While I realize that it can be deduced to Haskell (GHC), that isn't expressly stated in the introduction, which could be useful just to put yourself and the reader on the same page from the start.