Rendering the Earth

I recently came across a neat tutorial for Blender on how to create a realistic Earth image using real land and cloud images from NASA. Well, all right, we hit “realistic” about half way through the tutorial and went on to “slightly stylized,” but I like how it came out so I thought I'd share it here.

…and looking at it now, I realize you can't actually see any of the land textures because they're all on the dark side of the planet (except Hawaii, which should be visible, but I can't find it even in the full-size image). Huh. Not the greatest composition ever, is it? I do love the night lights effect though, that's pretty neat.

While I could have added a realistic starry backdrop, I was focused on finishing the rest of the tutorial and instead went for a simple random noise texture stretched to get stars. While playing around with different noise textures for the background I found the Voronoi noise texture and thought it was cool enough to save a picture of:

(Incidentally, in this picture you can actually see the land textures…)

Anyway, I now have a fairly simple (minus the stylized parts) way of rendering a pretty realistic Earth, and I bet NASA has similar textures for other Solar System bodies, so I may have to do something with that in the future…

Keeping Busy, with Art!

Well, here I am, unemployed for the first time in…over five years, I think? It's strange to have this much free time on my hands, but by no means unpleasant. Apart from the usual stuff of looking around for a new job, I'm relishing the chance to get back to doing something I haven't done as much of lately: art.

I've never really considered myself much good at doing real-world pencil-to-paper type art, so when I discovered programs like Inkscape, GIMP, and Blender, I was quite excited. I've spent many happy hours playing around with each of them, but with a full-time job I found I just didn't seem to have as much creative energy left at the end of the day. (Probably because I was often spending a lot of it at work coding!) With all this free time on my hands, though, I'm looking forward to getting back into creating more visual stuff again.

In fact, I got back into Blender this week, and came up with some simple but neat-looking stuff. I haven't used Blender for even longer than Inkscape or GIMP (due to occasionally having need of 2D image manipulation but not much need for 3D modeling) so I was worried I'd need to completely relearn it, but it looks like quite a lot of it is coming back to me already after a little brushing up. Monday I created a simple test of Blender's more physically-accurate renderer Cycles with a small brushed-metal model and some simple lights:

I think it came out pretty nicely for not having used Blender in a few year (and not really using Cycles even at that point, as it was pretty new back then). I'd like to really push my Blender skills and improve them while I have the chance, so I'll probably have more practice work to post on this blog in the next few weeks. And maybe some other stuff too, who knows? A hui hou!

Silver Fountain

Well, after making my silver fountain picture the other day and having gone to all the work to create a three-second-long fluid simulation for it (which probably took half an hour to calculate) only to use just one of the 75 resulting frames, I decided it would be fitting to make an actual video, since fountains are generally known more for their liquid nature.

I re-modeled the fountain itself since I wasn't quite satisfied with how it looked (which included making it an actual object with volume rather than a paper-thin shell), calculated a full five seconds of fluid simulation, then rendered the entire thing at 25 frames per second, 300 cycles per frame. I'm not sure how long the entire process took because I did it over night, only that it took less than ten hours.

I should explain that the reason I can get such realistically rendered water is because of Blender's new (as of version 2.61 from last December) internal rendering engine called Cycles. Unlike the original internal render mechanism, Cycles has no set “end” point. Instead, it simply renders a predetermined number of cycles, each cycle improving the picture. Each cycle does not, however, improve it by the same amount, leading to a point of diminishing returns after which additional cycles cost more in time than they are worth in aesthetic improvement. What that point is, of course, is entirely subjective.

A larger version of frame 43 from the video, 300 cycles. Took 13 minutes 50 seconds to render.

The existence of this point leads to a trade-off between “How good do I want this picture to look?” and “How much time am I willing to devote to making it look good?”. This becomes especially critical in video rendering when you're rendering hundreds of frames. For instance, for my single picture a few days ago, I let it render for 1,000 cycles, which took two hours and twenty-one minutes. Obviously I can't do that for each of the 125 frames I was rendering for this video, but I still wanted it to look decent, so I compromised and rendered 300 cycles for each frame. If you look closely (or maybe not all that closely, I don't know) you can definitely spot the difference.

Similar to above, but with 500 cycles. Not as dramatic a difference as I'd hoped, but you can see that the water in the basin and the shadow looks less noisy than the 300 cycles version. This one took 23 and a half minutes to render.

I do think that Cycles is a very cool development for Blender (although it forces a re-learning of the texturing and lighting process almost from the ground up), and it certainly works well for single pictures when you can let it go for a long time (it's not uncommon to see artists using such open-ended renderers letting them run for 5,000 cycles or more). I've got some ideas for future artwork of a more substantial kind, so keep an eye out for it in the future.

Composition in Silver

In honor of today being my parents' 25th wedding anniversary, I created this thematic composition on the theme of silver. I'm ashamed to admit that it sort of snuck up on me so I only had a few hours last night to put it together, but I think it came out alright.

I actually got the idea for this composition while taking a drink from one of the drinking fountains at the Vis, where the combination of silvery brushed metal, liquid water droplets splashing and frolicking about and a low sun angle added up to create a sublimely beautiful spectacle. The picture below from my phone is my best attempt to capture it, but it was much more beautiful than it appears here.

As mentioned, total work on this picture was only ~4-5 hours, a fair portion of which was waiting for Blender to render the image so I could see how it looked and what tweaks needed to be made (in fact the final render, which I left going overnight, took almost three-and-a-half hours by itself). The metal texture was made by me in GIMP, and I'm quite pleased with how it turned out for the amount of time I had to work on it. I was trying to replicate the white scratches you can see in the picture above. The modeling and fluid animation  were done in Blender, and rendered with the new internal Cycles engine that achieves much more physically realistic pictures. For instance, that water has the index of refraction of water at room temperature, and the light in the scene is realistically being refracted through it. I haven't done much fluid animation before and have had mixed results when I tried, so I'm really very happy with how this came out. I might even animate a short video of it splashing in the fountain in the near future.

Finally, happy silver anniversary Mom and Dad! Here's to another 25 years!

Daniel's Software Musings (Part 1)

As you can see, it's been a few days since I said I would do some reviewing. When I wrote that, I conveniently managed to forget that I tend to put off doing reviews because the philosophical concentration required to achieve an Aristotelian level of understanding of the review subject's essence is somewhat taxing. But I didn't forget, and it's been nagging me for the past two days, so here's the first part of what will most likely be a multi-part series over time of various programs.

To start off, I'm going to review some programs that I use fairly frequently. I'll start of by mentioning that all of the programs I'll be reviewing are free; most free in the sense of “freedom to use for your own purposes” (which usually means open source), all free in the sense of “no cost”. Also, because of that, most of these programs will run on pretty much any operating system.

Creativity

Blender

Blender is a 3-dimensional modeling program that I've been using since the autumn of 2008. I've posted pictures made using it before, most recently here. It is a very powerful program, and as such has a learning curve that may charitably be described as “precipitous”. However, for the persevering, there are a lot of resources online detailing how to use it, ranging from tutorials for those who have never opened it before to discussions of many of the highly advanced features that are hidden within. Blender may not be quite on par with similar commercial products, but it is still an incredibly versatile and powerful program, and is a couple of thousand dollars cheaper to boot. If you're willing to put the time into learning it, you can produce some amazing stuff. In addition to still renders, it can also do animation.

The GIMP

The GIMP, or GNU Image Manipulation Program as the full unimaginative name goes, is an incredibly powerful application on par with the much costlier Photoshop in many ways. And the ways it isn't are ones that typically only the most advanced of users will notice. I use this program for all my picture editing needs (such as the many panoramas I make) because it can handle anything from the simplest of tasks to projects of extreme difficulty and complexity. For instance, the picture in this post was made entirely with GIMP. I used it to make a few modifications to the Blender logo picture seen above, in fact. Like any new program it will take you a while to get used to, but it definitely has an easier learning curve than Blender by virtue of the fact that you're usually working with pre-existing pictures instead of making them from scratch, and you can get your feet wet doing small modifications before jumping into the deep end of photo editing.

Productivity

Python

If you need a powerful, multi-purpose programming language that is also simple to learn, you need Python. This elegant language is easy to pick up (I've learned everything I know about it by consulting its thorough and well-organized help documents) and yet has a wide variety of standard commands that provide most of what you need right away (“batteries included” is a phrase often used by fans to describe it). And if that function you need doesn't happen to come with the download, chances are, unless it's an extremely uncommon or unusual function, that someone, somewhere, has already coded it up and made it available. Ultimately, Python provides a nice “first language” for beginners that can be explored and extended to match the user's growing skill. It's well summed up in the unofficial motto: “easy things should be easy, and hard things should be possible.”

Recreation

Quod Libet

Quod Libet is a very interesting little music organization and playback program that plays a variety of common formats. It comes with Ex Falsa, a bundle of code that organizes your music based on tags. Or, rather, it lets you organize your music the way you want to using tags. It had a definite learning curve to it, but I find it easy and enjoyable to use now. It lets you search your music based on any tags it has: name, composer, album, artist, etc. Like other such programs, you can organize your music into playlists, which it will play very well. And it has one real stand-out feature I haven't seen before: by right-clicking on the Play button while a song is playing you can tell Quod Libet to stop after that song, which turns out to be a surprisingly useful feature when you only want to listen to one song in a playlist. It's one of those didn't-know-you-needed-it-but-now-you-can't-live-without-it functions. (If you're wondering about the name, “quod libet” is a Latin phrase meaning reoughly “whatever you like” or “what you will.” It comes from a phrase in logic “ex falsa, quod libet”, which means “from a falsehood, whatever you want” referring to the fact that if you begin with a false premise you can logically prove any (incorrect) thing you want. The developers liked the idea of “whatever you want” to describe the organization capabilities of Quod Libet, hence the name.)

Keep watching in the future for more informative and hopefully interesting reviews! A hui hou!

Galactic View

It's been a while since I posted any artwork up here, but today I finally have something to show. This particular picture (another in my space artwork series) represents far and away the longest time I've taken between start and finish on one project; this picture was actually begun in the summer of 2008! I'll explain the reason for the gap further down.


This picture is the result of me wondering what you might see from a planet in orbit around a star on the outskirts of a globular cluster. Globular clusters are roughly spherical collections of hundreds of thousands of stars that orbit galaxies (including our own; the Milky Way galaxy has ~180 known globular clusters). They're composed of old yellow or red stars, not the bright blue young ones that make the spiral arms of spiral galaxies so obvious. Speaking of spiral galaxies, they aren't easy to draw in any sort of half-way realistic way which is the reason it took me so long to finish this picture. It's actually two pictures: the sky is a 2D image I created using the GIMP, and the foreground is a 3D render made using Blender. The globular cluster was mostly complete back in 2008 (I did only minor retouching upon taking up the project again), but I got stuck trying to find a way to make a spiral galaxy that looked good. This was actually so long ago that it was before I discovered Blender, so I also had nothing really in mind for the foreground.

Sometime in March, I think, I came upon a script for creating spiral galaxies in GIMP, and while the galaxies it creates are a long way from the final image, it gave me enough of a starting point to get back to work on it. That was still the hardest part of the image to get looking satisfactory for me. Galaxies are a lot more complicated than I'm used to painting. I think it looks pretty good….at least the arms do (the core, I'm still not sure how to improve…). Realistically the galaxy wouldn't be anywhere near as bright as I've drawn it, and you wouldn't really be able to see the colors like that. Ah, well, I'm not trying for complete verisimilitude…(The galaxy appears to be a spiral, somewhat similar to our own galaxy.)

The lava lake idea came about because I found this link to the USGS webcam overlooking the active Puʻu ʻŌʻō crater on Kīlauea a while back and like to watch how the lava inside changes day by day. I've never seen it do anything dramatic like the picture, though. In the picture it was originally going to be a lake of water, but a) that planet looks way too cold for water to exist as a liquid on it. And b) it was just nowhere near as interesting as a lava lake. Getting it to glow and look half-way decent took me quite a while, and many, many, attempts.

Platonic Polyhedrals

A few weeks ago I found out that Blender comes with a script for generating various types of regular solids and decided to try it out when I got the chance. Being (a) a mathematician, and (b) a geek (not necessarily in that order), I opted to check out the Platonic solids by making a set of gaming dice (otherwise known as polyhedrals).

Here you can see all five of the Platonic solids. From left to right we have a tetrahedron, a cube or hexahedron, an octahedron, a dodecahedron, and an icosahedron. There are many ways you can define Platonic solids, but one of the simplest is that they are the solids you can make with every face being the same shape, having no gaps between them. It is easy to show that no more than five can exist in three dimensions; in four dimensions, there are six, and in all dimensions higher than four there are only three. Such dice have been around for a long time; icosahedral dice have been found from as far back as the second century AD (maybe the Romans were into role-playing games?).

The Platonic solids are the only regular solids possible, but there are other sets of solids that conform to certain rules, such as the Archimedean and Catalan solids. They're a bit more complicated so I won't get into them here, but I like this part of geometry because it's more visual than most of math.

The dice shown above are traditional in that the sum of opposite sides is equal to one higher than the number of faces (except for the tetrahedron, which doesn't have opposite faces). For example, on the cube, 1 and 6 are on opposite sides, to give a total of 7. Likewise, 2 and 5 and 3 and 4 are on opposite sides. In the same way the sum of opposite sides of the octahedron is 9, of the dodecahedron 13, and of the icosahedron, 21. I don't know what other conventions apply to the placement of numbers on dice besides that, so the exact layout is entirely my own.

(Warning: the section below is slightly technical, and more of a personal review of what I learned from this project. If you are not into computer graphic design, feel free to skip this portion. If you are, you may find it interesting and possibly learn something.)

I also used this project as an excellent opportunity to practice UV-unwrapping. UV-unwrapping is the process by which you take a 2-dimensional image and map it onto an object so that it affects various parts of the appearance of the object. The name comes from the fact that two coordinates are needed for the map, but since x, y, and z are already taken, u and v are used (they're commonly used in mathematics for such purposes). The UV-unwrapping in this case was fairly easy because I could mentally follow the cuts in the mesh necessary to lay them out flat with no distortion (even though I made it harder for myself by beveling the edges, which created more surfaces). Each die has two image maps on it: one for the color, a base dark-green color with the numbers in white, and one that causes the numbers to be depressed (the swirly light-green color comes from a distorted noise texture added on top).

 The depressed number are an example of what is known as bump mapping. If you were to examine the actual meshes used to create this image, you'd see every face as smooth as every other face. What the bump map does, is tell the renderer to treat certain areas as if they were elevated or depressed for the purposes of shading. This is an incredibly powerful tool for rendering high levels of detail that would be prohibitively hard to do by actually changing the mesh; combined with UV-mapping you can really bring a blank mesh to life.

Anyway, I've analyzed this enough, and I should probably get to bed now. Tomorrow  later this morning I'm taking my housemate Jonathan up to Mauna Kea on a summit tour for the first time. A hui hou!

Space Art

This week one of my space art pictures I created was published in the school newspaper. It was done back in 2008, when I first realized that due to open-source computer graphics programs, making high-quality graphical art was no longer just for dedicated professionals. The realization that I could, on my own, with a bit of work invested, create pictures (of astronomical subjects, primarily) that could hold their own in an art contest was quite intoxicating.

Fictitious Solar System 3: moon around gas giant.

At that point my work was entirely 2-D. I had had GIMP installed on my computer for a while but one day while browsing its website I came across a plugin that would automatically create simple planets. I was hooked. I figured out how to draw stars (and yes, every one of those stars is hand drawn) and nebulae and within a couple of weeks had turned out several (O.K., 3) pictures of fictional solar systems. That's basically what sparked the interest I have in computer graphics that I have to this day. It was only a few months later that I found out about Blender and graduated to 3-D modeling. I made one rather forgettable picture of an astronomical nature with it before getting busy with school and work, and running into snags with the next project I had in mind, which remains uncompleted to this day. Maybe I should brush it off and have another try sometime...

Edit: And here's that self-same project, finished a few months later!

Fiat lux (Let there be light!)

Last night I finally completed my multi-month long quest to get LuxRender to play nicely with Blender. (Blender, as I've mentioned before, is a 3-D modeling program, and LuxRender is "a physically based and unbiased rendering engine". It has a whole lot of nifty features, the most interesting for me being that it realistically models individual wavelengths of light, leading to realistic effects such as dispersion, or the ability to create a blackbody emission spectrum.)

I actually started looking at getting LuxRender to work back around Sping Break, but wasn't able to get it working. Blender is going through a serious reworking right now (going from version 2.4x to the new, radically redesigned and updated 2.5x series), and there didn't exist a version of LuxRender to go with the new Blender version. It seems to have appeared only recently, as I took another look at it after the semester ended and didn't find anything then.

Anyway, it took a complete wipe and clean install of the very latest Blender version (2.5 Beta), and of LuxRender (not quite sure what version it is, as it's still experimental as far as I can tell) to get Blender to acknowledge that LuxRender existed on my computer (quite the tedious manual installation. That should get better, in time).

Below is the first picture I've managed to get from it that looks halfway decent. A copper cube resting on some kind of...dirt...or maybe sand. I'm not sure which.

Nothing special, but I like the pretty copper color.

As both the Blender and LuxRender versions I'm using are still beta or experimental versions, it tends to crash every so often, so is not immediately useful right now. That will improve over time, though, and has inspired the artistic, creative part of my brain with some ideas for future projects.

At the moment, though, I have to study hard in order to try to test out of my basic Mechanics class on Monday and Wednesday. I have three hours allotted for each day, each of which will cover one half of the course (!). A hui hou!

To-do

Things to do over spring break:

• Install and start learning IRAF.
• Check out and start learning \(\LaTeX\). In progress.
• Brush up my (minimal) Python coding skills.
• Check out Maxima. Did, found it to be part of Sage.
• Download and begin learning Sage. In progress.
• Check out GeoGebra.
• Start learning Synfig. Started...
• Figure out new Blender interface.
• Figure out how to integrate LuxRender with Blender (possibly).
• Learn how to use Avidemux (possibly).
• Explore JavaScript a bit more (possibly). 
• Do Partial Differential Equations homework. Started. Got stuck.
• Do Electromagnetism homework.
• Finish up Gravitation and Cosmology homework. 
• Brush up on Calculus homework (possibly). 
• Finish Complex Analysis homework. Done!
• Fill out census form. Done!
• Do taxes. Done!
• Clean my room. Done!
• Figure out class schedule for next semester. 
• Update this blog a little more frequently. Pretty well.
• Finish moving all my music onto my new computer. Mostly done. 
• Check out possible scholarship/internship opportunities.
• Get up to the Vis at least once. Did!
• Go to the beach (if it warms up and stops raining).
• Do various other, as-yet-undetermined, things. 
• Enjoy life in Hawai`i, even if it is rainy and cold currently.

Oh, and sleep and eat should probably be on there somewhere as well.

More busyness.

I realize I haven't written anything for a couple days, mostly due to the sheer number of homework assignments that seem to have conspired to be due around now. Right now, I have Modern Physics homework due tomorrow (thankfully completed tonight), Partial Differential Equations homework due on Friday and a 2-page writing assignment in the same class due on Monday (it's a writing intensive course). I also have an Electromagnetism test due on Friday, as well as a currently unknown(!) number of Complex Analysis problems ("at least ten") due the same day. So that is why I haven't written anything recently.

However, before wrapping up tonight, I'd like to mention the results of a test I ran to compare the processor speed of my old and new computer. I found a rather large and detailed scene I'd made in Blender, a 3-D modeling program, and rendered it on both computers. The results? Old computer: ~2:05. New computer: 0:09.75, or a whopping 12.82 times faster. It's interesting that both processors are rated at the same speed, 1.6GHz, the main difference is that my new computer has 4 cores running at that speed compared to one in my old (there are a few other small differences as well). Now, the speed increase is probably not due entirely to the processor, since this computer also has 16 times more RAM and a 1GB video card (my old computer doesn't even have a discrete video card), but the processor is certainly a large part of it. So next time you're considering a new computer, take it from me: quad-core is the way to go. Dual-core, at least.


(I hope to eventually expand this into a sort of series of reviews doing point-by-point comparisons of my new and old computers in various categories. We'll see how far I get.)