Tube stand

Alright, everyone, welcome to the Misskaten Labs. The coats are hanging to your right, the disinfection room is just ahead, and if you would kindly follow me, we should reach the dissecting room in no time. Please proceed in an orderly fashion, no pushing, no shoving, yes, you can take pictures, but wait until we— Oh no, it's not far, we are almost there, really, just around this corner, and... Wait— don't touch that— 

Sigh. Days with no Blender accident: 0.

But since we've come this far, we can as well talk about another Blender project of mine: a tube stand. 

Inspiration for this piece came from my brief infatuation with chemistry. At the ripe age of mumble mumble, I went back to uni to study that amazing subject, only to realize that the instruction provided was very different from what I saw in online courses, and I was forced to admit that I was too old for that... stuff. But at least it gave me an opportunity to try my hand at modelling lab equipment. The idea was to create many pieces separately and then put them into a marvelously grandiose scene that would stun chemists worldwide with its intricate detail and lifelike quality. Well, that didn't happen. As I already revealed to you, I have created *two* pieces of equipment, first of which you might still be trying to wrap your head around.

In the beginning, the project went surprisingly well. The modelling part was a breeze — after all, there are not many structurally challenging bits in this one. I also started using background images, which helped me work a lot faster. But as many an aspiring artist would tell you, there is a teeny-tiny problem with these images. You'll find a ton of them (that's good) and none of them will be angled correctly (not so good). What it forces you to do is use your imagination — and a measuring tape. Yes, this is the fabled part of the project where you are spending your afternoon measuring pixel distances on screen and translating them into real-life dimensions.

With the stand modelled eventually, I sank my teeth into the vials. Not literally, of course, even though I might have had as well, for all the grief they had given me. You'd think that the shape of a tube is a simple one. Long, narrow, curved on one side, with an opening on the other. No big deal, right? Even with a background image it was hard. Especially the opening bit. And don't get me started on the inside of the tube. What inside, you ask? Well, there has to be thickness to the glass, doesn't it, you cannot just do the outer layer and hope for the best. It doesn't work. (Believe me, I tried).

After creating the final shape of a tube, I duplicated it five times, and I placed all the tubes into the alloted slots in the stand. And then, intoxicated with happiness from having just finished the bloody vials, I looked at them and thought: let's add a bit of realism. People like that sort of thing. A bit of realism in my case meant carefully repositioning the tubes so that they didn't stand perfectly upright. (Because they wouldn't under normal circumstances.) Satisfied with a job well done I decided to model the liquids inside. Some of you brighter lot may already be sniggering about what necessarily happened next. It is not a good idea to duplicate anything you haven't finished modelling, and it is a genuinely terrible idea to rotate it beforehand. Lesson learned: never fiddle with your tubes until you are fully finished.

Delete, delete, delete, and — yes, you've guessed it, delete, and start again. This time I got it right. The very solid liquid was modelled inside the original tube, and it even had little irregularities at the top to simulate the uneven surface. Then came another round of duplication and repositioning, followed by a sigh of relief because everything was looking good. At this point it was only a matter of setting up the liquid material correctly. I had previously tried and dismissed the option of using real liquid simulation, so I needed to look at properties of regular materials and use one that fitted the best. Which in my case would be glass. 

Now, I won't go into the details of researching the correct index of refraction for liquids, but rest assured it was long and arduous. Nevertheless, I got something I was relatively happy with, and I used the material on each liquid with appropriate changes in color and brightness. Finally, I added provisional lights and rendered the scene. Not much use beating around a bush here, it was a mess. The color of each liquid was significantly different from the one I set up for it, and there were some weird shadows on outer edges of the liquids. And thus the nightmare began.

In the effort to fix this, I recalculated the normals on each vial. Not only it didn't help, it actually made the matters worse. Then I tried rescaling the liquids. Much worse, again, because now the light was bouncing in the very real space between the liquid and the vial, creating unwanted shadows in the process. And so, after a long struggle with myself, the only other option I was left with was to delete everything and start from scratch. Model, adjust, duplicate, reposition, you know the drill. This time I was very precise when modelling the liquids. I made sure there was no extra space and the normals were pointing in the right direction; all in all, I created much cleaner topology. And it still didn't work.

Slowly realizing that I wouldn't be able to get rid of the shadows, I turned my attention to the tube stand. With freshly gained expertise on wooden materials, I quickly designed one in the node editor. (Because of how good I am, and totally not because I copied it from the table lamp project.) However, the tube stand differs from a table in one crucial aspect: everything. For start, it is not flat. Which means a lot more time spent on UV mapping. And when you have the object unwrapped (if you've never done it, imagine surface of a cube laid out in 2D so that you can cut it out from paper and glue it together to form the cube), you push it around the texture, and it doesn't fit at all. So you cut it into more pieces and push around those, and then you somehow lose one of the pieces and you end up with a distorted texture. Fun.

Not entirely sure how, but I made it work. The tube stand looked decent, the vials still did their nasty business, but I trained myself to be able to almost ignore it, and I was ready to finalize the scene. Selecting a surface material was easy — not much groove going on when it comes to laboratory tables, I hear. Simple, white, with a bit of reflection. Something that doesn't detract from the colors of the stand and evokes the right kind of atmosphere at the same time. Then the lights. I was lucky enough to place the first lamp in a way that created those long shadows you can see in the final render. As a wholly unexpected bonus, I got rid of most of the vial shadows. Still not sure how that worked but I'm certainly not complaining. 

One of the downsides was that the tops of the vials became very dark — to the point where they didn't look like glass anymore. But every time I moved the lights, something else fell apart. Either the front shadows disappeared, or the liquids didn't look transparent, or the stand wasn't catching enough light to give it that polished shiny look. In the end, I moved full circle to the initial lighting setup and stayed there.

And that about wraps up the adventures I enjoyed while working on the tube stand. Next time I'm going to show you the other lab equipment — my most daring and complex project yet — after which I'll finally be able to transition to real time commentary. Which means I'll have to complete a project every week. Yay me.