Saturday, 12 November 2011

Microscopic worlds - the movie, and how it was created

Microscopic Worlds - Life that we don't see, watch the video in HD full screen, follow the link to vimeo to do it.

   We are surrounded with various living creatures, but how often do we notice the tiniest ones and how small can they be? Such common but inconspicuous organisms like water fleas, seed shrimps, and hydras are less than a centimeter (0.4 inches) in size but they are very important components of the freshwater ecosystems. The vast majority of organisms are even smaller and they are completely invisible with naked eyes. Using sophisticated optic systems I am bringing even the smallest animals before your eyes. They can be magnificent and scary, fast-moving or hiding; most of them look nothing like animals we see every day… 

If you are looking for a version with subtitles, follow the link to Youtube, just notice that YouTube has slightly worse quality. However, YouTube might be a better option if you would like to embed the video.

   I am bringing my sincere apologies for the narration audio quality, but there’s nothing I can do to improve it at this point. It also is a bit too fast and some scenes change before you can focus on them; considering that it's my first video ever, please don't be too harsh when judging it.

What this post is about

   Overall I am not surprised that no one tried to produce a movie recorded through a microscope--it’s an insanely difficult task due to technical limitations.

   In this post I’ll describe what was behind the scenes, how I made this movie possible and what difficulties I encountered.

   If you are not familiar with my activity, please visit my site to know what kind of photography I'll be talking about:

Production time – nearly 4 months
Movie budget – 8 $

Technical limitations

Basically, two problems are the most significant when it comes to making pictures through a microscope: depth of field and transparency of microscopic organisms.

    Technical limitations make the depth of field of the image uselessly small under high magnifications. Essentially, only a tiny razor-thin plane is in focus and the rest is blurred. Those who are familiar with macro photography probably know better what I am talking about. In case of photography the solution is the same as in macro photography – stacking. Most of my micro photos are stacks (see one of the most spectacular examples). Z stacking requires moving the lens away from the object (or towards) and making pictures on every step and then software combines those pictures into one. With videography it does not work this way, well... theoretically it’s possible but not with my budget and gear.

   The second limitation is less obvious – most of microscopic organisms are very transparent, and some of them would be easy to miss when looking through regular light microscopes. For this situation microscopists invented contrasting techniques that use various optical effects to achieve high contrast between different transparent elements. A very ancient but cheap method that’s called phase contrast is still popular, but I never used it for photography or videography as it basically is very unattractive. It creates halos around objects and lots of optical artifacts and it is completely useless at high magnifications; yet, it’s better than no contrast at all. DIC is a modern approach and it uses polarized light to enhance contrast, it’s a very difficult technique to explain and you can read more in Wikipedia.

   Essentially, these two problems of microscopy are combined when it comes to videography and, as a result, focusing on inner parts of the objects allows seeing inside the objects, which is cool for some purposes, but making the picture really confusing for those who are not familiar with high magnifications.

   On top of that, working alone on video recording through the microscope does not work sometimes. I wish I had at least 6 hands to do the job--too many microscope parts that had to be continuously adjusted. If you could see me during the filming, most likely you would not be able to follow my hand movements as I had to act very quickly. By the end of every filming session I was very exhausted and my hands were in pain.

   Yet, I realized that with some patience it would be possible to get some nice footage and combine it into a video. 

The gear and the idea

   Hard to say how I came up with the idea of making it, but at the beginning I thought it would be impossible as I had:

1) An expensive and good microscope, but not designed for specimens like mine, it was bought for the team of geologists and the only thing that it had to observe living specimens was a 100x lens with DIC setup, which is an overkill for a movie. It did not even have 40x lens – the one with which I would have made nearly everything as it gives the optimal magnification to observe the majority of microorganisms.

2) No biodiversity in European part of Russia. Literally no biodiversity, even microscopic one. Only during spring it’s possible to spot something interesting, but I did not have a camera in spring. I could travel for hundreds of kilometers to see the same two boring species of ciliates and flagellates in a distant pond. To illustrate the problem, approximately one third of the footage shows Antarctic biodiversity.

3)Camera which is a good for photography (Sony NEX-5) but sucking as a camcorder, it does not have any manual adjustment and recording through the microscope is a painful and frustrating process, especially when it comes to dark field.

4)No connector for my camera. I used an old one designed for a small point-and-shoot camera and and after some duct tape fixes I managed to attach the camera to the scope. The major problem was actually the camera’s field of view, which was larger than the image from the microscope (so, I had a dark circle surrounding the picture), it required cropping off an area of about 30% that is a huge quality reducing step making FullHD impossible.

5)No budget, which at the beginning was not a problem at all, but turned out to be the most slowing down thing.

   With such a “brilliant” start I was definitely not serious about a movie, nevertheless, I was recording everything I could see and at some point while looking through the collection of my videos I said "that’s it! I can do it"!
Lacking the most essential lenses and contrasting setups I had no choice but to explore various options of lighting of the microscope, play around with all hidden settings, and I started with relatively large specimens like larvae and water fleas. I found out that I could cheat a little bit with the contrasting using the DIC prism designed for 100x lens to produce spectacular effects on various magnifications and a lot of other things. What I could not get at all, was a 40x lens.

   Just to give an idea, DIC is used in the footage of the inner structure of the rotifier. A DIC prism is the core element of the technique and for a 100x lens it costs couple of thousands of dollars.
Eventually, while advancing with photography, I also figured that larger organisms are so magnificent; especially hydras, and I found a lot of interesting objects. It was photography that made me think about making videos as it was clear that a still photo, even a stack, could not deliver everything that I wanted.

  No matter how spectacular the organisms were, photography just failed at everything. I could take pictures of only those things that could stop moving for a second or more and that happens really seldom. When they die and become still, they turn into ugly blobs very fast. That’s a fundamental property of life. To maintain shape a living organisms need to metabolite, or, in other words, be alive, as the degradation process of the organic matter is very fast.

  On the negative side, the wrong way of using the DIC prism resulted in the optical artifacts you can see in the background if you point your attention at the pattern.

Getting samples

   Filming was not my main occupation; I could only do that in my spare time. At first I thought that getting something to look at would be the most enjoyable process. I imagined going to the nearest pond, getting some water, and then observing it through a microscope and filming.

Shore fly larva
A shore fly larva, some people think that it's hard to find something as ugly as this creature

   But I was badly wrong. Very soon I realized that there’s nothing to look at. A water sample in the nearest pond had the same couple of species as a water sample very far away in a different biotope. At first I thought I was very bad at finding microscopic stuff and kept trying just to just realize that biodiversity is a word that does not exist at the place where I live. Which I knew, but never expected to be true for microscopic world. May be as someone who spent most of the life in the tropics I'm too picky as for me a normal biodiversity is when I can't find two similar bugs within a day, no matter how hard I try. Deep inside the continents it is normal for biodiversity to be low.The middle part of Russia is one of the most isolated parts of the world from the oceans and it makes sense when you think of ways how animals would get there during long times (I'm talking about millions of years). Of course, low temperatures at winter make it worse but every spring the life makes out for the low biodiversity and that was the time at which I should have done the recording.

   From certain point of view it was good for the movie: it is not overwhelmed with millions of different creatures, it was possible to think of a story and dedicate more time for the most spectacular objects. It also allowed me to explore more options of the microscope.

   Later, one of my research tasks involved looking into Antarctic samples for protists and I could get plenty of Antarctic samples from different regions of the continent as the lab in which I worked basically studies life in various extreme environments. And of course I decided to use this opportunity for my movie, though I did not expect a lot from Antarctica. I got giant rotifers, which are normally ten times smaller in average, and I got water bears that are so abundant in the continent, probably because they do not have a decent competition. Overall, including ciliates, nearly a third of my footage shows Antarctic biodiversity.

   Another interesting objects that I could record were bacteria from volcanic permafrost. Quite an exotic sample, but I was lazy to get a dense culture of bacteria on my own.

   The final part was macro footage. I didn't have a macro lens so I had to invert the 16mm pancake  using loads of duct tape and a flashlight for 8 bucks (remember the movie budget? That’s where it went!) With many problems and swearing every time the lens fell off, I could get some minutes of macro video which I used for the intro.
  I wish I had more time to construct a thin aquarium and film some "big" animals to give a better idea of sizes.

Getting everything together

   Gradually it became too cold for microscopic organisms and I gave up on filming, though the material I had for protists was exceptionally poor. Besides, I could not do the movie forever and no matter how little material I had, I needed to finish it and I definitely have other things to do in life.

   So, I started with thinking how to organize a video. After some weeks of work and reviewing all my video collection gazillion times I finally came up with a script. Learning how to use a video editor and organizing the movie took me around four days of work and then I got seriously stuck as I did not have sufficient spare time for the movie and, most importantly, I did not have anything serious to record narration.

   The video editing itself is an easy task with modern software. Though I regret choosing Cyberlink Powerdirector. Sony Vegas is much superior, for example. However, when I realized the mistake, I did not want to change it as it would result in another month of wasted time considering the work with audio and minor adjustments.

   It was so weird to learn that movie editors don’t have drawing features or something that could fix sensor dust on the footage. It’s such an elementary thing to do; but even Photoshop in video editing mode fails to offer something but frame by frame editing for which I don’t have time. All other options with masking the sensor dust just do not work in Powerdirector. It seems that some genius spend half of his life designing it to cut off hundreds of possible ways of removing sensor dust from a video. Overall, I spent several weeks on fruitless attempts to do something about that. Update: Photoshop CS6 allowed to automate sensor dust removal.

   At first, I was sure  about making a sort of music clip about wildlife with some text explanations, but consequently I came to the idea that video through a microscope is a hard thing for the majority of people to understand without at least some explanations. I did not have enough material or time to make proper animations explaining essential stuff like why the depth of field is so low and what’s phylogeny. After all, I wanted people to be curious about such things, so the only thing I could do was to make a video with minimal narration.
The most obvious reason why I did not want to narrate was because of my accent. As I am not a native speaker, it takes unbelievable amount of time to make the pronunciation more or less clear and to fix all major errors even within a couple of sentences. I wish I also learned how to be more expressive because the whole narration sounds awful.

   At the very start I thought I would do that with cheap headphones but it certainly was a waste of time. As I did not have any friends who could provide me with a decent microphone or any voice recording tools for a sufficient period of time, it took me a month of work, as bizarre as it sounds, to record 3 minutes of narration and to assemble it together. I ended up with countless files that did not match between each other in terms of frequency and quality. Most of them harmed with noise reduction process; and I had to spend the last five days matching different recordings and in some cases editing frequencies. I believe some of you experienced with sound editing would laugh reading that, but try it without sound recording tools and without money.  (Not even a desktop computer, though I would not complain about my laptop; laptops and sound recording just don't do well together). The only thing that makes me feel better is the fact that I've never done a video before or had any experience with audio editing.

   To be honest, I thought that getting music and applying it to the movie would be the hardest part, but that was one of the easiest and enjoyable parts. The only restriction was the selection of music as I could use only something available under creative commons license.

What I think about this experience

   First of all, it’s a process that can be done only by a team of people. Making such work alone is a stupid idea. Months of work can result only in minutes of high quality video. Update: I learnt how to be faster and more efficient. The next video was much better and took me about 2 weeks to accomplish.

   Second, I think that a video about microscopic life is a great idea and someday BBC or other guys would do a serious project. However, as far as I know, no one ever attempted to make a serious video through a microscope before me, may be because there are few people who can organize the process. Discovery made a clip about microscopic life but it was all CGI, spectacular one I must admit; but still, even they went only as far as making scanning electron microscopy of dead bugs and making 3d models out of scans. BBC sometimes uses footage made through microscopes but I am surprised that with their budget they get such an awful quality and basically even they tend to substitute it with cgi.

   Finally, I do not regret about doing all that, it was a great experience and though I leave it in a 95% finished state (missing good quality narration and having some imperfections), I am happy that I managed it, though I believe I could have done it much better if I had time. Especially with audio.


  I would like to thank those friends who were helping me in collecting samples and looking for cool things, visitors of forums for helpful remarks, the author of the Skeptic Wonder blog for providing me with the phylogenetic tree of eukaryotes, and of course, big eff you to my neighbor with his awful guitar skills and goat-like voice because of whom I had to delay the sound recording for another week.


  I made another video a year after this one. Check it out at my website. I seriously improved my techniques and still working on better image quality and smaller objects. In addition, I became a freelance videographer and license microscope footage. Most of my growing video collection is unpublished because producers generally prefer exclusive never-before-seen material. Contact me for details if you are interested.

No comments: