4 ms·

I've made a fair amount of these kinds of animations in the past, and it's much harder than it would seem to portray the random nature of interactions. Another

I've made a fair amount of these kinds of animations in the past, and it's much harder than it would seem to portray the random nature of interactions. Another common criticism is that a lot of these visualizations don't show how completely packed full of molecules/cells each of these environments are. You're tasked with the almost-impossible goal of clearly conveying accurate information, visually, in realms where a) though there are a lot of structural data, the visual relationship of things gets VERY complex, b) not everything is actually known, and c) the time and visual scales aren't naturally perceptible to humans. It's definitely not impossible (see Drew Barry's work, and from the glimpses of this project I've seen it looks pretty great actually), just difficult, and there's not a huge amount of money in it unless you want to work for pharma. Personally, I've been moving more towards 3D interactive content vs. linear videos because you can show a lot more layers, and having users drive what they're focusing in on can be a lot more powerful than trying to design one animation for every learning style.

busyant 7y ago
Just a small correction for anyone looking for this person via Google: it's Drew Berry https://en.wikipedia.org/wiki/Drew_Berry https://en.wikipedia.org/wiki/Drew_Berry
TeMPOraL 7y ago
RE random and packed, the first video on the page (about bacterial flagella) manages to pull out both by exploiting the time domain. First, it shows the usual but misleading view focused on the flagellum, but near the end, it fills in all the other missing activity in the displayed cutout - giving a sense of how it fits into the whole.
- 1auralynn 7y ago
  Yeah! I've been seeing bits and pieces of this project over the last few weeks, and it really does look great.
mncharity 7y ago
An educational graphic where some aspects are carefully accurate, and others utterly bogus, used by students unable to tell which are which, is going to prolifically foster misconceptions... when used in isolation. So, why do that? Software limitations, or visualization clarity, or pedagogical focus, or budget, may require a misleadingly empty cell. But why then use it in isolation? Why not start by showing a realistically crowded and packed representation? Briefly, like 2 seconds. Static. Drawn with crayons. By a 5 year old. Sorry, it's a pet peeve of mine. A problem with these recurring... I'll call them "excuses", is that they depend on a silent addendum. That something has been prioritized over reducing misconceptions, over slapping on secondary misconception-antidote representations. Sometimes it's "that wouldn't be pretty" (almost a quote), or sometimes it's couldn't be bothered, or just don't care. Sigh. For conveying the random violence of nanoscale, I like simulation of virus icosahedral capsid assembly. The panels are tethered together, so they can't wander off. They just keep variously smashing into each other. Disassembling, misassembling, and finally, sometimes, succeeding. The panel size/speed, and collision and attachment rates, make realism watchable. Maybe with some skipping ahead. Apropos 'not naturally perceptible' scale... I've a hobby project around teaching scale down to atoms for early primary. It's funny or sad or something, that it's not that hard to give at least some children, a firmer grasp of the size of cells, than that of at least some first-tier medical-school graduate students. It's a very low bar. There's a lot of "We tried teaching <some topic> really really badly, and wow, surprisingly, that didn't work out. So we drew the obvious conclusion: students aren't developmentally ready to understand so complex thing". But as you say, XR is a source of hope. And personalized instruction. And when MIT did a VR cell biology app, pulling in and interviewing researchers for domain knowledge, they reported a reoccurring challenge was... getting the enthusiastic researchers to leave and end the interviews. So there's seems hope for gathering the needed expertise as well.
- 1auralynn 7y ago
  I mean, why try to do anything? There are tons of science educators working on low-tech tools like you describe, with success. My point is, we can all be armchair critics, but from my experience when you actually try to make one of these visualizations "right", you run into a never-ending torrent of complex decisions that need to be made, both on the scientific end of things and pedagogical, most of which end in some kind of compromise that others can pick apart. There's also a severe lack of funding to support this kind of work. Totally agree on XR and scientists - biologists LOVE visualization, particularly XR. That's definitely not the problem. In fact, one of the things I've been trying to do is to make XR tools for scientists to make their own visualizations using their datasets (http://10k.systems http://10k.systems). There just aren't enough headsets yet, but it's possible Quest will change this.
hagy 7y ago
For those curious, here are some animations by Drew Berry, https://www.wehi.edu.au/wehi-tv https://www.wehi.edu.au/wehi-tv. They're visually stunning, accurate, and highly informative.

busyant 7y ago

Just a small correction for anyone looking for this person via Google: it's Drew Berry https://en.wikipedia.org/wiki/Drew_Berry https://en.wikipedia.org/wiki/Drew_Berry

TeMPOraL 7y ago

RE random and packed, the first video on the page (about bacterial flagella) manages to pull out both by exploiting the time domain. First, it shows the usual but misleading view focused on the flagellum, but near the end, it fills in all the other missing activity in the displayed cutout - giving a sense of how it fits into the whole.

1auralynn 7y ago
Yeah! I've been seeing bits and pieces of this project over the last few weeks, and it really does look great.

mncharity 7y ago

An educational graphic where some aspects are carefully accurate, and others utterly bogus, used by students unable to tell which are which, is going to prolifically foster misconceptions... when used in isolation. So, why do that? Software limitations, or visualization clarity, or pedagogical focus, or budget, may require a misleadingly empty cell. But why then use it in isolation? Why not start by showing a realistically crowded and packed representation? Briefly, like 2 seconds. Static. Drawn with crayons. By a 5 year old. Sorry, it's a pet peeve of mine. A problem with these recurring... I'll call them "excuses", is that they depend on a silent addendum. That something has been prioritized over reducing misconceptions, over slapping on secondary misconception-antidote representations. Sometimes it's "that wouldn't be pretty" (almost a quote), or sometimes it's couldn't be bothered, or just don't care. Sigh. For conveying the random violence of nanoscale, I like simulation of virus icosahedral capsid assembly. The panels are tethered together, so they can't wander off. They just keep variously smashing into each other. Disassembling, misassembling, and finally, sometimes, succeeding. The panel size/speed, and collision and attachment rates, make realism watchable. Maybe with some skipping ahead. Apropos 'not naturally perceptible' scale... I've a hobby project around teaching scale down to atoms for early primary. It's funny or sad or something, that it's not that hard to give at least some children, a firmer grasp of the size of cells, than that of at least some first-tier medical-school graduate students. It's a very low bar. There's a lot of "We tried teaching <some topic> really really badly, and wow, surprisingly, that didn't work out. So we drew the obvious conclusion: students aren't developmentally ready to understand so complex thing". But as you say, XR is a source of hope. And personalized instruction. And when MIT did a VR cell biology app, pulling in and interviewing researchers for domain knowledge, they reported a reoccurring challenge was... getting the enthusiastic researchers to leave and end the interviews. So there's seems hope for gathering the needed expertise as well.

1auralynn 7y ago
I mean, why try to do anything? There are tons of science educators working on low-tech tools like you describe, with success. My point is, we can all be armchair critics, but from my experience when you actually try to make one of these visualizations "right", you run into a never-ending torrent of complex decisions that need to be made, both on the scientific end of things and pedagogical, most of which end in some kind of compromise that others can pick apart. There's also a severe lack of funding to support this kind of work. Totally agree on XR and scientists - biologists LOVE visualization, particularly XR. That's definitely not the problem. In fact, one of the things I've been trying to do is to make XR tools for scientists to make their own visualizations using their datasets (http://10k.systems http://10k.systems). There just aren't enough headsets yet, but it's possible Quest will change this.

hagy 7y ago

For those curious, here are some animations by Drew Berry, https://www.wehi.edu.au/wehi-tv https://www.wehi.edu.au/wehi-tv. They're visually stunning, accurate, and highly informative.