Deepsky Without a Telescope

on Sunday, 26 October 2014 13:04.

Deepsky astrophotography requires many thousands of dollars of special equipment, including large telescopes mounted on fancy equatorial tracking platforms and specialized camera sensors. Or, at least, that's what I thought from my 40 years of experience as an amateur astronomer and my utter failure at astrophotography. And then, last night, I had a revelation: My stock Canon DSLR shoots pretty nice widefield starscapes and has a ton of resolution, so what if I was careful about technique and zoomed in on the photos? A few hours later, I had bagged dozens of Messier objects, galaxies, nebula, double stars, Uranus and an old Russian rocket. Not bad for living ten miles south of the glare of Disney in swampy, humid Florida.

NOTE: All detail photos here are lightly processed (e.g., noise reduction) and are cropped 1:1 in terms of pixels, so, basically, what you see is exactly what I shot. As a bonus, this means that all cropped images accurately show the relative size of these objects in the night sky!

I started the evening working with the lovely crescent moon and the rapidly descending Saturn, while also grabbing Mars in the waning light. Easy targets and nothing special here.

Mars was embedded in the heart of our galaxy, so I waited as long as I could for the darkness to build before the entire assembly was swimming in the thick atmosphere near the horizon. At this point, I knew I was going to have some success with this project, imaging a dozen Messier objects and a naked-eye invisible satellite (that I noticed later when editing) that I ID'd at Heavens-Above as an Okean 1-7 rocket body, launched in 1994 (click image to embiggen). 

I got a nice shot of the Andromeda galaxy (M31) and the barely visible fuzzy blob of M33, both of which are millions of light years away. From my years of dragging the ole telescope out, I can tell you that yes, this is what amateur astronomers see: lots of fuzzy blobs. These pictures are actually a lot more colorful that what the eye observes through a telescope (see the star trails below). Hunting objects down on your computer after the shoot is fun, but it still lacks the visceral thrill of using a red flashlight, paper star charts and scanning for an object you've never seen before at 3am in the dead of winter in Wisconsin. Hmmm. Maybe I'm getting old, but that doesn't sound as fun as I nostalgically remember it.

Jupiter was up, so I got a cute shot of the King and his entourage of moons, and then I went hunting for Uranus and Neptune. As I mentioned, the seeing from my location is horrible, and I couldn't even trace Aquarius or Pisces with the naked eye, but I used Google Skymaps and took my best shot. It took me nearly an hour to verify Uranus in one of my images back at my workstation and I never found Neptune, but it's probably there in one of the snaps!

I shot a bunch of open clusters, including the Pleiades. Two of my favorites of the night were the Ring Nebula, M57, and the double-double star in Lyra (although I could only resolve two pairs). Of course, I had to work with the Orion nebula for a bit (M42 - image at beginning of the article and below). Even the widefield was awesome (click to embiggen - the Horsehead is in the shot too if you know where to look!). Gorgeous.

 

Equipment: While I didn't use a telescope or tracking for any of these shots, I do have a nice camera (Canon 6D - $1,700) and the absolutely gorgeous glass light-bucket Canon 85mm f/1.2 ($2,100). My original plan was to do this project with my old T2i and a hundred dollar 50mm f/1.8, total value less that $500, but, frankly, while you can get most of the images here with that setup (especially the widefields - do go shoot that, please!), the difference in quality was stark.

My only technique was using a very fast lens, a low-noise camera sensor and limiting the exposure to a mere one second. Without special tracking equipment, even a six second exposure ends up with dramatic (and kinda cool) star trails.

With a minimum of $5,000 of dedicated astrophotography equipment, extremely careful tracking and extensive post processing combining dozens of long exposures, I'm sure I could get better shots...given a year of practice, at least. All told, I nabbed a couple of dozen Messier objects and had an enjoyable evening out under the stars. There are worse ways to spend an evening.

 
 
 

4K120 & the End of Video (NAB 2014)

on Thursday, 10 April 2014 23:59.

I just got back from the NAB Show 2014 in Las Vegas and here’s my personal takeaway from the event.[1] Note that this gathering is covered to death by industry pros and amateurs alike and so, yea, the Sony A7s prototype was cool, the Panasonic GH4 was neat, I want a Black Magic Ursa and maybe an AJA CION too...blah, blah, blah, well covered… go read the press releases. Instead, the best part of my NAB was the Technology Summit on Cinema on Saturday and Sunday, produced in part by the good folks at SMPTE (which I heard pronounced by the president Wendy Aylsworth and other board members as both "Ess Em Pee Tee Eee" and as "simpty" - so no video nerd bar bets won here).

Post is the New Production

One of my big ah-ha! moments was realizing that Post is the new Production. Panelists and speakers repeatedly illuminated how the dividing line between production and post is all but invisible today, with many traditional post tools being used on set and on location right next to the cameras and lights. Of course, there’s also a ton of content originating in post in the form of CG characters, sets, backgrounds and visual effects, but the intrusion of post work into the traditional production process is pervasive and expanding.

Internet > Television > Cinema

For more than a hundred years, the gold standard in motion picture distribution was the cinema or movie theater. Made-for-TV productions were always second class and Internet video was a distant, distant third, consisting mostly of poorly lit, shaky handheld, portrait home videos of cute cats and painful nutshots. Cinema was better than television, which was better than Internet video. The tables have turned. Although Hollywood cinematographers are still shooting with the best cameras and glass at 4K resolutions (or better), movies are still being finished at 2K to accommodate movie theaters with expensive and (relatively) recently purchased 2K projectors.

Today, however, original Internet video can be shot at 4K and displayed that way on YouTube, and all Netflix and HBO original programming (for example) is being shot and finished at 4K. Internet video is technically better than broadcast television, which is better than the cinema. Let that sink in for a minute.

4K and 5K Effects
While distribution to 2K projectors in theaters is one barrier to progress, Michael Cioni, CEO of Light Iron, noted that most visual effects work today is still 2K. I was rather surprised to hear that, especially since my friends over at Rampant Design Tools have been churning out drag-n-drop 4K and 5K VFx for a while now. I guess someone should tell Hollywood. Please, go buy their stuff so they can afford to pay me for my endorsements.

With consumer 4K displays being widely available (if not affordable), it’s looking like the living room can easily be a better viewing experience than the movie theater.[2] Combine that with 4K cameras for production, 4K VFx, 4K post and the Internet for 4K distribution and it looks like we’re right in the middle of the most radical change in motion picture technology since the advent of digital video a couple of decades ago, a change that is even bigger than the transition from SD to HD. The question is: Will it stick? Doesn’t this just mean 8K is around the corner and then 12K the year after that?

Better Pixels and the Human Vision System

I’d argue that 4K is an important advance that will be with us for some time, not because the technology won’t get better - it will - but because 4K is enough for our human visual system.

Dr. Jenny Read of Newcastle University presented an epic, whirlwind, brain-melting summary of research that showed that 4K images surpass the resolving power of our eyes at recommended viewing distances from the screen.

Of course, the television industry will someday convince us we need more, no matter what, but still: 4K can be indistinguishable from reality. Science!

Even so, Phil Squyres (Senior Vice President of Technical Operations for Sony Pictures Television) pointed out that, even today, content shot and finished at 4K looks better on 2K displays than 4K content finished at 2K (which is how movies are delivered right now). This suggests that content shot and finished at 8K will also look better than a native 4K workflow, even on pathetic 4K displays...but that’s not our problem sitting in our living rooms, right?

There’s more to motion pictures than just pixel count, however. We don’t merely want more pixels, we also want Better Pixels, for example:

(1) Color gamut is important and could be expanded some, but we’re pretty close to theoretical limits here too, with 12- and 24-bit color giving us the post-processing headroom to exceed our visual system on delivery.

(2) We could definitely perceive a wider dynamic range than is currently available in production, with perhaps 20 stops being ideal. That’s about a 105 contrast ratio. Today, we’re sitting at 14-16 stops on the most sensitive cameras and films. No doubt next-gen sensors will soon eclipse film once and for all and our displays are close, if not already hitting, million-to-one contrast ratios, so we’re close here, too. This is actually an extremely complex topic, since the human eye readily adapts from light-to-dark-and-back and contrast sensitivity is not as easily measured as resolving power. The science here as presented by Dr. Hans Hoffmann of the European Broadcasting Union and Dr. Touradj Ebrahimi of the Swiss Federal Institute of Technology relied on smaller samples of self-reported preferences of subjects in their respective labs, but their data was in agreement: million-to-one contrast ratios on 1000-nit brightness displays(!) in mostly-dark rooms (they argued about the definition of that too) was a comfortable sweet spot.

(3) Frame rate, however, is a very well-defined problem. In fact, not only is 24fps an arbitrary and artificial frame rate determined by the bare minimum technical needs of an almost 100-year old audio standard (yes: audio),[3] it’s also alarmingly close to the most humanly annoying rate anyone could possibly have picked, and which necessitated 48- and 72Hz shutters on modern projectors to prevent audience seizures and riots. (They used to call movies "flicks" because of the horrible flicker.) Dr. Read revealed that 60fps is about what we need for low-action acquisition and display, so 720p60 is a nice minimum, but for higher speed action, 120fps would really fill in all of the gaps and blurs.

And so we can conclude that 4K resolution, 12-bit color and 22 stops of dynamic range (and a million-to-one contrast ratio) at 120 frames per second is our ideal, perfect video format that would be indistinguishable from reality. You can whinge that it doesn't look like film, because reality doesn't, but the future is that we'll have oil paints and water colors and digital verisimilitude and it's all good. 4K may be here today, but start saving your pennies, boys and girls, because 4K120 laser illuminated projectors ain’t gonna be cheap.[4]

Footnotes:

  1. I also learned that drinking a bunch of Jameson neat with Guinness chasers before going on the New York, NY roller coaster at 1am is not a great idea if you want to attend seminars on the psychophysics of human vision at 9am the next day.
  2. So why go to the theater at all? Michael Karagosian, former technical consultant for the National Association of Theater Owners, related an anecdote of a survey they conducted in the 1980s to find out what people wanted most in a movie theater. Do they want better resolution? Better dynamic range? Nope. Cup holders. The survey revealed they wanted cup holders. Go figure. 
  3. Friends don't let friends shoot 24p.
  4. I hereby claim "4K120" as my invention. Sony can pay me later. 

Get Out and Shoot!

on Friday, 04 April 2014 15:42.

It seems like some people talk more about video production than they actually produce. Unfortunately, I might be talking about myself, but there are a lot of folks who are creating amazing video on a regular basis, simply for the sake of creating video - the best reason of all! One of my current favorites is producer and blogger (and friend) Matt Janowsky. He's working through a 52 videos in 52 weeks project right now. What I like about it is not that he's simply rolling tape (electrifying memory locations?), which would be easy, but he's working on specific skills an techniques he'd like to be better at AND he's sharing it with us all, including a breakdown of not just how, but why and what could have been better.

Another really fun way to get shooting without all the hassle of figuring out WHAT you're going to shoot (which is actually very challenging) is to participate in the Vimeo Weekend Challenge. This weekend's challenge is using Match Cuts. Here's an example of that technique.