Telescope Reviews: M101, obscene # of subs w/300mm 2.8 L IS

Here I was thinking I was all cool and stuff taking pics of M101, but several others beat me to the punch, mostly because of the absurdly obscene number of subs I collected on this target and the (relatively) incredible amount of processing time it took.

Happy to use my new 300mm f/2.8 L IS lens I gathered subs as the full moon was waning in the 1-2 hour period after astro-dark and before the moon ruined the subs with all the wonderful clear weather we had last week. Figures now that the moon is gone, we have wall-to-wall rain forecast. This is my first official non-test-shot astrophoto with the lens!

This was all automagically run unguided-- who needs guiding with 45 sec subs-- and EOS utility and the Canon 40D. Focusing was a breeze-- autofocus on a bright star, then use liveview to micro-adjust focus without even touching the lens. 90% of the time it was spot on anyway. That's the hidden benefit of live view and EOS utility- you can nudge the focus a teensy bit without having to make feather-touch adjustments by hand.

The 300/2.8 is surprisingly sensitive to temperature changes-- I literally threw out 100+ subs due to focus shift in less than an hour. I guess that's the price you pay to image at f/2.8-- I rarely had focus shift problems with the 400/5.6L, so the 2nd and 3rd nights I got smart and checked focus a couple of times until the air temp stabilized enough around midnight. Now I appreciate first-hand why Bert has the servo setup with the stepper motor.

Once I converted all my subs in Iris I was surprised to find the total number-- 470!! That's 3h 52m 30s of data. At f/2.8. And ISO1600. If shot through the SCT at f/10, that would equate to roughly 32 hours of integration time hahaha... I don't have that much clear weather in a month. Iris didn't like that many subs-- when I tried the sigma clipping stack, it took 24 hours to process it. I used up 100GB of disk space in temp files. And the result ended up being posturized! Grr! A simple "add" stack with the "normalize if overflow" checkbox checked worked in less than an hour and produced a much smoother and useable result. I guess with 470 subs, the noise and meteors and cosmic rays get buried once the stack is normalized at the end, because I didn't see any.

FWIW, I could barely make out M101 in my subs. The camera histogram was 4/5ths the way to the right in my Mag 5-ish skies @ 45sec @ ISO1600 (which works out to 3min @ f/5.6, which is my normal exposure length with the 400/5.6L). I'm very happy with this result.

Details:
M101, Taken 22-24 Apr 2008
Stack of 470 x 45sec subs, unguided, 3h 52m 30s total @ ISO1600
Canon 40D on a Canon EF 300mm f/2.8 L IS @ f/2.8 piggyback mounted on an LX200
Canon EOS utility for camera control and Live View for tweaking the lens autofocus
Calibrated, stacked and processed in Iris
Final processing in Photoshop CS2, using Noel's Astronomy Tools, GradientXTerminator and Fredmiranda.com's intellisharpen II plugin

This is the full-frame version resized for web. The 100% full frame image can be seen here.

--------------------
LX200 8" classic, f/10, Meade eq. wedge, .63x FF/FR
Canon 40D and 5D, unmodified
Canon EF 300/2.8L IS, 400/5.6L, 135/2L, 50/1.8, 85/1.8, 35/2, 24-70/2.8L, and Peleng 8mm fisheye
Orion Apex 102mm (4") Mak-Cass
Pimped out with accessories and bling
My DSLR Astrophotography Webpage and photo bucket with full equipment list