November 14. Last
night, I did some sky trials with the 200mm F2.0. Largely frustrating,
but, still and all, not bad results for what are comparative
snapshots. Fifty seconds? What kind of exposure is that for a deepsky
object?
In the early hours of deep twilight, the
western sky was mostly cloudy or at least hazy, but I tried one quick
frame of NGC 7000. Some unidentified tracking issues and clouds blowing
through meant I got only one marginal frame. Even allowing for these
problems, none of the images is all I would like. None is as clean as
I had hoped and expected. Most of the defect may be due to focus (micrometer,
live view, and magnifiers be damned; it's still hard to find the best
focus with this lens) or maybe there is cause for more subtle adjustment,
or maybe cause for a change in workflow. Based on the one decent frame,
here's the North America Nebula; 50 seconds at F2.0, ISO 800, Hap Griffin
modded Canon 20D, dodging clouds:
Next, I ran 90, 1-minute exposures of the
Pleiades. Focus was off. How in the — ? It was far enough off that
Maxim won't auto-align on stars (apparently 'cause it doesn't see any).
But not so far off that something couldn't be salvaged. Here are the
first five frames, manually aligned in Maxim and carefully stretched
in Photoshop:
<Sheepish grin> I hadn't appreciated the
degree
to which
those sharp spikes would improve the perceived sharpness of otherwise,
err, "marginal" images. That, and shrinking them for web display.
The original images of the Pleiades have
a FWHM point spread of 6.2 pixels (two or three would be better). Even at that level
of fuzz, we're getting on toward Mv 13.5 or so in 1 minute which suggests
that perfect focus might get us well into the 15's (I recall a pro survey
using similar lenses feeding efficient CCDs which was reaching 18 with
short exposures, so don't be too impressed).
With color images
of DSLR size, Maxim can take its sweet time, especially when manual
alignment is involved. I didn't think it worthwhile to spend all day
(or more) combining a large pile of poor data. Better stuff tonight,
weather willing. In the first batch, there's some potential for a nice
animation
as the Pleiades
passed through
the upper boughs of a pine tree, but that, too, is for another day.
Some notes:
1) The 200mm appears to suffer
some pretty awful coma and chromatic aberrations, at least when focus
is not spot on, which it seldom is;
2) maybe this
will be mitigated by a clip-in LPR filter (or at worst by a clip in
H-a filter) or at the
very worst you can probably align and even re-scale the channels seperately
and recombine;
3) it didn't offer to dew up, go figure,
but I'm making a better hood
for it anyway;
4) USE FLATS with this lens at F2.0 because it has a dreadful
hotspot,
even on an APS-size sensor;
5) get a focusing tool working on a notebook so you
can actually see to focus with the 20D, because a) I'm not sure that focussing
with the 50D and swapping bodies really works when things are this
critical and b) that won't work with an H-a filter because the 50D
is blind at that wavelength;
6) use ISO 1600 because you're going to
have plenty of images to combine with this fast a lens and it's all about
speed anyway; and
7) keep
at it because a) when all is said and done, that image doesn't
look that bad, and b) you know from embarassingly ancient experiements
with the CCD
that
it's really good for some subjects.
OK, now get to work for tonight. Make sure
the micrometer is working right, and find out how to cut Kydex neatly.
And shoot some flats. And move the mount to keep the Pleiades out of
the tree.
November 14. People,
it's been that kind of day. The Kydex I ordered to make a light- and
dewshield for the 200mm arrived, but it was much thicker and less flexible
than expected. So instead of a nice, round conventional dewshield,
I made a rectangular,
boxy one. And while cutting one sheet, I got confused and made the second
cut the wrong way, so instead of it being 12-inches deep, it's more like
9.5. That's plenty, as it turns out, but it's just the way things have gone.
I used some epoxy
putty to hold things together and then wrapped plenty of heavy-duty
black duct tape
around it. Then I applied ProtoStar's flocking paper to the inside. All
in all, it's a very sturdy, dark, shield. And ugly enough to work, too.
You know how honest to goodness observatory stuff looks -- all business
and no shelf appeal? -- it's like that. In six hours on a very humid
and cool evening, not a bit of dew formed on the lens although it is
soaking everything
else. Score one for me, in spite of myself.
I also rebuilt the micrometer focusser for
the 200mm. The bracketry gave it some slack last night and it has
none now. I put it on the other side of the lens, too, so that when I
use it with a guide-scope
I don't have to reach between them to focus.
The next project of the day: a focusing
mask similar to the one I use with the 5-inch as recommended by Ron Wodaski
in his CCD book. It doesn't work nearly so well, though -- too short a
focal length? Misaligned? Anyway, something
else to troubleshoot.
I verified that adding the IDAS LPR clip-in
filter DOES produce a significant focus shift with this lens (what doesn't?).
So I decided to shoot test frames with it in the 50D rather than in the
20D --
live view being critical to get focus close. This will be even more
important in H-a, but the 50D is blind to it. So I spent some time tonight
getting ready to outfit the Acer netbook as a focusing aid by using it
to examine the 20D's exposures.
Do not start shooting with the 200mm at F2
at ISO 1600 until you're sure it's dark. It looked pretty dark to me
when I shot a series of 100-second images of Sadr. Here's the best frame
as it came out of the camera and as it looks after a little work:
Not bad for such a brief exposure in twilight,
eh? See previous remarks for the relationship between perceived sharpness
and digital diffraction effects.
Here's a slightly confusing perspective on the 200mm (with
ugly but very effective hood) on the G11 under a not-quite-dark-enough
sky (10-22mm Canon EFs @ 10mm, 30s, F4.5, ISO 1250, Canon 50D):
More when I know more.