Jump to content


Photo

Question on all sky camera optics

  • Please log in to reply
7 replies to this topic

#1 sixtysomething

sixtysomething

    Sputnik

  • -----
  • Posts: 34
  • Joined: 02 Sep 2011

Posted 14 November 2012 - 10:26 AM

In the type of all sky camera where the camera lens looks down at a hemispherical mirror, is the overall f ratio set by the f number of the lens?

In other words if you have a DSLR with say a 50mm f1.8 looking down at a hemispherical mirror reflecting the entire sky into the lens is the overall f# still f1.8? Would the sky fog limit stay the same at the same ISO as if the lens was just pointed up?

#2 Sky Captain

Sky Captain

    Metal Whisperer

  • *****
  • Posts: 11238
  • Joined: 07 Nov 2004
  • Loc: Loc: Loc:

Posted 14 November 2012 - 11:19 PM

In the type of all sky camera where the camera lens looks down at a hemispherical mirror, is the overall f ratio set by the f number of the lens?



Yep.

Not sure about the sky fog limit though.

#3 sixtysomething

sixtysomething

    Sputnik

  • -----
  • Posts: 34
  • Joined: 02 Sep 2011

Posted 15 November 2012 - 11:41 AM

Intuitively it seems to me the sky fog limit or the histogram would stay the same, convex mirrors don't seem to photograph any darker than their surroundings that I've ever noticed, all the way down to little silver Christmas ornament balls.

But then intuition is so often wrong.

Anyway I'm doing a video based all sky camera for meteor detection mostly and the really wide angle lenses are fairly slow for the most part and good ones are expensive enough that I wouldn't want to leave one out in the weather even under a dome. I'm still suffering from sticker shock after looking at a top of the line Fujinon.

So a longer and faster lens pointed down at a convex mirror seemed like the shade tree way to get a wide angle but fast lens.

#4 mark cowan

mark cowan

    Vendor (Veritas Optics)

  • *****
  • Vendors
  • Posts: 4086
  • Joined: 03 Jun 2005
  • Loc: salem, OR

Posted 15 November 2012 - 05:43 PM

Cost kind of depends on the sensor size you want to use. I've been looking into building one of these and will probably go with a c-mount lens and one of these cameras - want several FPS continuous at high resolution... And a lens like this one.

BTW though, this is really only for daylight applications (cloud tracking). But the principle is the same in any case (disregarding sensor sensitivity).

Best,
Mark

#5 GlennLeDrew

GlennLeDrew

    Voyager 1

  • *****
  • Posts: 10644
  • Joined: 17 Jun 2008
  • Loc: Ottawa, Ontario, Canada

Posted 15 November 2012 - 08:00 PM

Indeed, assuming near to full reflectance, any mirror, no matter the form and degree of curvature, conserves image surface brightness. And so the exposure will be the same as that for the camera's lens alone.

Of course, the reduced image scale will be equivalent to using a smaller aperture (at the same f/ratio), and so point sources will record to a reduced level of faintness.

Do note that the image produced by the strongly curved mirror will lie on a curved surface, the center of curvature of which is shared by the mirror. For a convex reflector, the image lies somewhat below the surface of the mirror. To minimize defocus due to the curved image surface, stop down your lens somewhat. This also has the benefit of reducing astigmatism as well.

#6 nytecam

nytecam

    Voyager 1

  • *****
  • Posts: 11455
  • Joined: 20 Aug 2005
  • Loc: London UK

Posted 16 November 2012 - 01:10 PM

Aiming a 'regular' fl lens down onto a convex mirror facing upwards will produce a hemisheric 180deg fisheye image but image quality is compromised especially as the perfiery.

As the mirror is strongly negative the camera lens needs refocusing to ~1m or so for infinity to be in focus. Whilst the f/ratio of the camera lens remains unchanged, the 'negative' mirror greatly reduces the effective aperture of the lens to almost pinhole levels so only bright [naked eye] stars are recorded.

You could try this approach :grin:

#7 GlennLeDrew

GlennLeDrew

    Voyager 1

  • *****
  • Posts: 10644
  • Joined: 17 Jun 2008
  • Loc: Ottawa, Ontario, Canada

Posted 16 November 2012 - 11:59 PM

Maurice,
Just to be clear, it must be said that indeed the smaller effective aperture imposes a limit to faint star detection, just as would be the case for a fisheye lens. And again, as for a fisheye lens, extended objects would record just as well off the mirror as when imaged with the lens alone; only the image scale would be smaller and thereby reveal less detail.

#8 sixtysomething

sixtysomething

    Sputnik

  • -----
  • Posts: 34
  • Joined: 02 Sep 2011

Posted 17 November 2012 - 07:40 AM

That's a very nice unit nytecam, I'm sure the price tag is every bit as impressive as the images, albeit in a somewhat different manner.

The Mooonglow Technologies all sky camera has a 1.2mm f2.8 fisheye, that's not too far from being a pinhole lens in practical terms with a clear effective aperture less than 0.5mm in diameter.

I've seen f0.75 lenses in the 4mm range for the video camera I'm planning on using and f0.95 lenses are fairly common and not unreasonably priced, that's three stops faster than an f2.8 which should make a noticeable difference in needed exposure length for a given histogram.

I've also found a heated 8.5" convex spot mirror for big trucks at the auto parts store. My camera is quite small and putting it even fairly close to the reflector shouldn't block out that much sky.

It's been my experience that with a decent exposure my still camera has no problem seeing clouds at night either but my location has a moderate amount of light pollution. Here's a fifteen second wide angle shot I got the other night at f2 and ISO 3200, the clouds are obvious.

Given that my horizon is mostly obscured by trees some distortion at the periphery of the image isn't a major concern, if I can get 150 degrees at the equivalent of f1.4 I'll be more than happy.

Attached Files








Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics