Now, the next BIG contest is ON!
in [Digital Cameras]
|
Prev: Here Comes the 3-D Camera: Revolutionary Prototype Films Worldin Three Dimensions
Next: Why the Nikon Coolscan V ED is so expensive event on vintage market ?
From: Martin Brown on 16 Jun 2010 09:55 On 15/06/2010 13:34, J. Clarke wrote: > On 6/15/2010 2:10 AM, David J Taylor wrote: >>> And I am trying to point out that a 200mm f/2.8 lens at f/2.8 will >>> capture more light from a star than a 50mm f/2.8 lens at >>> f/2.8. >>> >>> It's patently obvious that a 200mm f/2.8 lens has a larger >>> diameter than a 50mm f/2.8 lens. >>> >>> -Wolfgang >> >> More light (photons) in a given period, when other things are similar. And crucially for a star in good focus the light all falls on the same physical sensor area for a given focal ratio. This f ratio gives a certain brightness for any extended object irrespective of focal length, but for an unresolved point source like a star it puts all the light from the larger aperture onto a same sized spot. > I don't see Wolfgang's post, but it contains the usual misconception > about photographic vs astronomical imaging. > > The 200 will "capture more light" in the sense that more light will > enter the lens. However for a given sensor size, a smaller proportion of > that light will strike the sensor. If you make adjustments so that both > cover the same field of view, for example by adding two 2x > teleconverters to the 50, _then_ the 200 will show its "greater light > gathering", and this will be reflected in the photographic formulae, as > at that point the 50/2.8 will be functioning as a 200/11.2. This is true but misleading. The 200mm f2.8 wins out on unresolved objects because of the physics of diffraction. Camera lenses are just faster than normally used in astronomy prime focus imaging device. This is largely because slower optics are easier to make diffraction limited. Although it is true that astronomical optics tend to be f5 or slower (with the exception of Schmidt cameras which are about f2 and they are most definitely *cameras* - no prospect of visual observing). The Palomar survey Schmidt being one of the early ones: http://www.nps.gov/history/history/online_books/butowsky5/astro4f.htm#photos f2.5 with 120" focal length and 48" working aperture capable of diffraction limited imaging on a 7 degree field of view using 10" or 14" photographic plates. > Astronomers generally work with a magnified field to begin with--in > other words they almost always use some kind of teleconverter--so that > they are working at effective focal lengths longer than that of their > primary lens or mirror. In that situation, larger aperture always > results in a brighter image. Photographers, on the other hand, hardly > ever use a teleconverter these days, so effect of the diameter of their > lens is accurately reflected in the f/ratio. This is nonsensical gibberish. The point about the 200mm f2.8 lens is that it will form an image of an unresolved point source like a star with the same linear dimensions as the 50mm f2.8 lens but with 16x as much light hitting the same patch of the sensor. The gain is *real* provided that the optics are truly diffraction limited which usually holds valid for short exposures with working apertures of 4" or less. Extended objects also get 16x as much light but on an area that is 16x bigger so that the illumination is the same. Stars are a special case! Don't take my word for it! Go out and take a few shots of the constellations with different apertures and focal lengths. Nothing beats doing the experiment. Guiding will be a problem for modest focal lengths at just a few seconds so you need to choose very bright stars. Regards, Martin Brown
From: Martin Brown on 16 Jun 2010 11:12 On 16/06/2010 14:28, David J Taylor wrote: > "whisky-dave" <whisky-dave(a)final.front.ear> wrote in message > news:hvah80$fuj$1(a)qmul... > [] >>> More light (photons) in a given period, when other things are similar. >> >> Well that makes sense, but I'd have thought it obvious..... >> that 200mm f/2.8 lens would have a larger diameter than a 50mm f/2.8 >> lens. > > Indeed, by definition of f/number. > > What some people seem to be having difficulty with is that - given > nominally identical lenses but with everything scaled by four - is that > the light collected would be spread out over a larger area with the > 200mm lens, even though the number of photons per unit area would be the > same, and the implications of that. That is true for extended objects, the f ratio tells you how bright the resulting image will be for any lens - which is why it is so useful. But it is not true for an unresolved point object like a star. If you quadruple the aperture at constant f ratio the image of an unresolved point object stays exactly the same linear size at the sensor but gets 16x as much light delivered to it. This is a key property of unresolved objects and focal ratios. And it holds good until the aperture or the exposure time prevents the system from being diffraction limited (typically >6" and a few seconds). Regards, Martin Brown
From: David J Taylor on 16 Jun 2010 12:22 "Martin Brown" <|||newspam|||@nezumi.demon.co.uk> wrote in message news:xR5Sn.142419$0M5.125298(a)newsfe07.iad... > On 16/06/2010 14:28, David J Taylor wrote: >> "whisky-dave" <whisky-dave(a)final.front.ear> wrote in message >> news:hvah80$fuj$1(a)qmul... >> [] >>>> More light (photons) in a given period, when other things are >>>> similar. >>> >>> Well that makes sense, but I'd have thought it obvious..... >>> that 200mm f/2.8 lens would have a larger diameter than a 50mm f/2.8 >>> lens. >> >> Indeed, by definition of f/number. >> >> What some people seem to be having difficulty with is that - given >> nominally identical lenses but with everything scaled by four - is that >> the light collected would be spread out over a larger area with the >> 200mm lens, even though the number of photons per unit area would be >> the >> same, and the implications of that. > > That is true for extended objects, the f ratio tells you how bright the > resulting image will be for any lens - which is why it is so useful. > > But it is not true for an unresolved point object like a star. If you > quadruple the aperture at constant f ratio the image of an unresolved > point object stays exactly the same linear size at the sensor but gets > 16x as much light delivered to it. > > This is a key property of unresolved objects and focal ratios. And it > holds good until the aperture or the exposure time prevents the system > from being diffraction limited (typically >6" and a few seconds). > > Regards, > Martin Brown Yes, I was referring to normal photographic use, not astro-photography. May I ask - in astrophotography, how important is it that the lens is diffraction limited, considering how bad atmospheric disturbance can be during long exposures? Or is there some "optimum" combination of stacking, exposure time, and nearness to diffraction limits which depends on the seeing, lens focal length etc. etc. Perhaps a feeling one gains with experience? Cheers, David
From: Martin Brown on 16 Jun 2010 15:28 On 16/06/2010 17:22, David J Taylor wrote: > > "Martin Brown" <|||newspam|||@nezumi.demon.co.uk> wrote in message > news:xR5Sn.142419$0M5.125298(a)newsfe07.iad... >> On 16/06/2010 14:28, David J Taylor wrote: >>> "whisky-dave" <whisky-dave(a)final.front.ear> wrote in message >>> news:hvah80$fuj$1(a)qmul... >>> [] >>>>> More light (photons) in a given period, when other things are similar. >>>> >>>> Well that makes sense, but I'd have thought it obvious..... >>>> that 200mm f/2.8 lens would have a larger diameter than a 50mm f/2.8 >>>> lens. >>> >>> Indeed, by definition of f/number. >>> >>> What some people seem to be having difficulty with is that - given >>> nominally identical lenses but with everything scaled by four - is that >>> the light collected would be spread out over a larger area with the >>> 200mm lens, even though the number of photons per unit area would be the >>> same, and the implications of that. >> >> That is true for extended objects, the f ratio tells you how bright >> the resulting image will be for any lens - which is why it is so useful. >> >> But it is not true for an unresolved point object like a star. If you >> quadruple the aperture at constant f ratio the image of an unresolved >> point object stays exactly the same linear size at the sensor but gets >> 16x as much light delivered to it. >> >> This is a key property of unresolved objects and focal ratios. And it >> holds good until the aperture or the exposure time prevents the system >> from being diffraction limited (typically >6" and a few seconds). > > Yes, I was referring to normal photographic use, not astro-photography. Just to stress the point this *only* works for unresolved stars. Extended nebulae images cannot made brighter with a bigger working aperture because they have 4x the light spread out over 4x the area. I'm sure you know this, but other posters here seem horribly confused. > > May I ask - in astrophotography, how important is it that the lens is > diffraction limited, considering how bad atmospheric disturbance can be > during long exposures? Or is there some "optimum" combination of > stacking, exposure time, and nearness to diffraction limits which > depends on the seeing, lens focal length etc. etc. Perhaps a feeling one > gains with experience? Its actually still a topic of serious ongoing research! At 8" and above the diffraction limiting deep sky imaging for amateurs is less important and the scopes are referred to in a slightly derogatory way as "light buckets". Planetary observers are more demanding and their kit tends to be truly diffraction limited on axis so that on those rare moments of good seeing they have the detail. For most amateur deep sky imaging a good rule of thumb is that you will struggle to do any better than 2" arc fwhm (1/1000 solar diameter) which is equivalent to 3" working aperture. But with devilishly cunning pro hardware and selective shift and sum of the images with the highest contrast ground based results on larger apertures can now beat the HST for small bright targets. The technique is called "Lucky Imaging" and has revolutionised amateur planetary imaging work. This can be done small scale with a humble webcam, very good optics and a lot of patience. Ideally you want good seeing! http://www.ast.cam.ac.uk/~optics/Lucky_Web_Site/ The main effect so far seems to demonstrate that things hitting Jupiter with a big wallop are more common that we once thought. A couple of weeks back some of the top amateur planetary imagers caught another impact (or rather someone else spotted it in his images). And he already had one notch for finding a thing hits Jupiter event. http://www.impactlab.com/2010/06/06/amateur-astronomers-capture-collision-on-jupiter/ This stuff is typically right on the limits of the diffraction limit for the scope and CCD in use and they use specialised software to shift and add the images to get the optimum result. The good thing is that now there is usually more than one amateur imager pointed at the same planet so any transient phenomena can be confirmed after the event. The main difference between telescope optics and camera lenses these days is that the scope trades filling the whole field of view with a good image for perfect diffraction limited behaviour over a smaller central sweet spot. This is in part an artefact of the most easily mass produced SCT telescopes which have a slightly curved focal plane. Regards, Martin Brown
From: David J Taylor on 17 Jun 2010 02:24
"Martin Brown" <|||newspam|||@nezumi.demon.co.uk> wrote in message news:7B9Sn.37958$7d5.31094(a)newsfe17.iad... [] > Its actually still a topic of serious ongoing research! Trust me to ask an awkward question! [good stuff snipped] > http://www.ast.cam.ac.uk/~optics/Lucky_Web_Site/ [more snipped] > The main difference between telescope optics and camera lenses these > days is that the scope trades filling the whole field of view with a > good image for perfect diffraction limited behaviour over a smaller > central sweet spot. This is in part an artefact of the most easily mass > produced SCT telescopes which have a slightly curved focal plane. > > Regards, > Martin Brown Thanks for that Web site pointer, Martin. Most interesting, and a demonstration of something which can be done now that digital imaging and powerful computers are readily available. BTW: Nikon had/have a "Best Shot Selector" option for some of their cameras whereby they would take a sequence of images, and select the one with the largest JPEG file size as the one with the best detail (i.e. least blur). Just slightly related. BTW: some years back I had connections with both Cambridge and Edinburgh astronomy circles. My wife used to bend glass plates to fit that curved focal plane! Cheers, David |