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 4 Jun 2010 10:42 On 04/06/2010 14:36, J. Clarke wrote: > On 6/4/2010 8:55 AM, Martin Brown wrote: >> On 03/06/2010 09:33, DanP wrote: >>> On 2 June, 16:44, Wolfgang Weisselberg<ozcvgt...(a)sneakemail.com> >>> wrote: >>> >>>>>> Lenses are imperfect, there's diffraction (and air for us down >>>>>> here) and stars are definitively larger than 0, just very far away. >>>>>> Stop waffling and try it. >>>>> So there is no source of light with perfect parallel rays. >>> >>> Your previous quote: "stars are definitively larger than 0, just very >>> far away." is incompatible with parallel rays. >> >> They are close enough to parallel as makes no difference (apart from our >> sun which subtends about half a degree). The largest stars are a few >> tens of milliarcseconds across as viewed from the Earth. It is right on >> the limits of our best optical telescopes to resolve them. Most stars >> are considerably smaller in angular size. >> >> Betelgeuse is just about resolved by the HST it is 20,000 smaller in >> diameter than the width of the sun as seen from Earth. >> >> http://www.spaceimages.com/atofbet.html >> >> To all intents and purposes at all ordinary apertures available to >> photgraphers and amateur astronomers they are ideal point sources with a >> bit of annoying phase screen in front from the atmosphere. >>> >>>> Both are f/4 at maximum aperture. Does that change your claim? >>>> Or does "the amount of light let in depends only of exposure >>>> time and f number" hold? >>> >>> No, both set at f/4 will let in the same amount of light. >> >> That is clearly an insane position to take. What you say is only sort of >> correct for an image formed by extended objects that are resolved. >> >> The key point about stars is that they are unresolved. >> >> A star remains unresolved even in the best lenses or telescopes and so >> all the light that hits the aperture ends up in the image. The larger >> lens captures more light and gives a smaller image Airey disk. The rough >> rule of thumb breaks down a bit when the scope is much bigger than 10" >> aperture as seeing starts to be the limiting factor and the star images >> stay roughly constant angular size after that point. >> >> Astronomers always want bigger apertures to capture more light. >> http://www.scopecity.net/LightGrasp.php >> >> It doesn't help on extended objects the laws of physics see to that but >> for a point source like a star the bigger aperture always wins. > > I think that you need to investigate the assumptions being made with > those limiting magnitude calculations. Comparing telescopes to > photographic lenses is comparing apples to oranges. Photographic lenses > are used with the primary lens projecting directly on the sensor, > astronomical telescopes almost always use secondary lenses. So are most modern telescopes - CCD cameras are now commonplace in amateur astronomy. Professional telescopes have almost always been operated as cameras with film or plates recording images and now CCDs. The main effect is that because CCDs have a much better quantum efficiency than the eye and can integrate for longer the photographic limiting magnitude is about 3 astronomical stellar magnitudes better. http://adsabs.harvard.edu/abs/1958LowOB...4...83J (using Kodak 103a-O plates) On time exposures it is more a matter of how well you can guide. The basic rules of optics are that at a fixed focal ratio of say f4 A resolved object like a brick wall if you make the lens 2x bigger the focal length is also doubled and although it collects 4x the light because it forms an image that is 2x bigger in linear scale so the brightness is constant. 4x as much light spread over 4x the area. An unresolved object is a point so that if you make the lens 2x bigger it collects 4x the light and dumps it all on the same sized spot. It is as a result of this that a larger aperture lens or telescope can always see fainter stars. The only assumption is that the optics are operating at close to their diffraction limit without excessive aberrations. This condition is true for good quality telescopes and usually true for most decent camera lenses when stopped down to f5.6 or slower. Fast camera lenses used wide open tend to render off axis stars less than perfectly but should be reasonably good on axis. Regards, Martin Brown
From: DanP on 4 Jun 2010 20:00 On 4 June, 13:55, Martin Brown <|||newspam...(a)nezumi.demon.co.uk> wrote: > On 03/06/2010 09:33, DanP wrote: > > > On 2 June, 16:44, Wolfgang Weisselberg<ozcvgt...(a)sneakemail.com> > >> Both are f/4 at maximum aperture. Does that change your claim? > >> Or does "the amount of light let in depends only of exposure > >> time and f number" hold? > > > No, both set at f/4 will let in the same amount of light. > > That is clearly an insane position to take. What you say is only sort of > correct for an image formed by extended objects that are resolved. Well I am not the one to claiming a camera lens .0001mm diameter can be set to f/4. Start Quote-------------------------------------------------------------------------------------------- >> So, talk about lens diameter. Will a lens diameter of 10.000km >> result in a darker star image than one of 0.0001mm? > At maximum aperture the 10.000 km will give a brighter picture. > But set at the maximum aperurte of the .00001mm the results will be > the same. Both are f/4 at maximum aperture. Does that change your claim? Or does "the amount of light let in depends only of exposure time and f number" hold? End Quote-------------------------------------------------------------------------------------------- So, if I shoot stars with 2 different diameter lenses, both set at same f number, focal length, exposure time and ISO, will the results be the same? For the sake of the argument say they resolve the same. > The key point about stars is that they are unresolved. > > A star remains unresolved even in the best lenses or telescopes and so > all the light that hits the aperture ends up in the image. The larger > lens captures more light and gives a smaller image Airey disk. The rough > rule of thumb breaks down a bit when the scope is much bigger than 10" > aperture as seeing starts to be the limiting factor and the star images > stay roughly constant angular size after that point. Thanks, I thought I was actually seeing the size of the stars at pixel level. Just curious, would that change at a lower aperture (higher f number)? I think it would although I see no advantage in it. It helps to see brighter stars spread over more pixels. > Astronomers always want bigger apertures to capture more light.http://www..scopecity.net/LightGrasp.php > > It doesn't help on extended objects the laws of physics see to that but > for a point source like a star the bigger aperture always wins. On telescopes aperture is lens/mirror diameter. On cameras no one cares about lens diameters and aperture refers to f stop. I agree with you on both cases. DanP
From: LOL! on 4 Jun 2010 22:29 On Fri, 4 Jun 2010 17:04:09 -0700 (PDT), DanP <dan.petre(a)gmail.com> wrote: > >Only laser light is parallel. Stars have an angular size (larger than >0) so they cannot have parallel rays. "Coherent light" doesn't mean parallel light rays. All lasers are rated in milliradians (mRad) of divergence (or convergence). There is no such thing as an absolute 100% collimated laser. The parallelism of the nearest starlight (not counting the sun) would far surpass anything that can be created by human hands. You have no idea how funny this is watching them trying to explain the merest basics of optics and light to you. You *can* find the shutter button on a camera, I hope? (not) LOL!
From: Ray Fischer on 5 Jun 2010 01:52 DanP <dan.petre(a)gmail.com> wrote: >On 3 June, 23:09, Wolfgang Weisselberg <ozcvgt...(a)sneakemail.com> >wrote: >> DanP <dan.pe...(a)hotmail.com> wrote: >> > On 2 June, 16:44, Wolfgang Weisselberg <ozcvgt...(a)sneakemail.com> >> > wrote: >> >> >> Lenses are imperfect, there's diffraction (and air for us down >> >> >> here) and stars are definitively larger than 0, just very far away. >> >> >> Stop waffling and try it. >> >> > So there is no source of light with perfect parallel rays. >> > Your previous quote: "stars are definitively larger than 0, just very >> > far away." is incompatible with parallel rays. >> >> And yet the rays are parallel. �Fancy that. > >Only laser light is parallel. I think you'll find that lasers are much less parallel than is starlight. > Stars have an angular size (larger than >0) so they cannot have parallel rays. If you want to get anal then there is no such thing as parallel rays since there will always be some error. >But I can concede that is practically true for short focal lengths. Even for long. It's not hard to calculate. Take a star of diameter 1,000,000 miles (largish, but not very). Put at a distance of five light years, or 5.9e12 miles. One divided by the other, inverse sin, and you get a divergence maximum of 9.740283e-06 degrees, or about 0.17 microradians. -- Ray Fischer rfischer(a)sonic.net
From: DanP on 5 Jun 2010 08:38
On 5 June, 06:52, rfisc...(a)sonic.net (Ray Fischer) wrote: > DanP <dan.pe...(a)gmail.com> wrote: > >On 3 June, 23:09, Wolfgang Weisselberg <ozcvgt...(a)sneakemail.com> > >wrote: > >> DanP <dan.pe...(a)hotmail.com> wrote: > >> > On 2 June, 16:44, Wolfgang Weisselberg <ozcvgt...(a)sneakemail.com> > >> > wrote: > >> >> >> Lenses are imperfect, there's diffraction (and air for us down > >> >> >> here) and stars are definitively larger than 0, just very far away. > >> >> >> Stop waffling and try it. > >> >> > So there is no source of light with perfect parallel rays. > >> > Your previous quote: "stars are definitively larger than 0, just very > >> > far away." is incompatible with parallel rays. > > >> And yet the rays are parallel. Fancy that. > > >Only laser light is parallel. > > I think you'll find that lasers are much less parallel than is > starlight. > > > Stars have an angular size (larger than > >0) so they cannot have parallel rays. > > If you want to get anal then there is no such thing as parallel rays > since there will always be some error. > > >But I can concede that is practically true for short focal lengths. > > Even for long. > > It's not hard to calculate. Take a star of diameter 1,000,000 miles > (largish, but not very). Put at a distance of five light years, or > 5.9e12 miles. One divided by the other, inverse sin, and you get a > divergence maximum of 9.740283e-06 degrees, or about 0.17 > microradians. Fair enough. I thought the star covers a number of pixels due to its size Martin cleared that one. DanP |