Single lens objective refractor

"Wfoley2"
.. But, then, seeing was much
better in 1946 than now.
Clear, Dark, Steady Skies!
(And considerate neighbors!!!)

You sure have that right! For street lighting we had a bare light bulb
at each block intersection, not a mercury or sodium vapour lamp scattering
light all over hell at four house intervals. Energy conservation, my BUTT!
:-( The only really dark place now is the far side of the Moon during the
lunar night!
Ed

(Maurice Gavin) wrote in message ...
On 23 Feb 2004 11:09:37 -0800, (Lurking Luser) wrote:

you could build a single objective lens refractor and fix
chromatic aberration....
Another dumb idea brought to you by,
James King

[snipe]

I've proved it works with sample colour spectra at

http://home.freeuk.com/m.gavin/chromab.htm

but will it 'fire' the Forum???

Actually I have been to your web page. That is where I got the idea.
Surplus Shed has a 127mm lens for only $6.50. The focal length is very
short though, only f/3.6. I think it might make a very good
spectrometer. As a telescope it would be terrible. But for six bucks
it sound like it is worth playing with.

Do you think adding a convex mirror at the end of the tube might help
to spread the spectra out? I was thinking of putting it at a 45-degree
angle so that the light comes out the side.

Great web page it really got me thinking,
James King

"Lurking Luser"
Do you think adding a convex mirror at the end of the tube might help
to spread the spectra out? I was thinking of putting it at a 45-degree
angle so that the light comes out the side.

Great web page it really got me thinking,
James King

Why not buy an in-expensive thin film type holographic grating from
Learning Technologies Inc. I have used such a grating to secure bright
fireball meteor spectra. The grating is mounted in front of the camera lens
acting as an objective dispersive element. Not as efficient as an expensive
blazed replica grating but it works for bright fireballs.
Ed Majden - AMS Spectroscopy
http://members.shaw.ca/epmajden/index.htm
Follow the lead to the AMS Web site for more information.

On 24 Feb 2004 10:24:12 -0800, (Lurking Luser) wrote:

(Maurice Gavin) wrote in message ...
On 23 Feb 2004 11:09:37 -0800, (Lurking Luser) wrote:

you could build a single objective lens refractor and fix
chromatic aberration....
Another dumb idea brought to you by,
James King

[snipe]

I've proved it works with sample colour spectra at

http://home.freeuk.com/m.gavin/chromab.htm

but will it 'fire' the Forum???

Actually I have been to your web page. That is where I got the idea.
Surplus Shed has a 127mm lens for only $6.50. The focal length is very
short though, only f/3.6. I think it might make a very good
spectrometer. As a telescope it would be terrible. But for six bucks
it sound like it is worth playing with.

Do you think adding a convex mirror at the end of the tube might help
to spread the spectra out? I was thinking of putting it at a 45-degree
angle so that the light comes out the side.

Great web page it really got me thinking,
James King

How's your optics? Let us know how you get on :-)

On Mon, 23 Feb 2004 20:43:11 GMT, Chris L Peterson
wrote:

On 23 Feb 2004 11:09:37 -0800, (Lurking Luser) wrote:

OK, here is another one of my harebrained ideas that is probably
flawed in some obvious way that you fine people will point out to me.
But here goes.

It seems to me, with the advent of digital cameras and computer
graphics, you could build a single objective lens refractor and fix
chromatic aberration with PhotoShop. You could take three pictures.
The first focused in the blue part of the spectrum, the second in the
mid frequencies, and the last in the red zone.

One problem is that color sensors have crappy filters in front of them. A result
of this is that lots of red makes it through the blue and green, etc (or similar
for CMY). This is one reason that color sensors don't do a very good job with
most astronomical targets. For terrestrial imaging, there is lots of image
processing going on to reconstruct something like accurate color. But anything
that makes it through the wrong filter in your case is just going to contribute
to blurring, which isn't really fixable.

People using conventional three filter imaging with B&W sensors or film get very
good results without needing an apochromat by doing just what you suggest- but
their filters are much higher quality. Normally, you use an achromat since the
more complex design minimizes other aberrations as well. A single element
objective that is well enough corrected for good imaging would need to be
aspheric, and would probably end up costing more and performing worse than a
basic achromat.

_______________________________________________ __

Chris L Peterson
Cloudbait Observatory
http://www.cloudbait.com

Wouldn't a single element lens need to be refocused for each RGB image
and thus be of a difference image scale eg blue image 'smaller' than
red image? CYM wouldn't work as each filter transmits two primary
colours - each coming to a different focus.

On Tue, 24 Feb 2004 20:28:28 GMT, (Maurice Gavin) wrote:

Wouldn't a single element lens need to be refocused for each RGB image
and thus be of a difference image scale eg blue image 'smaller' than
red image?

Yes, but the original poster proposed doing just that. You can fix the image
scale during post processing.


CYM wouldn't work as each filter transmits two primary
colours - each coming to a different focus.

CYM wouldn't work as well because the filter bandwidths are wider, that is
correct. Everybody who uses this technique with tricolor imaging and achromats
is using RGB filter sets.

_________________________________________________

Chris L Peterson
Cloudbait Observatory
http://www.cloudbait.com

"Ed Majden" wrote in message news:YTN_b.609110$X%5.601260@pd7tw2no...

[snipe]
Why not buy an in-expensive thin film type holographic grating from
Learning Technologies Inc. I have used such a grating to secure bright
fireball meteor spectra. The grating is mounted in front of the camera lens
acting as an objective dispersive element. Not as efficient as an expensive
blazed replica grating but it works for bright fireballs.
Ed Majden - AMS Spectroscopy
http://members.shaw.ca/epmajden/index.htm
Follow the lead to the AMS Web site for more information.

Actually I am looking at some grating from Edmunds Scientific, and I
have played with just plain old CDs. But the CDs are not nearly good
enough for stars. Fine for the sun, but not the stars. I'll check out
Learning Technologies.

I think I read somewhere that a grating of just 300 etchings is the
best for stars. Is this true?

Thanks for the tip,
James king

This encourages me to publish my braindead idea...

Build a big reflector with bad surface and, say f/1 speed. (I like it
fast and don't fear coma...)

Measure the point spread function PSF for each pixel or for certain
pixel areas of the CCD with a laser beam.

Recalculate the "horrible" raw image with FFT and Inverse PSF.

How about that?

Or more theoretically:
Which image errors can be corrected by image processing?

Wilfried

"Lurking Luser"
Actually I am looking at some grating from Edmunds Scientific, and I
have played with just plain old CDs. But the CDs are not nearly good
enough for stars. Fine for the sun, but not the stars. I'll check out
Learning Technologies.

I think I read somewhere that a grating of just 300 etchings is the
best for stars. Is this true?

James:
A grating with fewer grooves are generally a bit more efficient than one
with more grooves but the dispersion is less. The best diffraction replica
gratings have a "blaze angle". Most of the light is concentrated into a
first order spectrum on one side of the zero order. For meteor spectroscopy
I use a 600 g/mm B&L precision transmission diffraction replica grating.
These are rather expensive however, in the $800 US$ range for a 52X52 mm
sized grating. Thin film type gratings don't offer this feature. For
stellar spectra using a telescope you could consider using the eyepiece type
gratings sold by Rainbow Optics. See: www.StarSpectroscope.com
Reflection type diffraction gratings can often be found on eBay. Select one
that is blazed for the visual region of the spectrum, around 5000A or 500.0
nm. You can make a nice star spectroscope using one of these. Do a search
with "google" for amateur stellar spectroscope" or "spectrograph" to get
some idea of how to do this. Transmission gratings are better for meteor
work as they are mounted if front of a normal camera lens and you secure
spectra over a wide region. It is possible to use large reflection gratings
to do this. The camera looks down at the grating so it must be large enough
to accept all of the light entering your camera lens. Dave Airey from the
B.A.A. did this and it worked very well. If your lucky you may find such a
large grating on eBay at much less than what a new one costs.
I hope this helps.

Ed Majden - Amateur Meteor Spectroscopy
Courtenay, B.C.

Or more theoretically:
Which image errors can be corrected by image processing?

Wilfried

Well....technically probably ALL of them....

But that requires at least 2 things....knowing what the errors are very
accurately....and measuring what you image the first time very accurately....

The first one probably isnt TOOO much a problem....which then leaves imaging
error from seeing...but you get that with ANY telescope so no big deal....

The second part is whats gonna kill you though (imho)......imagine your star
image is spread over many pixels due to all kinds of abberations/errors
......(and besides you would need a large number of pixels to well sample every
star image for your correction schemes to work in the first place...

Well, since your star image is very spread out...the intensity level is very
low for each pixel.....and at low light levels there is a good bit of
measurement noise (ie what you measure....vs what you would measure if you had
a perfect, noise free dectector)....

Guess what.....that noise is gonna propagate through ALLL those calculations
you do to correct for the errors.....and it wouldnt suprise me it it got bigger
as it propogated vs smaller....

ie bad data in = bad info out....

Plus, the computations will be a bear....

Give it another 20 years when we have some REAL computing power and nearly
perfect detectors and it might actually be a sometimes used technique....

of course I could be way off base here...so who knows

take care

Blll