Poor quality of lunar images with 20D/C90

[dylan]

Archived from groups: rec.photo.digital (More info?)

"jess" <jgates@erols.com> wrote in message
news:1127598136.442095.138800@f14g2000cwb.googlegroups.com...
> Nominally it's about 1000mm f11, but I have read that the effective
> aperture is more like f12.5 (as Ray Fisher suggests below). I'd
> believe you about the exposure, but I read ISO 1600, 1/1800 from the
> EXIF data and if anything the image looks underexposed.
>

1/1800 @ 1600 ISO
= 1/900 @ 800 ISO
=1/450 @ 400 ISO
=1/225 @ 200 ISO
=1/112 @ 100 ISO

from my figures of 1/250 @ f8 using 100 ISO
your f number = approx f 16, so 12.5 sounds about right.

try setting 100 or 200 ISO for better quality.

and use the mirror lock facility in the camera.

Cheers

 

[Guest]

Archived from groups: rec.photo.digital (More info?)

Now that you mention it, most of the webcam images I've seen are of
relatively small objects (as viewed from Earth), and all I really
wanted was a decent picture that images the entire moon, full and/or in
various phases..

Someone in the thread mentioned the Series IV 'Stiletto' Focusers by
Stellar Technologies international' The link is

http/www.stellar-international.com/

This (or something like it) is what you may have been recalling. The
Ronchi Screen units seem interesting. Can you comment on the relative
advantages and disadvantages of each type?

I have enjoyed your posts and web site immensely; your discussions of
"digital vs. film" are a big reason I decided to take the DSLR
plunge. In fact, it was initially the two photos of the moon by
yourself and Annika (???) that led me to try the moon photo in the
first place. Now if only I could afford those long "L" lenses . .
..

 

[dylan]

Archived from groups: rec.photo.digital (More info?)

> from my figures of 1/250 @ f8 using 100 ISO
> your f number = approx f 16

D'oh ! try f11, makes your f12.5 nearer.

 

[Guest]

Archived from groups: rec.photo.digital (More info?)

Well, Mr. Clark. did indicate that the 20D is the CAMERA of choice for
deep sky work. Both webcams and DSLRs are cameras, but I think it
would be difficult to justify calling a dedicated CCD imager a
"camera", at least in the ordinary vernacular sense. And, as Mr.
Clark points out webcams are rather ill-suited to DSOs.

I don't know enough about the subject to know whether the 20D
specifically is the camera of choice, although I have read that there
are some problems with the Nikon D70 having to do with the noise
reduction circuitry, as I recall. It does, moreover, seem possible to
allow taking decent DSO images with a 20D and related DSLRs. See

M31: http/www.schweifstern.de/images-pages/deepsky/m31_ef200_ws.htm
M31: http/www.schweifstern.de/images-pages/deepsky/m31_20D_450.htm
Pipe nebula: http/azastrophotography.com/Pipe%20Region.htm
Lagoon nebula:
http/www.photo.net/photodb/photo?photo_id=3708346&size=lg
M31, M41, Horsehead Nebula,Great nebula in Orion; M33;
http/web.canon.jp/Imaging/astro/pages_e/13_e.html
Numerous (with Canon 10D):
http/www.ricksastro.com/Gallery/htm/10D.htm
Numerous (with Canon 10D):
http/www.astrosurf.com/afernandez/gallery/wo-10d/WO_10D_astrophoto.htm
Numerous (1DS Mk2):
http/www.whirlpoolgalaxy.com/firstlight_canon_1dsmii_table.html

There are a lot more. Now, these may not be as good as those taken by
professional astronomers, and they may not be as good as those taken
with the most modern equipment specifically designed for deep space
astrophotography. But I would say that calling DSLRs "not
suitable" for guided imaging of DSOs is difficult to justify on the
basis of these images.

As to the Hubble, well, yeah, it takes really, REALLY good images of
DSOs. However, the $6.2 billion cost is a bit out of my price range.
My last name is Gates, but my first name isn't Bill.

 

[Guest]

Archived from groups: rec.photo.digital (More info?)

jess wrote:

> Now that you mention it, most of the webcam images I've seen are of
> relatively small objects (as viewed from Earth), and all I really
> wanted was a decent picture that images the entire moon, full and/or in
> various phases..
>
> Someone in the thread mentioned the Series IV 'Stiletto' Focusers by
> Stellar Technologies international' The link is
>
> http/www.stellar-international.com/
>
> This (or something like it) is what you may have been recalling. The
> Ronchi Screen units seem interesting. Can you comment on the relative
> advantages and disadvantages of each type?

Yes that's it! Thanks for finding it.
The Ronchi ruling would be easier to focus because you don't need to
position the star exactly on a knife edge. The more expensive one
with the 300 lines/inch would probably work better. At 300 lpi, that
is a 12 micron spacing. The diffraction spot diameter of an
f/11 system is 14 microns. So with a faster system, the star image could
fall in between the rulings giving a somewhat less precise
focus. But I'm not sure if it would be enough to make a difference.
It might with the cheaper lower 180 lpi model.
The knife edge is very precise, but harder to work as precise
positioning of the star at the knife edge is necessary. A clock
drive would be necessary with the knife.

> I have enjoyed your posts and web site immensely; your discussions of
> "digital vs. film" are a big reason I decided to take the DSLR
> plunge. In fact, it was initially the two photos of the moon by
> yourself and Annika (???) that led me to try the moon photo in the
> first place. Now if only I could afford those long "L" lenses . .

Thank you. Good luck.
Roger

 

[Guest]

Archived from groups: rec.photo.digital (More info?)

john_doe_ph_d <john_doe_ph_d@yahoo.com> wrote:
>jess wrote:
>
>> As I indicated above, I'll try focusing on a single
>> star before trying the moon the next time. I'll also try to let the
>> moon get higher in the sky (assuming the clouds ever break up).
>>
>If you think focusing on the moon is hard, wait until you try to focus
>on a tiny star. I was thinking of doing the opposite: waiting until the
>moon was in view so I could focus at it and then move over to also do
>some star trails.

Stars tend to be easier because they focus to a tiny point. The moon
doesn't.

--
Ray Fischer
rfischer@sonic.net

 

[Guest]

Archived from groups: rec.photo.digital (More info?)

Roger N. Clark (change username to rnclark) <username@qwest.net> wrote:
>Ignore the people saying the 20D is not good for astrophotography.
>In fact, it is the camera of choice these days for deep sky.

No, it isn't. Sorry. The 20D is not suitable for deep-sky objects
that require 30-120 minute exposures. Cooled imagers with built-in
autoguiders still work much better.

As an aside, the Hubble space telescope did a deep-sky photo of the
object around the Andromeda galaxy. The combined exposure time was
about four days. THAT is deep sky.

>Here is a moon shot, limited by lens and seeing:
>http/www.clarkvision.com/galleries/gallery.astrophoto-1/web/moon-JZ3F3658-60-c-5x-700.html

Nice enough moon shot, but no relation at all to deep-sky photography.

--
Ray Fischer
rfischer@sonic.net

 

[Guest]

Archived from groups: rec.photo.digital (More info?)

Ray Fischer wrote:

> Roger N. Clark (change username to rnclark) <username@qwest.net> wrote:
>
>>Ignore the people saying the 20D is not good for astrophotography.
>>In fact, it is the camera of choice these days for deep sky.
>
> No, it isn't. Sorry. The 20D is not suitable for deep-sky objects
> that require 30-120 minute exposures. Cooled imagers with built-in
> autoguiders still work much better.

You mean like this:
Veil nebula: 140 minutes with 20Da (20D with IR filter removed):
http/astrophotography.aa6g.org/Astronomy/Astrophotos/ngc6992_20da.html

or this:
The Pleiades, M45: 180 minutes with a 20Da:
http/astrophotography.aa6g.org/Astronomy/Astrophotos/m45_20da.html

See this for technical specs for low light astro work:
EOS 20D CMOS detector characterization noise and dark current
http/astrosurf.com/cavadore/APN/carac/EOS20D

Thousands and thousands of stunning deep sky astrophotos are being taken with
DSLRs. Latest data from the 20D camera is that it has 3 electron
RMS read noise at ISO 1600, an incredibly low value. No other DSLR camera
has achieved that. Most cooled CCD imagers have around 20 electron read noise.

> As an aside, the Hubble space telescope did a deep-sky photo of the
> object around the Andromeda galaxy. The combined exposure time was
> about four days. THAT is deep sky.
>
>
>>Here is a moon shot, limited by lens and seeing:
>>http/www.clarkvision.com/galleries/gallery.astrophoto-1/web/moon-JZ3F3658-60-c-5x-700.html
>
> Nice enough moon shot, but no relation at all to deep-sky photography.

This thread is about photographing the moon!

If you would go to the gallery and look at other photos (or simply click the
previous button on the above page), you would see deep sky astrophotos
taken with DSLRs.

Roger

 

[Guest]

Archived from groups: rec.photo.digital (More info?)

Roger N. Clark (change username to rnclark) <username@qwest.net> wrote:
>Ray Fischer wrote:
>
>> Roger N. Clark (change username to rnclark) <username@qwest.net> wrote:
>>
>>>Ignore the people saying the 20D is not good for astrophotography.
>>>In fact, it is the camera of choice these days for deep sky.
>>
>> No, it isn't. Sorry. The 20D is not suitable for deep-sky objects
>> that require 30-120 minute exposures. Cooled imagers with built-in
>> autoguiders still work much better.
>
>You mean like this:
>Veil nebula: 140 minutes with 20Da (20D with IR filter removed):
>http/astrophotography.aa6g.org/Astronomy/Astrophotos/ngc6992_20da.html

10 minute exposures.

>or this:
>The Pleiades, M45: 180 minutes with a 20Da:
>http/astrophotography.aa6g.org/Astronomy/Astrophotos/m45_20da.html

10 minute exposures.

The 20d is a fine camera and you can do some excellent astrophotography
with it. The 20Da has some additional benefits. But it still has
more noise than cooled CCD imagers dedicated to the job.

--
Ray Fischer
rfischer@sonic.net

 

[Guest]

Archived from groups: rec.photo.digital (More info?)

Ray Fischer wrote:

> Roger N. Clark (change username to rnclark) <username@qwest.net> wrote:
>
>>Ray Fischer wrote:
>>
>>
>>>Roger N. Clark (change username to rnclark) <username@qwest.net> wrote:
>>>
>>>
>>>>Ignore the people saying the 20D is not good for astrophotography.
>>>>In fact, it is the camera of choice these days for deep sky.
>>>
>>>No, it isn't. Sorry. The 20D is not suitable for deep-sky objects
>>>that require 30-120 minute exposures. Cooled imagers with built-in
>>>autoguiders still work much better.
>>
>>You mean like this:
>>Veil nebula: 140 minutes with 20Da (20D with IR filter removed):
>>http/astrophotography.aa6g.org/Astronomy/Astrophotos/ngc6992_20da.html
>
>
> 10 minute exposures.
>
>
>>or this:
>>The Pleiades, M45: 180 minutes with a 20Da:
>>http/astrophotography.aa6g.org/Astronomy/Astrophotos/m45_20da.html
>
>
> 10 minute exposures.
>
> The 20d is a fine camera and you can do some excellent astrophotography
> with it. The 20Da has some additional benefits. But it still has
> more noise than cooled CCD imagers dedicated to the job.
>

You have conflicting statements:
1) "The 20D is not suitable for deep-sky objects"
2) "The 20d is a fine camera and you can do some excellent astrophotography"

The facts of multiple astounding astrophotos prove #2 is correct.

It matters little if you do one long exposure or multiple added exposures.
Multiple short exposures average the read noise and minimize thermal noise.
The fact that DSLRs with a different processing method get within a factor
of about 2 of a cooled dedicated $10,000 CCD is astounding, and at
more than 5 times lower cost. In fact the multiple short exposure
method has advantages: guiding errors in one frame can be thrown
out. Other events like airplanes, bumping the telescope, etc
only ruin one frame of many. CCD imagers are starting to use this
method too. The only time a cooled CCD gets MUCH fainter is when
a black and white image is done and one uses no filters over
the CCD.

But for color images, the difference between cooled CCDs and
DSLRs is very small. The DSLR is much easier to use in my opinion, as
you get full color with each image. And among DSLRs, the 20D has
the lowest measured noise and overall best noise performance
among amateur astronomers who are reporting test results on cameras.

Roger

 

[Guest]

Archived from groups: rec.photo.digital (More info?)

Bob wrote:

> On 24 Sep 2005 11:05:23 -0700, "jess" <jgates@erols.com> wrote:
>
>>I have a relatively new DSLR and an old (~ 1980, black tube) Celestron
>>C90. The other light I tried taking a picture of the full moon with
>>the camera at prime focus (tripod, ASA 1600, 1/800, remote release
>>after ~10 sec of MLU), and the results were reasonably well-exposed
>>but very fuzzy (and noisy: maybe ISO 200 or 400 would be better).

I suspect the focus is off. You might find it worth practicing in the
daytime on a distant object first (or even the moon).

Full moon is more difficult to photograph well than first or last
quarter where there are sharp shadow details on the terminator to aid
focussing. You might also try putting a shadow mask with two holes at
oppposite sides in front of the scope objective as a focussing aid.

When the two spots merge the image is in focus. If you are stuck
focussing on a crude glass screen the other option is to take several
shots refocussing between each and hope that you hit the sweet spot.

My Pentax istD works pretty well with my old MTO f10 1000mm mirror lens.

Try asking on sci.astro.amateur chances are someone there has exactly
the same hardware combination and can advise on the pitfalls.

Regards,
Martin Brown

 

[Guest]

Archived from groups: rec.photo.digital (More info?)

Roger N. Clark (change username to rnclark) <username@qwest.net> wrote:
>Ray Fischer wrote:
>> Roger N. Clark (change username to rnclark) <username@qwest.net> wrote:
>>>Ray Fischer wrote:
>>>>Roger N. Clark (change username to rnclark) <username@qwest.net> wrote:

>>>>>Ignore the people saying the 20D is not good for astrophotography.
>>>>>In fact, it is the camera of choice these days for deep sky.
>>>>
>>>>No, it isn't. Sorry. The 20D is not suitable for deep-sky objects
>>>>that require 30-120 minute exposures. Cooled imagers with built-in
>>>>autoguiders still work much better.
>>>
>>>You mean like this:
>>>Veil nebula: 140 minutes with 20Da (20D with IR filter removed):
>>>http/astrophotography.aa6g.org/Astronomy/Astrophotos/ngc6992_20da.html
>>
>>
>> 10 minute exposures.
>>
>>
>>>or this:
>>>The Pleiades, M45: 180 minutes with a 20Da:
>>>http/astrophotography.aa6g.org/Astronomy/Astrophotos/m45_20da.html
>>
>>
>> 10 minute exposures.
>>
>> The 20d is a fine camera and you can do some excellent astrophotography
>> with it. The 20Da has some additional benefits. But it still has
>> more noise than cooled CCD imagers dedicated to the job.
>>
>
>You have conflicting statements:
> 1) "The 20D is not suitable for deep-sky objects"
> 2) "The 20d is a fine camera and you can do some excellent astrophotography"

There's no conflict.

>The facts of multiple astounding astrophotos prove #2 is correct.

Where are all of the 120-minute exposures?

>It matters little if you do one long exposure or multiple added exposures.

Of course it matters.

--
Ray Fischer
rfischer@sonic.net

 

[Guest]

Archived from groups: rec.photo.digital (More info?)

In article <dhah4n$r06$1@bolt.sonic.net>, rfischer@bolt.sonic.net
says...
> >The facts of multiple astounding astrophotos prove #2 is correct.
>
> Where are all of the 120-minute exposures?

Does it really matter if you do twelve ten minute exposures or one 120
minute exposure? Methinks not, as the amount of light collected should
be the same.

> >It matters little if you do one long exposure or multiple added exposures.
>
> Of course it matters.

See above.
--
http/www.pbase.com/bcbaird

 

[Guest]

Archived from groups: rec.photo.digital (More info?)

Brian Baird <no@no.thank.u> wrote:
> rfischer@bolt.sonic.net

>> >The facts of multiple astounding astrophotos prove #2 is correct.
>>
>> Where are all of the 120-minute exposures?
>
>Does it really matter if you do twelve ten minute exposures or one 120
>minute exposure?

Sometimes, yes.

> Methinks not, as the amount of light collected should
>be the same.

Except for the pixels that end up black because not enough photons
were detected to exceed the minimum.

--
Ray Fischer
rfischer@sonic.net

 

[Guest]

Archived from groups: rec.photo.digital (More info?)

In article <dhd4ih$l0d$1@bolt.sonic.net>, rfischer@bolt.sonic.net
says...
> > Methinks not, as the amount of light collected should
> >be the same.
>
> Except for the pixels that end up black because not enough photons
> were detected to exceed the minimum.

Which is why you shoot at high sensitivities when you stack images.

Really, this isn't rocketry.
--
http/www.pbase.com/bcbaird

 

[Guest]

Archived from groups: rec.photo.digital (More info?)

On Tue, 27 Sep 2005 04:21:18 GMT, Brian Baird <no@no.thank.u> wrote:

>In article <dhah4n$r06$1@bolt.sonic.net>, rfischer@bolt.sonic.net
>says...
>> >The facts of multiple astounding astrophotos prove #2 is correct.
>>
>> Where are all of the 120-minute exposures?
>
>Does it really matter if you do twelve ten minute exposures or one 120
>minute exposure? Methinks not, as the amount of light collected should
>be the same.

Beyond the time required to fully record the image, keeping
the lens open on a digital camera carries no benefit -- the photons do
not accumulate over time as they do on film. By making ten shorter
shots, you can superimpose and average out the information from
multiple "as exposed as they can get" shots.

>
>> >It matters little if you do one long exposure or multiple added exposures.
>>
>> Of course it matters.
>
>See above.

 

[Guest]

Archived from groups: rec.photo.digital (More info?)

kashe@sonic.net wrote:
> On Tue, 27 Sep 2005 04:21:18 GMT, Brian Baird <no@no.thank.u> wrote:
>
>
>>In article <dhah4n$r06$1@bolt.sonic.net>, rfischer@bolt.sonic.net
>>says...
>>
>>>>The facts of multiple astounding astrophotos prove #2 is correct.
>>>
>>>Where are all of the 120-minute exposures?
>>
>>Does it really matter if you do twelve ten minute exposures or one 120
>>minute exposure? Methinks not, as the amount of light collected should
>>be the same.

This is correct. If you get 100 photons (electrons) per 10 minute exposure,
then after 12 exposures added together you have a total of
1200 photons (electrons). The problem with multiple exposures is read noise.

Say the read noise is 3 electrons (the read noise of the 20D),
then the final noise accumulated read noise is 3 * square root 12
= 10.4 electrons. But the noise from 1200 electrons is
square root 1200 = 34.6 electrons, so photon noise dominates.
Total noise is sqrt(10.4^2 + 34.6^2) = 36.1 electrons.
>
>
> Beyond the time required to fully record the image, keeping
> the lens open on a digital camera carries no benefit -- the photons do
> not accumulate over time as they do on film.

This is correct. Film loses charged state of the photosensitive grains,
so photons do not accumulate linearly. This is called reciprocity
failure. Digital sensors do not suffer from that problem so are
quite superior in that regard.


> By making ten shorter
> shots, you can superimpose and average out the information from
> multiple "as exposed as they can get" shots.

Huh?

>>>>It matters little if you do one long exposure or multiple added exposures.
>>>
>>>Of course it matters.
>>
>>See above.

Yes, the statement
"It matters little if you do one long exposure or multiple added exposures."
is correct, assuming read noise is small, which it is on good cameras
like the 20D.

Jeez, I was out for 2 days (calibrating a NASA sensor) and this thread
is still going.

Roger

 

[Guest]

Archived from groups: rec.photo.digital (More info?)

Ray Fischer wrote:

> Brian Baird <no@no.thank.u> wrote:
>
>>rfischer@bolt.sonic.net
>
>
>>>>The facts of multiple astounding astrophotos prove #2 is correct.
>>>
>>>Where are all of the 120-minute exposures?
>>
>>Does it really matter if you do twelve ten minute exposures or one 120
>>minute exposure?
>
>
> Sometimes, yes.
>
>
>> Methinks not, as the amount of light collected should
>>be the same.
>
>
> Except for the pixels that end up black because not enough photons
> were detected to exceed the minimum.
>

As long as you digitize the noise there is no minimum. In any
one exposure, the signal can be less than one DN (data number)
per exposure. Then given enough exposures you can dig the signal
out of the noise by averaging multiple exposures. We are doing this
right now with the NASA Cassini mission to measure faint rings.
The Cassini VIMS instrument is limited to about 1 second exposures
or the infrared channels (5 micron wavelengths) saturate due
to thermal emission inside the instrument running at 140 kelvin.
We do hour plus integrations averaging thousands of exposures.
Signal levels 0.1 DN/exposure and lower have been extracted to give a
signal-to-noise ratio good enough to identify composition from the
spectrum and produce images that show the extent of the rings.

Roger