| |
Main
Date: 30 Aug 2006 20:04:23
From: SW
Subject: Canon Rebel XT, Nikon D50 or Pentax K100 best?
|
The low end SLRs are getting very reasonably priced and I am ready to make
the plunge into digital SLR astrophotography.
Which of these is most sensitive (highest S/N) for low light astronomy?
Does the K100 use the same Sony sensor as the Nikon D50?
Does the K100 smooth out the RAW images like the D50?
SW
|
|
| |
Date: 30 Aug 2006 19:20:45
From: atasselli@hotmail.com
Subject: Re: Canon Rebel XT, Nikon D50 or Pentax K100 best?
|
atasselli@hotmail.com wrote:
> SW wrote:
> > The low end SLRs are getting very reasonably priced and I am ready to make
> > the plunge into digital SLR astrophotography.
> >
> >
> > Which of these is most sensitive (highest S/N) for low light astronomy?
>
> Between 350D and D50, the D50. I have no info on the K100 performance.
>
This might help deciding:
http://www.sfu.ca/%7Edmunro/D50.html
Andrea T.
|
| |
Date: 31 Aug 2006 01:15:25
From: Chris L Peterson
Subject: Re: Canon Rebel XT, Nikon D50 or Pentax K100 best?
|
On Wed, 30 Aug 2006 20:04:23 -0400, SW <SW@!aol.com > wrote:
>The low end SLRs are getting very reasonably priced and I am ready to make
>the plunge into digital SLR astrophotography.
>
>Which of these is most sensitive (highest S/N) for low light astronomy?
I had the chance a couple of months ago to do some Milky Way shots with
a D50. The images were quite a bit noisier than my 300D. Since the 350D
(XT) is less noisy than the 300D, I'd say you don't want the D50. I
don't know about the Pentax.
_________________________________________________
Chris L Peterson
Cloudbait Observatory
http://www.cloudbait.com
|
| | |
Date: 31 Aug 2006 10:33:11
From: SW
Subject: Re: Canon Rebel XT, Nikon D50 or Pentax K100 best?
|
>>
>>Which of these is most sensitive (highest S/N) for low light
>>astronomy?
>
> I had the chance a couple of months ago to do some Milky Way shots
> with a D50. The images were quite a bit noisier than my 300D. Since
> the 350D (XT) is less noisy than the 300D, I'd say you don't want the
> D50. I don't know about the Pentax.
>
> _________________________________________________
>
> Chris L Peterson
> Cloudbait Observatory
> http://www.cloudbait.com
>
Two different opinions, I was afraid of that!
Were these Milky Way shots done with the same lens?
The ISO, exposure length, temperature all affects the S/N.
Perhaps the other sources are measuring the D50 noise with the median
filtering in effect?
SW
|
| | | |
Date: 31 Aug 2006 17:13:24
From: Chris L Peterson
Subject: Re: Canon Rebel XT, Nikon D50 or Pentax K100 best?
|
On Thu, 31 Aug 2006 10:33:11 -0400, SW <SW@!aol.com > wrote:
>Two different opinions, I was afraid of that!
>Were these Milky Way shots done with the same lens?
No, but the lenses were similar and were used at the same focal length
and focal ratio.
>The ISO, exposure length, temperature all affects the S/N.
These were the same in both cases (180 seconds, ISO 400). More to the
point, however, I shot dark frames with both cameras. Sampling the
central million pixels, the standard deviation with the 300D was 387,
and the standard deviation of the D50 was 572, or about 1.5X the noise.
This was quite visible in both the dark and light frames.
My 300D does not require a dark frame, but the D50 has a very bright
corner that produces a significant gradient. It is necessary to subtract
a dark frame to remove that (an operation that adds some noise to the
final image).
>Perhaps the other sources are measuring the D50 noise with the median
>filtering in effect?
Possible. Such filtering can make the noise measurements lower, although
at the expense of also filtering out detail.
_________________________________________________
Chris L Peterson
Cloudbait Observatory
http://www.cloudbait.com
|
| | |
Date: 01 Sep 2006 03:13:54
From: SW
Subject: Re: Canon Rebel XT, Nikon D50 or Pentax K100 best?
|
I haven't seen many people using Pentax DSLR in astronomy,
This new camera might be as good as a Nikon D50 as
it boasts ISO up to 3200.
Any reason why one shouldn't go with Pentax for astrophotography?
SW
|
| | | |
Date: 01 Sep 2006 13:57:11
From: Chris L Peterson
Subject: Re: Canon Rebel XT, Nikon D50 or Pentax K100 best?
|
On Fri, 01 Sep 2006 03:13:54 -0400, SW <SW@!aol.com > wrote:
>I haven't seen many people using Pentax DSLR in astronomy,
>This new camera might be as good as a Nikon D50 as
>it boasts ISO up to 3200.
Keep in mind that ISO values are largely meaningless for astroimaging.
Raising the ISO on a camera doesn't really increase its sensitivity in
most cases. Astroimaging is about collecting and counting photons, and
the camera receives the same number whatever the ISO setting. There is a
reason that cameras designed for astroimaging don't usually have
variable gain.
Raising the ISO is useful when you need to make short exposures. But
that is something you don't do with DSOs, where you need many minutes or
hours of total exposure time to get acceptable S/N. There is no
advantage to increasing the gain since it increases the noise right
along with the signal.
With the 300D, the best noise performance is at ISO 200, so that's where
I leave it for astroimaging. I expect the sweet spot is similar for
other digital cameras.
>Any reason why one shouldn't go with Pentax for astrophotography?
Only because it seems to be largely untested for that purpose.
_________________________________________________
Chris L Peterson
Cloudbait Observatory
http://www.cloudbait.com
|
| | | | |
Date: 01 Sep 2006 14:43:19
From: Paul Schlyter
Subject: Re: Canon Rebel XT, Nikon D50 or Pentax K100 best?
|
In article <ihegf29pp3nklnlph18r5pqdfaok2e3il2@4ax.com >,
Chris L Peterson <clp@alumni.caltech.edu > wrote:
>On Fri, 01 Sep 2006 03:13:54 -0400, SW <SW@!aol.com> wrote:
>
>>I haven't seen many people using Pentax DSLR in astronomy,
>>This new camera might be as good as a Nikon D50 as
>>it boasts ISO up to 3200.
>
>Keep in mind that ISO values are largely meaningless for astroimaging.
>Raising the ISO on a camera doesn't really increase its sensitivity in
>most cases. Astroimaging is about collecting and counting photons, and
>the camera receives the same number whatever the ISO setting. There is a
>reason that cameras designed for astroimaging don't usually have
>variable gain.
>
>Raising the ISO is useful when you need to make short exposures. But
>that is something you don't do with DSOs, where you need many minutes or
>hours of total exposure time to get acceptable S/N. There is no
>advantage to increasing the gain since it increases the noise right
>along with the signal.
>
>With the 300D, the best noise performance is at ISO 200, so that's where
>I leave it for astroimaging. I expect the sweet spot is similar for
>other digital cameras.
>
>>Any reason why one shouldn't go with Pentax for astrophotography?
>
>Only because it seems to be largely untested for that purpose.
>_________________________________________________
>
>Chris L Peterson
>Cloudbait Observatory
>http://www.cloudbait.com
The failure if ISO film sensitivity values to be meaningful at very
long exposure time is called "recipricity failure". For several
traditional films there are data available about the recipricity failure
factor for various exposure times - if that factor is accounted for,
ISO values become meaningful even at very long exposure times.
Anyone knows whether the reciprocity failure of CCD chips is worse,
or better, than for film?
--
----------------------------------------------------------------
Paul Schlyter, Grev Turegatan 40, SE-114 38 Stockholm, SWEDEN
e-mail: pausch at stockholm dot bostream dot se
WWW: http://stjarnhimlen.se/
|
| | | | | |
Date: 01 Sep 2006 15:03:29
From: Chris L Peterson
Subject: Re: Canon Rebel XT, Nikon D50 or Pentax K100 best?
|
On Fri, 01 Sep 2006 14:43:19 GMT, pausch@saaf.se (Paul Schlyter) wrote:
>The failure if ISO film sensitivity values to be meaningful at very
>long exposure time is called "recipricity failure". For several
>traditional films there are data available about the recipricity failure
>factor for various exposure times - if that factor is accounted for,
>ISO values become meaningful even at very long exposure times.
Yes, but I'm talking about digital sensors, not film. ISO values are
meaningful in film because (ignoring reciprocity failure) a high ISO
value is essentially equivalent to a higher quantum efficiency. That is
not the case with digital imaging.
>Anyone knows whether the reciprocity failure of CCD chips is worse,
>or better, than for film?
There is no reciprocity failure with CCDs. They simply count photons.
Indeed, good quality, cooled CCDs approach being perfect devices for
long exposures- that is, they have nearly zero noise and they count
nearly every photon.
_________________________________________________
Chris L Peterson
Cloudbait Observatory
http://www.cloudbait.com
|
| | | | | |
Date: 01 Sep 2006 17:18:29
From: William Hamblen
Subject: Re: Canon Rebel XT, Nikon D50 or Pentax K100 best?
|
On Fri, 01 Sep 2006 14:43:19 GMT, pausch@saaf.se (Paul Schlyter)
wrote:
>Anyone knows whether the reciprocity failure of CCD chips is worse,
>or better, than for film?
CCD chips have essentially no reciprocity failure. The problem with
long exposures is that the number of random thermal electrons
increases right along with the electrons produced by starlight. Some
digital cameras are inherently less noisy than others. Astronomical
CCDs are cooled to reduce thermal electrons. It is possible to
perform statistical analysis on the digital data to measure the amount
of noise objectively. I haven't seen studies like that on different
digital cameras.
Bud
--
The night is just the shadow of the Earth.
|
| | | | |
Date: 02 Sep 2006 20:48:59
From: Iordani
Subject: Re: Canon Rebel XT, Nikon D50 or Pentax K100 best?
|
Chris L Peterson wrote:
> Keep in mind that ISO values are largely meaningless for astroimaging.
> Raising the ISO on a camera doesn't really increase its sensitivity in
> most cases.
This is news to me. Can you explain what the ISO setting does?
A pointer to some more info would also be valuable.
I tried google but found nothing appropriate.
|
| | | | | |
Date: 02 Sep 2006 19:36:23
From: Chris L Peterson
Subject: Re: Canon Rebel XT, Nikon D50 or Pentax K100 best?
|
On Sat, 02 Sep 2006 20:48:59 +0200, Iordani <somewhere@earth.net > wrote:
>This is news to me. Can you explain what the ISO setting does?
>A pointer to some more info would also be valuable.
>I tried google but found nothing appropriate.
The ISO value is a measure of the gain applied to the output amplifier.
In all cases I'm aware of this is done after the sensor is read, but in
principle it could be done inside the sensor, which could provide some
moderate advantage in terms of readout noise.
Astronomical cameras have a gain chosen to provide a good match between
the number of electrons each pixel can store and the number of bits the
A/D converter uses. In the simplest case, you want a full pixel to
produce a full count from the A/D.
In astronomical image, the intrinsic noise is determined by the number
of photons collected (the noise is the square root of the photon count).
When you increase the gain, you increase the signal and the noise the
same amount. That's why unfiltered, high ISO images are noisy. It also
means that any bright sources in your image will saturate much earlier.
High ISO astronomical images may show larger stars because of this. When
you are processing a raw image, it makes no difference if a pixel has a
value of 100 ± 10 or 400 ± 40. You could as well multiply all the pixels
of an ISO 100 image by 4 after the fact as to shoot it at ISO 400.
Well, almost. The difference is that if you can reduce the number of
subframes you need, you can reduce the total contribution of readout
noise. By boosting the gain at the sensor you can pull the lowest level
signal a little higher above the quantization noise.
With the 300D, I've done detailed noise measurements. What I found is
that there is no difference in noise between a 300s exposure at ISO 100
and a 150s exposure at ISO 200. But a 75s exposure at ISO 400 has more
noise, and a 37.5s exposure at ISO 800 has even more, and this trend
continues at ISO 1600 and ISO 3200 (the highest the camera goes).
So clearly the best place to operate this camera is ISO 200. However,
the noise at ISO 400 is only a little higher, so I use that for shots
where a longer exposure might be a problem.
A good camera is an efficient photon counter, recording 50% or more of
all that hit it. From a signal processing standpoint, there is no reason
to artificially multiply that number by anything. You just need to find
the value for a specific camera that gives the best results- probably
between one and two counts per photon.
_________________________________________________
Chris L Peterson
Cloudbait Observatory
http://www.cloudbait.com
|
| | | | | | |
Date: 02 Sep 2006 21:55:33
From: Pierre Vandevennne
Subject: Re: Canon Rebel XT, Nikon D50 or Pentax K100 best?
|
Chris L Peterson <clp@alumni.caltech.edu > wrote in
news:29mjf2ta9hbma4lhvehp0g2uk6u7b6mr29@4ax.com:
> In all cases I'm aware of this is done after the sensor is read, but in
> principle it could be done inside the sensor,
How? Just curious.
|
| | | | | | | |
Date: 02 Sep 2006 22:43:34
From: Chris L Peterson
Subject: Re: Canon Rebel XT, Nikon D50 or Pentax K100 best?
|
On 02 Sep 2006 21:55:33 GMT, Pierre Vandevennne
<pierre@datarescue_ns.com > wrote:
>Chris L Peterson <clp@alumni.caltech.edu> wrote in
>news:29mjf2ta9hbma4lhvehp0g2uk6u7b6mr29@4ax.com:
>
>> In all cases I'm aware of this is done after the sensor is read, but in
>> principle it could be done inside the sensor,
>
>How? Just curious.
The charge might be multiplied before the readout amplifier, which would
mean that the readout noise wouldn't change with ISO setting. There are
some so-called zero readout noise sensors that use a similar technique
(but the devices are currently exotic and mostly experimental; I doubt
anything like this is done in any consumer cameras).
_________________________________________________
Chris L Peterson
Cloudbait Observatory
http://www.cloudbait.com
|
| | | | | | | | |
Date: 03 Sep 2006 17:18:08
From: Pierre Vandevennne
Subject: Re: Canon Rebel XT, Nikon D50 or Pentax K100 best?
|
Chris L Peterson <clp@alumni.caltech.edu > wrote in
news:u32kf2t9l2dtt7nqcsjgrrkrbvivbcp5jc@4ax.com:
> The charge might be multiplied before the readout amplifier, which would
> mean that the readout noise wouldn't change with ISO setting. There are
> some so-called zero readout noise sensors that use a similar technique
> (but the devices are currently exotic and mostly experimental; I doubt
> anything like this is done in any consumer cameras).
I have seen some stuff in a Marconi presentation, but I was under the
impression that it introduced its own "multiplier" noise and that it was
quite complex. Agree that it sounds way beyond the current technology
deployed in digital cameras.
|
| | | | | | | | | |
Date: 04 Sep 2006 05:56:27
From: duncan
Subject: Re: Canon Rebel XT, Nikon D50 or Pentax K100 best?
|
Sony sensor specs for the SXV-M25. My testing shows that the D50 will
meet the below specs at the same temperature. Spectral response of a
IR-filter removed D50 should be the same as below.
#
CCD type: Sony ICX413AQ SuperHAD CCD with ultra low dark current, Bayer
RGB matrix and vertical anti-blooming.
#
CCD Full resolution Pixel data: Pixel size: 7.8uM x 7.8uM, Image
format: 3024 x 2016 pixels
#
CCD Image area: 23.4mm (Horizontal) x 15.6mm (Vertical).
#
CCD quality: Grade 1 or better - No bad columns, no dead pixels, no
more than 50 'hot' pixels (saturated in <10 seconds).
#
Spectral Response: QE max at 540nM (~60%), 50% roll-off at 400nM and
650nM.
#
Readout Noise: Less than 12 electrons RMS - typically ONLY 7 electrons!
#
Full-well capacity: Greater than 25,000 e- (unbinned)
#
Anti-blooming: Overload margin greater than 800x.
#
Dark current: Dark frame saturation time greater than 100 hours. Less
than 0.02 electrons/second @ + 10C ambient.
cheers
Duncan
|
| | | | | | | | | | |
Date: 04 Sep 2006 14:02:07
From: JJ
Subject: Re: Canon Rebel XT, Nikon D50 or Pentax K100 best?
|
How does the Canon Rebel XT sensor specs. compare with the Sony ICX413AQ?
Does the Pentax K100D also use this sensor?
I'm a little relunctant to buy the D50 because of Mode 3 requirements.
Can it work without a laptop in the field?
SW
duncan <dunmunro@shaw.ca > wrote in news:fePKg.500767$Mn5.417147@pd7tw3no:
> Sony sensor specs for the SXV-M25. My testing shows that the D50 will
> meet the below specs at the same temperature. Spectral response of a
> IR-filter removed D50 should be the same as below.
>
>
> #
>
> CCD type: Sony ICX413AQ SuperHAD CCD with ultra low dark current, Bayer
> RGB matrix and vertical anti-blooming.
> #
>
> CCD Full resolution Pixel data: Pixel size: 7.8uM x 7.8uM, Image
> format: 3024 x 2016 pixels
> #
>
> CCD Image area: 23.4mm (Horizontal) x 15.6mm (Vertical).
> #
>
> CCD quality: Grade 1 or better - No bad columns, no dead pixels, no
> more than 50 'hot' pixels (saturated in <10 seconds).
> #
>
> Spectral Response: QE max at 540nM (~60%), 50% roll-off at 400nM and
> 650nM.
> #
>
> Readout Noise: Less than 12 electrons RMS - typically ONLY 7
electrons!
> #
>
> Full-well capacity: Greater than 25,000 e- (unbinned)
> #
>
> Anti-blooming: Overload margin greater than 800x.
> #
>
> Dark current: Dark frame saturation time greater than 100 hours. Less
> than 0.02 electrons/second @ + 10C ambient.
>
>
> cheers
>
> Duncan
>
|
| | | | | | | | | | | |
Date: 04 Sep 2006 20:57:35
From: Duncan Munro
Subject: Re: Canon Rebel XT, Nikon D50 or Pentax K100 best?
|
I looked at a K100D DF. It has less noise than an Ist DS2 but more than
a D50. I don't think the K100D is a good choice for astrophotography.
Pentax seems to have implemented the amplifier in a much more noisy way
than Nikon or Minolta.
cheers
Duncan
JJ wrote:
>How does the Canon Rebel XT sensor specs. compare with the Sony ICX413AQ?
>Does the Pentax K100D also use this sensor?
>
>I'm a little relunctant to buy the D50 because of Mode 3 requirements.
>Can it work without a laptop in the field?
>
>SW
>
>
>
>duncan <dunmunro@shaw.ca> wrote in news:fePKg.500767$Mn5.417147@pd7tw3no:
>
>
>
>>Sony sensor specs for the SXV-M25. My testing shows that the D50 will
>>meet the below specs at the same temperature. Spectral response of a
>>IR-filter removed D50 should be the same as below.
>>
>>
>>#
>>
>>CCD type: Sony ICX413AQ SuperHAD CCD with ultra low dark current, Bayer
>>RGB matrix and vertical anti-blooming.
>>#
>>
>> CCD Full resolution Pixel data: Pixel size: 7.8uM x 7.8uM, Image
>>format: 3024 x 2016 pixels
>>#
>>
>> CCD Image area: 23.4mm (Horizontal) x 15.6mm (Vertical).
>>#
>>
>> CCD quality: Grade 1 or better - No bad columns, no dead pixels, no
>>more than 50 'hot' pixels (saturated in <10 seconds).
>>#
>>
>> Spectral Response: QE max at 540nM (~60%), 50% roll-off at 400nM and
>>650nM.
>>#
>>
>> Readout Noise: Less than 12 electrons RMS - typically ONLY 7
>>
>>
>electrons!
>
>
>>#
>>
>> Full-well capacity: Greater than 25,000 e- (unbinned)
>>#
>>
>> Anti-blooming: Overload margin greater than 800x.
>>#
>>
>> Dark current: Dark frame saturation time greater than 100 hours. Less
>>than 0.02 electrons/second @ + 10C ambient.
>>
>>
>>cheers
>>
>>Duncan
>>
>>
>>
>
>
>
|
| | | | | | | | | | | | |
Date: 08 Sep 2006 00:22:21
From: SW
Subject: Re: Canon Rebel XT, Nikon D50 or Pentax K100 best?
|
How much more noise would you guess the K100D have compared to the D50?
Twice as much?
One can always stack images, even non dslr can be usable.
SW
Duncan Munro <dmunro@sfu.ca > wrote in news:3r0Lg.504161$iF6.436827
@pd7tw2no:
> I looked at a K100D DF. It has less noise than an Ist DS2 but more than
> a D50. I don't think the K100D is a good choice for astrophotography.
> Pentax seems to have implemented the amplifier in a much more noisy way
> than Nikon or Minolta.
>
> cheers
>
> Duncan
>
|
| | | | | | | | | | | | | |
Date: 09 Sep 2006 20:32:13
From: duncan
Subject: Re: Canon Rebel XT, Nikon D50 or Pentax K100 best?
|
The actual dark current is many times higher on a K100D than on a D50:
Pentax K100 Dark Frame info (21c)
Model: PENTAX K100D
Date/Hour: 2006/08/10 18:28:52
Exposure time: 365.500 s
Aperture: f/13.0
Focal length: 33.0 mm
ISO: 200 ASA
Format: 3040x2024
*** Red layer ***
Mean: 327.1 Median: 231
Sigma: 238.0
Maxi.: 4095.0 Mini.: 0.0
*** Green layer ***
Mean: 290.7 Median: 240
Sigma: 177.2
Maxi.: 4095.0 Mini.: 0.0
*** Blue layer ***
Mean: 408.6 Median: 326
Sigma: 210.3
Maxi.: 4095.0 Mini.: 0.0
Basically a K100D sensor would completely saturate in about 70min. This
is not very good.
This is the stats from a 30min ISO200 D50 DF (20c):
=== Image: d:\d50astro\asv\irdarks\200d.nef ===
Format: 3039x2014
(0,0)-(0,0) - Binning: 1x1
26/05/2006 - 10:59:01 - IT: 1799.60
Model: NIKON D50
Date/Hour: 2006/05/26 10:59:01
ISO: 200 ASA
Format: 3039x2014
*** Red layer ***
Mean: 5.3 Median: 1
Sigma: 19.1
Maxi.: 4041.0 Mini.: 0.0
*** Green layer ***
Mean: 7.1 Median: 2
Sigma: 23.9
Maxi.: 3864.0 Mini.: 0.0
*** Blue layer ***
Mean: 9.5 Median: 3
Sigma: 27.2
Maxi.: 4095.0 Mini.: 0.0
The DF saturation time is about 100 hours. This is the lowest DC I've
seen in a DSLR. Here's a comparison of a D50 and a Canon 1DMK11:
http://www.clarkvision.com/imagedetail/long-exposure-comparisons/index.html
However, at ISO200 I think the D50 turns the tables on the Canon.
Having to use Mode 3 is a PITA, but using 30min exposures mitigates
this, and once you automate Mode 3 this is no longer an issue and the
D50 becomes a superb astrocamera.
http://www.sfu.ca/%7Edmunro/D50.html
Cheers
Duncan
SW wrote:
> How much more noise would you guess the K100D have compared to the D50?
> Twice as much?
>
> One can always stack images, even non dslr can be usable.
>
> SW
>
>
>
> Duncan Munro <dmunro@sfu.ca> wrote in news:3r0Lg.504161$iF6.436827
> @pd7tw2no:
>
>
>>I looked at a K100D DF. It has less noise than an Ist DS2 but more than
>>a D50. I don't think the K100D is a good choice for astrophotography.
>>Pentax seems to have implemented the amplifier in a much more noisy way
>>than Nikon or Minolta.
>>
>>cheers
>>
>>Duncan
>>
|
| | | | | | | | | | | | | | |
Date: 13 Sep 2006 20:23:46
From: SW
Subject: Re: Canon Rebel XT, Nikon D50 or Pentax K100 best?
|
These are fantastic numbers for the D50.
How long does it take the Canon Rebel XT (dark frame) to saturate at 21C?
Doe you have similar statistics for the XT?
SW
duncan <dunmunro@shaw.ca > wrote in news:hxFMg.521506$Mn5.81736@pd7tw3no:
> The actual dark current is many times higher on a K100D than on a D50:
>
> Pentax K100 Dark Frame info (21c)
>
> Model: PENTAX K100D
> Date/Hour: 2006/08/10 18:28:52
> Exposure time: 365.500 s
> Aperture: f/13.0
> Focal length: 33.0 mm
> ISO: 200 ASA
> Format: 3040x2024
>
> *** Red layer ***
> Mean: 327.1 Median: 231
> Sigma: 238.0
> Maxi.: 4095.0 Mini.: 0.0
> *** Green layer ***
> Mean: 290.7 Median: 240
> Sigma: 177.2
> Maxi.: 4095.0 Mini.: 0.0
> *** Blue layer ***
> Mean: 408.6 Median: 326
> Sigma: 210.3
> Maxi.: 4095.0 Mini.: 0.0
>
>
> Basically a K100D sensor would completely saturate in about 70min.
> This is not very good.
>
> This is the stats from a 30min ISO200 D50 DF (20c):
>
> === Image: d:\d50astro\asv\irdarks\200d.nef ===
> Format: 3039x2014
> (0,0)-(0,0) - Binning: 1x1
> 26/05/2006 - 10:59:01 - IT: 1799.60
>
> Model: NIKON D50
> Date/Hour: 2006/05/26 10:59:01
> ISO: 200 ASA
> Format: 3039x2014
>
> *** Red layer ***
> Mean: 5.3 Median: 1
> Sigma: 19.1
> Maxi.: 4041.0 Mini.: 0.0
> *** Green layer ***
> Mean: 7.1 Median: 2
> Sigma: 23.9
> Maxi.: 3864.0 Mini.: 0.0
> *** Blue layer ***
> Mean: 9.5 Median: 3
> Sigma: 27.2
> Maxi.: 4095.0 Mini.: 0.0
>
> The DF saturation time is about 100 hours. This is the lowest DC I've
> seen in a DSLR. Here's a comparison of a D50 and a Canon 1DMK11:
>
> http://www.clarkvision.com/imagedetail/long-exposure-comparisons/index.
> html
>
> However, at ISO200 I think the D50 turns the tables on the Canon.
>
>
> Having to use Mode 3 is a PITA, but using 30min exposures mitigates
> this, and once you automate Mode 3 this is no longer an issue and the
> D50 becomes a superb astrocamera.
>
>
|
| | | | | | |
Date: 02 Sep 2006 23:39:49
From: Iordani
Subject: Re: Canon Rebel XT, Nikon D50 or Pentax K100 best?
|
Chris L Peterson wrote:
<snipped >
Thank you so much Chris!
Very very informative reading.
|
| |
Date: 30 Aug 2006 18:00:05
From: atasselli@hotmail.com
Subject: Re: Canon Rebel XT, Nikon D50 or Pentax K100 best?
|
SW wrote:
> The low end SLRs are getting very reasonably priced and I am ready to make
> the plunge into digital SLR astrophotography.
>
>
> Which of these is most sensitive (highest S/N) for low light astronomy?
Between 350D and D50, the D50. I have no info on the K100 performance.
>
> Does the K100 use the same Sony sensor as the Nikon D50?
Yes.
> Does the K100 smooth out the RAW images like the D50?
Don't know.
Andrea T.
|
| |
Date: 31 Aug 2006 15:50:03
From: atasselli@hotmail.com
Subject: Re: Canon Rebel XT, Nikon D50 or Pentax K100 best?
|
Chris L Peterson wrote:
> On 31 Aug 2006 13:55:50 -0700, "atasselli@hotmail.com"
> <atasselli@hotmail.com> wrote:
>
> >I'm reading it under Iris. The lowest value is 0 ADU and the highest is
> >4095 (hot pixel) so the dynamics is all there.
>
> I'm suspicious. Such low dark current as you report is not consistent
> with any silicon devices at room temperature, nor for Sony's specs on
> sensors of this type. I think a baseline is being subtracted somewhere
> (regardless of the apparent 0-4095 range).
>
> The numbers I see in the raw array seem to make much more sense.
>
If it's a clever trick (removing the amplifier bias on the fly) then it
must be done in firmware. Checked both bias and light frame and they
are both consistent with the dark frame. A 5min exposure has a
background value of about 300-350 ADU (100mm f/2.5 lens with my mag.5
skies) and the saturated stars have a peak value slightly shy of 4096
ADU. The bias frame has an average of 1.2 ADU. and no hot pixels.
Whatever it is there is no time to create a consistent dark to remove
it before downloading the picture on the drive when I switch off the
camera.
Andrea T.
|
| |
Date: 31 Aug 2006 13:55:50
From: atasselli@hotmail.com
Subject: Re: Canon Rebel XT, Nikon D50 or Pentax K100 best?
|
Chris L Peterson wrote:
> On 31 Aug 2006 11:58:10 -0700, "atasselli@hotmail.com"
> <atasselli@hotmail.com> wrote:
>
> >On a 180s exposure (to equal yours) at 400ISO the brightest
> >100 pixels of the spot have an average value of 21 ADU!
> >The same dark frame has an average value of 2 ADU.
>
> That's really odd. My 180s dark has an average pixel value of about
> 1200. I'm pulling the data right out of the raw array; how are you
> extracting data from the raw?
>
I'm reading it under Iris. The lowest value is 0 ADU and the highest is
4095 (hot pixel) so the dynamics is all there.
Andrea T.
|
| | |
Date: 31 Aug 2006 21:33:05
From: Chris L Peterson
Subject: Re: Canon Rebel XT, Nikon D50 or Pentax K100 best?
|
On 31 Aug 2006 13:55:50 -0700, "atasselli@hotmail.com"
<atasselli@hotmail.com > wrote:
>I'm reading it under Iris. The lowest value is 0 ADU and the highest is
>4095 (hot pixel) so the dynamics is all there.
I'm suspicious. Such low dark current as you report is not consistent
with any silicon devices at room temperature, nor for Sony's specs on
sensors of this type. I think a baseline is being subtracted somewhere
(regardless of the apparent 0-4095 range).
The numbers I see in the raw array seem to make much more sense.
_________________________________________________
Chris L Peterson
Cloudbait Observatory
http://www.cloudbait.com
|
| | | |
Date: 31 Aug 2006 22:57:22
From: duncan
Subject: Re: Canon Rebel XT, Nikon D50 or Pentax K100 best?
|
Chris L Peterson wrote:
> On 31 Aug 2006 13:55:50 -0700, "atasselli@hotmail.com"
> <atasselli@hotmail.com> wrote:
>
>
>>I'm reading it under Iris. The lowest value is 0 ADU and the highest is
>>4095 (hot pixel) so the dynamics is all there.
>
>
> I'm suspicious. Such low dark current as you report is not consistent
> with any silicon devices at room temperature, nor for Sony's specs on
> sensors of this type. I think a baseline is being subtracted somewhere
> (regardless of the apparent 0-4095 range).
>
> The numbers I see in the raw array seem to make much more sense.
>
> _________________________________________________
>
> Chris L Peterson
> Cloudbait Observatory
> http://www.cloudbait.com
The D50 does have a much lower dark current generation than practically
any other DSLR, it also has much lower amplifier glow. The d50 uses a
modified Sony ICQ413 CCD sensor, which has inherently low dark current.
This sensor also has relatively large pixels for good sensitivity.
Compared to a 300D the D50 has a much better overall SNR. The D50 is not
the same as a D70/70s as indicated by this review (in Japanese, but the
pictures on the last page tell the story):
http://www.eanet.com/kodama/astro/blog/hn200510-nikonreview.pdf
The D50 has a completely revised, low noise processing system, that
gives it an almost unbeatable SNR and DR.
cheers
Duncan
|
| | | | |
Date: 01 Sep 2006 17:12:47
From: Michael McCulloch
Subject: Re: Canon Rebel XT, Nikon D50 or Pentax K100 best?
|
On Thu, 31 Aug 2006 22:57:22 GMT, duncan <dunmunro@shaw.ca > wrote:
>The D50 has a completely revised, low noise processing system, that
>gives it an almost unbeatable SNR and DR.
Review information I've seen on the D50 show significantly higher
noise in the red channel, especially at the 1600 ISO setting. The blue
and green channels are comparable to other DSLRs.
http://www.dpreview.com/reviews/nikond50/page17.asp
Plus, how does the noise spec compare if the IR filter is removed from
both the Nikon and Canons? Is the far red even noiser in the Nikon?
I guess the proof is in the results. Who can point me to a page of DSO
astrophotos taken with a D50?
---
Michael McCulloch
|
| |
Date: 31 Aug 2006 11:58:10
From: atasselli@hotmail.com
Subject: Re: Canon Rebel XT, Nikon D50 or Pentax K100 best?
|
Chris L Peterson wrote:
> On Thu, 31 Aug 2006 10:33:11 -0400, SW <SW@!aol.com> wrote:
>
> >Two different opinions, I was afraid of that!
> >Were these Milky Way shots done with the same lens?
>
> No, but the lenses were similar and were used at the same focal length
> and focal ratio.
>
> >The ISO, exposure length, temperature all affects the S/N.
>
> These were the same in both cases (180 seconds, ISO 400). More to the
> point, however, I shot dark frames with both cameras. Sampling the
> central million pixels, the standard deviation with the 300D was 387,
> and the standard deviation of the D50 was 572, or about 1.5X the noise.
> This was quite visible in both the dark and light frames.
That the Nikons have a higher thermal noise than the Canons is no
mistery (Buil quote the same figure, 1.5x wrt the 350/20D). What it is
lower is the read-out noise and what it is higher is the sensitivity,
especially in the red. That is, wrt to the 350D. Ah, forgot, it has
also a deeper well. Given that you can remove the effect of thermal
noise by a very large extent it shouldn't enter the evaluation, IMO.
>
> My 300D does not require a dark frame, but the D50 has a very bright
> corner that produces a significant gradient. It is necessary to subtract
> a dark frame to remove that (an operation that adds some noise to the
> final image).
My D50 has a weak lighter spot (roughly with a radius of 150 pixels)
near a corner but I would't call it bright by any reasonable meaning of
the word. On a 180s exposure (to equal yours) at 400ISO the brightest
100 pixels of the spot have an average value of 21 ADU!
The same dark frame has an average value of 2 ADU. For all practical
purposes I rarely remove the dark (but I do remove the hot pixels) for
exposures up to 6 min. BTW the read-out noise is less than 10 ADU,
which makes it comparable to high end CCD cameras.
And the addition of nose by using a suitably made dark frame is
risible.
>
> >Perhaps the other sources are measuring the D50 noise with the median
> >filtering in effect?
>
> Possible. Such filtering can make the noise measurements lower, although
> at the expense of also filtering out detail.
I use only the so called "Mode 3" in my tests/exposures.
Andrea T.
|
| | |
Date: 31 Aug 2006 20:05:08
From: Chris L Peterson
Subject: Re: Canon Rebel XT, Nikon D50 or Pentax K100 best?
|
On 31 Aug 2006 11:58:10 -0700, "atasselli@hotmail.com"
<atasselli@hotmail.com > wrote:
>On a 180s exposure (to equal yours) at 400ISO the brightest
>100 pixels of the spot have an average value of 21 ADU!
>The same dark frame has an average value of 2 ADU.
That's really odd. My 180s dark has an average pixel value of about
1200. I'm pulling the data right out of the raw array; how are you
extracting data from the raw?
_________________________________________________
Chris L Peterson
Cloudbait Observatory
http://www.cloudbait.com
|
| |
Date: 01 Sep 2006 11:04:26
From: laura halliday
Subject: Re: Canon Rebel XT, Nikon D50 or Pentax K100 best?
|
Paul Schlyter wrote:
> The failure if ISO film sensitivity values to be meaningful at very
> long exposure time is called "recipricity failure". For several
> traditional films there are data available about the recipricity failure
> factor for various exposure times - if that factor is accounted for,
> ISO values become meaningful even at very long exposure times.
True, but that's not how it works with CCDs.
When a CCD records an image, photons knock electrons in to
the CCD wells during the exposure. When the exposure is complete,
the electronics read how many electrons there are in each well.
Changing the ISO value means scaling this raw value before
digitizing it. The CCD sensor's response to light doesn't change.
> Anyone knows whether the reciprocity failure of CCD chips is worse,
> or better, than for film?
As others have pointed out, CCDs have zero reciprocity
failure. This is well-known. I've verified it myself with my
Digital Rebel, doing the same reciprocity test I do with
film.
Laura Halliday VE7LDH "Que les nuages soient notre
Grid: CN89mg pied a terre..."
ICBM: 49 16.05 N 122 56.92 W - Hospital/Shafte
|
| |
Date: 06 Sep 2006 17:16:34
From: nytecam
Subject: Re: Canon Rebel XT, Nikon D50 or Pentax K100 best?
|
SW Wrote:
> The low end SLRs are getting very reasonably priced and I am ready to
> make
> the plunge into digital SLR astrophotography.
>
>
> Which of these is most sensitive (highest S/N) for low light
> astronomy?
>
> Does the K100 use the same Sony sensor as the Nikon D50?
> Does the K100 smooth out the RAW images like the D50?
>
> SW
Most low end DSLRs use Sony sensors so results, apart from
manufacturer's software tweaks, will be similar. Canon use CMOS
sensors and have been more thoroughly investigated for astro use -
Google Christain Buil's website including removal of IR block filter.
That said I'm very happy with my unmodded Canon Rebel [=300D] and
regularly record stellar spectra which gives an instant look at
spectral response etc in a way that even sensor or body manufacturers
don't - amongst many other astro targets.
Nytecam 51N 0.1W
--
nytecam
|
|
