Introduction.
Over
the last couple of years I have been a member of QCUIAG
and involved with the hardware modifications of webcams and video
cameras for use in ‘non-planetary’ astronomy.
18 months
ago it was enough to be able to record any thing
at all of deep sky objects. However
with every development the unconventional imaging devices get a little
closer to the performance of conventional astronomy cameras.
The 1004JG modification by Jon Grove brought us similar
sensitivity to standard astronomical CCD and Etienne
Bonduelle brought the webcams close by substituting the colour CCD
in a webcam for an unfiltered one.
M42 taken with the ICX424 modified webcam through a ETX70.
Here
is a tale about 2 types of CCD cameras converging.
Webcams and dedicated astro CCD cameras are very different beasts.
As costly astro CCD's are optimised for deep sky imaging their
performance on bright targets (Moon and planets) is compromised.
So many (or most) of the best amateur pictures taken of Jupiter,
Saturn and the moon are taken with webcams
The
differences between these two types of camera fall in to a numbers of
areas.
|
Webcams
|
Astronomical CCD's
|
CCD Size
|
¼"
|
1/3" upwards
|
Pixel Size
|
5.6x5.6 um
|
Usually 7-10um
|
Binning
|
No
|
Variable 1x1,2x2 etc
|
Cooling
|
No
|
Peltier Cooling
|
ADC
|
10 bit (8bit images)
|
12-16 bit
|
Unfiltered CCD
|
No
|
Yes (No)
|
Exposure control
|
1/10000 to 1/5 sec
|
1/10 sec to hours
|
Transfer to PC
|
USB fast
|
Parallel usually (some USB)
|
Cost
|
Cheap
|
Expensive
|
The
size of the CCD determines both the field of view and the number of
photons which can be collected. For planets this is not an issue
as there are small bright objects. However, when imaging large
faint objects, a large CCD is preferable.
The
pixel size determines the resolution and to a large extent the
sensitivity of the camera. For planets small pixel size and high
resolution is good while sensitivity is the most important criterion
for deep sky imaging and is met with larger pixels.
Binning
is the technique of combing imaging pixels into larger virtual pixels.
When done in hardware before digitisation this offers nearly all
the advantages of even larger pixels and more sensitivity at the
expense or resolution; again only useful when deep sky imaging.
As
well as collecting signal from arriving photons CCD's also develop
signal from thermal electrons. This dark signal is temperature
and time dependant. To minimise the effect of longer exposures
the temperature should be reduced. Also different types of CCD's
have very different amounts of dark signal.
Planetary
disks tend to be quite bright over their whole surface while deep sky
objects such as nebulas and galaxies can have very bright cores and
outer regions which are many orders of magnitude fainter. To
capture all the 'dynamic range' of a planet in one go a 8 analogue to
digital converter (ADC) is largely sufficient and in line with the
detail our eyes can pick up. However to generate the kinds of
images of galaxies where detail is observable in both arms and core
much more dynamic range is required and 12-16 bit (ADC).
Using
CCD's with built in filters provides instant colour images. This
again is important to planetary imaging where rotation can seriously
limit the time available to collect individual coloured frames.
Of course there is a trade off. The filters on the surface of a
CCD reduce its sensitivity markedly and also impact on its resolution.
Exposure
is controlled differently in the webcams and astro cameras. The
webcams electronic shutter makes very short but accurately timed
exposures possible which can have the effect of 'freezing' the seeing.
However there hardware limits the maximum exposure to 1/5th
of a sec which is too short to start recording much detail in deep sky
objects.
Webcams
contain some very high tech hardware to transfer images to a PC and at
the PC end the CPU of modern computers has been optimise to handle
high speed picture data. By making use of this webcams can
record many more images per sec than a standard astro camera.
The downside is some compression is used which can have a negative
impact on quality.
Finally
cost. Webcams can be manufactured on very large scales as
consumer items which brings down the price. Compared to
astronomical CCD cameras they are very good value. Astro cameras
start at about 10 times the price of a webcam and easily go up to 100
time the price. For this kind of money you do get a nice camera
but its still not as good as a webcam for planetary imaging!!!
So
if you want to take a picture of Jupiter though a telescope than you
need a webcam. If you want to use the same webcam for deep sky
imaging then you loose out to astro camera users in all areas except
cost!
Members
of QCUIAG and Astrocam have been increasing the webcams performance
for deep sky imaging over several years. Progress has come in a
number of jumps which relate well to the categories describe.
The first deep sky capable webcan was the Connectix B/W as modified by
David Allmon. As I don't know too much about this camera
discussion will be limited to stages in the development of the current
Mod.
The
first development was that of video integration. The stacking of
a large number of frames. People important in the development of
this technique include Jürgen
Liesmann and Bev
Ewen-Smith. There developments were primarily aimed at increasing
sensitivity and limits of detection, this we now achieve with
different methods. The main advantage this technique offers the
current mods is to increase image depth or dynamic range. By
stacking large number 8 bit frames a final image can have a dynamic
range of 12 or more bits. The ability to generate images with
sufficient dynamic range is extremely important in deep sky imaging
and provides a method to compete with astro cameras with better
analogue to digital converters.
The
best method to increase sensitivity is to take longer exposures.
The technique used to get the webcams to take controlled
exposure pictures is describe here and
is now very well established and supported by a very wide range of
software.
This
document describes the process of replacing the ICX098 CCD in the
Phillips Toucam, vesta (untested!) and Logitech QC3000/4000 with a ICX424.
The new CCD is a 1/3" device with square 7.4 um pixels. The QE of this CCD are shown in this graph from Mathias
Rimkus
All charts show only typical values. Differences
of more than 25% between several
chips are possible. The ICX sensitivity function is extracted from
absolute Qe statements of other companies and relative diagrams
of SONY. Matthias Rimkus
The
ICX424 in yellow can been seen to have a very similar curve compared
with other CCD used in astronomy KAF400,KAF400E
= Kodak (z.B Audine, SBig), TC256 = Texas Instruments (z.b Cookbook camera),
ICX084,ICX424,ICX285 = Sony (z.B MRxxx cameras).
In addition the ICX424 has less dark frame noise and hot pixels
than the other CCD listed here.
Also
described in the instructions is a method to allow 2x1 on chip binning which can deliver a
2 fold increase in sensitivity. Taken together with development
in the last few years on the software side and the hardware control of
exposures the ICX424 webcam is very much comparable with some
astronomical CCD cameras and still has the advantage of being able to
stream video at 5 frames per sec to help with finding and focusing.
|
Webcam ICX424
|
Astronomical CCD
|
CCD Size
|
1/3"
|
1/3" upwards
|
Pixel Size
|
7.4x7.4 um
|
Usually 7-10 um
|
Binning
|
Yes 2x1
|
Variable 1x1,2x2 etc
|
Cooling
|
Up to you!
|
Peltier Cooling
|
ADC
|
10 bit (8bit images but 12+bits with video
integration)
|
12-16 bit
|
Unfiltered CCD
|
Yes (No)
|
Yes (No)
|
Exposure control
|
1/10000 to hours
|
1/10 sec to hours
|
Transfer to PC
|
USB fast
|
Parallel slow (some USB)
|
Cost
|
Still cheap
|
Expensive
|
M1
Taken with a ICX424 Toucam using 2x1 binning.
As
I very much hope that these developments can form the basis of a
commercial product which should bring down
the cost of CCD astronomy to the non DIYer, I do need to
protect this work
I
order to obtain the instructions to make this modification for
yourself I ask that
you agree to the following terms and conditions.
- You
will not distribute these instructions.
- You
will not place information derived from these instructions into
the public domain. If
you have any improvements or tips you would like to give others I
will be happy to add these to the instructions distributed,
and acknowledge your input.
- You
will not sell a camera modified in the way described for any more
than the value of the parts, unless all extra money is
passed to MSF.
- You
will not sell a camera modified as described as part of a larger
deal.
- The
instructions are provides 'as is' with no guarantees of being
correct, accurate or even working.
You agree to be responsible for any injury or damage
that results.
If
you can not agree to these terms I quite understand.
If
you can agree, and would like the instructions then
download the files by clicking the agreement below.
I
agree to the terms and conditions for the SC3 instructions.
PS
If the ICX424 CCD is not large or sensitive enough for you have
a look at Greg Beekes' ICX414 modification
http://www.cosmicshed.com/
A
ICX424 QC4000 waits for dark
|