View Full Version here: : RH200 Test Images
23-06-2012, 09:47 AM
Had a bit of clear sky so I collected some data with the wood camera restraint. THe adjustable metal one is being made as we speak.
This first image is a region just to the right of Alpha Cent. 10x8 min exposures in 3nm NII. x1.5 times native pixel size. 13 MB
Note how the stars are fairly good to the corners The adjustable camera holder will eliminate even this small variance. It shows what seems to be a SN remnant?
This second image of M16 region is 3nm NII to red and HA to green. Exposures 10x4 min for each. 8MB
Even at only four minutes exposures really faint NII and HA are being recorded.
The SMC is still low in the sky so I limited the RGB to 15x2 minute exposures. The NII was 7X8 minute. Here is a blend of NII and RGB. 4 MB
I was not going to put these up as they are only tests of the system. They are a good indicator though of what the system is capable of.
Here is an animated gif of a crop from the SMC image enlarged x3 compared to the same area from a mosaic made with the 300mm lens. 2MB
23-06-2012, 10:06 AM
I can just imagine when this thing is maxed out. Comet shots would be unreal.
24-06-2012, 10:25 AM
While these images show the kind of FOV this system produces they do not yet in my view show the image quality potential.
When you look at full frame versions, all the larger stars seem to have a fuzz surrounding them. Does not look like a focus issue either.
There is also a large reflection ring in ACthingy.jpg (http://d1355990.i49.quadrahosting.com.au/2012_06/ACthingy.jpg) . You sure your NII filter in in the right way around ?
The SMC one still has fairly egg shaped stars on top left quadrant as well. Hopefully your steel adjuster takes care of this.
Not criticism just observation. I am hoping you will get this system dialled in as it has HUGE potential. I am just not seeing it yet other than FOV.
These fast systems with wide fov and big chips sure seem like a challenge. When my AG10 arrives I am going to be going through same pain no doubt with my STX 16803
24-06-2012, 10:53 AM
I uploaded a quick image showing what I am talking about in terms of oblong stars as well as the fuzz I speak of around larger stars.
(This is a crop at 100% saved as 90% quality jpg to ensure no artifacts introduced by saving)
24-06-2012, 11:43 AM
You are correct Chris the image train flexes differently depending on where the mount is pointed. The wood frame has eliminated about 80% of this but flexes itself in certain orientations. The aluminium version with adjustable nylon flat pads on all four sides of the camera should fix this.
A solid bracket that holds the filter wheel is also on the drawing board. It will attach from the three threaded holes in the CFW (see picture) to the top of the camera holder. The difference in focus positions due to filters is less than 100 micron so this will not cause any problems.
I am sure this problem of image train flexure occurs in all systems but only becomes obvious with fast systems. The Atlas focuser weighs 1.3kg (3lb), the filter wheel weighs 3.1 kg (7lb) and the PL16803 weighs 3.0 kg (6.7lb). For a total of more than 8kg (18lb) with adapter.
My best guess is I have flexure in the mounting plate of the RH200 and in the filter wheel interfaces. I have measured this with a dial indicator and it smoothly varies by up to 0.5 to 0.7 mm at the sensor position depending on mount orientation. This corresponds to 77 pixels movement vertically from the dovetail! There is also an orthogonal movement to this or laterally on the sensor. In the real world there is no such thing as a rigid body.
This is why so many swear by the use of OAG's rather than put up with the so called flexure of their guidescope. The flexure is in their image train!
The thingy image does not show this as Alpha Cent was at the zenith and the geometry did not change much during the exposure run. I suspect the 'ring' which is produced by Alpha Cent. is due to light leakage around the baffles. I made sure to insert all filters the 'correct' way. I intend to extend the dew shield so the optics only 'see' the area being imaged on the sensor. This should also increase contrast by eliminating scatter of light not contributing to the image. The haloes around bright stars are partially or wholly caused by diffraction from the dew shield inner lip. A laser cut aperture in the dew shield extension should minimise this. I may even make it a square aperture with rounded corners and curved sides aligned with the sensor orientation. The movie industry has known about this for years have a look at the elaborate lens hoods they use.
I am sure there are things I have not thought of. The real fun is taming this system so it produces the results I want. This is a totally new game and we do not understand all the rules yet!
24-06-2012, 02:25 PM
I should also say I process my data that shows very dim stuff down to the noise while not saturating the bright stuff. It is called tone mapping. This will also make faint haloes more noticeable. It is all a matter of taste.
If you know anything about diffraction you will realise that at long focal lengths and high focal ratios it is lost in the noise. The spikes produced by secondary vanes being the exception as the edges are straight. There are two sorts of diffraction, Fraunhofer and Fresnel, one far field and one near field. Just Google it. In a complex optic like the RH200 both apply and should be taken into consideration when designing the optic. This can lead to effects not seen in a pure far field situation such as a long focal length refractor or reflector with a simple plane secondary mirror.
A simplistic way to think of it is Fraunhofer with a telescope and Fresnel with a microscope including an electron microscope..
As I said these are test images to give me an idea of what is really happening not a polished final product for general consumption.
All your insights are gratefully accepted.
We are all just slightly evolved apes groping in the dark and hopefully heading for the light!
26-06-2012, 07:11 AM
I forgot to say that the NII thingy image at 10x8 minutes or 80 minutes total is equivalent to 10x22 minutes or 220 minutes total with an optic at F5 with the same focal length. At f10 88 minute exposures would be needed to get close. With about a third the exposure the time dependant noise associated with any CCD is also cut to a third which further enhances the signal to noise.
This is the whole point of this system. With the very fast optics it is now practical to image very dim stuff with 3nm NB filters. It is not surprising that magnitude 1 stars in or near the image field will be a problem.
Contrary to popular belief a 3nm NB filter does not attenuate a fast wide field.
Below is a screen grab of a NII flat with a line profile. This flat was taken before the system was aligned and held by the wood frame hence the slight assymetry.
04-07-2012, 07:55 AM
Luke Bellani dropped by today with the prototype camera restraint or optical train stabiliser.
There are a few pictures below.
It is not finished yet as the two Al blocks will be adjustable and lockable top and bottom on the camera.
The sides will also have adjustable plates. The camera will slide on nylon pads for focus change.
The frame is mounted on two dove tails for easy lateral and axial centring.
What this device really means is that orthogonality of the sensor to the optical axis is adjustable by tweaking the frame pads. The RH200 if not unduly disturbed will keep it's collimation at factory setting perfection.
vBulletin® v3.8.7, Copyright ©2000-2013, vBulletin Solutions, Inc.