bird
09-11-2005, 02:50 PM
I know a few people are interested in understanding how to get high detail in their planetary images so I thought it worth starting a thread to discuss.
High-res Planetary Imaging is all about the weakest link. Whatever the weakest link in your setup is, thats going to determine the result because most of the time the weakest link is far and away worse than the other parts of the system. Improving is all about identifying and then slowly controlling or eliminating these factors:
1. collimation and optics
2. thermal control and equilibrium
3. tracking and alignment
4. image acquisition
5. processing
There are things that we can't control (i.e. seeing) but for the rest we have to try and do the best we can so that on nights of good seeing we can take the most advantage.
1. Collimation and Optics
------------------------------
There's plenty of info around about how to collimate almost any type of scope. Don't underestimate it's importance - see this link:
http://www.astrosurf.com/legault/collim.html
Buy yourself a decent set of collimation tools, and make sure the centre spot on your primary (assuming a reflector!) has a nice hole in the middle. The Jim Fly collimation tools are very nice, that's what I use, but there are other ones around.
You need the following as a minimum:
- Cheshire / sight tube
- Laser
- 2x barlow
- Jim Fly reflective centre spot (red triangle, much easier to see than white or black).
If you have a 2" focusser then I highly recommend that you also get the Jim Fly 2" BlackCat and 2" AutoCollimator, just makes life so much easier.
Learn to use the barlowed laser approach to align the primary - this is really nice and is not affected by errors in the secondary or focusser. Then you can use the autocollimator to align the secondary in isolation.
One of the biggest problems to overcome in collimation is to learn how to adjust each of these things accurately, and not get confused by secondary errors throwing out the alignment of the primary and vice versa.
2. Thermal Control and equilibrium
-----------------------------------------
There is another thread around here that talks about measuring the temperature of the mirror and tube. Something like that is needed so that you know what sort of situation you are in. If you're lucky enough to live somewhere with very stable temps than you might be ok, but if you live somewhere with big day/night swings (like me) then you're in trouble.
Tube material also plays a big part in this, as is discussed in the "To Dew or not to dew" thread in this forum. I can reccommend using unpainted aluminium as about the best thermal material. I have my tubes made by Gary Mitchell in Sydney, he's an astronomer who happens to run an aluminium workshop and he does great work.
Only try to get the best images when you know that everything is in thermal equilibrium, otherwise you'll be killed by the tube currents or boundary layer effects. If you build the temperature logger and a cooling unit (described elsewhere) then you can take some control of this, but otherwise you have to limit your imaging to parts of the night when nature has done the work for you - most likely between 2am and sunrise.
Set the scope up outside at sunset and leave it there to equalise. I do this even if I am not going to start imaging until 4am.
3. Tracking and alignment
-------------------------------
I know that Mike The Iceman has done some great work on imaging without tracking, but there are limits to how much detail you can see like this. There is still a lot of motion blur as the image moves through the FOV, and also the tube and mount will still be ringing from the nudges that you give them for several seconds.
To get a serious jump in image quality requires that the scope be tracking so that the target sits in the FOV for at least a couple of minutes. It doesn't have to be perfect, and it can be wandering around a little - that doesn't matter too much as long as you can leave it alone for at least a minute or two before it comes close to one edge. It only takes a rough and ready polar alignment to get this close.
Not only is this good for imaging, but it lets you do a much better job of focussing and judging the seeing. Also lets you use a higher magnification than otherwise.
4. Image Acquisition
-------------------------
Save your images in the highest resolution format available. Avoid saving the images as jpeg or other compressed formats because you're immediately throwing away some of the high res data.
Just keep saying to yourself, "disks are cheap" as all your gigabytes vanish...
ToUCam users:
When the seeing is poor you'll find that higher framerates will give you more good frames and less blur, and when the seeing is good you'll find that lower framerates give you higher quality images. This comes to trial and error and lots of practice.
Firewire Users:
Always capture at the highest framerate you can and store the data in a lossless format. Firewire is either 400 or 800 mbit, fast enough for potentially 100 fps with no image compresion and 16 bit resolution per pixel.
Colour v Monochrome + filters:
You will always get a better result if you use a monochrome camera + filterwheel and RGB filters, no question about it. BUT it's not really worth making the investment until you feel that you're already getting good results with the colour camera. If you're getting "good" images and want to make a step up in quality then it's time to check out one of the firewire cameras from places like www.theimagingsource.com or www.ptgrey.com.
I would recommend against usb2 cameras, as they are very finicky and most of the units on sale at present are MS Windows only. Firewire has a serious advantage that all firewire cams stick to a published specification that lets them work with a range of both free and commercial software, on all platforms. This also makes them future proof.
5. Processing
Processing is all about information, knowing when to preserve information and knowing when you can afford to lose it. In general try to keep all the raw data in a lossless format, and always store your intermediate results from Registax or Astra Image in a high precision format like 48 bit TIFF. Store your final result in the same high precision 48 bit TIFF format and make a JPEG copy to post onto the web.
Try lots of different processing steps, it sometimes helps to have Registax, Astra Image etc all open at the same time so you can go back and forwards between them to try different things.
Processing is a really large topic...
Lets get some discussions going about this stuff... I'd love to see a whole bunch of aussies producing great images!
Bird
High-res Planetary Imaging is all about the weakest link. Whatever the weakest link in your setup is, thats going to determine the result because most of the time the weakest link is far and away worse than the other parts of the system. Improving is all about identifying and then slowly controlling or eliminating these factors:
1. collimation and optics
2. thermal control and equilibrium
3. tracking and alignment
4. image acquisition
5. processing
There are things that we can't control (i.e. seeing) but for the rest we have to try and do the best we can so that on nights of good seeing we can take the most advantage.
1. Collimation and Optics
------------------------------
There's plenty of info around about how to collimate almost any type of scope. Don't underestimate it's importance - see this link:
http://www.astrosurf.com/legault/collim.html
Buy yourself a decent set of collimation tools, and make sure the centre spot on your primary (assuming a reflector!) has a nice hole in the middle. The Jim Fly collimation tools are very nice, that's what I use, but there are other ones around.
You need the following as a minimum:
- Cheshire / sight tube
- Laser
- 2x barlow
- Jim Fly reflective centre spot (red triangle, much easier to see than white or black).
If you have a 2" focusser then I highly recommend that you also get the Jim Fly 2" BlackCat and 2" AutoCollimator, just makes life so much easier.
Learn to use the barlowed laser approach to align the primary - this is really nice and is not affected by errors in the secondary or focusser. Then you can use the autocollimator to align the secondary in isolation.
One of the biggest problems to overcome in collimation is to learn how to adjust each of these things accurately, and not get confused by secondary errors throwing out the alignment of the primary and vice versa.
2. Thermal Control and equilibrium
-----------------------------------------
There is another thread around here that talks about measuring the temperature of the mirror and tube. Something like that is needed so that you know what sort of situation you are in. If you're lucky enough to live somewhere with very stable temps than you might be ok, but if you live somewhere with big day/night swings (like me) then you're in trouble.
Tube material also plays a big part in this, as is discussed in the "To Dew or not to dew" thread in this forum. I can reccommend using unpainted aluminium as about the best thermal material. I have my tubes made by Gary Mitchell in Sydney, he's an astronomer who happens to run an aluminium workshop and he does great work.
Only try to get the best images when you know that everything is in thermal equilibrium, otherwise you'll be killed by the tube currents or boundary layer effects. If you build the temperature logger and a cooling unit (described elsewhere) then you can take some control of this, but otherwise you have to limit your imaging to parts of the night when nature has done the work for you - most likely between 2am and sunrise.
Set the scope up outside at sunset and leave it there to equalise. I do this even if I am not going to start imaging until 4am.
3. Tracking and alignment
-------------------------------
I know that Mike The Iceman has done some great work on imaging without tracking, but there are limits to how much detail you can see like this. There is still a lot of motion blur as the image moves through the FOV, and also the tube and mount will still be ringing from the nudges that you give them for several seconds.
To get a serious jump in image quality requires that the scope be tracking so that the target sits in the FOV for at least a couple of minutes. It doesn't have to be perfect, and it can be wandering around a little - that doesn't matter too much as long as you can leave it alone for at least a minute or two before it comes close to one edge. It only takes a rough and ready polar alignment to get this close.
Not only is this good for imaging, but it lets you do a much better job of focussing and judging the seeing. Also lets you use a higher magnification than otherwise.
4. Image Acquisition
-------------------------
Save your images in the highest resolution format available. Avoid saving the images as jpeg or other compressed formats because you're immediately throwing away some of the high res data.
Just keep saying to yourself, "disks are cheap" as all your gigabytes vanish...
ToUCam users:
When the seeing is poor you'll find that higher framerates will give you more good frames and less blur, and when the seeing is good you'll find that lower framerates give you higher quality images. This comes to trial and error and lots of practice.
Firewire Users:
Always capture at the highest framerate you can and store the data in a lossless format. Firewire is either 400 or 800 mbit, fast enough for potentially 100 fps with no image compresion and 16 bit resolution per pixel.
Colour v Monochrome + filters:
You will always get a better result if you use a monochrome camera + filterwheel and RGB filters, no question about it. BUT it's not really worth making the investment until you feel that you're already getting good results with the colour camera. If you're getting "good" images and want to make a step up in quality then it's time to check out one of the firewire cameras from places like www.theimagingsource.com or www.ptgrey.com.
I would recommend against usb2 cameras, as they are very finicky and most of the units on sale at present are MS Windows only. Firewire has a serious advantage that all firewire cams stick to a published specification that lets them work with a range of both free and commercial software, on all platforms. This also makes them future proof.
5. Processing
Processing is all about information, knowing when to preserve information and knowing when you can afford to lose it. In general try to keep all the raw data in a lossless format, and always store your intermediate results from Registax or Astra Image in a high precision format like 48 bit TIFF. Store your final result in the same high precision 48 bit TIFF format and make a JPEG copy to post onto the web.
Try lots of different processing steps, it sometimes helps to have Registax, Astra Image etc all open at the same time so you can go back and forwards between them to try different things.
Processing is a really large topic...
Lets get some discussions going about this stuff... I'd love to see a whole bunch of aussies producing great images!
Bird