This is my second night using astro snap for capturing bmp frames. Astro Snap's main advantage at this point is the fact that it can capture sub-frames. It does not write the full 640 by 480 pixels away, but only the pixels containing an image of Mars - resulting in much smaller captured bmps as well as much quicker stacking processing time.

I am much more happy with this results compared against the first nights results.

I am using K3 CCD Tools to stack the bmps and select the top 100 frames.

I learnt that making the sub-frame too small is a BAD idea. K3 CCD Tools unfortunately picks the WORST cut off frame to do it stacking on. So I had to manually go through 1000s of frames to create each picture on this page. I deleted the worst centered and cut off bmp files before doing the stacking. This was VERY time consuming. I had to do this frame by frame. If I watched the AVI too fast I miss cut-off frames.

So from tomorrow night I will make the sub-frame larger - this will add to processing time stacking the images but it is better than manually deleted poor frames one-by-one - out of 1200 frames per picture.

Those stacked images that INSISTED on being cut off I still added onto this page. From tomorrow there should hopefully be no longer any images here of Mars being cut off somewhere.

I concentrated on testing with donuts since this is a big problem for me. There are only two relatively donut-clear areas on my ccd chip. While capuring bmp frames, I spend most of my time slewing Mars around the ccd - chip - trying to steer clear of the donuts.

This requires full concentration for the whole time when each set of bmps for one image is captured - 3 minutes at a time.

The other problem I have is that Mars stays in view around 30 seconds most of the time. So while trying to avoid donuts, I must also try to keep it in view and away from the viewing edges - quite a challenge.

If this is not challenging enough, I sometimes use Astro snap to only capture quality frames. So I keep my third eye!!! on the quality settings. When not enough frames are being captured per second I lower my frame capture standard - while keeping the image of Mars away from the donuts and also keeping it away from the side - running of the screen.

Suggestion: if you have one or two astrobuddies to help that would help a lot. One astrobuddy could hold the remote control for the telescope (if you have one) to try and keep Mars from the donuts while the other focus on the quality settings.

Having a 100 dust free webcam is my goal but unrealistic. There will also be a donut somewhere. So last night I 'invested' in learning how bad their effect really is.

I am working on improving my telescope's periodic error and would also like to have Astro snap's autoguiding working. Then capturing frames of Mars will be less stressful and demanding. I want to enjoy myself too. While capturing frames I am so busy I cannot even afford to let my attention wander to look at Mars itself.

So, I talked enough for now, here are those pictures...

For the picture of Mars on the right I kept the webcam image of Mars directly on many dark dust donuts for 100 captured frames. At six points during the capturing, I moved the image of Mars to another area with only dark dust donuts. Finally I kept the image of Mars on a dust free area for at least 100 frames.

The good news about this test is that many relatively faint dust donuts have no influence on image quality - none that I can easily see anyway.

It seems that the total / average stacked image of Mars overwhelmed the faint dust donuts totally.

So when capturing frames of the planet Mars we can rest assured that a faint dust donut here and there is nothing serious. You must not keep the image of Mars on the same spot throughout the whole bmp or AVI capturing session, since then the dust donuts will have enough time to gather enough signal detail on many frames to leave its mark on the final stacked image.

I also did another test run with very few faint donuts. This would have tested whether few faint donuts is acceptable. However, even this test with many, but faint donuts proved that they are not too bad. So I assume the very few faint donuts will have no noticeable effect whatsoever: many faint donuts is still not a problem - provided you do not linger on then the whole capturing session.

Info below is for picture of Mars on the right.

Just notice the range of colors visible in the image on the left - light yellow, a darker yellow, orange and tints of red and maroon. This is in addition to white at the South Pole (top), and blue at the bottom - North Pole. There is a brown collar surrounding the South Pole. I did not do any color processing - these are the colors directly from the raw stacked bitmap frames. The colors might not be realistic/natural, but it sure it pretty!

These images should be compared the those below, where Astro Snap captured only the webcam frames of Mars that were higher in quality.

Unfortunately I did these 2 sets of tests 2 hours apart ... so we cannot compare Astro Snap picking the top frames versus K3 CCD Tools picking the top webcam frames.

Based on this 2 images and the 2 images of Mars below, it seems that clearly Astrosnap should not do the picking since the images below is clearer, sharper and has more detail.

It might be that while Astro snap is doing the picking while it is writing the captured images away, the quality of the images it writes away suffers. More tests will have to be done.

I would have preferred to let Astro Snap do the quality assurance up front. Then at the end of a night, I only have to stack around 200 quality images to get one quality stacked image. If Astro Snap does not do the quality assurance while it is capturing frames, I end up with around 1200 frames for every 3 minutes. Processing this takes 6 times longer than 200 quality frames.

In both cases I had only 600 instead of 1200 frames to process.

These 2 images are the worst 2 images of Mars on this page. So Astro Snap's image selection funtionality does not look too promising at first glance. Even the 2 ugly dust donut images right at the top of this page look much better - detail wise. The donut images were captured without image selection functionality running.