Monday, July 30, 2007

20070730 Jupiter

C5 at f/30, DBK 21AF04. Seeing was quite good at 6-7/10, transparency not bad for Hong Kong at 5-6/10, but not as great as early last week.

2205 (GMT+8):-

2209 (GMT+8), this one is captured with BY8 (device) and Y800 (file) and Debayer'ed by Registax, color balance of the camera will be disabled, the saturation setting is also disabled, got to do the color afterward:-

2214 (GMT+8), use RBG24 again:-

2217 (GMT+8), 120% resized:-

I found that if the NGF-S is placed at the 2" visual back, the system will be too rear heavy such that the whole setup could fall down! I need to find a better way, therefore.

From a Philips ToUCam Pro to a TIS DBK 21AF04

Notice the noise level, the ToUCam has a blue-ish background with a lot of random noise, while the DBK despite having the same CCD, has a clean dark background. Working at f/22.

The exposure was adjusted to about the same. Gain is hard to "normalize" across two systems, so it's done by eyes. The ToUCam Pro has a 1/25s of exposure, and around 75% gain while the DBK has 1/23s exposure, and around 80% gain.

We are isolating the effect of frame rate in this test, the DBK can go up to 60 fps uncompressed while the ToUCam Pro can only go up to 10-15 fps at most without introducing too much compression artifacts.

Let's include two raw images from the above AVIs:-

You can see that the ToUCam Pro's one has compression, more noise, and some sharpening artifacts, the DBK's one is cleaner.

Sharing the NGF-S with two scopes

My Borg 45ED II needs the NGF-S for fine focusing and also to reach focus. My C5 can be used without the NGF-S, but it will be better with it especially for high power imaging, however, the C5 is a bit too small for the NGF-S so that its base cannot fully sit on its SCT thread adapter due to the original focuser knob.

I just found that I can solve the problem like this:

1. Borg 45ED II: to use a 2" visual back behind the rough focusing draw tube of the Borg
2. C5: to use a 2" visual back which will raise the base to avoid contact with the original focuser knob

So, to attach the NGF-S to the above two setups, I simply use the SCT thread adapter, together with a 2" to SCT adapter, so now the NGF-S can be placed inside any 2" visual back! In that case, I can easily remove the NGF-S from one scope to another, just like swapping an eyepiece, and the C5 issue can be solved at the same time. I found that my C5 has enough focus travel, of course, the downside is that the C5 will be rear heavy and with the added length of the NGF-S + 2" visual back, it will away from its optimal optical configuration, it will be like f/11 or even a bit more due to the moving primary.

But so far, it seems the best workaround.

Sunday, July 29, 2007

20070729 Sun

There is a new small active region AR965, this time I am too lazy that I didn't open the window glass, for my daughter is also inside doing her summer homework and playing piano (with air conditioner), don't want her to suffer the heat with me. Without her, I won't even turn on the fan to avoid any local seeing issue.

I'm just setting my Borg 45ED II on the 410 head on a small tripod, so no tracking.

This is an excellent example to show that window glass is a great blur filter.

1520 (GMT+8), APM Herschel Wedge, ND3, Baader CaK, DMK31AF03:-

1534 (GMT+8), SM40/BF10, DMK31AF03:-

You cannot even focus correctly across the whole focal plane.

Saturday, July 28, 2007

20070727 Jupiter

Seeing was pretty poor, I rate it as 1-2/10, see below:

Like a Chinese 湯丸 (soup oval dumpling?) boiling in the soup. Transparency was a bit worser than last night, since I only pushed the gain a bit more from default (saved last night), so it's like 4/10.

2230 and 2235 (GMT+8), at f/30:-

Thursday, July 26, 2007

20070726 Jupiter

Seeing 1-2/10, mostly 1/10, and transparency 4-5/10. C5 at f/30, DBK 21AF04. I pulled the camera out and observe with a 7mm Ortho, pretty great view indeed, I don't feel much difference with a C8 at this image scale.

Taken at 2244 (GMT+8), notice that the atmospheric dispersion is so serious that RGB re-align does not help much as you can see from that Jovian satellite in the view, I shall take Jupiter earlier but my daughter didn't sleep early enough:-

2246 (GMT+8) and 2252 (GMT+8):-

The atmospheric dispersion is very serious indeed.

Wednesday, July 25, 2007

20070725 Jupiter

Didn't take Jupiter for a few days, so try again tonight. I did try to use f/30 at first, but found that it couldn't really reach focus, I blamed the seeing and went back to f/20, and it could reach focus easily, I found it not as bad as f/20, so I wonder why. Anyway, I took several f/20 clips.

Going back to f/30, the problem came again. I wonder why 50% more of image scale would be that far worse. I checked the CCD, remove the dust, turn the barlow, and I finally found why. The NGF-S reached the end of its focus travel. It could reach very near the optimal focus, but it's a few hairs away. I turned the C5 focus knob, and the fine tube again with the NGF-S in the other direction, wow, it's sharp again.

It's so confusing that since I could even make out the Jovian satellites when the focus travel is just a few hair away, and my near parfocal eyepiece also told me that it's close to focus indeed. Luckily, I got it found before it's too late. So, I've a few f/30 clips too.

ALl taken with 1/15s, f/20 clips at 550 gain and the f/30 clip at ~800. Transparency dropped today to around 5/10, and seeing was 3-4/10, I estimated it at 2/10 before that, but it's actually 3-4/10, not very bad indeed.

2147 (GMT+8), f/20:-

2149 (GMT+8), f/20, alternative color style:-

2150 (GMT+8), f/20, 150% resized, this one is sharpest of the batch, so I resized it to 150%:-

2203 (GMT+8), f/30:-

2205 (GMT+8), f/30:-

2207 (GMT+8), f/30:-

2208 (GMT+8), f/30:-

Additional Remarks: on the next night I found that the NGF-S cannot fully sit in its SCT thread adapter in a C5 OTA, there was a gap there since the focuser has a fat base and so preventing the focuser from sitting all the way in. This could cause collimation problem.

Mount stability

I still remember that several years ago when I use my Giro mount with my C8, the mount is rock stable and it didn't move a bit when I turned the focus knob. Last night, I was using the LXD55 with my C5, but the image jumped crazily when I touched (not turned) the focus knob, it takes slightly more than 1s for it to settle down. Therefore, to do focusing, I turn it roughly to focus the image, and the tap on side of the focuser knob to fine tune it, the action is repeated again and again until the image looked sharp. Problem is seeing, you will need to wait for around 10s at least to wait for better seeing to know whether the focus is correct for not, and this has to be repeated many times to verify and find the best position.

With my NGF-S, the problem can be solved entirely, but sometimes, I'm too lazy to remove it from my Borg 45ED II and to re-attached it to the C5. Somehow I think I shall leave the NGF-S on the Borg during weekend when I can do solar imaging, and put it on my C5 during week days, for doing planets or lunar imaging.


The sun is moving to another position, maybe I can use my C5 for indoor solar imaging soon. Previously, there's only very few sunshine going directly inside my playing room (home@walnut), but now it's getting more and more.

Also, the bus station where it was completely flooding with sunshine in the morning (the bus station which I waited for a bus to office), now it's hidden in the shade of a close by building, the sun has turned!

Let's hope.

Tuesday, July 24, 2007

20070724 Lunar Observation (DMK+C5)

DMK31AF03, C5 at f/30, transparency 7-8/10, seeing 2-3/10. Seeing was very bad such that I can see the wave front of heat! It's like ripples in form of a curved line, scanning through the image stream, not exactly periodic but it came again and again, ruining a lot of frames.

2149 (GMT+8), near the center of the image, we have a swallow crater called W. Bond, and the small but deeper one on the left corner of it is called Epigenes, there was another similar one on the far right, and it's called C. Mayer:-

2151 (GMT+8), the famous 20 cents (Hong Kong):-

2153 (GMT+8), Clavius:-

2158 (GMT+8), Plato:-

Monday, July 23, 2007

Debayering in TIS DBK/DFK cameras

DBK/DFK allows us to reconstruct color information (debayer) in different places, but it is not well documented (** see below), I spent a whole afternoon to figure them out how it works and I hope it will be useful to you. Here're the three possibilities:

1. On camera: it's done on the hardware, I call it the "device codec" which one can set in the device property.

By setting to Y800/BY8, the Debayering is switched off and you can do 60fps, you can store the data in Y800 and the file size will be smallest. By doing that, Registax can handle more frames with its ~2G limit. With this consideration, even if you are not doing 60fps, you still benefit from this approach for you can capture a longer AVI and can still be processed by Registax in one shot without cutting the AVI into several 2G pieces.

By setting to UYVY, the Debayering is switched on hardware level, and you can do 30fps at most. Data size is 3 times as in Y800/BY8.

2. In IC Capture: if you choose to switch off Debayering in hardware by setting Y800/BY8 in device codec, you can do it by the image capturing software. IC Capture allows you to Debayer with different algorithms on the fly.

For slower machines, Debayering by IC Capture might lead to lower frame rate, since CPU cycles are used for Debayering. Debayering in IC Capture will generate the same file size as in UYVY device codec and RGB24 file codec.

Debayering by IC Capture enables one to generate color AVIs in RGB24 but still enjoy the power of 60fps.

3. In Registax: it's done by Registax during image processing. Debayering are disabled in the camera and also in IC Capture, and this job can be left for Registax.

In this situation, you can get the smallest file size by saving in Y800 file codec, and it saves CPU cycles for Debayering during image capture, too.

Due to the smallest file size generated, one can have the most number of frames in a single AVI file and still within the ~2G file limit of Registax.

** I was wrong, it's well documented now, check below:

A discussion of frame rate and shutter speed

I've received a couple of emails to ask about how to get good planet images. Frankly, I ain't very skillful but if you care to know how I think, here it is.

Let's forget what you hear about frame rate, shutter speed, or whatever in your mind first. Now, let's start from scratch about what contributes to a good planet image:

1. High quality raw frames: it simply means clean and correctly exposed image. We do stacking for this reason, to reduce noise.

2. More high quality raw frames: hey, please don't trick me, no I ain't. Some people said higher fps is always good, because we have more frames to choose from, to freeze poor seeing, etc. However, I would say it's only part of the story. What we want is not millions of frames, but instead, we want hundreds of good frames!

Under pooor seeing, one would say, we need to get more frames to choose from, to freeze seeing. But this usually mean that we will need to use faster shutter. If you want to use 30 fps, you will need 1/30s or faster. That means underexposure or higher gain. Underexposure means less signal, which we won't want. Higher gain means more noise, which we won't want too.

So, how to combat seeing? Let's watch the following video:

Very bad, right? One can say we can get more frames so that we can pick the best from! So, grabbing more frames means a higher chance of good frames. That's right, but consider that, we are collecting a lot of rubbish frames too. And the worst thing is, those best frames are compromised at the same time, for they're either underexposed or noiser than what we can get with slower shutter and lower gain!

The net effect is, we generates a lot of rubbish, and we compromise also our good frames!

One would say that we have no choice, but it's the best compromise under poor seeing condition. I would say, no, that's not. If you spend time in the night sky, you will know poor seeing is not a night long event usually, you can get good moments sometimes. Patience is the key. My friend Mr. Eric Ng said that, he ever waited for four hours before first clicking the record button, for what? to wait for the best seeing.

So if you want to get image, the key is patience. If you just want to get the best image without patience, go for the compromised route, but you won't get the best possible image.

Sunday, July 22, 2007

Casio UV/IR blocker testing: M7

Fujinon 75mm f/1.8 wide open, 13.75s at 520 gain, 28 frames stacked, DBK 21AF04. This time with UV/IR cut which is removed from my Casio QV2800, the image looked sharper and free of those large blue/red halo, I shall try to stop down the aperture next time to see if it helps or not:

20070722 Lunar Observation

Seeing was very poor at 1-2/10, transparency around 6-7/10. I used C5 at f/30 for high power close up, DMK 31AF03. Full disc shot was done by Borg 45ED II at prime focus, DBK 21AF04 with "Casio" IR/UV blocker.

1939 (GMT+8), Upper one is Aristoteles and the lower one is Eudoxus, the one very swallow one on the left to Aristoteles is Egede.

1944 (GMT+8), Upper Right is Cassini which is the main target of this shot, the large empty "volcano" like hole near the bottom is Aristillus, the bright patch on the top left is Mons Piton.

1948 (GMT+8), Cyrillus and Theophilus, 2 points alignment, feather = 8 used, but the boundary still left.

1950 (GMT+8), Vallis Alpes, actually, it's just the left of the first shot (1939).

Finally, 2002 (GMT+8), taken by Borg 45ED II with DBK 21AF04 with "Casio UV/IR" filter installed.

Lunar imaging is prohibitly hard due to poor larger scale seeing condition, with a small telescope, planet imaging is still possible since it's not very hard to get planet size good seeing cell, but lunar imaging is another story, since a single crater can be way larger than a planet (angular size).

Adding IR/UV filter to DBK

After taking some deep sky testing shots with the DBK, I found the UV/IR signal is ruining the image, making red and blue halo to nearly all stars.

So, I placed a small piece of UV/IR blocking filter to the DBK in front of its CCD, this filter is removed from my Casio QV2800, it matches perfectly.

I've tested it with day time object and it can reach infinity focus, and white balance is correct, too. Originally, I fear that by adding a UV/IR filter on the CCD, some C mount lenses might no be able to reach focus at infinity, luckily, it still can. That UV/IR filter is more than 1mm thick, so the optical path length should be more than 1mm which could give problem to C mount lenses.

Hopefully the sky would be good enough tonight for me to test again.

20070722 Absolute Minimum

The sun is absolutely silent today, with only a few small prominence, very few short and subtle QRF, no active region except a small brighten patch, nothing.

Combined with two exposures, one for surface detail and one for prominence. Borg 45ED II, SM40/BF10 at prime focus, DMK31AF03.

Another shot with an alternative method:-

Saturday, July 21, 2007

20070721 DBK M7

Transparency is quite good tonight, so why not give a try on deep sky? DBK 21AF04 has smaller chip (1/4") than DMK 31AF03 (1/3") and so the field of view is more restricted, and also due to its bigger pixel size, the image scale is also smaller. Anyway, it's in color, more convenient for lazy people like me. I don't think that I can spend much time in DSOs with my current lifestyle, so fast food type is good to me.

The first attempt is by Fujinon 75mm f/1.8 wide open, 11.55s at 350 gain. Chromatic abberation can be seen easily, all the stars have blue halo. Since DBK has no UV-IR block, maybe it's out of focus UV and IR? 32 frames are obtained.

Second attempt is to close down it, to see if the problem can be reduced or not. I close the aperture stop by 2-clicks (i.e. f/4), gain is now 605, exposure is 16.36s. Chromatic abberation persists, more or less the same as before, so it's more like IR/UV issue. 10 frames acquired only, for it's more like an experiment:-

Third attempt is a Cosmicar 150mm f/3.2 wide open, 13.75s at 655 gain. I think I need to use an UV-IR cut for deep sky imaging.

20070721 DBK Jupiter

Again, seeing is 3-4/10, transparency is 7-8/10, first shot is 1/8s and ~500 gain, second one is 1/15s and ~700 gain, a test between frame rate and exposure, third shot is re-focused, I didn't use NGF-S tonight.

2244 (GMT+8), f/30:-

2246 (GMT+8), f/30, my best C5 Jupiter ever, beat even all my previous C8 shots hands down, maybe due to the fact that I'm doing imaging indoor, more aperture does not really help, but optimizing smaller aperture is even more fruitful:-

2249 (GMT+8), f/30:-

20070721 PST experiments

The sun went to a very bad position from inside my home and so I can't really test my C5-PST setup. Until today, it seems not too bad, but real observation is still impossible, so I just try it despite part of the aperture is blocked.

I placed the Cheapy-ERF on the C5, and test the temperature at prime focus, it's not even warm, the Cheapy-ERF is very efficient to cut down the temperature of the focused solar disc. At least my C5 won't die due to the exccesive heat.

IR-UV has no guarantee here.


Second experiment was to test with a Herschel Wedge. Herschel Wedge cannot be used with compound telescope normally, since the unfiltered light collected and focused by the primary mirror, will kill the secondary mirror and the baffle tube. However, with the Cheapy-ERF in place, the above is proven safe.

I placed the Herschel Wedge there with the Cheapy-ERF, and then a ND3, I found it works perfectly. Of course, the optical quality will depend on the quality of the Cheapy-ERF, so whether it's better than a piece of Baader film is yet to be verified. However, at least it works and I can throw away my Baader film based filter.

20070721 Solar

Got good sky, it's very blue, very hard to come by in Hong Kong these years, but seeing is no good, I guess 2-3/10 in best moments, transparency is superb at 8-9/10 with little cloud around. It's a great prominence day.

Full disc captured by Borg 45ED II at prime focus, and then go for high power for prominences with 3x barlows, I usually use 2x but this time I used 3x due to the good transparent sky. High transparency sky enables me to use faster shutter and lower gain, thus lower noise at high effective focal ratio.

1417 (GMT+8), full disc shot, Borg 45ED II at prime focus:-

1418 (GMT+8), full disc prominence shot, Borg 45ED II at prime focus:-

1422, 1423, 1424 (GMT+8), close up of a dragon like prominence, Borg 45ED II with 3x barlows:-

The following two shots are taken at 1425 and 1427 (GMT+8) respectively, showing the surface detail and the dragon like prominence:-

Featured prominences of the day, 1430, 1432, 1434 (GMT+8):-

Two close up shots targeted on those subtle, small and short QRF, taken on 1436 and 1438 (GMT+8):-

Finally, a CaK full disc by Borg 45ED II at prime focus, APM Herschel Wedge and Baader CaK, it shows nothing essentially, taken at 1442 (GMT+8):-

The prominences of today are pretty big indeed, but my SM40 is really too small to really kick up the image scale.

Friday, July 20, 2007

20070720 DBK Jupiter

First two are by 2x, then by 3x. Seeing was 3-4/10 at first, worsen to 1/10 sometimes and went back to 2-3/10 finally, transparency was 5-6/10 with a lot of cloud floating around.

Taken on 2219 (GMT+8), f/20, 150% resized:-

Taken on 2237 and 2247 (GMT+8), f/30, 150% resized:-

Taken on 2249 (GMT+8), f/30, 100%:-

Finally, I packed up my camera, took out my 7mm Ortho to observe briefly, the Jupiter ball is so contrasty and lovely! Don't forget to have a look.