Tuesday, April 29, 2008
Those 60mm scopes are having singlet objective, which is quite hard to understand at first. But then I guess I know the reason. Since this scope is offered for both H-alpha and CaK, I believe that a doublet which corrects (like spherical aberration) at both wavelength would need to be custom made, that would mean a special cost.
By using a singlet in the objective, and then to use the collimator lens before the etalon to correct the aberration, it would be cheaper to do. Why?
Firstly, the collimator lens is to make the light ray parallel before passing the etalon, that means it will be slightly different in design for different wave length, since different light focus slightly different and thus the lenses should be different for H-alpha and CaK. Therefore, this lens has to be custom made anyway.
And I guess you know what I mean here, since the collimator lens is to be made specifically for H-alpha and CaK, it's thus more economical to design them also to correct for other aberration!
So, let's break down the situation like this:-
1. If you use a doublet as the objective, you will need a special doublet which corrects for both H-alpha and CaK. And in that case, you will also need special collimator lens for H-alpha and CaK respectively. Therefore, we need to have three different custom made lenses.
2. If you use a singlet as the objective, you can have the same one for H-alpha and CaK, since all the aberrations were left to be corrected by the collimator lens. And in this case, you will need a special collimator lens for H-alpha and CaK respectively, but it's just the same as the above. Therefore, we only need to have two different custom made lenses.
This means cost saving. Of course, it might not be the case, I just guess.
Saturday, April 26, 2008
The sun has moved to another location with respect to my home window direction, I've spent over 10 minutes to invent a new way to enable my small Borg 45ED II to point at the sun, guess what, it's to use a binocular L bracket to extend it a little bit further out, and since it's 90 degrees, I can rotate the RA of the LXD55 so that it's pointing more outside, therefore, the 2cm extension does not help much, but with the 90 degree angle, it helps for more than 10cm in total, I believe. Since a L-bracket is used, you cannot expect to use anything heavier than my small Borg. It's truely a very versatile scope for environment limited people like me.
Here're the results:-
1352 (GMT+8), as you can see the artifacts left by the moving cloud, but they're somewhat averaged out; the nice filament is still visible here:-
1353 (GMT+8), there's a very nice prominence, but the sky is too bad for me to push up the image scale:-
I tried to test my Borg 45ED II with different barlows and eyepieces.
My first target is a 2" Pentax SMC XL 40mm... the Pentax is so big that it's bigger than the objective end, the field is nice and bright as well as flat, no vignetting to my eye. Very sharp and no color, it's 8.125x...
Next is my 13mm Nagler Type 6. It reaches focus in the same 2" diagonal as well, the field again is nice and bright and flat, too. However, the Nagler is not really a day time eyepiece, the view is not easy to hold and you can see blackout rather easily. There were some color at the edge of very high contrast object, like the edge of some white wall, it's very low, and you can't see it unless you specifically look for them. It's 25x now.
Then I inserted a 5x Powermate there, it's nearly parfocal, I mean when I drop the 13mm Nagler back there, it's basically in focus. I tweaked the focus with my JMI controller, and it's very sharp and contrasty. There is no obvious color, but the exit pupil is getting small so that you can see floaters. Image is still solid, no sign of breaking down. Tweaking the focus reveals a tiny amount of image shift from the NGF-S, not a big problem here. It's 125x, i.e. 0.36mm exit pupil.
I went crazy and I inserted a 3x barlows between the 5x and the Nagler. To my surprise it's like parfocal too, tweaking the focus a little bit and the image is still quite solid!!! I can still see detail in the image, of course, it's too dim now and all the dust, floater jumped into the view. Again, no more color than before. It's 375x!!!! For that small objective, it means 0.12mm exit pupil only!
Modern objectives are so great that you can push as far as your eyes allow!
Case A: The simplest one, debayer'ing is done at hardware level, and so the data coming from the camera is already in full color. It uses the most Firewire bandwidth, as well as hard disk space and bandwith, but this is the simplest one. Since debayer'ing is done by the camera, it's good for those who has slower CPU.
Case B: By choosing Y800 for the device codec, the hardware sends original data which is not yet debayer'ed, that saves some Firewire bandwidth, and then the debayer'ing is done at software level by IC capture (see the pull down menu), and since the data stream is already debayer'ed, you will also need RGB24 to store the information. It uses less Firewire bandwidth, but it uses the same hard drive bandwidth and space as Case (A). In that case, more CPU cycle is used for Debayer'ing, so if you got a relatively slow CPU, you will get less maximum fps.
Case C: We have chosen Y800 for the device as above, so the data coming from the camera is still not yet debayer'ed, and therefore, it uses less Firewire bandwidth. And Debayer'ing is also disabled in IC Capture, meaning that Debayer'ing is not done on-the-fly as well, and so it requires less CPU cycle. And since the data is not debayer'ed, we can use Y800 to store the data as well, that uses less hard disk space as well as bandwidth.
As you can see, this combination uses the least system resources, so if you got a slow CPU, a slow hard disk, this combination will give the highest possible fps. Debayer'ing must be done on Registax, however, but this is not a problem. Since Registax needs not to do things in real time, a slow CPU won't affect the quality or the frame rate (obviously).
Remember that you can only use Y800 as the storage codec in the last case, since for case A and B, the data is already debayer'ed and in color, and thus, if you use Y800, you lost information and it couldn't be recovered afterward.
Thursday, April 24, 2008
某天放工回家的時候，收到 TIS 職員的一封電郵，內容是說他們開設了一個中文版的網誌。當時實在感到有點意外，因為畢竟對於業餘天文觀測來說，大部份人都是來自歐美地區。因此，一般國外生產商都不會提供中文版的文檔，更惶論以中文提供的售前或售後服務！
我自 2006 年 3 月起開始使用 TIS 的產品，那時他們還未有專注於天文上的應用。基本上他們的對像都是工業用途的，但就硬體規格來說，看來就相當適合天文應用。我的第一部 TIS 相機是 DMK 31AF03，那時只提供到 30s 曝光時間，但對於我來說，我的主要目的是窄頻太陽攝影，所以就是有 1s 曝光已經足夠有餘，而 30s 曝光對我來說，也可以讓我一嘗拍攝較光亮的深空天體。另外，31AF03 也提供比一般 webcam 更高的解像度，亦有較大的 CCD，實在非常適合太陽攝影！當然，30 fps 未經任何壓縮的能耐都是極其吸引！就衡量了這些東西之後，我就訂購了一部。那時他們的代理網絡還小，我自己是經由一間澳洲的代理買下的。
當然，這部 DMK 31AF03 並沒有叫我失望，相反，他更叫我喜出望外！這部小小的藍盒子，雜訊很低，而 frame rate 亦真的能夠達到每秒三十格！坦白說，對於我的年老 PC 來說 (Duron 1G, 六歲多了!) ，這實在有點意外！Firewire 的能力實在不能低估！同時間為了一嘗拍深空天體的願望，我特別買了一些閉路電視攝像鏡頭 (CCTV lens)，它們都很便宜 (從 ebay 買的, 數十元美金一支)，而光圈也很大，十分便利於拍攝。正因我也沒有很好的赤道儀，又沒有很黑的天，基本上 20s 以內畫面就全白了，所以當時的 30s 極限也足夠有餘。我發現在市中心地區，一部 DMK 加幾支細小的攝像鏡頭，比起用肉眼 + 8" 鏡還要好很多！這算是意外驚喜！
相關的相片可從這個網誌中看到，只是選 label 中的 DSO 就可以了。
後來，TIS 說可以為我的 DMK 31AF03 升級 firmware 做 DMK 31AF03.AS （要檢查 serial number 以確定硬件是否可以支援)，這實在是一個好消息！但這樣長的曝光時間我還沒有機會試。單從 dark frame 來看，似乎也很不錯呢！
在 2007 年 6 月我購入了一部 DBK 21AF04.AS 來拍行星。因為我一向都是在家中拍攝，所以大氣穩定度都不太好。以前我是用 C8 + ToUCam Pro 來拍的，拍的也可以，但水準都只能在普通以下。改了用 DBK 後，我第一次拍的木星已經要比我以前拍的要好很多了！但有一點是更重要的，就是那次我只是用一支 C5! 沒錯，那一次用 C5 拍的要比以往我們 C8 拍的好！當然，我要強調的是，那是因為我的拍攝環境所限，未能完全發揮 C8 的烕力吧了。但無論如何，DBK 確能讓人更輕易的發揮出望遠鏡應有的水準！
Thursday, April 17, 2008
First of all, I was told that the 450D can be controlled completely via a single USB cable now. That means you don't have to buy a separate cable to control the shutter in bulb mode. It saves $80-$100 USD., and of course, it also saves the trouble. Good news indeed.
Secondly, the Live View capability will remove the need for focusing software as well, that's another like $50 USD.
Thirdly, the new kit lens offer decent performance with IS, so I've no urgent need to buy a 10-22mm which saves like 5K HKD.
Next, from the limited results now we have on the Internet, the noise of a 450D improves. Actually, people find it the best for 300D and 350D but then when going to 400D, the noise was getting worse. It's a good sign to see that it's back on track now.
Finally, this will be my first purchase:-
1. Canon 450D with the kit lens for daily, and wide field shots
2. Canon 50mm f/1.8 to play with large aperture
3. Canon 200mm f/2.8 to play with medium focal length deep sky
4. One more set of backup battery
5. 4G class 6 SDHC
Here is a link for more information:-
Wednesday, April 16, 2008
I'm glad to hear that. TIS is a really special company. I still remember two years ago I was among the first few amateur astronomers to start using it, and soon after that, they are fully prepared for the astronomy market. Not just they are now including a 1.25" adapter out of the box, they also support longer exposure, I have even received a new version of IC Capture where you have a reticle overlaid on the screen, etc.
Now, they even take care the Chinese community. Great product, great customer services!
Saturday, April 12, 2008
I tested the frame rate of my newly arrived USB 2.0 on my slow Sony Vaio U3, I could get around 12 fps when the built-in hard drive is used to store the AVI.
Then, I tried to use something else which should be faster than my hard drive, what I used was a 2G SD card inside a USB 2.0 card reader. I guess a fast USB thumbdrive should perform better, but at least it should be faster than a mechanical hard drive.
The result is 7.4fps... surprised? Not really.
With the DMK 21AU04.AS, the USB 2.0 bus was nearly saturated, especially given a slow CPU. The USB 2.0 specification is far more CPU demanding than that of the Firewire interface, and therefore, the DMK already sucked up all the juice.
To add one more USB 2.0 device to the same bus will only make the situation worse.
Therefore again, if you have a slow computer like I do, go for Firewire version, it will offers better performance as my former experiments indicated.
Wednesday, April 09, 2008
I shall test it for white light with my APM Herschel Wedge, and if the color is more than desirable, I guess a simple color filter, like a green one is going to heal most of it. And I guess the performance of it with a Baader CaK should be outstanding.
If that's the case, this would be useful for H-alpha as well. Of course, I shall test it with my Cheapy-ERF + PST etalon module combination, I guess we will need a barlow both to reach focus, and to get slower focal ratio for the PST etalon module. Finally, if it works out, but not so well due to the internal reflection issue, I shall go ahead and buy a 75mm Lunt's filter. With 75mm, and at 820mm focal length, it's going to be powerful enough.
Of course, if it works with the Cheapy-ERF, that would be the best in terms of resolution and also the cost.
And it will work as a wide field deep sky instrument, it just won't work for planets and the moon.
Sunday, April 06, 2008
Telescope: C5, Borg 45ED II
Filter: PST/BF10, SM40/BF10
Camera: DMK 31AF03
The shots taken by the Borg were all too bad, zero detail essentially so I didn't process them. So, all the shots belows were taken with the C5/PST based setup.
Saturday, April 05, 2008
Telescope: Borg 45ED II
Filter: Coronado SM40/BF10
1336 (GMT+8), I truely love this full disc shot myself:-
1339 (GMT+8), with 3x barlows:-
Friday, April 04, 2008
Seeing 1/10, transparency also 1-2/10.
Telescope: Borg 45ED II
Filter: APM Herschel Wedge with ND3
Camera: TIS DMK 21AU04.AS
Since the chip (1/4") is smaller than the DMK 31AF03 (1/3") and it does not even enough to cover the full disc in one shot, so this is a mosaic of two images stitched together:
Finally, the sky cleared up a little bit, but still very foggy feeling, when I looked into the eyepiece, it's like looking through muddy water, anyway, here's a H-alpha shot with SM40/BF10, again two frames stitched together to form the full disc:-
Notice that very nice prominence bended back to the solar disc and presented as a filament there. Poor sky it was, otherwise it should be a very nice view!
- Duron 1GHz
- 768M memory
- 7200RPM 160G hard drive
- Windows XP
Frame rate test:
(Test 1) DMK 31AF03.AS, i.e. Firewire, 1024x768
=> full 30 FPS without any problem
(Test 2) DMK 21AU04.AS, i.e. USB 2.0, 640x480
=> only 24-25 FPS
I have removed all USB devices before the test, similarly for the Firewire test, the DMK is the only device attached. All test conducted with IC Capture 2.0
I guess it's because USB 2.0 requires more CPU intervention and thus the slower CPU has become the bottleneck. Therefore, if you have a relatively old machine, go for Firewire version is better.
(Test 3) DBK 21AF04.AS, i.e. Firewire, 640x480
=> 56 fps!
When compare to (Test 2), the speed for Firewire is nearly double of the USB 2.0 counterpart! Therefore, Firewire is superior than USB 2.0 in slower systems.
When you plug in the camera, the system detects it and the hardware wizard comes out asking for driver. Normally, you shall click cancel and use the CD that comes with the camera to install the driver. However, the driver installation software failed to detect the camera and keep asking you to connect the device!
The system detects the camera as a USB device, however, it should be detected as a video device.
You use hardware wizard to install the driver. However, don't choose automatic but instead, go for manual install. When you are asked to choose the type of device, choose video device instead of USB! If you choose USB, the wizard will again fail to detect the device and thus failing to find the suitable driver for it. However, if you choose video, it will detect and proceed to install!
Thanks Stefan from TIS, he's the best technical support people ever!
Thursday, April 03, 2008
Therefore, let me report the performance on my Sony Vaio U3 first, here's the configuration:-
1. Crusode 933MHz CPU
2. 512M memory
3. 20G hard drive
Since it didn't come with USB 2.0 port, I used a PCMCIA USB 2.0 card and it works without problem.
I was using Y800, and the highest frame rate I managed to get is 12-13 fps, I believe that the bottleneck should be the hard drive and the CPU.
Wednesday, April 02, 2008
After plugging in the hardware, I got the following with
[695584.136000] Linux video capture interface: v2.00
[695584.180000] uvcvideo: Found UVC 1.16 device DMx 21AU04.AS (199e:8101)
[695584.180000] uvcvideo: No valid video chain found.
[695584.180000] usbcore: registered new interface driver uvcvideo
[695584.180000] USB Video Class driver (v0.1.0)
Then we will need to install the driver:
#sudo apt-get install build-essential linux-headers-generic
Then remove any existing driver:
#sudo rm /lib/modules/2.6.20-15-generic/kernel/ubuntu/media/usbvideo/uvcvideo.ko
Now, you're ready to download and compile the driver:
#sudo wget http://www.astronomycamerasblog.com/wp-content/uploads-extra/linux-uvc_trunk_arne.tar.gz
Then extract the tarball:-
#sudo tar -zxvf linux-uvc_trunk_arne.tar.gz
Get inside the directory, compile it and install it by:-
#sudo make install
If everything work fine, the driver is ready. Now, it's time to install the software. You can use apt-get to install the software, simply check out the instruction here to add the repository first:-
Remember to complete the authentication part. Once it's done, do the following:-
#sudo apt-get update
#sudo apt-get install ucview
So everything is ready, do this to start the application:-
This is my third blue ice-cube, and it's the first USB model.
I tested it with a machine with the following configuration:
- Celestron 2G CPU with 512M RAM
- Windows XP Home
- Maxtor 60G 5400 RPM hard drive
Installation is straight forward, and IC Capture.AS fired up without problem. I set the codec to Y800, set the shutter to faster than 1/60s so that I can test its frame rate.
After 10.1s, the I got 481 frames, i.e. around 47.6 fps, I tried several times and the figure is around 44-47 fps. Not bad for a 5400 RPM hard drive.
Since I've no lens/telescope on hand, what I can do is to test dark frame noise. If I set gain to zero, I got absolutely black screen except 7 hot pixels. If I set gain to maximum, the noise is not too bad for most of the screen, but I could see there are quite a lot of amplifier glow at the upper left corner, and this is similar with my DMK 31AF03.AS as well as my DBK 21AF04.AS.
I am going to test this camera again in my home desktop which has a Duron 1G plus 768M memory but with a 7200 RPM hard drive, so I expect the frame rate could be a bit higher. And then I shall test it with my notebook which is a rather old Sony Vaio U3, and I guess I could get around 30 fps at least.
Thank TIS for this camera!
Tuesday, April 01, 2008
29 July/Tue: Gather at Ürümchi
30 July/Wed: Go to Yi Wu by bus
31 July/Thu: Stay in Yi Wu and have some sight seeing
1 August/Fri: Solar Eclipse!!! And travel back to Ürümchi by train
2 August/Sat: Back to Hong Kong from Ürümchi
So now we will have two nights in Yi Wu, accommodation will be those modern Mongolian style tent.
HKAS will organize a talk on solar eclipse imaging in mid-June.