View Full Version : I just received my DMK 21AF04.AS and.....

June 20, 2007, 20:55:21
...I would have a couple of questions regarding the camera operation. Before my question, let me describe my setup:
1- Scope: LX-200 10" f/10 Classic with motorized focus
2- Power Mate: Teleview 2.5x
3- Laptop: HP, with P4 3.6GHz processor and firewire interface built-in.
4- Camera: DMK 21AF04.AS (the one...)

I think this is everything. If any other detail needs to e clarified, please let me know. Now my questions:

1- The contrast slider action in the capture tool bar seems to me a little odd as it just adjusts the background brightness not changing the actual image's contrast. Is this a situation normal?

2- When I set the camera to 30fps to capture a 40secvJupiter's AVI using a clear UV filter and uncompressed y800 video, the image comes with a type of onion ring effect very strong. I am adjusting the camera's gain to occupy the whole 8 bit space and gamma to 15 units on the slider. In addition to the onion effect mentioned, the image looks washed out with low contrast. Am I doing something wrong?
When I set the camera to 60fps, increasing the gain to keep the average pixel and keeping the gamma in 15 units, the onion effect improves some but it is still there. However, capturing through RGB filters the onion effect almost entirely gone, but the contrast is still poor. Does this fact suggest anything?

3- Is there any good tutorial about the use of the camera and respective software?

The image I posted here is an example of what I got yesterday. The image doesn't have any kind of post processing, it was just cropped and converted to jpeg. It was generated by stacking each AVI file (RGB) in Registax (with a very light wavelets sharpening) and then registered and combined in MaxDL. The seeing was about 3 out 5.

That is it. I would appreciate any help.

Best Regards,


June 21, 2007, 16:17:32
The camera has no contrast. It's implemented with DirectShow as Gain.

Please post pix that shows the "onion ring effect" or send them to me directly...john.berryman@theimagingsource.com

The best tutorial to date seems to be the this forum, the yahoo group, and AstronomyCamerasBlog.com.

June 27, 2007, 14:45:55
I am copying an e-mail that I already sent to john.berryman@theimagingsource.com with a sample image containing the onion ring effect.
================================================== =====
Dear Sir,

My name is Fernando Pinheiro Guimaraes. I posted a message on the Support Forum with some questions about my DMK 21AF04.AS. I am still getting that onion ring effect as you can see in the images attached. The Jupiter_onion file is an unprocessed image as it comes out of the camera after being aligned and stacked in Registax. The 20070617_2314UT file is a final one after processing. The onion effect is still there.

Do you think I am doing something wrong in terms of camera's setting or that is its normal behavior?

Best Regards,


================================================== =======

Stefan Geissler
July 2, 2007, 07:52:40

> ... or that is its normal behavior?
No, that look strange. We never got images with such kind of rings.

>... something wrong in terms of camera's setting
That's very unlikely but frankly, since we never encountered this problem we cannot be sure. But it's really unlikely.

July 2, 2007, 08:28:57
Fernando, I saw a defect like yours in old webcam images were the gain was set lower. Probably it is due to dynamic/gain loss. I think you can try one more time increasing the dynamic/gain. When I use DMK my first job is to establish the larger dynamic using integration time slider and gain one. In any case, I think, this is a "wrong" DMK setting and not for a DMK/ICcapture *problem*. Hope this can help you.
Clear skies.

July 2, 2007, 20:44:45
Thank you guys for the reply, I really appreciated that.

Marc, my major activity with the camera is to image Jupiter. For that I am using a manual filter wheel with RGB filters (Astronomik) in a f/30 system (a f/10 scope plus a 3x Barlow).
In this situation at 30fps I have to work with the gain's slider above 75% of its maximum value to fill the 16 bits bit space with about 80%. With this setting the gamma is adjusted to 14 units to maximize the dynamic range. I tried other values for the gamma correction, but 14 seems to be the best one.
In my application I can do two things otheer than the settings I am trying: either to push the gain to 100% and work with the histograms close to 100% or to decrease the gain to have a lower histogram. Both extremes to me don't make sense, but we never know.

Would you have a procedure to test maybe the signal to noise ratio or something to evaluate the camera's parameters?

Would you have any additional thoughts on this matter?

Best regards,


July 3, 2007, 08:21:13
Fernando, I read your reply and somethings sounds strange to me. Even if Jupiter is very luminous, I think that at f30 it is not so much possible to capture at 30fps. I usually try to obtain the biggest dinamic possible using exp. times and gain (no more than 50-60%) but if the dynamic is not larger enough I lower the frame rate too. I usually capture with gamma and brightness at their minimum, just playing with exp. times, gain and frame rate. Another try would be to decrease the frame rate.
Clear skies.

July 3, 2007, 14:04:35
Thanks for your considerations. On your reply I have the following remarks:

"I think that at f30 it is not so much possible to capture at 30fps. "

It is true that to go with this settings at f/30, the gain will have to be about 85% of its maximum. I am now using 853 units for gain and 12 units for gamma. It looks like the onion ring effect has gone with these values.
I did last night a bunch of tests using several combinations of gain and gamma and the one I just said was by far the best. I agree that I wouldn't be at the very best setting noisewise, but 30fps give me enough frames to average out the noise very reasonably in the final stacked image.

"Another try would be to decrease the frame rate."
Last night I tried 15fps to be able to set less gain, but the final stacked image presented a worse signal to noise ratio than the image at 30fps as described above. The problem here is that due to the fast rotation of the planet, one is able to capture for no more than 40 seconds and 15fps doesn't give enough frames for stacking.

But anyway, thanks for the suggestions. I will be using this thread to place my progress in imaging Jupiter with the DMK.

Best Regards,


July 3, 2007, 21:15:09
Fernando, for sure you are right.... high frame rates permit us to freeze the seeing and give us more time in planets like jupiter with a fast rotation; last but not least ... more time to change filters for color images. Anyway, I prefer to take all the time with a mono filter like red or green instead of color images. Personal opinion for sure. Anyway... the important thing is to have a good result and we'll have.
Clear skies

July 11, 2007, 16:00:01
Hi folks,

I found the best stting for the camera in order to get rid off the onion rings effect:

1- The gamma adjustment should be set to the minimum value (10)

2- The frame rate should be adjusted between 30fps and 15fps. The right value is the one the fills the histogram, also the gain should be keept below 70% of its maximum value.

Sometimes, depending on the seeing, the frame rate will be as small as 10fps. That is OK, because, analyzing some images, I concluded that is better to have less frames bad that more frames good.
I got some very good images of Jupiter at 10fps with bad seeing doing this way and capturing 40sec for each (RGB) filter.

I would like to have some comments from people who use to image Jupiter.

Best Regards,