My CCDStack demo expired and I did not have a chance to use it because my telescope was out in Celestron for repairs and it took them 3 months to fix...
When I try to combine the images at the end, using my single color shot camera, nothing happens...Is this because the Demo expired ?
Jorman

That should not be a problem. Expiration should only affect the ability to save files. Tell me exactly what you are doing.

Stan

I open the files containing the image, I do the calibration procedure using the masters that I have on file, then I click on the single color shot option and then I pretty much follow the Prodcess sequence and everything seems to work until I get to the combine stage and when I click on combine nothing happend...

What version? (refer to Help)

When you click on combine does a sub menu appear for "Sum", "Mean", "Median", etc.?

Is the "Combine" menu option grayed out?

Check Image Manager to be sure that there are at least 2 "included" images. Otherwise there are not enough images to combine.

What option(s) do you use in the "Convert Bayer" dialog?

It would be helpful if you could take a screen snapshot (<ctrl>PrtScn then paste into Paint or other program to make a jpeg). Be sure that the "Image Manager" window is open and showing.

(sorry for all the debug instructions but I do not have a computer that I can use to test the expired state)

Stan

This message has been edited. Last edited by: Stan,

Hi Stan,

I am going to give you a step by step as I do it :

Open file and select 20 images from my folder
Calibrate using dark and bias master for 30 seconds ( image exposure) and hit apply to all button
Adjust display : click on DDP, apply to a;; and auto scale buttons.
Convert Bayer : Check first radio button on the right under Bayer Pattern option
Click apply to all. At this point the image turns blue
Click Register, at this point the screen turn pinkish...
Selec medium bright stars around the image using Star Snap function
Click Align All
Click Apply
Click bi-cubic spline and wait for the process to complete. At this point the image turns back to blue when the interactions are finished'
Normalize image using Auto
Do Data Reject, Std sigma reject and hit Apply to All. At this point, when finished, the swcreen turns back to grayish...
Enter Combine Mean ( there are 20 images in the Image Manager)The screen is now back to blue, but nothing happens........The image looks like any of the individual subframes, but blue
I have attached a shot of the final Image Manager screen

By the way there is nothing to be sorry for, you have given me so far more attention and info than a lot of other sites...Keep up the good work I appreciate it Thanks.

Image_Manager_screen.doc (90 KB, 10 downloads)

Hi Jorman,

I suspect that the combined image is actually created but the images have not been correctly de-Bayered and/or balanced. So first, let’s look at a single image: Open 1 image and calibrate it. Next, de-Bayer it (Convert Bayer). Does it look OK? If not then keep reading. If it looks OK then please take another screen shot of the problem (this time show the full screen, not just the Image Manager).

It is necessary for the program to know the correct Bayer matrix pattern. Also, the Bayer filters are intrinsically unbalanced and so it is necessary to apply filter ratios to the de-Bayered image. DSLR RAW files contain metadata that documents the Bayer matrix pattern and “neutral” color ratios. Unfortunately, most Bayer FITS files do not document either.

CCDStack will attempt to select the correct Bayer pattern if it is present in the Header data. But for FITS that crucial information is probably not available to CCDStack and so you must figure it out by camera documentation or by trial and error. Currently CCDStack does not even look for color ratios in the Header and so the user must initially enter them (the “convert Bayer” form has a place to enter the ratios). The ratios are rarely (as in never) documented in FITS and are likewise never in the camera documentation.

After you have selected the correct Bayer pattern and de-Bayered the image then you need to find the approximate color ratios.
Open the "color"; "adjust" form via main menu:

1) press the "set background" button and drag-out a rectangle over an area of sky background. Accept the result.

2) drag-out a rectangle over a galaxy, star cluster, or other feature that should be somewhat "white" and press the "Balance" button.

Now the image should be recognizable. Try stacking.

Stan

Hi Stan,

I am not sure what you mean by do the images look OK. They look just like the original shot only that depending on which Bayer pattern I select , the image looks either blu, red, green or a grayishgreen color !!!
Jorman

Please post a screenshot of a "correctly" de-Bayered image.

Stan

This image has been calibrated and debayered using the first option in the color menu

Jorman

cirrus_nebula_30_secs_with_Hyperstar.FIT (1,022 KB, 5 downloads)

Hi Stan,

Did you receive the image? It looked like it did not go through when i hit the Post Now button

Jorman

The "Image Manager" screenshot proves that the Combine worked and created an image called "median cirus". BTW, it is not advisable to use Median combine in conjunction with data rejection; Mean is preferable.

that FITS is defective, probably because its file size exceeded the limit for attachments. Only a part of the Red layer survived.

So, the original problem is solved - the images are being combined. I can’t tell from anything that you have sent if the deBayering is working correctly (I only suspected a problem based on your description of the images turning colors). If you are happy with what they look like then as far as I can tell – nothing is wrong(?)

Stan

This message has been edited. Last edited by: Stan,

Hi Stan,

I guess I did not define the problem well from the beginning. Assuming that everything is working fine, the problem is that the combined image does not look any better than any of the individual images. These images have been taken with a Hyperstar lens ( focal ratio 1.9 ). 15 images at 1 minute each are equivalent to approximately 75 minutes long exposure at FL 6.3 . So why do not the combine images look any better than a single image ? Is it my settings or maybe I just do not know how to manipulate the combine image....Since my Demo expired I can not save the final image so that you can see what I mean, but thrust me. There is no appreciable difference. Any other thoughts ?

Jorman

Jorman

Hi Jorman,

You should be able to see a significant difference between a single image and the combined image by aggressively scaling (e.g. lower the gamma &/or DDP). But if not then here are some possible causes:

* Be sure to Normalize. The best way to normalize is “Control”; “Both”.

* After data rejection (but before closing the DR form) - blink the stack to verify that rejections are as desired.

* Combine via Mean (instead of median).

* You may be seeing some color errors that are not removed because data rejection only operates on the Lum layer (CCDStack internally represents color images via LAB not RGB). The most robust rejection is achieved by de-Bayering to single colors and processing each color separately then combining the single-color masters. This is a bit more work but will produce the very best results. A future version will more easily facilitate multi-color rejections.

But the main problem may be a misconception that you have regarding f-ratio and S/N (a very common misconception): S/N is primarily determined by aperture, not f-ratio. The “Signal” in S/N is the number of photons from an object (e.g. star) and the “Noise” is the associated noise from the object (Poisson) and the area it occupies (sky and detector noise). For a camera lens, changing f-ratio actually changes the aperture. Changing the focal length only (as you did) has a only a small effect on S/N because photon statistics are unaffected (the same number of photons are collected from a star regardless of the FL).

I must prepare for AIC. If you are attending then please come by the CCDWare booth. If not, then we can resume this thread Monday or later. Drop me an email so that I can instruct you on how to send me some of your raw frames – stan_ccd at yahoo dot com.

Regards,
Stan

P.S. It is often a good practice to create a simple combine via Mean immediately after registration. The image may have many blemishes but it represents the highest possible S/N. Then process the stack via normalization and data rejection to generate another Mean. Compare the two Mean images to confirm that the processed image has good S/N.

Hi Stan,

Enjoy the show. No I am not going I live in Miami, Florida.

Stan with all do respect to your extreme expertise in this subject, I am not sure that I agree with your statement about S/N to FL. From a practical point of view I can tell you that if I use my C14 scope at prime focus and take a 30 sec exposure of say a nebula, then I take it at 6.3FL and then at 1.9 FL the results go from I do not even see the nebula to I see the nebula to Woh! I a see a great shot of the nebula.....It seens to me the collection of photons as a function ot time increases tremendously from f/11 to f/1,9 with the same telescope. Anyway I do not want to turn our discussion into a theoretical discussion of photon acquisition.
Please look at the attachment, Should not the combine mean image at the bottom show 450 seconds exposure instead of 30 seconds ? I still think that there is something wrong with my Demo or I am doing something fundamentally wrong....

Jorman

Image_Manager_for_Stan.doc (341 KB, 8 downloads)