Eta Carinae
- Thread starter Jharvey
- Start date
OriolLehmkuhl
Well-known member
Very nicely done Jack! Congrats :wink:
Regards,
Oriol & Ivette
Regards,
Oriol & Ivette
Carlos Milovic
Well-known member
Hi Jack 
Very nicely done I agree with Maxi that the background is too dark... specially for this zone of the Milky Way. And, btw, the image is mirrored
Very nicely done
I agree with Maxi that the background is too dark... specially for this zone of the Milky Way. And, btw, the image is mirrored
Tremendous image jack. It's saying "deconvolve me"
Or maybe wavelets instead of deconvolution. Anyway, both techniques require a good star mask that isn't easy to build with StarMask alone.
I'll post a new technique to build star masks that can save your day It's a *killer* combination of HDRWaveletTransform and StarMask that is giving us extremely good results.
Here is an excerpt of a private email that I sent to Thomas W. Earle a month ago. In this email I describe an experiment with one of Thomas' wonderful wide field CCD images: a very difficult Ha image of the M42/HH region. We were looking for a way to build a good star mask for this image, and that's how I developed this new technique :
*********
1. Here you see a preview on the image with a suitable STF active. The image is of course linear at this point:
http://pixinsight.com/tmp/m42/01.jpg
I have extracted this preview as an independent image (by dragging the view selector).
2. After resetting the STF, I have stretched the preview image with HistogramTransform:
http://pixinsight.com/tmp/m42/02.jpg
A quite aggressive transformation, as usual with linear CCD images.
3. Now I have applied a wild HDRWaveletTransform instance to flatten the image at a wide range of scales:
http://pixinsight.com/tmp/m42/03.jpg
Note the unusually large amount of iterations (eight). By iterating in this way, the resulting image is so flat that M42 and M43 are no longer a problem
4. And now it's time to build a star mask. Once we have flattened the image thanks to HDRWT, all is much easier...
http://pixinsight.com/tmp/m42/04.jpg
The resulting mask isn't bad, considering that I've made just a couple tries. It can be improved by fine tuning parameters.
With this M42 image we have to face a dynamic range problem; that's why it's so difficult. The huge dynamic range due to the M42 region is a serious obstacle to StarMask. The HDRWaveletTransform process allows us to remove the dynamic range problem completely. Once the image is basically flat in terms of large-scale contrast, StarMask performs extremely well, as expected.
*********
As you see, this technique is so powerful that it allows us to build a star mask that protects even all stars inside M42 and M43! I'll write a detailed tutorial as soon as possible.
Or maybe wavelets instead of deconvolution. Anyway, both techniques require a good star mask that isn't easy to build with StarMask alone.
I'll post a new technique to build star masks that can save your day
It's a *killer* combination of HDRWaveletTransform and StarMask that is giving us extremely good results.Here is an excerpt of a private email that I sent to Thomas W. Earle a month ago. In this email I describe an experiment with one of Thomas' wonderful wide field CCD images: a very difficult Ha image of the M42/HH region. We were looking for a way to build a good star mask for this image, and that's how I developed this new technique :
*********
1. Here you see a preview on the image with a suitable STF active. The image is of course linear at this point:
http://pixinsight.com/tmp/m42/01.jpg
I have extracted this preview as an independent image (by dragging the view selector).
2. After resetting the STF, I have stretched the preview image with HistogramTransform:
http://pixinsight.com/tmp/m42/02.jpg
A quite aggressive transformation, as usual with linear CCD images.
3. Now I have applied a wild HDRWaveletTransform instance to flatten the image at a wide range of scales:
http://pixinsight.com/tmp/m42/03.jpg
Note the unusually large amount of iterations (eight). By iterating in this way, the resulting image is so flat that M42 and M43 are no longer a problem
4. And now it's time to build a star mask. Once we have flattened the image thanks to HDRWT, all is much easier...
http://pixinsight.com/tmp/m42/04.jpg
The resulting mask isn't bad, considering that I've made just a couple tries. It can be improved by fine tuning parameters.
With this M42 image we have to face a dynamic range problem; that's why it's so difficult. The huge dynamic range due to the M42 region is a serious obstacle to StarMask. The HDRWaveletTransform process allows us to remove the dynamic range problem completely. Once the image is basically flat in terms of large-scale contrast, StarMask performs extremely well, as expected.
*********
As you see, this technique is so powerful that it allows us to build a star mask that protects even all stars inside M42 and M43! I'll write a detailed tutorial as soon as possible.