[pinholecam]

Perhaps I misunderstand what is trying to be accomplished but I was wondering the same thing myself.

A couple of my single frame shots...

I can see that you have some serious firepower there.
Nice shot.

That's focus stacking, where each shot has a different focus. This is high-resolution stacking, for increasing resolution/dynamic range and reducing noise. See Take Better Photos, reduce noise, increase quality and resolution of digital photographs. Leading superresolution technology. for another product that does this.

Thank you all for the feedback and compliments

Wildman - FANTASTIC shot. You must have used a longer lense or a telescope...which makes me jealous in the slightest bit



The lens was a Quantaray Tech-10 lens...aka the Sigma 70-300 f/4-5.6 DL Macro Super. I bought it new in 1998 (ish) for $199.99. Its been a good learning lens, and my partner LOVES it...I just think it lacks a little sharpness in the images shot from 200-300mm. My copy also has a small stick in the zoom ring about 200mm. This series was shot at 300mm, 1/160s at f/8.0, ISO 200 on my trusty K10. In fact, this was a 21 stack salute to the K10...as it was the last major image project I took with it before I retired the K10 to 'backup body' in favor of my new K-5.



baro-nite - Good little read. As you mentioned there are a couple of different reasons to stack images. Hi-res stacking offers the ability to improve existing resolution of an image with few drawbacks. For me, Ive always had a bit of fascination with astrophotography, but dont have the financial ability to invest in another hobby (cameras and computers suck up enough of my spare cash). Purchasing telescopes, equitorial mounts, tracking motors, etc, is beyond what I have interest/finances for at this point in my life. Image stacking though offers the chance to get an excellent image WITHOUT the need for thousands of dollars in gear just to shoot picures of the heavens.

Will this work for landscapes? Absolutely! Ive tried stacking other images as well and have had some success. One key though is that it works best when you stack UNIQUE images of the same scene - ie - set your camera up on a tripod, and shoot multiple images of a static scene. You can stack multiple copies of the SAME image as well, but the gains there are significantly reduced - especially when using images that natively came from a 10+ MP camera. In fact, stacking the same image seems to increase noise and artifacting almost as much as it does resolution. Again - Im new at this so YMMV.

Thanks guys.
Its an eye opener and I've certainly learned something new.
Appreciate it.

[philbaum]

That's focus stacking, where each shot has a different focus. This is high-resolution stacking, for increasing resolution/dynamic range and reducing noise. See Take Better Photos, reduce noise, increase quality and resolution of digital photographs. Leading superresolution technology. for another product that does this.

Thats where my interest lies as well. Whereas the usual software gets rid of noise by identifying and smearing it, this method, i think, gets rid of it by averaging the results.

I wonder if this would get rid of motion, like some moving grass or bush or tree in a landscape scene. Even if one has a high resolution lens, it would seem that this method might allow even more refinement for improved enlargement???

I assume one can muck about with it and wouldn't necessarily have to use 20 shots. All the top Pentax dslrs have a 9 shot multiple image menu choice but i don't think its exactly equivalent to this software.

Thanks for the discussion!

[dosdan]

Thats where my interest lies as well. Whereas the usual software gets rid of noise by identifying and smearing it, this method, i think, gets rid of it by averaging the results.

In 2009 I did a test of 14-shot hand-held stacking for NR in PhotoAcute with the K20D at ISO3200. 14 PEF shots is the depth of the K20D shot buffer in Continuous Shooting (Hi) mode.

https://www.pentaxforums.com/forums/digital-processing-software-printing/7470...king-test.html

But according to DxOMark, the K20D raw files are "cooked" (i.e. have some non-defeatable NR applied in-camera) from ISO3200 and up. (Even so, the chroma noise is pretty bad in a PEF from this camera at IS03200.) See:

DxOMark - DxOMark review for Pentax cameras

So the stacking of uncooked raw files may result in either better NR, due to the noise being less correlated (i.e. more random), or perhaps in better sharpness.


Dan.

[PentaxScott]

Should be reduced all the way to zero. There's no new information to be gained this way.

This is true. I did use the wavelets to sharpen the image, but aside from that there was a SMALL improvement in visible resolution. Not sure if it was more of a trick of the eyes, or if it was because I was playing with 8 MP JPG images from my Mother-in-laws Canon ultra zoom P&S. Mathematically you are correct though...

Thats where my interest lies as well. Whereas the usual software gets rid of noise by identifying and smearing it, this method, i think, gets rid of it by averaging the results.

I wonder if this would get rid of motion, like some moving grass or bush or tree in a landscape scene. Even if one has a high resolution lens, it would seem that this method might allow even more refinement for improved enlargement???

I assume one can muck about with it and wouldn't necessarily have to use 20 shots. All the top Pentax dslrs have a 9 shot multiple image menu choice but i don't think its exactly equivalent to this software.

Thanks for the discussion!

This does also work for noise. I also took a 5 shot sample of the Orion Nebula to see what kind of results I could get. With a 300mm lens and no tracking, I was limited to a mere 1.3s per exposure to eliminate star trails. Needless to say, that wasnt nearly enough time for my K10, however, I did get enough of an exposuse to see a faint "cloud" in the image. Unfortunately, the rest of the image looked like a explosion of colored confetti from all the chroma noise. I ran the 5 image through and eliminated the noise to less than 10% of what it was, and improved the image substantially! The image was still not worth a grain of salt, but at the time I wasnt trying to get that...I was merely testing the program to see how successful it could be.

You wouldnt need 20 shots in order to stack using Registax, however, the more images you work with, the more resolution/dynamic range you can create in your image, and the better the noise reduction. There will always be a point of diminishing returns, but that will vary on the image, and the needs of the photographer. Ive also been told that working with different exposure times (and thusly ISOs/f-stops) will also improve things - ie - for a wide angle landscape of the sky at night, stack 10 at ISO 100 at 30 sec, plus 10 at ISO 400 at 8 sec, plus 10 at ISO 1600 at 2 sec, etc (values just pulled out of my head).

[KevinR]

Thanks for the useful information on using Registax. I've been wanting to play with this on the moon, and your advise makes great sense.