Jump to content

SkySurveyBanner.jpg.21855908fce40597655603b6c9af720d.jpg

'Lucky' Imaging - effect of exposure duration and stack percentage


Recommended Posts

I took my Moon images last week with exposures of 5 and 10mS and stacked them with various pertcentages in Autostakkert 3 to see the effect. The seeing wasn't great and the Autostackert quality graph showed that with the curve quickly falling to around the 50% mark. I unfortunately deleted the videos before taking a screenshot of the graph.

The videos were at 8 bit, 30s long at 30 fps. Here's the result of altering the stacking percentage of 5mS exposures at gain 150 on ASI178MM. All images were given the same sharpening amount in Imppg.

It's a png to avoid jpg artifacts. Click to view at full size. The 10% is noticeably sharper and not visibly any more noisy than the others.

1535796447_Stackpercentage.thumb.png.6bf1bf48775eef9486ae4d3036d47fb4.png

Here's the difference between different exposure lengths. The camera gain was halved on the 10mS to give a similar image brightness. I was surprised at how much sharper the 5mS image was. In times of good seeing there may not be so much difference. The 5mS is again not visibly any more noisy due to higher gain, as stacking just a relatively few frames greatly reduces the noise anyway.

602228480_Exposurelength.thumb.png.7d67ec1cd91d0a6480dc71910b9e32a0.png

As expected, it's best to go for as short an exposure as you can, even at the expence of putting more gain in the camera, and to stack as few images as you can get away with before the noise starts getting noticeable.

As I mentioned previously in another thread, according to an article by Craig Stark, once you stack at least 50 images you greatly reduce the quantization noise and effectively gain 4 bits of resolution (as long as the image contains noise) so a 50 image stack at 8 bit is effectively the same as a 16 bit image stack, as the CMOS planetary video cameras are generally only 12 bit. I've never found a 16 bit stack planetary image to be any better than an 8 bit one. At 8 bit you'll halve the storage required and also get twice the frame rate if your using the full sensor area. :smile:

Alan

  • Like 3
Link to comment
Share on other sites

  • 2 weeks later...
2 hours ago, nfotis said:

Interesting.

What does happen if you use 12-bit images instead? Do you get any perceptible difference?

N.F.

I've not found any difference in stacked 8 or 12 bit videos. Recording 12 bit just reduces your frame rate for full frame recordings as it is output from the camera as 16 bit, and also takes up twice the storage space.

Alan

Link to comment
Share on other sites

Interesting stuff which kind of validates what I do / have found but never tried to prove. Keeping the exposure as short as possible keeps the frame rate up which helps in grabbing lots of frames in moments of better seeing. I experimented using the Mono 16 bit captures at one point but could not see sufficient benefit to offset the longer capture times and additional storage requirements. I've also played around with stacking different numbers of frames but find 50ish is about the sweat spot for what I do.

Thanks

  • Like 1
Link to comment
Share on other sites

4 hours ago, symmetal said:

I've not found any difference in stacked 8 or 12 bit videos. Recording 12 bit just reduces your frame rate for full frame recordings as it is output from the camera as 16 bit, and also takes up twice the storage space.

But surely if you used a frame depth of twelve bits and stacked the same number of frames as in your 8-bit images you'd also gain four bits of resolution and end up with the equivalent of a 16-bit image?

It's certainly an interesting experiment to have done though.  It would be interesting to see how far it extends with the planets, where you also need to choose your exposure time to get sufficient dynamic range.  I've taken to imaging the Moon in Ha because otherwise it's so bright that even the fastest frame rate on my camera still blows out parts of the image.

James

Link to comment
Share on other sites

1 hour ago, JamesF said:

But surely if you used a frame depth of twelve bits and stacked the same number of frames as in your 8-bit images you'd also gain four bits of resolution and end up with the equivalent of a 16-bit image?

If you had noiseless images, (or very low noise) then the 16 bit image would be better. As soon as you stack a few images containing a "reasonable" amount of inherent noise then the inherent noise quickly becomes the dominant factor and overwhelms the quantization error 'noise'.

From Craig Stark article for his test camera which had a noise Standard Deviation of 1.87,

"The quantization noise in a 10-length stack of 8-bit images is smaller than the inherent noise in a 200-length stack of 16-bit images"

From his graphs, as long as the Standard Deviation of the inherent noise is greater than 0.4 the inherent noise dominates the quantization noise for a stack of at least 50 frames when comparing 8 or 16 bit images.

I suppose it's worth measuring the noise of your camera of a dark frame at your usual gain and exposure, for 8 and 16 bit and see what the Standard Deviation is. That would indicate whether it's worth using 16 bit images when stacking. 😉

Alan

Edited by symmetal
Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...

Important Information

We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue. By using this site, you agree to our Terms of Use.