Creating a multi exposure length HDR composite

[admin]

So let's say you are imaging a very high dynamic range object such as M42. You can't quite get a good signal without taking a long exposure, but that blows out the core. So what do you do? You take two exposures! One shorter exposure to bring out the trapezium, and a longer exposure to show the rest of the fainter complex.

StarTools comes with a super easy way of compositing two (or more) exposure lengths.
All you need to do is process each exposure as normal, as usual recovering the max amount of detail.
Then, when you're done, simply launch the Layer module, open the processed alternative exposure length, set Filter Type to 'Min Distance to 1/2 Unity' and increase the Filter Kernel Radius until you got a smooth blend. Easy as 1-2-3!

m42hdr.jpg (195.38 KiB) Viewed 24927 times

Data courtesy of Rowland Cheshire

[ML7087373]

I was so excited to see this tutorial and after looking at it I find it will not work for my situation as far as I can see. I tried this with 2 stacks of M31 data taken on different nights, different years actually, and the layers won't combine because the images are of different dimensions.

I thought about the possibility of capturing different exposure lengths on the same night, but I believe that cropping the stacking artifacts would cause the same result. Is there a way to do this with StarTools that I'm not seeing?

[admin]

Using your stacking solution (for example DSS), try stacking one data set using the other as a reference. This way you'll have perfectly aligned images and the same dimensions. Does this help?

[ML7087373]

Using your stacking solution (for example DSS), try stacking one data set using the other as a reference. This way you'll have perfectly aligned images and the same dimensions. Does this help?

I don't know. I'll have to look into that. My first impression is that, as I dither my lights, the stacking artifacts in each stack will always be different yielding different dimension lights. I'll have to give it a try and see if I can use that angle to make it work.

[Alvinillo]

I have a question, isn't it better to stack all those exposures in a single fit? DSS already does that using groups

Regards

[ML7087373]

I have a question, isn't it better to stack all those exposures in a single fit? DSS already does that using groups

Regards

I don't think so. The whole idea of using different exposure lengths is to take advantage of the advantages of them. A shorter exposure will expose detail in brighter areas that will saturate a longer exposure. An example would be the Triangulum region in M42. A longer exposure will exposure the more faint aspects of an object, sacrificing the detail in the brighter parts due to sensor saturation. If you stack long and short exposures together the blending of those images will not do the best for both aspects. If you use the method described in the OP you can get the best of both.

[Alvinillo]

Thanks for the reply ML, I know what you are talking about. You have made me to re-read DSS help and I was wrong, I though that different groups were useful for working with different calibrations AND for combining different exposures. In DSS help they do not mention anything about different exposures.

Anyway I've found some interesting stuff here http://deepskystacker.free.fr/english/technical.htm

Entropy Weighted Average (High Dynamic Range)
This method is based on the work of German, Jenkin and Lesperance (see Entropy-Based image merging - 2005) and is used to stack the picture while keeping for each pixel the best dynamic.
It is particularly useful when stacking pictures taken with different exposure times and ISO speeds, and it creates an averaged picture with the best possible dynamic. To put it simply it avoids burning galaxies and nebula centers.
Note: this method is very CPU and memory intensive.

Anyway, when you are stacking you are already doing some sort of HDR and I think it would be logical that those different exposures come together at stacking phase. I understand that MaximDL already does it
http://www.cyanogen.com/help/maximdl/Us ... osures.htm

When imaging an object with high dynamic range on a DSLR camera, it may be necessary to combine images taken with different exposure lengths, and different ISO settings to reduce the saturation of brighter portions of subjects, while keeping fine detail in dimmer regions. The ISO level can be controlled through the Sensitivity selector.

[admin]

Using your stacking solution (for example DSS), try stacking one data set using the other as a reference. This way you'll have perfectly aligned images and the same dimensions. Does this help?

I don't know. I'll have to look into that. My first impression is that, as I dither my lights, the stacking artifacts in each stack will always be different yielding different dimension lights. I'll have to give it a try and see if I can use that angle to make it work.

You should not have any trouble;
Most stackers should have settings to use just the reference frame for dimensions. This is DSS for example;

Screenshot-Stacking Parameters.png (74.91 KiB) Viewed 23500 times

Leave the HDR compositing to StarTools and don't use DSS' compositing algorithms. The reason being that the latter makes the data non-linear and makes noise propagation readings impossible, so process them separately.

[christian]

Hello,

i´ve got a question about this, i found where in DSS i could set the Reference picture to alingn at.
For example i have 15x 90 sec and 15x 270 sec
So i stack the first 15 with 90 sec and then i should take the result of the stack to the 270 sec stack and set as reference? The 90 sec stack won´t be included in the new 15x 270 sec stack? I´ve searched for a option to take the last star alignment, but didn´t found an option, in the DSS documentation i couldn´t find an answer too (eventually i searched false) .

I hope i have an little error in reasoning and it´s not so complicated, eventually someone could explain and/or could post a screenshot of this option to take.

Oh and if this is important i dither my pictures. If this is important.

Very much thanks.

Christian

[admin]

Hello,

i´ve got a question about this, i found where in DSS i could set the Reference picture to alingn at.
For example i have 15x 90 sec and 15x 270 sec
So i stack the first 15 with 90 sec and then i should take the result of the stack to the 270 sec stack and set as reference? The 90 sec stack won´t be included in the new 15x 270 sec stack? I´ve searched for a option to take the last star alignment, but didn´t found an option, in the DSS documentation i couldn´t find an answer too (eventually i searched false) .

That's the idea, yes - using one stcak as an alignment source for the other. You may or may not get better results when using the longer (or shorter) exposure stack as a references - I haven't tried this.
Dithering (glad you're doing this ) should not really matter.

Do let us know how you get on!