AEB Layering HowTo

Are there any tutorials or simple instructions for a newbie trying to use GMIC to do automated stacking of three AEB layers to create an HDR photo?

Welcome to the board!
AEB = Automatic Exposure Bracketing, I trust.
The words ‘newbie’ and ‘simple instructions’ are inclined to steer my advice toward Darktable, a FOSS tool that is well represented here at pixls.us. This video tutorial may be helpful: How to make an HDR image in darktable. I think if you are new to these tools and want to merge your stack and get on with life, then that is the way to go.

G’MIC - the command line tool - is eminently qualified to slice-and-dice the dynamic range of your images in enough ways to make your eyes water, but if you are not familiar with accessing the command line tool of your platform, then, I fear, we are outside of the range: ‘simple instructions’.

That said, here is a basic introduction to the G’MIC command line tool. If you’re not intimidated by shells and have some specific and exacting compositing criteria that you are not sure how to implement, then ping here. Hope this helps.

1 Like

For exposure fusion of bracketed JPEGs - I strongly recommend enfuse
For merging bracketed raws to a high dynamic range raw - I strongly recommend HDRMerge

Are you talking about the enfuse plugin for Adobe Lightroom? If so, this is a bit off topic as I was trying to use free software like GMIC.

I was trying to use the GMIC plugin for GIMP rather than the command line tool. Darktable does seem to be much simpler. The only problem I found with Darktable is that it won’t create an HDR from a stack of JPEGs, or at least I couldn’t work out how to do that. I found a tutorial suggesting it should be possible but it didn’t say how.

You may be (inadvertently) ignoring JPEGS. The reason for this is that JPEGS alone are a terrible source for High Dynamic Range work. It is a lossy format (information is thrown away in the interest of compactness in compression) and very little information about image dynamics survives. Cameras often create “JPEG Previews” of raw format images so that conventional photo viewers can present something about a raw-formatted image for identification purposes, but not for use in a workflow. For this reason, Darktable can be set to ignore JPEGS, assuming them to be for preview purposes only, and that you really wish to import raw-formatted images, which are rich and detailed enough for HDR work.

If all you have is a “JPEG” stack, then I fear you won’t have very satisfactory results no matter what tool you use. Most of what is useful for HDR work has been compressed out of your images. However, something is better than nothing - check the Darktable user guide for enabling JPEG import. If you go the Darktable route, ask questions in the Darktable section of this site; I am, by no means, a Darktable expert - but over there, you have the real pros.

I would also thumbs-up @Entropy512 suggestion regarding HDRImage. it is worth a look and there is a dedicated forum here as well.

That plugin is just a wrapper around a well known FOSS algorithm implementation that has existed for over a decade. (Enfuse is typically bundled with hugin but can be used standalone)

I would disagree on this for the use case of bracketed JPEGs into enfuse - it was specifically designed for this exact purpose and works quite well. So well that it’s been (ab)used to implement a general tonemapping algorithm by feeding it “synthetically bracketed” shots from an HDR raw (see Google’s HDR+ and NightSight pipelines, both depend on a slightly modified variation of Mertens exposure fusion that feeds the algorithm with synthetically bracketed shots.)

The only other option for JPEG inputs might be the Debevec approach, discussed in Python Image Processing in Computational Photography | ToptalÂŽ - but I would strongly advise against this if it can be at all avoided. Either enfuse or merge raws.