Image stacks

In the past we've discussed techniques for HDR and also increasing depth of field that involved stacking 2 images and then selectively combining the 2 layers. Let's say I took 3 or more shots of a scene with the focal point set at near, mid, and far points. I then want to combine all the images into one scene to get everything in crystal clear sharpness. With 2 images I would just stack the 2 then process away. Now my question is what about more than 2 images in a stack--should you combine the first 2, then flatten them and then add the third image, etc.? Or, should you just keep adding them until you run out of RAM?

My initial feeling is that by flattening the images after each stack is made you cut down on the number of layers you're working with and therefore the RAM overhead and processor load is reduced. But, are there other issues that I'm not taking into consideration here? I guess the same issue comes up when taking multiple shots and combining them into a composite image like a mosaic or a panorama with the exception that these are just being stitched together and not necessarily overlaid. Thanks--Larry

Larry, I think you are spot on here - if you are limited on RAM/drive space. The real benefit to not flattening the layers would be for saving a flexible unflattened image, so you can go back and make alternations if needed. I always save a version of my image that is still layered, for just that purpose. But if it is a simple blend, I think you could easily get away with your idea of combining two layers, flattening & then bringing in additional layers one at a time & flattening before the next imported layer. Hope this helps.

I haven't done any HDR work, but for panoramas I always flatten the image right after it finishes the auto-stitch/auto-blend and before I do any processing work. If I don't, then the blend stitch lines usually become visible when I try to do certain things, such as a simple resize, crop, etc. I've read articles recommending not to flatten the image, but they never explained why, nor can I see any advantages to keeping the layers when you have defined edges between the stitched photos.

That said, I can see potential small advantages to keeping the layers for an HDR shot where you have overlapping areas, and if you don't need to flatten them for more processing power, then don't.

Thanks guys--I guess I could have been a little more explicit. I do most of the RAW processing on each image to be used first, then stack them as layers, then adjust more to feather the edges/overlap areas, then flatten, followed by any last minute tweaks before printing. My main reason to flatten them is that at 25 MB's per image and working at 16 bits with multiple adjustment layers it is easy to hit 300+ MB's fast. Since I save the original RAW edits individually I can go back to that point if need be and re-stack, etc. Maybe once I move up to a 64 bit machine, running at 3 GHz, and about 8 GB ram I won't be so stingy about space but with 3 images and all the various adjustment layers I could see files easily exceeding 500 MB!

Larry