Taking Panorama images with a large EV range

[wilko61]

Hi all,
I wasn't sure whether to post here or over within the Panorama section (but that was more for critique) I'm wondering when taking a number of images to be used for a panorama how does one account for the possible large EV range. I was looking to take a series of images for a panorama yesterday in the Blue Mts. The setting were going to be Manual Focus, ISO 100, f/16 and however the shutter speed would have range from 1 sec (in the shadows) to 1/30 (in the valley).

Taking the Images
I think this would be a range of around 5 f-stops, so I was thinking of setting the shutter speed to 1/8, but then bracketing at +/-2 f-stops. This would have given me 18 images (6 frames x 3 images)

Post Processing using HDR (or blending)
When I process the 18 images, I would take each set of 3 images and create 6 individual HDR images (or blends). Take the 6 individual HDR images (or blends) and turn these into the Panorama.

I suppose I have two 2 questions:-


How does the photographer deal with such a range of light when trying to capture a panorama?
If the images captured are bracketed, how to process these frames into a panorama?


Maybe I think too much about a problem, rather than just having a go!!

Thanks for any assistance / direction

Wilko

[ricktas]

shoot in full manual. Set your ISO, aperture and shutter speed. Do not change them between frames for the shoot. Also do not use a polariser as polarisers change intensity based on angle to the sun so as you pan, the angle changes and you will end up with a different polariser effect in each frame.

I shoot my pano's on full manual, camera tipped into portrait aspect and use a tripod.

Yes you are thinking about it to much! I would scrap the HDR idea, go out find my location, set to manual, shoot. go home and merge them into one pano. No HDR, or doing sets of photo merges and them merging those with others. Just shoot. Just merge them all in one go

[wilko61]

Wow what a quick response Rick, Thanks. I was going to do all that you said, this I have read over in the Panorama forum. So with the large range in EV, I should just select a shutter speed to give me something in the "middle range" and just shoot away :). My concern, I suppose would be that one side of the image would be dark (under-exposed), correct exposure for part and the other side would be bright (over-exposed). I don't know enough about the stiching-software (CS6) to know whether it would try and smooth out the exposures.

Thanks for your support.

Wilko

[ricktas]

you can lighten/darken areas of your photo once it is merged. Yes, pick an ISO Aperture Shutterspeed combination that gives you a good overall exposure across the scene.

[ameerat42]

Yes. Once it is merged. It is important to do all the photography, processing (really just RAW to jpeg), and stitching at the same settings.
This is mainly to get good and consistent results in the stitching program, and at worst, to stop it from producing rubbish.
I agree strongly with ditching the bracketed exposures idea as impractical. The ONLY way it would work is to take the first series of shots at one setting,
then repeat the series at other settings. That would take too much time and that means the lighting conditions would change uncontrollably.
Just take the one sequence as quickly as possible before the light changes - ie, sun goes behind clouds.
Am.

[tduell]

A belated comment on this...I only just discovered the thread.
I guess it depends a lot on the capabilities of the software you use to do your stitching/blending.
I use Hugin, and if I was shooting a pano where there was a large EV range, I would shoot bracketed shots.
When loading the images Hugin will automatically allocate stacks (current hugin source, earlier release versions may require stacks to be set manually) and then stitch with "exposure fused from stacks". This usually gives a good result.

Cheers,
Terry

[ameerat42]

OK, since we're at counterpoles on this - and I concede your expertise in this area, TC - I will say that my comments
were based on a 2009 version of Hugin which would balk at any moderate difference in sky tone (mainly sky tones).

I notice too that you say "usually gives a good result". I never used the method you describe so it would be interesting to try next time
I do this in such circumstances. Now I have the latest version of Hugin. (I couldn't run it on my other machine.)

Am.

[tduell]

I will be interested to hear how you get on.
The automatic detection of exposure stacks will depend on the EV differences in the bracket, which will need to be shot with a manual setting. When I said "...the current source" automatically allocates the stacks I meant the current development source. I don't think that hugin-2014.0.0 (the current release) does the automatic stack detection, I can't remember, as I do a lot of development testing and haven't used the 2014.0.0 release for quite a while. I'll have to check that and report back, but regardless, it isn't difficult to allocate stacks to the loaded images.
Please have a moan (direct to me if you like) if you have any problems with hugin, as the project is always keen to fix problems and provide help where possible.

Cheers,
Terry

[tduell]

I don't think that hugin-2014.0.0 (the current release) does the automatic stack detection, I can't remember, as I do a lot of development testing and haven't used the 2014.0.0 release for quite a while. I'll have to check that and report back,
Terry

Just to clarify, the auto stack detection will be in the next release. Not sure when that be, but as there have been some nice improvements/additions since the code was branched for the 2014.0.0 release, hopefully it won't be too long.

Cheers,
Terry