Is my plan for HDR panoramas correct?

[spotmeter]

I am planning my yearly trip to Yosemite in May and want to do some HDR stitched photos with my Canon 5D and my 85,135 and 200mm lenses in order to produce some large-sized prints when I return home.

My plan is to first take several wide-angle photos of a scene (i.e. Half Dome at sunrise) and use the histogram to find the dynamic range. I will shoot several exposures with the same f-stop, varying the shutter speed until I can get all the values within the histogram.

Let's say that the scene requires 4 exposures (1/200-1/25th of a second to get all the shadow detail and avoid any blown highlights.

Let's say that I want to then make a final panorama of five vertical frames, giving me a total of approx. 40 megapixels (some loss for overlap).

I will then shoot the first left frame at the four exposures that I found included all the detail from the scene. I will then pan to the right to shoot four more exposures, overlapping the first frame by 20%.

I will repeat this process for all five frames.

Then, once I get home I will make a stitched image of each exposure (one image of all shots at 1/200th, another of all shots at 1/100th, etc.) and then combine them in Photoshop to get an HDR image.

Am I on the right track here?

If so, what stitching program do you recommend? Do you recommend the HDR function in Photoshop, or is there a better one? I use a Mac.

I shoot in RAW. Should I convert the files to 8 or 16-bit TIFF files?

Thanks in advance for your help and advice.

[Coot]

Your methodology sounds interesting. I would test it out before you leave...and report back to us.

[wood]

I am planning my yearly trip to Yosemite in May and want to do some HDR stitched photos with my Canon 5D and my 85,135 and 200mm lenses in order to produce some large-sized prints when I return home.

My plan is to first take several wide-angle photos of a scene (i.e. Half Dome at sunrise) and use the histogram to find the dynamic range. I will shoot several exposures with the same f-stop, varying the shutter speed until I can get all the values within the histogram.

Let's say that the scene requires 4 exposures (1/200-1/25th of a second to get all the shadow detail and avoid any blown highlights.

Let's say that I want to then make a final panorama of five vertical frames, giving me a total of approx. 40 megapixels (some loss for overlap).

I will then shoot the first left frame at the four exposures that I found included all the detail from the scene. I will then pan to the right to shoot four more exposures, overlapping the first frame by 20%.

I will repeat this process for all five frames.

Then, once I get home I will make a stitched image of each exposure (one image of all shots at 1/200th, another of all shots at 1/100th, etc.) and then combine them in Photoshop to get an HDR image.

Am I on the right track here?

If so, what stitching program do you recommend? Do you recommend the HDR function in Photoshop, or is there a better one? I use a Mac.
I shoot in RAW. Should I convert the files to 8 or 16-bit TIFF files?

Thanks in advance for your help and advice.
[a href=\"index.php?act=findpost&pid=112365\"][{POST_SNAPBACK}][/a]

Remember, with RAW you have 4 points to play with exposure to the right and left.

I have used many stitching prog.  and the more ease and with a good results is "PTGui"

If you want big files, 16bit.

Wood

[Monito]

Always convert to 16 bit files for quality, especially since that is what stitching and HDR are about.  Practice the post-processing before you make the trip.

There are two approaches to making the HDR exposures.  One, suited to 3 exposures using Automatic Exposure Bracketing (AEB), is to get the best single exposure for the scene and then AEB +/- 1 1/3 stops or +/- 1 2/3 stops.  Another way is to find the briefest exposure that will record the highest tone that you want to record and then step down from there in a series separated by 2/3 stop.  I used nine exposures to make this one:



I shot a 13 panel x 3 HDR exposure picture in Yosemite.  It took me 56 seconds to make the 39 exposures in two rows since I had prepared for it.  I used RRS equipment to avoid parallax and AEB.

I first tried to combine the HDR images first and then stitch, but the blend was not very good.  I used PSCS3beta to make the HDRs and then the stitch.  PSCS3beta is very good for stitching (vastly improved over CS2) but I don't think I gave it good source material.



I think there are improvements I can make to the consistency of the combination of the HDRs to improve the stitching.  Also there was tree sway that may have confused the stitcher.  Here is a stitch I did of two HDR panels (six exposures):



I have been doing the HDR combination first and then doing the stitching.  It might be best to do it the other way around, but I think you would require stitching software that allows you to adjust control points and record them so that you can get three perfectly consistent stitches.

I am still collecting reports about whether AutoPano Pro or PTGui have the capability and which is better.  My goal ultimately is perfection in stitching architectural photos.

[AJSJones]

The use of Photomatix is preferred over CS2's HDR tools and a Pano program that can use the same stitching points for each exposure is desirable, at least according to the discussions  here and here....