This panoramic image of Llandudno Bay is stitched together from nine overlapping exposures, using the almost magical qualities of the Photomerge function in Adobe CS5.
It was a fantastic evening, with the full moon rising just off the end of the Little Orme as the sun set and twilight crept across the sky.
I'd been planning on producing a panoramic sweep of Llandudno Bay from this vantage point on the Great Orme for some time, and this was definately the night to be there.
To give myself the best chance of producing a decent final image I took my time setting up the camera, following a routine that I've developed over time when shooting panoramas.
This is what I did.
Step 1 - Set up the tripod. One of the secrets to shooting images for a panorama is to use a carefully levelled tripod, placed on firm ground and set at a height that allows the camera to pan around the panorama without the need to point the camera up or down. Keeping the camera level reduces lens distortions and makes it easier for the software to stitch the images without having to crop too much off the top and bottom.
Step 2 - Select a lens. In general I try and avoid wide angle lenses due to the field of view distortions inherent with these focal lengths. Also, I will try and use a prime lens for the same reason, as the optimised optical design of a prime lens results in less distortion, vignetting and chromatic abberations that the equivalent zoom lens. For this particular panorama I used a 100mm prime lens, the Canon f2.8 macro, which also has the benefit of being bitingly sharp.
3. Build the system. Having located and levelled your tripod and selected your lens the next step is to fix the camera to the tripod and level the tripod head so the camera is level. I use a spirit level to achieve this. Once the camera is level I tighten up all the tripod and head clamps except the clamp that allows the camera to rotate laterally. This clamp is tightened enough to allow the head to rotate without being sloppy. Then I do a sweep around the view that I want to capture, looking through the viewfinder to check that the horizon stays central and level though the scene. Small corrections to the levelling of the tripod and head can be made at this stage to get the best result possible.
Step 4 - Set up the camera. The camera settings I almost always use when shooting images for a panorama are as follows: File capture - RAW. White balance - Daylight (definately not auto). Exposure mode - manual. Mirror lockup - enabled. Shutter release - remote. Focusing - manual, with the camera pre-focused onto the most important part of the scene. In this case the lights along the promenade.
These settings result in the sharpest images possible, with no blur due to camera shake or focus hunting.
Step 5 - Set the exposure. When I can I will always use an aperture of f8 or f11 as these mid range apertures usually give the sharpest results, using only the central portion of the lens to project the image onto the camera's sensor. These images were shot at f8. ISO is normally set to the camera's base setting to minimise noise, ISO100 in this case. When it comes to selecting a shutter speed I aim the camera at the brightest part of the panorama and take a few test shots, altering the exposure manually until the brightest part of the scene is just shy of blowing out (over exposing). I then lock this exposure manually and use it for all the images that will make up the panorama. This usally means that some of the panorama is under-exposed, but you can't vary the exposure as you go along or the final stitch will look very silly, even if the software can achieve it.
Step 6 - Take the images. Starting at one end of the scene you want to capture, take images with at least a 30% overlap between each one, using the remote release to avoid touching the camera.
These are small thumbnails showing the original RAW images that were taken of the scene. Note the amount of overlap between the individual exposures and that the horizon is placed centrally in the frame to avoid distortion.
Step 7 - Processing the RAW files. I use Adobe Camera RAW (ACR) to process the RAW files. The important thing when doing this is to make sure that any changes you make are applied to all the files ('select all' and 'sychronise' options activated), so that a consistent look is achieved. Otherwise, when the converted files are stitched together there may be differences in white balance or exposure across the panorama.
In this case, the only change I made to the images was to warm them up a little with the white balance adjustment, as the originals were shot with daylight white balance set in the camera and were a bit too blue. I leave all other adjustments to later on in PhotoShop, but it's important to adjust the white balance at the RAW stage as you're working on files with only a single luminance channel. Once the RAW file is converted to an RGB file it becomes much harder for the software to alter the white balance correctly as there are now three colour channels to adjust instead of just the one.
Step 8 - Exporting files from ACR. Once the white balance has been set in ACR, I then save the files as 16bit TIFFs in the ProPhoto RGB colour space. This is one of the largest colour spaces available and I use it to make sure that I'm not throwing away any of the colour gamut captured by my camera in the original RAW files.
For a more detailed techincal discussion about the merits of various colour spaces check out this article by Michael Reichmann.
I use 16 bit TIFF files to ensure I'm not losing any data by compressing the files into 8 bit JPEG format. This is important to give me more leeway in PhotoShop for colour and luminosity enhancements later on.
Step 8 - Stitching the panorama. Having got nine overlapping images into a useable 16bit TIFF format the next step is to load them into Adobe PhotoShop's PhotoMerge facility.
PhotoShop CS2 and earlier versions are hopeless at merging images, but CS3 and later versions do a grand job. I use CS5 with vignette removal and distorion corrections both enabled.
I don't know how PhotoShop does what it does to stitch panoramics, but it's fascinating to watch the process unfurl on your computer screen until up pops a fully merged pano.
This is where you find out if your original shooting technique was up to scratch, because if there's mis-alignment or distortions in the original files then yuor resulting stitched panoramic will be horrendously curved (I've had a few!) to the point where you have to crop most of it away, or use the warp tool to try and pull it back to shape - not clever!
Fortunately, as you can see here, my stitch ended up pretty good, with only a small amount of cropping needed top and bottom to remove the transparent pixels. It's always worthwhile when shooting your original images for later panoramic stitching to allow a reasonable margin for cropping at the top and bottom of your frame, otherwise you might find yourself having to crop out the very things you wanted to show in your final composition.
To finish off this stage of the process I flatten all the layers, crop the image as large as I can get it while removing the transparent pixels around the edges, and re-save the file as a 16bit PSD (Photoshop) file, still in the ProPhoto colour space ready for some enhancements.
Step 9 - Image enhancements. There are as many ways to enhance an image as there are photographers out there, so I'm not going to say my way is best. I'm just going to descibe what I did to this particular image.
When working on an image in PhotoShop I always use a new individual layer for each stage of the operation, building up a master file that can be re-visited and changed in the future. This means that all the changes I make can be reversed if things go horribly wrong, so no harm done.
When starting work on an file in PhotoShop I'll have a pre-visualised, finished image in mind which gives me a target to aim at. I'll then change the colours, tonality, cropping, shape, in fact just about anything that gets me to my final goal for a particular composition.
For this panoramic view of Llandudno Bay I didn't really want to change much. Just enhance its visual appeal through subtle changes to colour, tonality and sharpness, nothing more radical that that.
Starting off, here's a screenshot of the building detail at 100% with no enhancements applied.
The first operation was to copy the original layer and to brighten the shadows somewhat using the 'shadow/highlight' adjustment tool.
Next I copied the new layer again and applied a small amount of sharpening using the 'Smart Sharpen' filter with radius set to 1.0, strength set to 100% and mode set to 'Lens Blur'. Only this small amount of sharpening was required in order to overcome the effects of the anti-aliasing filter fitted to my camera's sensor. Thanks to using the shooting techniques described earlier the native sharpness of the original RAW files from the camera was very good. It certainly helps to get it right in camera.
Finally I added a levels adjustment layer, moving the black point slider until it was just touching the left hand edge of the histogram to make sure I had a true black point for the image. I left the white point alone as I didn't want to blow out the fine details in the highlights of the lights on the buildings. Adding my copyright logo was the finishing touch.
All content copyright © Howard Litherland 2009-2019 unless otherwise stated.