Vertoramas / Discuss

Panorama Paul    says:

Quite a few of my contacts have recently asked me to describe the steps I take to create one of my hand-stitched, double-image vertoramas.

So I’ve typed up this little tutorial...

Motivation: My main motivation for creating my vertoramas is to show more of the scene than can be captured with only one frame.

The next most important reason is that with a vertorama, I can correctly expose for the sky and for the foreground. Unless I use a (carefully aligned) graduated neutral density filter over my lens when shooting a landscape scene, no matter how much I fiddle with my camera’s settings, I’m either going to get an underexposed foreground, or an overexposed sky!

I could of course process my images using Photomatix, but that’s another discussion that I don’t wish to enter into now. Other than HDR, there is simply no other way (that I know of) to achieve a balanced exposure throughout the entire image.

Software: To process my vertoramas, I use Paint Shop Pro (version 11) and a collection of plug-in filters called nik Color Efex Pro 2.0 - Traditional. Although I’ve used Paint Shop Pro to process my photographs, there is absolutely no reason why my tutorial won’t work for Photoshop, or any other photo editing software that supports layers and third-party plug-in filters.

I now no longer use any panorama stitching software to align and stitch my images, since I firmly believe that I can do a better job by hand! I’m not saying that I’m perfect however... and I’ll happily review my opinion as soon as someone can show me any stitching software that is able to recognise the objects that it is stitching (and treat them accordingly (like keeping the horizon straight)).

Preparation: Before I take the photographs for any vertorama, I first decide where the best place to stitch and blend the two images will be. This is easy when I have a clear horizon with cloudy skies, but a little more tricky when the scene includes objects that span vertically across both images. Since I’m mostly shooting with my 11mm ultra-wide lens these days, there is usually lots of perspective distortion at the edges of the frame. I always look out for that, and try to minimize any distortion where possible.

These days I also always shoot at least 20 to 30 images for each vertorama. I do this at slightly varying apertures and shutter speeds, and with small adjustments in viewing angle. With the ultra-wide lens, I’m constantly amazed how moving the lens a degree or two in any direction can present a completely different perspective. Although there are often only minor differences between any of the images when I sort through them later, there are usually one or two images that clearly stand out as being better than the rest. Besides a quest to capture the perfect photograph for any scene, my main reason for shooting off so many different shots is that... it usually costs me some time and effort to get into the position where I’m ready to take the photograph, but it doesn’t cost me anything extra to take a few dozen more shots!!

Stitching and Blending: Okay, after selecting the sharpest, best exposed, and most interesting two photographs, the next step is to seamlessly blend them together into one image. The first part of this step is to get the two images to match as closely as possible in terms of perspective distortion and exposure. It’s usually very easy to blend two images somewhere in the cloudy sky, especially if there is very little difference in their exposure... so that’s why I’ve selected the following two images for this tutorial. This is what they look like before I start any processing (straight out of the camera)…





First I create a new canvas that is a little bit larger (both vertically and horizontally) than the two images stacked together, and then I open each of my original images, copy them to the clipboard, and paste them as new layers onto my new canvas. I usually paste the sky layer first and after pasting that layer, I move it up to the top of the canvas. When I’ve pasted the ground layer on top of the sky layer, I set the layer transparency to about 50% (so that I can see through it) and then drag that layer down to fit together with the sky layer as closely as possible. When I’ve moved the layers into position, I reset the upper layer’s opacity back to 100%.

Sometimes I need to do some perspective correction, rotation and resizing to get the layers to fit perfectly, but sometimes (like this time) it’s not really that important to get an exact fit... blending clouds can be very forgiving!

Having moved and adjusted the two layers until they match as closely as possible, the next step is to rearrange the order of the layers and to remove either the ground portion from the sky layer, or the sky portion from the ground layer. In this example, I’ve placed the ground layer above the sky layer (in the layer palette order), and I’ve blended its sky together with the lower layer’s sky.

I realise that all this sky/ground, upper/lower, layers business might sound confusing to some, but I’m confident that this will make perfect sense to most people familiar with working with layers. When I refer to sky and ground layers, I am always referring to the top and bottom parts of the vertorama, and when I refer to the upper and lower layer, I am always referring to their order in the layer palette.

But before I attempt to blend the layers together, I must first attempt to get the exposures in the blending area to match as closely as possible. Blending two layers with perfectly matching exposures is often as simple as dragging a soft eraser brush over the hard edge of the upper layer. But the task is usually a little more difficult when the exposure of the area that you want to blend doesn’t quite match.

In this example, the exposures did not match perfectly, so I had to lighten the sky layer a little and then apply nik’s amazing Graduated Neutral Density plug-in filter to the ground layer. I’ll explain how I apply these filters in the next part of this tutorial, but for now I’ve got to say that this filter has got to be the most useful plug-in filter around... it gradually either darkens or lightens either the top or the bottom part of any layer (with all degrees of control). In this tutorial, I slightly darkened the sky part of the ground layer, until the exposure of the two layers matched perfectly.

So having tweaked the exposures to match as close as possible, the next step is to blend the two layers together, so that not even a trained eye (at the highest magnification) would be able to find the stitch. This is where practice makes perfect... your first blend might not be seamless, but I can promise you that you will soon surprise yourself with how easy this actually is!!

To blend the images I always use the eraser brush tool, with a range of sizes, transparencies and hardness. I start off by setting a very soft-edged brush, and completely erasing (no transparency) the line that defines the join between upper and lower images. Then I increase the brush transparency and continue to brush away the part of the image in the upper layer where I want image in the bottom layer to shine through. Some parts of the image often require small hard brushes, while other parts might only need a large soft brush for a seamless blend.

Sometimes the sky and ground layers have such widely differing exposures (like when shooting directly into the sun) that it is impossible to blend the skies of the two layers together. In these cases, I usually select (with the magic wand tool) and remove the entire sky part of the ground layer, and then I use a very small brush at high magnification (zoom) to carefully erase all remaining traces of sky from the edge of the horizon.

Each vertorama will present you with a new challenge and will require slightly different blending techniques. But the feeling of satisfaction that you will experience when a blend is flawless, is always worth the time and effort that you’ve put into it (in my opinion)!

Although not flawless (at high magnifications), my blended vertorama now looks something like this...



Processing Tricks: Now that I’ve blended the two images to my satisfaction, I merge all the layers into one layer. Then I duplicate it and turn off the visibility switch of the original layer (to preserve a copy of the image before I started messing around with it). From now on, I will make each adjustment on a duplicate copy of the original layer, finally merging all layers into one when I’ve made all my adjustments and when I’m happy with the way that each adjustment has affected all the others.

Rather than trying to get the exact look that I’m after with only one or two plug-in filters, I’ve discovered that applying tiny little bits of many different filters and effects, achieves a much more realistic (and pleasing) end result. Instead of applying an effect directly to a layer, I will always apply the effect to a duplicate copy of the layer, and then adjust the transparency of the effected layer until I can only just make out the difference between the original and the effected layer (by toggling the effected layer’s visibility switch). Then I merge the (mostly transparent) effected layer with the original untouched layer.

One of the things that I always take great care in avoiding while processing my images is applying some effect that discards any of the image’s details. I’m very, very careful not to do anything that burns out any of the original highlights, or further darkens the shadowed areas!

Although I usually process each image slightly differently to all the others (as the situation requires), over the past few months I’ve discovered that the following effects and filters give me exactly the kind of look that I’m trying to achieve...

Firstly, I adjust the Highlights, Mid-tones, and Shadows to try to reveal any hidden detail and to try to balance the tones.

Next, I may (or may not) use nik’s Graduated Neutral Density filter again, to darken the sky a little more, or to lighten or darken the foreground a little.

Then, I’ll always use a little bit of nik’s Sunshine filter, which although difficult to describe in words, can transform and breath light into even the dullest of images.

Then, to add a little bit more warmth (when required), I’ll often use a miniscule amount of nik’s Skylight filter (this one can really bring out all the gorgeous sunset colours).

Then, I might also add a bit more brilliance to the primary colours by applying a little bit of nik’s Brilliance and Warmth filter... not too much, because this one can really destroy all your highlights!

When you view a merged layer of all the effects that you’ve applied so far, you’re going to be in for a shock... it’s made everything too dark, and completely over-saturated!! This is when I use nik’s Contrast Only filter, which despite its name, will allow you to adjust the brightness, contrast AND saturation. I always decrease the saturation, and I usually increase both the brightness and contrast using this filter.

Next, I apply some (soft light) high-pass sharpening with the radius set to between 4 and 6 with 60% strength. This is another stage of my processing workflow where I’m always extremely careful not to destroy any image data by over-sharpening and introducing haloes.

Finally, I use nik’s Lighten/Darken Center filter to ever-so-slightly lighten the center and darken the edges. This subtle vignette effect helps to give any image a sense of depth, and can often make the world of difference to an otherwise flat image!

So, having applied all my image adjustments, the vertorama in my tutorial now looks like this...



Okay, now I’ve just spent one entire evening sharing all of my image-processing secrets with everyone... Why did I do that? Well, if there is only one person who feels inspired enough to try what they’ve read here today, then I feel that it will have been worth every minute that I’ve spent typing this!!

Date: November 2007
Originally posted at 8:54PM, 5 November 2007 PST ( permalink )
Panorama Paul edited this topic 21 months ago.