The “Very Large Montage” – How was it done….
I have had a couple of requests today asking how I put together the montage last week entitled “Very Large Montage” …. what an appropriate name. The finished image can be seen above.. I am going to explain this as a step by step guide from the very beginning actually shooting the images.
It was quite impulsive to be honest. I was stood composing an image of the bridge, it has quite a photogenic look to it. I saw the woman walking across the bridge and thought of an opportunity to do something quite different. Keeping the focal length the same I decided to shoot half a dozen images of her actually walking towards me. I tilted the camera to create different angles, all the time making sure I had her in the frame. Samples of these standalone images are shown below.
By the time the knowing actress had passed over the bridge I had captured at least two dozen images of just her. I then made sure that I kept my feet rooted to the same spot and making sure that the focal length of the lens, exposure, shutter speed and ISO were all kept the same I then proceeded to capture the base image for the huge photostitch. I wanted to emphasise the tall bridge stanchions in the final image so I shot plenty of frames that contained this part of the structure. Again, tilting the camera to grab them at different angles, doing it this way makes it much more interesting when doing a photostitch. I also made sure that every image I shot overlapped quite significantly and that I had also obtained enough to complete the full expanse of the scene.
And that is it. All the images from the location are complete, it’s just a case of bringing them home and letting your skill and the power of the modern PC do the rest. There are a number of ways that an image like this can be put together. Some of these are listed below.
1. To manually place all the images on a blank canvas and let them overlap slowly building a complete picture. This is very “David Hockney” and looks quite surreal but fantastic when its complete.
2. Automate the background image using software designed for stitching and then place the figures on manually using Photoshop to add layer masks and blending modes
3. Use a stitching program to complete the whole task, cross your fingers and hope it doesn’t get confused seeing the same figure multiple times.
I suppose there are many more ways an artist can portray this type of image but for the sake of this blog we will stick to these three. For this stitch I used option 2 and let Microsoft ICE do the donkey work and create the background in automatic mode. This would enable me to concentrate on editing the individual images of the figure and matching up the backgrounds to look acceptable. A point worth noting is that I never intended for the edges to be straight, they always were going to be uncut and “shabby”… That was the look I wanted to portray. As with anything, there are no rules when you do this yourself, try a little creativity and come up with something really left-field, its fun.
As you can see, the composite image that MS ICE created has included the original figure from the first frames. When creating the stitch, ICE sources the images from a folder on your hard-drive. I made sure that there were no other images that contained the young lady. This would have only complicated matters as the program reads the images and decides where to place them. I think a moving figure would have confused it somewhat. Anyway, this was a great baseline to work from. I started from the back just placing the first image of the girl so as many of the edges matched with the background as possible. It was never going to be possible to get them all to match but as long as it looked workable then I would be happy. The trick now is not to work on that new layer you have just inserted with the figure on, but to lock that layer in place, make it invisible and move on to the next. This step is then repeated until all the final figure is in place. Now you can make all the layers visible again and you get a good idea of what the final piece is beginning to look like.
It’s now time to work on each figures layer and decide what you want to remove using layer masks. As you can see from my final image, I chose to leave the bottom half of the figures unmasked. I actually spent over 30 minutes masking each figure around each foot, leg and arm and then decided that it didn’t look quite right. It looked somewhat “impossible” without the slightest hint of surrealism. One of the great little tools in Photoshop called the masking tool enables you to remove what you have already masked by merely changing the colour of the brush from black to white. So, I decided to put back all the edges except the ones next to the bridge structure. Playing around with this for about 35 minutes I arrived at something acceptable.
I left all the outer edges of the background as they were when I had put them together and then copied the whole composite on to a black square background. Converted it all to monochrome and added a title. Voila !
Thank you for reading, if you have any questions then please do not hesitate to drop me a line.
Posted on April 7, 2013, in Blogging & Wordpress, Composites, Flickr, Photography (General), Photoshop and tagged composite, how it was done, mark winterbourne, photoshop. Bookmark the permalink. Leave a comment.