That Venice pic - a workflow example
Discussion
As mentioned in my other thread, here is a description of how I did this shot:
It's not all that often that I produce something I'm really happy with, so when it does happen I tend to analyse what went right in the hope that I can apply some of the lessons next time. I'm sure forum regulars will let me know if this turns out to be useful, or just spectacularly self-indulgent.
Step 1: What to shoot?
Anybody else watching a series on BBC1 called Damages? The titles are a series of B&W shots of New York - lots of contrast, interesting detail on buildings etc. With this kind of thing in the back of my mind I was looking for something vaguely similar from my Venice trip.
Step 2: Location
We had a wander round on the evening of our arrival, and whilst St. Mark's Square was impressive, I couldn't immediately see what I could do with it. The arcade around the edge looked more interesting - arches for framing, interesting stonework, shiny floors.
Step 3: Light / Timing
It's a busy place, so to avoid huge crowds getting in the way it was going to have to be early or late in the day. As it turned out, I was there at sunrise so the square was almost empty, and the clear sky made this particular location work well.
Step 4: Capture
A couple of years ago when I was shooting film this step would have been called "taking the picture" and I would have used grad NDs to try to nail the exposure there and then. These days I reckon that the decisions about the picture are split pretty evenly between the point of shooting and the computer work afterwards. My priority at the time of shooting is now to capture everything I'll need to make the picture I want later.
In this case that meant a tripod and a number of bracketed exposures. The bright sky and the dark ceiling were never going to work in a single exposure so I did a series from too dark to too light, checking the histogram on the back of the camera to make sure everything was recorded. I used a shutter speed bracket to ensure that they would be easier to combine later - if I'd changed the apeture then out-of-focus areas wouldn't be the same across the images.
Step 5: Luck
Just as I was finishing I spotted two Carabinieri walking around the square. It's always good for composition to have a strong focal point so I waited until they came past - it's not a problem as everyone in Venice has a camera and nobody pays much attention to you.
As it turns out their presence made the picture, and I can't take the credit for that. It was just luck that they passed, and that they were in that recognisable uniform. All I can say is that I put myself in a position where something lucky could happen, and I caught it when it did.
Step 6: Processing
Back at home I opened the files up in CaptureOne and decided that all the information could be seen in two exposures. I processed these to show the relevant detail, but kept the appearance as similar as possible to make it easier to combine them. They were processed to 16-bit Tiffs to make it easier to extract shadow detail later.
In Photoshop I layered the images on top of each other, and checked the alignment by reducing the opacity of the top layer and moving it about. When I was sure they were lined up (shooting with a tripod really helps here) I put the opacity back to 100% and linked the layers together.
To get the ceiling in I'd had to point the camera up, which resulted in converging verticals. Diagonals and angles add interest and drama, but I was after a calm, composed feeling for the end image. To sort this out I applied the grid, selected all and transformed the image with a combination of perspective, skew and distort until the verticals across the image were close enough to the grid lines. As the layers were linked, the transformation was applied to both of them.
The next thing was to sort out which layer was to contribute what. It's possible to use an HDR application like Photomatix to do this, but as I didn't particularly want that 'HDR feel' I did it using a layer mask. It was applied to the top (darker) layer, and looked like this:
So where the mask is white the layer shows 100% and where it's black the layer is completely masked.
Step 7: The other layers
The final image is composed of these layers:
In order, from the bottom up, these do the following:
Lighter: The lighter of the two component images
Lighter SSharpened: As Lighter, but with some Smart Sharpen filter applied to bring out detail
LH top Corner lights: A Curves adjustment layer masked to act only on the highlights of the lamps in the corner. The bright highlight was distracting, drawing the viewer's eye away from where it should be.
Levels 1: Adjustment layer to increase contrast of shadow details
Darker: The darker of the two component images
Curves 3: Curves adjustment layer, masked to darken foreground and some other details
Curves 2: Darkens areas visible through the arches
Curves 1: A mild brightness increase for the central area. but not the arches
Hue/Sat 1: Set to colorize: Hue 45, Saturation 10, Lightness 0. Converts to monochrome, with a subtle tint. The slight warmth of the colour helps the mood - I tried it with blue, but it just didn't work.
...or Colorize: Is a layer group containing the Hue/Sat layer. When I was deciding what effect to use I created several groups, each with a different effect - this is the last one standing.
Crop?: A mask indicating the likely crop. As cropping permanently loses information, I don't apply them the last possible moment.
This series of layers evolved as a result of looking at the image and deciding where it was lacking. It would probably have been possible to get the same result with fewer layers, but I wasn't sure where I was going until I got there.
The final step was to remove the small group of people from the background using the clone stamp. Apart from this and the initial transform, every other change is non-destructive, and could be reversed if necessary.
The final version is too big to post, but here are two 100% crops:
I'm certainly not suggesting that this level of effort is essential for every picture, and in fact it's by far the most involved of the ones I did in Venice. It's more an example of how having an idea of what the end result should look and feel like can affect the decisions made at capture and in processing.
It's not all that often that I produce something I'm really happy with, so when it does happen I tend to analyse what went right in the hope that I can apply some of the lessons next time. I'm sure forum regulars will let me know if this turns out to be useful, or just spectacularly self-indulgent.
Step 1: What to shoot?
Anybody else watching a series on BBC1 called Damages? The titles are a series of B&W shots of New York - lots of contrast, interesting detail on buildings etc. With this kind of thing in the back of my mind I was looking for something vaguely similar from my Venice trip.
Step 2: Location
We had a wander round on the evening of our arrival, and whilst St. Mark's Square was impressive, I couldn't immediately see what I could do with it. The arcade around the edge looked more interesting - arches for framing, interesting stonework, shiny floors.
Step 3: Light / Timing
It's a busy place, so to avoid huge crowds getting in the way it was going to have to be early or late in the day. As it turned out, I was there at sunrise so the square was almost empty, and the clear sky made this particular location work well.
Step 4: Capture
A couple of years ago when I was shooting film this step would have been called "taking the picture" and I would have used grad NDs to try to nail the exposure there and then. These days I reckon that the decisions about the picture are split pretty evenly between the point of shooting and the computer work afterwards. My priority at the time of shooting is now to capture everything I'll need to make the picture I want later.
In this case that meant a tripod and a number of bracketed exposures. The bright sky and the dark ceiling were never going to work in a single exposure so I did a series from too dark to too light, checking the histogram on the back of the camera to make sure everything was recorded. I used a shutter speed bracket to ensure that they would be easier to combine later - if I'd changed the apeture then out-of-focus areas wouldn't be the same across the images.
Step 5: Luck
Just as I was finishing I spotted two Carabinieri walking around the square. It's always good for composition to have a strong focal point so I waited until they came past - it's not a problem as everyone in Venice has a camera and nobody pays much attention to you.
As it turns out their presence made the picture, and I can't take the credit for that. It was just luck that they passed, and that they were in that recognisable uniform. All I can say is that I put myself in a position where something lucky could happen, and I caught it when it did.
Step 6: Processing
Back at home I opened the files up in CaptureOne and decided that all the information could be seen in two exposures. I processed these to show the relevant detail, but kept the appearance as similar as possible to make it easier to combine them. They were processed to 16-bit Tiffs to make it easier to extract shadow detail later.
In Photoshop I layered the images on top of each other, and checked the alignment by reducing the opacity of the top layer and moving it about. When I was sure they were lined up (shooting with a tripod really helps here) I put the opacity back to 100% and linked the layers together.
To get the ceiling in I'd had to point the camera up, which resulted in converging verticals. Diagonals and angles add interest and drama, but I was after a calm, composed feeling for the end image. To sort this out I applied the grid, selected all and transformed the image with a combination of perspective, skew and distort until the verticals across the image were close enough to the grid lines. As the layers were linked, the transformation was applied to both of them.
The next thing was to sort out which layer was to contribute what. It's possible to use an HDR application like Photomatix to do this, but as I didn't particularly want that 'HDR feel' I did it using a layer mask. It was applied to the top (darker) layer, and looked like this:
So where the mask is white the layer shows 100% and where it's black the layer is completely masked.
Step 7: The other layers
The final image is composed of these layers:
In order, from the bottom up, these do the following:
Lighter: The lighter of the two component images
Lighter SSharpened: As Lighter, but with some Smart Sharpen filter applied to bring out detail
LH top Corner lights: A Curves adjustment layer masked to act only on the highlights of the lamps in the corner. The bright highlight was distracting, drawing the viewer's eye away from where it should be.
Levels 1: Adjustment layer to increase contrast of shadow details
Darker: The darker of the two component images
Curves 3: Curves adjustment layer, masked to darken foreground and some other details
Curves 2: Darkens areas visible through the arches
Curves 1: A mild brightness increase for the central area. but not the arches
Hue/Sat 1: Set to colorize: Hue 45, Saturation 10, Lightness 0. Converts to monochrome, with a subtle tint. The slight warmth of the colour helps the mood - I tried it with blue, but it just didn't work.
...or Colorize: Is a layer group containing the Hue/Sat layer. When I was deciding what effect to use I created several groups, each with a different effect - this is the last one standing.
Crop?: A mask indicating the likely crop. As cropping permanently loses information, I don't apply them the last possible moment.
This series of layers evolved as a result of looking at the image and deciding where it was lacking. It would probably have been possible to get the same result with fewer layers, but I wasn't sure where I was going until I got there.
The final step was to remove the small group of people from the background using the clone stamp. Apart from this and the initial transform, every other change is non-destructive, and could be reversed if necessary.
The final version is too big to post, but here are two 100% crops:
I'm certainly not suggesting that this level of effort is essential for every picture, and in fact it's by far the most involved of the ones I did in Venice. It's more an example of how having an idea of what the end result should look and feel like can affect the decisions made at capture and in processing.
Edited by S. Gonzales Esq. on Friday 28th March 00:01
MaceV6 said:
What sort of time frame did that work take you?
Probably 2-3 hours all in, spread over a day or two. Quite a lot of that time was spent looking at it and thinking about what to do next, as well as thinking about composition and how the viewer was going to look at it. I wanted the important parts of the picture to get attention first, but then provide a lot of detail that rewarded further study.Hopefully, your eye is drawn along the arches and then across the views through the arches to the people. From there you probably move down to the highlight on the floor, and only then start looking at the details.
Actually, I'd be interested to know if this really is how people view it - it's all very well me trying to influence this kind of thing, but I don't know if it works for anyone else.
Edited by S. Gonzales Esq. on Saturday 29th March 23:19
Gassing Station | Photography & Video | Top of Page | What's New | My Stuff