sinwp rss feed for PI Articles

articles/Nature/scoreandperformance

The Score and the Performance

by

Even now after spending over six years producing my prints digitally I am often bewildered by the way in which people complicate the matter. Photoshop as a tool is wonderful and it is often best to bear in mind that hardly anybody (if anybody) know how to use all the controls within it and more importantly why! I still recall the first time I sat down and tried to get to grips with all the menus' dialogue boxes and tools and making a complete hash of the image on the screen. I also remember a point were I was no longer looking at the image I was working on, but staring at the controls because I was just trying to make sense of it all.

This spooked me. I felt like some geek sliding into the dark side! Oh no I was becoming a techy and the artistic photographer in me was dying. More importantly I was not really enjoying what I was doing. This brought me right down to earth and I shook myself up and reminded myself that the image is what it is all about and not the process getting there. I fully understand that when we embark on a new learning curve we will take time to understand how the master the new tools and methods but what Photoshop was doing to me, and from what I have seen, does to a lot of people is to take over and become the focus of the activity. I was only interested in doing what I had always done to produce an image that delivered a visual statement of what I experienced when I was stood there in the landscape.

All I wanted to achieve was what I did in the dark room without any of the distractions of buttons and menus. Don't get me wrong, all of the buttons, commands and menus had perfectly useful functions but not for me and I have never been impressed by anybody who knows how to do things in Photoshop in 10 different ways. WHY?

So what did I do in the dark room, then? I used to darken parts of the image and I used to lighten parts of the image. I would maybe straighten a horizon, retouch dust particles or raise and lower contrast. More importantly I used to enjoy it. I enjoyed watching the image being created and the moment being re-born which is essential to making a good photograph. I returned to the keyboard, mouse and monitor and set myself a challenge. To learn and utilise only the functions and tools that would give me the control I needed and nothing more. If in the future I came across a problem that Photoshop could help me with then I would surely learn a new skill, but the greatest problem people have with Photoshop is trying to learn it all at once.

In the dark room I used pieces of card with holes cut out which would let light through from the enlarger which would darken that part of the image that was allowed to be exposed. Conversely I have narrow strips of wire with small pieces of card glued to them in different shapes. When I held these below the enlarger light they would create small areas of shadow on the photographic paper and therefore render that part of the image lighter. If you held either of these closer to the enlarger lens you would create a greater penumbra and so have a softer edge to your image adjustments. The opposite would be holding your piece of card closer the base-board of the enlarger creating a narrower penumbra or harder edge to you image adjustment.

This is known as the wonderful process of 'dodging and burning'. To raise and lower contrast I would introduce or reduce magenta as required in the enlargers' beam of light. Another habit I transferred from the darkroom was breaking up the image into 'working zones'. By this I am not referring to the Zone System, but a means of grasping what individual areas of the image need working on without just diving in and seeing what comes out the other end. For this image of Glen Coe there are four clear working zones. The sky, the sunlit mountain flanks, the loch and the reed bed in the foreground. During the printing process I may indeed make small selections within the 'working zones' but what defines these are natural breaks on the tonal range of the image and if I bear this in mind I can essentially segment my workflow.



Now that I knew what I wanted to emulate in Photoshop I began to explore the controls available to me and began a journey of simplifying things. After some time I immerged triumphant with my new slim-line tool-kit that consists of the lasso tool with feather, quick mask and curves adjustment layers. Not very techy eh! But works a treat for me and does all I need it to.

Stage 1

The first area I wanted to adjust contained two 'working zones' which are the mountain flanks and the sky. When we make a selection using the lasso tool we have to commit to setting a feather size. It is worth noting two facts. The feather size is in pixels and so the size of the feather in you image will be directly proportional to the image size. For example if we set a feather of 50 pixels this feather would appear to be far larger (softer edged) in an image of 500 pixels by 500 pixels compared to an image of 5,000 pixels by 5,000. For this reason alone there is not much point in asking somebody what feather they have used for a particular selection unless their image was identical in pixel dimensions.

In this example I have made a selection of the top of the image then clicked on Quick Mask in the tools palette. What this shows is a red area and an area that look normal. The red area is protected and the curves adjustment layer will affect the area that looks normal. The transition from red to normal is he size of the feather.

Stage 2

Having seen our feather size in Quick Mask I then turned it off returning to a normal image without the red masked area and now showing the marching-ants. I then selected a curves adjustment layer from the layers palette to make my adjustment of the sky and mountain flanks.

At this stage consider it is what you want to do. The reason I made this selection with two 'working-zones' is because I wanted to darken it, overall. I could have selected each area individually but if two workingzones can be altered at the same time then this reduces the risk of the 'patchwork' effect of many different selections in one image.

I then decided on what areas of tonality I wanted to effect. Although I wanted the paler greys in the sky to be 'calmed' slightly, the most important adjustment was the deepening of the shadow values. I selected the part of the curve line that closest represented the shadow 'grey' and pushed the line upwards. (The opposite direction in RGB). As I did this I kept an eye on the image at all times. Remember to do things gradually

The next area I wanted to concentrate on was the sunlit mountain flanks. With the last adjustment curve these values had been slightly darkened by the curve, but this was essential to introduce good shadow depth. What I needed to do now was to maintain that shadow depth but bring out the sunlit highlights. I made a rough selection of the areas of the mountain flanks leaving the shadowed mountain tops alone as I was happy with their values.

It is important to not get too hung up with super-accurate selections unless absolutely necessary. I have seen many photographers with their noses almost pushed against their monitor selecting with the magic wand around every tree branch. This is of course a useful tool in some cases but I would recommend stepping back and learning about exposure control before this becomes your standard practice as the results often look odd to say the least!

Stage 3

With my curve dialogue box open I then selected part of the curve line closest the highlight values of the sunlit flanks and pulled the curve downwards until the sunlit areas were much improved. I then chose the shadow end of the curve line and pushed it upwards, back to where it had been pulled from and in doing so retained those all important deep shadows. The sun was beginning to come out!

Stage 4

Now I gave my attention to the loch in the foreground of the image. When I was there, the water looked black and deep with the reeds at my feet, glistening in the low sun. I made a selection once again with the lasso tool of the bottom half of the image. Having brought up my curves dialogue box I knew straight away that I would anchor a point on the curves line towards the darker values and push upwards seeing the water darken in the image as I adjusted.

Whilst doing this is was clear that the lighter values of the curve line were also being pushed darker which would lessen the sunlit effect on the foreground reeds. I then anchored a point on the curves line for the lighter greys and 'pegged' this back to the original position so reintroducing the glisten in the reeds.

Stage 5

Whilst working on an image do not be taken over by the small sections but try to maintain an overall impression of what you are 'creating' as a whole. I advise that, when pushing and pulling curve lines, have your entire image in view on your monitor as the tonal adjustment you are making has a direct relationship to its neighbouring tone in the image and also the overall 'balance' of the image. When working on this image there was a selection of the top left sky that was bright which slightly upset the balance of this composition. Because I was intending to make a selection that would inevitably go right across the centre of the sky I set a large feather and made a rough selection of the area that needed calming. I then tapped 'Quick Mask' (the letter Q) to double check my feather was soft enough.

Stage 6

Using this selection I wanted to slightly mute the paler sky and introduce some more of the subtle mid-greys. With a curves adjustment layer I anchored a point on the curves line just above the mid-greys and pushed the line up slightly, quickly achieving the desired result and balancing the sky in the image.

I do not think it always pays to return to an existing curves selection to 'give it a little more' later in the proceedings. This is because you will find that further refining will not need to be applied to the same overall selection area.

This was the case with the sunlit mountain flanks. I felt they needed a little further 'lift' the emulate what I saw when I was there with my camera, that evening. Once again I made a new, smaller selection of the lower, sunlit flanks with the lasso tool and, with a curve, I pulled the highlights out a little further as well as 'pegging' the shadows back to retain their depth and impact. A small adjustment, but at this stage subtleties are very important and this is where many can ruin the image or lose track of what they are doing.

Stage 8

Continuing with slight refinements I now turned my attention to the foreground reeds at the very foot of the image. It was these reeds that still needed some life returning to them and more importantly separating their bright values from the dark water. Once again using a large feather I selected the reed I felt needed attention and applied a curve which pulled out the wonderful highlights and then I pegged back the darker water values behind making the reeds glisten once more.

Stage 9

The last adjustment was one to help maintain the viewer's eye within the bounds of the image. In the top left corner of the image was an area of grey that I felt was ever so slightly too pale and offered an escape for the roaming eye. This was easily fixed with a quick lasso selection and a curve in which I pushed up the white end of the curve line on its axis, gradually removing white pixels and muting the area to a matching darker grey.

And there it is. My simple method of what I practiced for over twenty years in my darkroom and emulated in front of the monitor in the office. My last advice is to take you time and keep the selections simple and the adjustments gradual. Always look at the overall image during each adjustment and please try not to do it in big, bold steps. After all this entire image was created using one selection tool, one adjustment tool and seven layers. That's it! Not much to learn but what we have seen is raw information being the 'score' and the finished image becoming the 'performance'.

See also the discussion on soft proofing for perfect black and white on your monitor in the next section.

Soft Proofing for Monochrome - Is it Possible

A few things came together to make this couple of pages possible. We stumbled across the Eric Chan icc profiles for ABW (see previous issue's barytas article), we had some discussions with Simon Prais at Colour Confidence and finally we had in front of us a freshly Fogra-certified, quality monitor. In theory, at least, we should have been in a position to do some accurate soft proofing.

Monochrome proofing is different to colour proofing. Normally you are looking for a colour match between either two print samples (say press and proofer output) or between a screen and a print. Variations in luminance may be masked by variations or similarities in hue and saturation when viewing colour. When it comes to monochrome (and assuming you do not have a colour-biased print) the only thing you have to go on is the monochrome luminance - quite a difficult proposition.

Initially we found it difficult to discern differences between two soft-proofed views, one proofed to a bespoke colour profile and the other to an Eric Chan monochrome, ABW profile. We therefore devised the following test and had four people participate in the trials. The tests were carried out 'blind', that is all the testers worked by viewing just the print and the soft-proofed screen view; they were unable to see the numerical values of their adjustments as they made them using either sliders or arrow keys to ratchet the values up and down. We applied either a curve, HSB move or levels adjustment, as appropriate, to vary the luminance of the print, its gamma value (in Levels) or the bow of the curve.

Changing the curve proved to be too difficult to master without seeing the palette and so we ended up concentrating on just the HSB Brightness slider and the gamma of the Levels dialogue. The testers were presented with a very dark or very light screen image and asked to adjust the values until they felt that the screen soft proof matched the print standing in a viewing booth. The viewing booth was set at 2,000 lux and 5300°K; tests outside the viewing booth were in mixed light of low strength. The screen was set at either 75, 80 or 160cd/m2, but mostly at 75cd/m2. A significant advantage of the ColorNavigator software from Eizo is that you can easily switch between profiles and thereby change the screen luminance.

An initial finding was that the testers improved dramatically, both in speed of adjustment and accuracy, in the first few attempts. One tester dropped from an initial error of 17 points down to consistent errors centred around ±2 points over 10 attempts. The quality of the viewing environment also had an influence; darker viewing was related to darker judgement settings. The initial errors were eliminated from some of the statistical analysis.

Using a correct illumination viewing booth, the testers' results varied both above and below the correct luminance value suggesting that errors were in judgement, not a bias in the viewing conditions - if, for example, the viewing conditions had been too bright we would have expected errors biased in that direction. The judgement errors varied between ±7 points in brightness value equivalent to -10RGB points and +6RGB points (ie a 16 RGB points variation). To put some perspective on this value we print the target here, set up at ±5 points and ±12 points so you can see the difference.

We repeated the test using the gamma slider of the Levels adjustment to check out the influence and ability to detect changes in contrast. This too was variable and showed no bias either way.

Visually the print looked crisper and more tonally separated in the viewing booth than the screen soft proof. Most observers, we feel, would describe the actual print as having more 'quality' than the screen view. The differences between a 'full-colour' and an ABW print were just possible to detect with the prints in hand and were confirmed as different when measured by densitometry (the ABW/Dark print was at 15.8% luminance in the top left corner; the full-colour print was at 14.9% luminance. The file was around 12% luminance). The full-colour print had more density than the ABW print in the shadows but the overall impression of the prints was that they were visually identical when viewed from a normal distance although there was a subtle difference in the highlight tone contrast levels (see graph).

We did manage to detect the difference between the highlight rendering using soft proofing of both colour and ABW/Dark but we would not like to be betting the mortgage on our ability! However, we consistently struggled to differentiate either prints or soft-proof views, even with the images side by side. This is, of course, good news, they were, after all, excellent prints made singly by two routes, of almost equal merit.

On-screen differences were slightly masked by subtle differences in rendering of the actual screen; one print looked warmer at the top but then the result was reversed if the other print was soft proofed at the top of the screen. In hand, the full-colour print looked slightly magenta and the ABW looked slightly greener. In reality the ABW was closer to the base white and the full colour was indeed about 1 Lab point towards magenta.

After a number of hours of testing we came to the conclusion that you can just about soft proof with either the full-colour or ABW profile but that differences in the viewing 'experience' mask any differences between the actual prints made by the two methods.

The aim of this process was to test if it is possible to tonally adjust a print on screen and then create a print that is both accurate and satisfying to the viewer. We often express this as delivering 'no surprises when the print pops out'. Our gut feeling is that this should be possible and that both ABW soft proofing and a good full-colour profile would deliver equal performance. The errors we measured are shown as examples here but may not reproduce with complete precision. In hand, the inkjet prints were readily differentiated at the ±12 level and just about at the ±5 level (on brightness). The level achieved by our testers (nominally ±7) therefore indicates that soft proofing is possible as a control measure but only if care is taken; it would not take much sloppiness for the whole thing to go belly up!

We had to work hard to get to this stage and we are indebted to Toby Herlinger of Fotospeed, who provided us with enough sample sheets to get through our testing. Most of the analysis was conducted on the Canson Baryta Photographique 310gsm paper; it has certainly caught our eye for its delightful quality.



Updated 27/04/2026 16:44:22 Last Modified: Monday, 27 April 2026