Week 9: The Constructed Image

Basic Digital Photography with Matthew Swarts

Charlie White.

Barbara Kruger.

Adam Fuss.

Vik Muniz.

Sandy Skoglund.

Zeke Berman.

Hannah Hoch.

Laurie Simmons.

Laurent Millet.

Nicholas Kahn and Richard Selesnick.

Joel-Peter Witkin.

For Next Week:

Download and review the slides from this weeks’ lecture: arts1850 slides the constructed image

Read: this Interview with Vik Muniz

Read: this Interview with Barbara Kruger

Read: this Interview with Adam Fuss

Whether or not you have been aware of it, up until this point in the semester we have been operating under quite a few rules and assumptions about what a photograph must be.  This week is about breaking the rules!   

Make 6-8 images or objects that challenge photography’s formal, historical, theoretical, and/or technical assumptions.  

Please upload (small jpeg) representations of your work to Dropbox!

To get you started, consider making work that exposes the truth or falsity of the following statements:

1. There is a correct way to expose a digital image.

2. There is a correct way to adjust and color correct.

3. There is a correct way to make a print.

4. Photographs are two-dimensional (i.e. flat!) objects.  They are never three dimensional.

5. Photography is a single image endeavor.  You must say it all with a single frame.

6. Photography is serious, not light-hearted or comical.

7. Photographs cannot be cropped down, resampled, or rephotographed.

8. You must present your work each week as either a slide show or a series of prints.

9. A photograph represents something that exists in the world.

10. A digital photograph must be made with at least a 5.0 megapixel camera.

11. The tools of Photoshop are not intended for the kind of digital photography this class is about.  (You shouldn’t paint all over the pretty pictures…)

12. A scanner is not a camera; nor, for example, is a Xerox machine.

13. You must interact with the computer to make your images.

14. Images made by a digital camera are RGB color images and they should be left that way.

15. Photographs should look like other photographs, for there is a long history to acknowledge and honor.

Using Filters

Photoshop has a unique collection of filters, or mathematical algorithms that change the ways that pixels are displayed, that can be applied either to an entire layer or any active selection. To see how this works, let’s look at an image:

With no selection active, any filter I apply to this base image will affect all pixels equally. From the Filter menu at the top of the Photoshop toolbar, I will select Filter>Stylize>Find Edges:

The Find Edges filter happens to be a one step filter, meaning there are no additional adjustment sliders available to fine tune the filter. When I apply it, my image immediately changes:

As with any editing operation, if this effect is too garish or wrong for you, consider applying a fade (Edit>Fade) immediately after applying the filter, as it gives you more options in terms of how the filter can appear:

With a small amount of fine tuning, I can fade the filter’s effects a little to create a hybrid image that combines the filter and the original image a bit more subtly:

And that’s it! I can apply any additional filter afterwards, or edit the image in any way we’ve previously discussed. However, the sad truth about using filters in this most simple way is that they are destructive to the pixels that they effect. Only by undoing them or moving into previous history states can you eliminate their effect. Editing sometimes becomes very difficult when you are working with filters. Luckily, there are better ways to filter, by creating something called a smart object out of the layer or selection you are interested in filtering.

Smart Filtering

In the Filter menu there is a better option for negotiating your filtering needs. With a selection active (or an entire layer you want to apply filter to selected in the layers palette), pull down Filter>Convert for Smart Filters:

A dialogue box will appear that will remind you about the fact that you are about to create a smart object:

A smart object is just that: a re-editable, math-based object that can now be filtered in such a way as to not destroy or affect underlying pixel information. When you create a smart object (ready in this case for smart filtering) in this way your layer information will be displayed slightly differently, with a graphic notation that denotes, well: smartness!

Now, with our selection active and converted to a smart object, we can really explore filtering with some of the freedom of adjustment layers. That is to say: we can edit, re-edit, filter, re-filter, alter masks, etc. without affecting the underlying pixel information permanently. To see what your options might be, once you have a smart object created and selected in the layers palette, pull down Filter>Filter Gallery to display access to most of Photoshop’s filters in a single interface:

When I did the above, my screen interface changed to bring up the filter gallery itself:

Each of these filters can be applied independently to your smart object, or in combination. In my case, I began my exploration of filtering possibilities for this image by selecting the Cutout filter from the menu. Note how the interface shows slider controls for altering the filter in the upper right hand corner of the window. The filter you select is displayed in a palette in the lower right hand section of the window, much like information in the layers palette:

When you adjust the sliders available to you in the upper right hand corner, the filter’s effect will change in the preview window in the center of the screen. Adjust the sliders until you are content with the way your image looks:

For each filter, the specific sliders will change. But once you play with a few of them, you will begin to understand your options. In my case, just the boy was selected in my smart object, so when I applied filters, my image changed as follows. Here’s Poster Edges, for example:

The important thing to conceptualize is that you can apply many (or just one) filter from within the filter gallery interface. To apply another filter to your already filtered image, just select the new filter icon (looks like a blank page) from the lower right hand corner of the filter gallery window. As you click on this new filter layer icon, you will begin to stack filter effects in the palette above. Filters will be applied in the order that they rest in the palette, with the bottom one being the first filter applied, and all others that rest on top of it will be applied subsequently:

Obviously, to delete a filter layer, just drag it to the trash icon at the bottom of the palette. When you are satisfied with how your image looks, simply click OK and you will return to the normal Photoshop interface.

Take note of the way a smart filter layer looks in the layers palette:

Clicking on the icon to the right of the filter name you applied will draw up the controls for that filter, so you can re-edit with ease. Clicking on the mask that denotes the selection you made prior to smart filtering will draw up the Properties window for that mask, which can also be edited in terms of feathering and density concerns, just as with adjustment layers!

If you are going to apply many filters to your image, using a variety of selection masks, it is likely a good idea to make a simple copy of your background (or base) pixel layer, to preserve that layer for future edits and selections, as it is a bit more tricky to edit the selection of a smart filter layer. To copy your background layer, simply drag it to the new layer icon at the bottom of the layers palette, then turn off the original:

Remember along the way that your filtering should be applied with the same scrutiny as your adjustment layers and other operations. To put it simply: a lot of damage can be done to your image very quickly with filters, not physical damage in the case of smart filtering, but psychological damage in how the valence of the image will read to your viewers. Apply filters in final form only when they significantly improve the effect you are trying to achieve with your viewers. If applying a filter simply makes your image look ‘filtered’ you are probably stalking the wrong set of operations for improving your work. We now operate in a visual culture where the language of Photoshop is becoming a part of the vernacular. People recognize, in other words, when something has been Photoshopped, mostly from abuse of filtering with images. Using filters effectively, in other words, is a subtle game. Use them wisely!

Using Text

Adding text to an image is relatively easy in Photoshop, and most of the controls involved should be familiar from word processing operations. To begin, let’s consider adding text to the following image:

In the example above, I have already selected the Type Tool, and the Photoshop interface has responded appropriately. The Type Tool is located in the lower third of the vertical toolbar on the left hand side of the screen, and looks like this:

The Type Tool is actually a cluster of type tools, be certain you are selecting the right one, as there are horizontal and vertical versions:

Take note of the horizontal toolbar at the top of the screen, as it will change to reflect the Type Tool interface:

Moving from the left hand side of the horizontal toolbar, the font is displayed and selectable from a pulldown menu in the left hand side of the toolbar:

The font style is the next option over, and is also selectable from a pulldown:

Next, select the approximate font size from the next pulldown over:

Note that if you also simply input a number into the font size box, the interface will prepare vector based type at that size, so don’t fret if you can’t find an appropriately small (or large) enough font size. In my case, when I inputted 150pt type into the font size box, my cursor changed to reflect the Type Tool insertion point, and I was able to type the word “listen” over the pixelated representation of my image, after selecting white as the text color from the color picker:

Once you place your text, your layers palette will reflect the new type layer appropriately:

To move the text layer, simply click on the Move Tool and rearrange where the text rests over your pixel layer.

You can also add effects to the text via the layer styles option, which is selectable by clicking on the following icon at the bottom of the layers palette:

Clicking here will bring up the layer styles window:

In my case, I wanted to add a gentle Drop Shadow to my text layer. I selected the appropriate checkbox in the left hand section of the Layer Styles window and then adjusted the appropriate sliders until the shadow effect looked the way I wanted it to:

Note how when I performed this operation, the notation reflected the new layer style application in my layers palette:

At any time, I can click on the [T] icon in the layers palette to re-activate the Type Tool and edit anything about my type layer. In my case, when I did so, the type in my image highlighted as follows:

With such a simple operation I was able to decrease the font size by inputing a new value in the appropriate input box in the horizontal toolbar:

I mentioned already that the Type Tool lays down vector based information. This means that the representation of type in your image is purely math, not pixels. It is size-independent from the underlying pixel layer, and can be scaled up or down without any degradation to the quality of the type. Unfortunately, all of Photoshop’s powers (for example: filtering!) are also unavailable to vector based information.  To edit the type layer as if it were an image like the underlying pixel information, it first needs to be rasterized, or converted to pixels. To do this simply select Type>Rasterize Type Layer.

Once your type is rasterized, it will appear in the layers palette as pixelated information. The type part of it will no longer be editable (so check for spelling and grammar of course before rasterizing anything!). The good news is that you can now treat the type layer as if it were an image, and filter, adjust, paint, etc within that layer. In my case, I simply applied a small gaussian blur (Filter>Blur>Gaussian Blur) to the layer to make the type a little fuzzy:


Basic Color Management Issues

In an ideal world, all of the digital equipment we use in class would be color calibrated to perfection with professional color calibration tools. Color calibration is of vital importance when it comes to printing digital images, as without a properly color calibrated system, there will be very frustrating differences between what your monitor displays and what your printer creates. Color calibration eliminates this frustration by checking the color accuracy of both the monitor and the printer, and establishes a link between these two disparate ways of representing images. When a monitor is calibrated correctly, you can be 100% certain that it is representing color accurately. Color calibration tools differ widely, but essentially, a device is used to measure the color emanating from the monitor. The device checks the color the monitor produces against known constants, and makes adjustments to the way the monitor displays information accordingly.

Color Calibrating a LCD Monitor.

Once the monitor is properly calibrated, the printer’s color must be measured against known standards as well, and a profile must be written by special software for each specific type of paper used.  Again, a device is used to measure how accurately the printer is reproducing color:

This profiling process is exacting and tedious, and must be repeated at regular intervals and whenever ink is changed in the printer. As you can imagine, this type of care requires round the clock maintenance, something the College is unable to currently staff. As such, our system is less than ideal but will still produce satisfactory results when appropriate care is taken along the way. Assume for now that you are working on a calibrated system, as the following steps for ensuring printing accuracy still apply.

Sizing Images for Ideal Output

The first step in preparing your image for output (printing) is to size the image for proper dimensions and print resolution. To do this, simply select Image>Image Size from the Image menu at the top of the Photoshop toolbar:

The following window will then appear on your screen:

The essential information in this dialogue box for our purposes is the print Resolution, which for photographs must be at minimum 300 pixels per inch. Ideally, your document size dimensions should physically fit on your paper of choice at this resolution. (Note: higher resolutions are fine, just do not attempt to print photographic information at print resolutions of less than 300 pixels per inch, as the results will be substandard and frustrating.)  In the example above, the ‘Resample Image’ checkbox is left unchecked. This means that Photoshop will only use the pixels available to the file, without interpolating any information to make up for less resolution. Interpolation is a mathematical process that involves software-based estimation of tonal values, and therefore involves some error. If you can avoid interpolation, you should. If you can’t there are interesting ways around some of the challenges it presents.

Let’s say you are working with an image whose minimum resolution falls short of 300 pixels per inch at the document size you want to print. In this case, try checking the ‘Resample Image’ checkbox:

Next, consider adjusting the interpolation method. By default, Photoshop will select ‘Bicubic Automatic’. As you can see, however, from the available options in the pulldown menu, in some instances it will be better to select ‘Bicubic Smoother’ (for enlargement) or ‘Bicubic Sharper’ for reduction:

Once you establish the interpolation method you will use to arrive at a minimum print resolution of 300 pixels per inch, click OK and whatever interpolation method you have chosen will begin. Depending on the size of your output, this is a process that can take anywhere from a few seconds to many minutes to accomplish. The important thing to realize is that in any case where interpolation occurs, Photoshop will use estimation methods to determine how to fill in the necessary pixels. For small leaps in resolution, this process is remarkably clean, and you can ‘get away’ with small jumps in resolution quite beautifully with program’s native power. With operations involving larger leaps in resolution, expect some error in the form of jaggedly represented information, blur, and image degradation.

Sharpening for Print Output using Unsharp Mask

Before printing any digital image it is a good idea to sharpen the available pixels. This is best accomplished via a filter in the filter menu called, surprisingly, Unsharp Mask. In my case, I wanted to prepare the following image for printing:

First, I made sure I was working in the background (or main pixel) layer:

Then, when I selected Unsharp Mask from the Filter>Sharpen menu, the following dialogue boxes appeared:

The main interface window for the Unsharp Mask filter will appear and look similar to this:

I made certain that both the image preview and my main image magnification were set at 50%, which is the recommended viewing magnification for applying sharpening:

Sharpening using this interface is a trial and error process. The best way to sharpen is to make a test print and judge the print, not the monitor, for sharpness, but the method described above will give you some options in terms of adjusting the amount, radius, and threshold of the unsharp mask filter. Keep in mind that how a file appears on the monitor during sharpening may or may not have anything to do with how it will appear as a print, but adjust the sliders and view the apparent sharpness as best as possible on the monitor. Files that are properly readied for printing may sometimes appear jagged and poor on the monitor but may print just fine. The only way to really tell is to make a print! Sometimes the best way to sharpen involves reapplication of the unsharp mask filter. Keep in mind as well that you can always travel backwards in the history palette if your sharpening takes a destructive turn. When you are satisfied with the way your file looks on the monitor, simply click OK to commit the filter’s effect to your pixel layer.

Making A Print

When your file is completely adjusted, color corrected, and sharpened, select File>Print from the horizontal menu at the top of the photoshop interface. A dialogue box like the following will appear:

This dialogue box controls how the printer will respond to your request for output, and its appropriate checkboxes must be managed carefully to ensure the most accurate color and tonality representation possible. First, under the Printer Setup dialogue box, be certain that you have selected the correct printer name from the pulldown menu. (In my case my printer at home is an Epson Stylus Photo R2880, so it is selected above). Next, make sure you are only printing (1) copy of your image, for wasting paper is one of the main drawbacks of the printing process. Do everything you can now to ensure that you waste as little paper and ink as possible. Check the layout (or the orientation of the print) and be certain your image is fitting on the paper properly, then click the Print Settings button to bring up the controls for your specific printer:

In the printer’s dialogue box there should be an option for examining the Print Settings. Pull down the appropriate menu and your screen will likely change (it is different for each printer) to a window that looks like the image above. First, establish that you are using the correct Paper Size from the appropriate pulldown. Next, check the Media Type (or Paper). In my case (above) I am using Epson Ultra Premium Photo Paper Luster, the paper type and surface that most approximates that traditional look and feel of a color photograph. (For more information on paper surfaces and options, click here.) Make sure your dialogue box looks similar to the one above, and check that the Color Settings reflect that No Color Adjustments will be applied. This is one of the main points of confusion, for when we print from Photoshop, the program manages colors, not the printer. Unless you turn off the printer’s method for adjusting colors, your print quality will be very poor. There is usually an option in this printer dialogue box to check the color matching and color adjustment. For my system it looks like this:

Lastly, under Print Quality, (see the Print Settings dialogue window, two images above) select the “most superlative” or best Photo option. (In my case it is SuperPhoto @ 5760 dpi). This will ensure that the printer’s heads are outputting at their maximum resolution. It’s generally a good idea to turn off High Speed options when printing photographs, as they may sometimes also lead to poor print quality. When your dialogue box looks approximately like the one above, click on Save to return to the main Photoshop print dialogue box.

Be certain, most importantly, that the pulldown for Photoshop Manages Colors is selected. As mentioned above, this option also ensures that the Photoshop software (and not the printer) controls how colors are represented in your final print. Next, select the Printer Profile for your particular type of paper. In my case it was SPR2880 Premium Luster, the Epson generated profile for the paper I am using to print. (Note: if you were working on a properly calibrated system, here is where you would select the custom created profile for your particular paper.) If your file is 16 bit, you may check the Send 16 bit Data box. More important, however, is that the rendering intent is selected to Relative Colormetric with Black Point Compensation properly checked. This rendering intent tells the program how to translate the tonalities of your file into a printed photograph. Relative Colormetric has been found to be the best overall setting for printing photographs.

Note that this interface window shows a preview of your image on a piece of paper like the one you have selected. There are also checkboxes in the lower left hand section of your screen to check how the print colors will be matched. Unchecking and rechecking this checkbox will show you a soft proof (or a monitor based proof) of your image, and is one way of estimating how your print will look when it is finished.

When you are satisfied that your dialogue box is properly checked and your image properly prepared, simply select Print and the printer will get to work!

Artist Spotlight: Joel-Peter Witkin

Joel-Peter Witkin.