Workflow Notes
Have Jeffrey Sward
address your group
about technical subjects, or
photographic style,
or
selected examples of photographic style.
“Save As” new file after each phase
Wacom tablet makes tedious jobs such as dust removal much easier.
Photographs are generally printed on photographic paper by labs at 300 pixels to the inch. Multiply
the desired print size by 300 to obtain the needed number of pixels. Example: 8x10 print is 2400x3000 pixels. Scan
negatives and slides to a size slightly larger than the desired size. Larger images may always be resampled to a
smaller size .
Recommended plug-ins
Color Efex Pro, and Sharpener Pro from Nik at
http://www.nikmultimedia.com/
. Tests with the latest version of Sharpener Pro indicate that for lab prints the default setting produce optimum
results for print viewing. Always make a separate file for the lab using "save as." Do not be dismayed
by the harsh appear on the print file on the monitor. This is caused by the monitor resolution of 72dpi compare
to lab prints 300dpi.
Sho professional and Gem professional from ASF, now part of the jolly yellow giant at
http://www.asf.com/
Genuine Fractals from
http://www.ononesoftware.com/
is useful for enlarging images.
If the annoyance of masking is needed, Mask Pro may be useful, also from
http://www.ononesoftware.com/
. However, often masks are more easily drawn by hand alternating the pencil and paint bucket tools. Also, the new
Quick Selection Tool often works better and faster than Mask Pro.
Two good noise reduction products are Noiseware from
Imagenomic
and Noise Ninja from
Picture Code
.
Not a plug-in, but an extremely useful thumbnail viewer, with configurable color profile viewing
is Thumbs Plus at
http://www.cerious.com/
.
|
Making Your Own Custom Photoshop Workflow
Some objectives for setting up a custom workflow:
-
Best possible picture quality.
-
Minimize labor (effort hours).
-
Minimize duration (elapsed time).
-
Be able to return to any point in the edit cycle and revise the edits.
-
Be able to identify changes made to images at a later date.
-
Be able to use tools with no adjustment layer available such as plug-ins or filters at any
point in the process.
-
Be able to apply any edit, plug-in, or filter against any selection or applied mask at any
point in the process.
-
Procedures are repeatable, predictable, and leave an audit trail.
These objectives are contradictory. For example, the best possible picture will inevitably use the
most amount of time. Since there will be inevitable trade-offs in any workflow, the goal should be a set a compromises
which is best suited to the nature of the images, end results desired, and limitations of time and money.
Here are some relatively uncontroversial steps which have mostly benefits:
-
Have and use a consistent naming structure for directories. Directories are often named by
subject, job, or date.
-
Have and use a consistent naming structure for the photographic files. Often a suffix is useful
to identify iterations.
-
Save all working copies in a lossless format such as psd.
-
Only crop when making
"final" images for web or prints.
-
Keep at least two backups of all files.
-
When performing the same commands in the same sequence in the workflow, set up the sequences
with an action. Action can be flagged to allow operator intervention. Even with operator intervention on every
step in the action, time will be saved by avoiding invoking the steps through menus.
-
Always show the histogram. After each step or action, check the histogram. When a blob appears
on the left, detail has been lost in the shadows. When a blob appears on the right, details has been lost in
the highlights. This loss of detail is call clipping. Avoid clipping shadows or highlights.
These suggestions are somewhat more controversial:
-
Place early in the workflow extremely tedious but vital steps, especially if they are unlikely
to be redone. Dust and/or digital artifact removal is the obvious item in this category.
-
All edits are sequentially applied. Each edit affects all subsequent edits. This makes ordering
of operations extremely important. For example, if edits are saved in adjustment layers, and the adjustment layers
are reordered by dragging, different results are achieved. Therefore, segment edits into groups, with a "save
as" with a new file name between each group. For example, fix dust first, save as, fix levels next, save as,
fix color and luminescence next, etc. Refer to the paragraph below on the great "adjustment layers vs. save
as" debate.
-
Sharpening must be applied after each resize. Alternatively, do not sharpen the original image
ever, but only sharpen "final" resized images for web or print use.
-
Whenever possible, scan and/or start with an image which contains far more pixels than the
final result requires. Perform all edits and corrections on the large image. After the large image is edited,
perform a "save as" for the desired size, resize, and sharpen. For example, an 8x10 print at 300 dpi requires
2400x3000 pixels. Much better results are achieved by editing a 4000x6000 pixel file and resampling, than working
with 2400x3000 pixels throughout the process.
-
The trade-off between speed and quality becomes extreme when publishing
"digital proofs," especially for web use. The best results are achieved with a full scale edit of each image,
followed by a resize. However, this involves a tremendous amount of time, especially since a "final"
result will likely not be needed for each proof. It is much faster to resize a set of original files to a smaller
size and perform quick edits, often with the assistance of Photoshop batched commands. However, any time
spent on proofing is that much less time which can be spent on the ultimate images. No easy answer here.
There appears to be an amusing growing controversy about the techniques used to archive these objectives
(from the list above):
-
Be able to return to any point in the edit cycle and revise the edits
-
Be able to identify changes made to images at a later date
-
Be able to use tools with no adjustment layer available such as plug-ins or filters at any
point in the process.
-
Be able to apply any edit, plug-in, or filter against any selection or applied mask at any
point in the process.
-
Procedures are repeatable, predictable, and leave an audit trail
Achieving these objectives has created the "save as" camp and the "adjustment layer" camp. These
two approaches are summarized in the chart below:
|
Item
|
"save as" camp
|
"adjustment layer" camp
|
Basic strategy
|
Group edits into repeatable blocks. Edit the main background image.
Perform a "save as" with a new file name between each group. The history list is used for redos between
saves.
|
Never change the background image. All edits are performed with new
adjustment layers. There is generally only one master file per picture.
|
Redo of edit technique
|
Open the file in the sequence which is one step before the edits
which need to be redone. Perform a "save as" with a new file from this point forward.
|
Make changes to the selected adjustment layer. If the adjustment
layer is early in the sequence, a cascading effect may be caused which will make it necessary to change subsequent
adjustment layers also.
|
Advantages
|
-
Avoids the complexity of multiple layers, especially when many edits are performed.
-
Plug-ins or filters which must operate on the entire image at once and cannot be isolated
to their own adjustment layer can be used at any point in the process.
-
Each file version also acts as an additional backup.
-
A visual history of the change sequence is visible simultaneously in a thumbnail tool
such as Thumbs Plus by viewing the saved files.
|
-
Each adjustment can be changed at any time by changing the layer.
-
Each adjustment can be completely activated or deactivated at any time by making the
layer visible or invisible.
-
History of changes is apparent from inspection of each adjustment layer.
-
There is only one master file, a simplification.
-
A visual history of the change sequence is visible sequentially by making adjustment
layers visible and invisible directly in Photoshop.
|
Disadvantages
|
-
Interpretation of the file naming convention and/or a separate document is necessary
to determine which edits were performed on which file version.
-
A particular edit cannot be individually post-adjusted. Only groups of edits can be
post-adjusted.
-
Any change to adjustments early in the process requires the recreation of all subsequent
edit files in all cases, a labor load.
-
Disk space is wasted with multiple file versions.
|
-
Many plug-ins and filters must act on a complete image and cannot be isolated to an
adjustment layer. Therefore, some scheme of intermediate copy layers is necessary. However, once an intermediate
copy layer is created, the ability to have changes to early adjustment layers apply to all subsequent layers
is lost.
-
Management of many layers introduces complexity.
-
A cascading effect may be caused by a change to an adjustment layer early in the sequence,
resulting in the need to change subsequent adjustment layers. It is questionable whether effort or elapsed
time is saved under these circumstances.
-
Disk space is wasted by the creation of layers, especially duplicate image layers.
-
A loss of the one master file loses all versions of the image.
|
|
All written content of this web site is solely the editorial opinion of Jeffrey
Sward. All images, graphics, and written content of this web site, including the html files, are creative products
covered by copyright law. All content copyright Jeffrey Sward 1975-2019. All rights reserved. No portion of this
web site or its constituent elements may be reproduced in any form, by any means, without prior written permission.
So there.
|