digital workflow

This page explains my workflow, for clients, especially those with a limited knowledge of digital photography.

Just a few years ago the production of a booklet, newsletter or annual report involved hiring the specialist skills of a photographer, a graphic designer, a repro house and a printer. What each of these people did was largely a 'black art' as far as the layman was concerned. The client just had to insist that what was produced met their requirements and aspirations.

Digital photography and desktop publishing have blurred these boundaries and roles and made it customary for clients to have a much more hands-on or DIY approach. On the plus side, a publication can move from concept to realisation very much faster, and usually much cheaper. On the downside, despite clever software it is necessary that some effort is made to understand the processes involved, if expensive mistakes are to be avoided.

In particular photography is a source of problems. Digital has made photography miraculously accessible and easy to achieve without technical knowledge, but every designer I work with complains about receiving images from clients that are of poor quality or simply unusable. Often these are contributed by amateurs or members of staff but inadequate quality from professionals is also a significant industry issue. The Universal Photographic Digital Imaging Guidelines are an attempt to improve standards that should not be necessary if professionals were more expert or less prepared to cut corners in order to compete on price.

Why shoot RAW?

Clients sometimes query why I shoot RAW instead of JPEG. RAW requires considerable processing to produce a usable image, which adds expense for the client. And that expense usually rules out requests like 'just put all the images on a CD'. Neither aspect is convenient nor endearing to clients, and I am sure some believe I am being difficult or trying to shake them down for extra money. After all, they quite often use digital cameras themselves and can burn all their images to CD easy enough.

The answer is quality. This is best illustrated by example, but if you don't know what the difference is between JPEG and RAW, here is a brief explanation.

Virtually all digital cameras are capable of producing JPEG's. Although results are usually pretty good there can be problems when shooting in conditions that diverge from the ideal bright outdoor 'snapshot' situations. Bright sunlight, flash, dull or mixed light sources all present special difficulties. Typically these are errors of white balance and exposure which lead to colour inaccuracy and 'blown' highlights (bright areas which lack all detail) or 'blocked' shadows (detail-less dark areas).

Shot as a JpegShot as a Jpeg, this image struggles to maintain detail in bright and shadow areas, and colour balance veers to blue in the shadows. There is simply too much contrast in the original scene. The eye is able to view such scenes and compensate for these aspects, but the camera cannot.

JPEG files are produced by intelligence within the camera. Good though it is, this can be overwhelmed by lighting conditions. Moreover there is a rigid limit to how much contrast can be accommodated in a JPEG that is often exceeded by real life. And allowing the camera to make decisions about the picture necessarily takes away some creative control from the photographer.

Additionally, JPEG is a lossy compression method which results in a blocky image structure if enlarged too much. Sharpening is also applied to the JPEG, to counteract softness that arises as part of the digitisation process.

Saving the image as a JPEG, in the camera irremovably fixes all these problems into the image file. Lost information has been lost forever and artefacts added. Unfortunately there is little that can subsequently be done to correct or improve things without degrading the image and causing worse artefacts.

Most professional and some advanced amateur cameras permit the recording of RAW files instead. A RAW file is a complete data dump of everything the camera can record at the set exposure, without any of the automated decision-making imposed. Those decisions are deferred until later, so that the photographer can make them in the 'electronic darkroom'.

Shot as RAW: Shot as RAWShot as RAW. Before processing adjustments are made, the image does not look promising. It is dull and dark, a result of being deliberately exposed to preserve highlight detail. This is deceptive. Unlike a JPEG, the RAW file contains far more information than can be seen.

A RAW file is somewhat similar to a film negative. As with a print from a negative, careful human processing will usually yield far better results than a machine can achieve.

The downside of RAW is that it requires properly calibrated hardware and specialised software, skill and time. Typically RAW files will require adjustment of exposure, white balance, levels, gamma, saturation, hue, localised dodging and burning, and sometimes - as has been done with this image - compositing of differently 'exposed' and colour balanced versions of the image from the same RAW file. Captions will also be added to the IPTC fields.

raw final compositeThe final file from RAW, after processing and post production. This is the photograph I saw when I took it. It is now a JPEG, but saving as a JPEG is the final stage after all other processing has been completed, so as to minimise losses and artefacts.

Inevitably this work represents an additional small cost to the client. However the quality is so superior to camera JPEG's that RAW is the only workflow I am prepared to offer professionally. It is the only way to guarantee good, usable images from uncontrolled lighting.

In my experience, professional photographers who produce camera JPEG's either have a very good reason (eg news photography time pressure, where images must be wired immediately), or simply do not know or care what they're doing, so leave it to the equipment. They may save themselves time and their clients money by doing this, but too often the results will be disappointing.

I do not supply RAW files to clients because, exactly like hand-printing from negatives, interpreting them to produce a final TIFF or JPEG is part of the creative photographic process. If a client has the skills and tools to process RAW files well, they must be expert photographers themselves and are unlikely to need my services. And if they're going to do it badly, I don't want my name on it.

Proofs

Because I shoot RAW, as described above, it would be wasteful of my time - and more importantly client's money - for clients to order more final production files than they actually need for their project. Clients who are used to photographers who shoot JPEG often expect all images from a shoot to be supplied on CD. This is not feasible when shooting RAW.

For this reason I usually supply an edited selection of proofs from which files for production may be selected, both as' contact sheets' containing low-resolution versions of the images, and as a web gallery within the secure client section of this site.

Contact sheets usually contain 20 images per A4 sheet, though you can have between 2-40 if you prefer (cost per sheet is the same). They are usually sent by email and may be printed locally, although I can print and post if required. Proofs are partially corrected for colour and exposure to make them more 'readable' than unprocessed RAW, and are good enough to be used as positionals.

The web gallery for each shoot is maintained indefinitely, so forms an archive from which you may order further production files at any time. Additional repro fees may be required for new usages outside the scope of the licensed rights agreed at the time of commission.

Resolution of ordered production files

I need to know the expected maximum size of reproduction when preparing files. '300dpi' is nowadays generally quoted for print, since this is more than adequate for a good Q factor in high quality litho printing, but it means nothing without the print size in inches or centimetres. Digital images really only have dimensions in pixels, and that is what I need to work out.

EG 7.5"x5" repro@300dpi is (7.5x300) by (5x300) = 2250pixels x 1500pixels

I do not apply unsharp masking to production files because of the artefacts that will cause problems during size adjustment. Images consequently look slightly soft as supplied. Sharpening must be carried out by your printer to suit their press.

Colour management

Unless specifically ordered for web use (ie sRGB), the final production files (TIFF or JPEG) will generally be in tagged Adobe 1998 RGB space. They will appear correctly in Adobe Photoshop but may look desaturated when viewed in PC applications that do not understand ICC tags such as web browsers. You should avoid re-saving the files in such applications as this will destroy the tagged colour space information that will be needed for accurate reproduction in print. If you require advice about colour management issues for your intended use, please ask.

I am occasionally asked to produce files in CMYK for litho printing. This is not a good idea. CMYK must be tailored to the press characteristics, paper and inks used and generic CMYK will give poor results. If you really want me to produce CMYK I will need the precise parameters from your printer, and this will be like getting blood from a stone as the printer will want to keep control of their end of the production process. All things considered it is far better that the printer does the RGB to CMYK conversion as they will know exactly what needs to be tweaked to get the best from their process.