Good Picture 2011: Symposium Review

By Mike Christianson, Symposium Chairman

N.B. Copies of the material used by some of the presenters during their talks can be found via the links on the main Good Picture 2011: Output Revisited webpage.

The Imaging Science Group of the Royal Photographic Society organised the ninth of its successful series of one day annual symposiums under the title “Good Picture” on Saturday 3rd December 2011 at the University of Westminster in Regents Street. The meeting was entitled “Good Picture – 2011: Output Revisited” and covered subjects in this very important part of the imaging chain. The weather this year was very much milder than the arctic conditions we had in 2010 and this attracted about 65 delegates, these included both RPS members and non-members with diverse photographic interests ranging from keen amateurs and club enthusiasts through medical and industrial photographers to professional photographers.

The day began with Dr Tony Kaye who, as usual, gave one of his very entertaining and informative talks on the so-called “pixel race” between camera manufacturers and its significance to the amateur and professional photographer; this was entitled “How Many Pixels?” . He began by reviewing the human optical system and how it senses light and colour. This was followed by a review of Visual Acuity and a discussion of the resolving power of the eye and he showed, by examples, that it was possible to get higher “visual” resolution from a print that had lower overall sharpness due to the action of the eye. Viewing distance and print size were also considered; pixel requirements vary in proportion to the area not the side length of prints e.g. a 8×12 inch print requires 4X the number of pixels as a 4×6 inch print. The correct viewing distance is considered to be when the print subtends at your eye the same angle of view as that of the original camera system. This results in an ideal 14.4 inch viewing distance for an 8×12 inch print. His overall conclusions were that for computer or projected images use of 2 megapixels is sufficient unless zooming is required. For printed images 8 to 10 megapixels should be sufficient provided the optics of the camera system are good. Poor optics can reduce resolution more than can be gained with more pixels and are often the limiting factor in the resolution of a particular camera.

Andy Johnson, National Digital Sales Manager of Calumet Photographic gave a talk entitled “Overview of the Camera, Printer and Storage Markets” which discussed present and likely future trends in these areas. Since the it’s inception in the 1990’s the camera market has been a race to increase resolution by increasing the number of megapixels; the previous talk indicated that this may now be replaced with an effort to improve the camera optics. The byword for the future camera market is “convergence” with any one piece of equipment being able to carry out a number of functions. For example the ability of still cameras to shoot video; high quality Nikon and Canon SLR cameras can be converted to professional quality video cameras at a fraction of the cost of bespoke professional equipment. Mobile phone, laptop and tablet computers etc. now have higher quality digital cameras incorporated and can share still and video images via the internet where use of online systems has expanded the sharing of images e.g. via social media and “cloud” storage. The future of the printer market will see more use in consumer systems of characteristics that previously were only used in high-end professional printers. For example the use of self-calibration whereby a reference print is made and read to produce an ICC colour profile. Wide gamut printers are now more available in the consumer market which use many inks to improve colour and tonal range e.g. Canon now sell a 12 ink printer with 5 different blacks. All these high quality, high megapixel devices produce huge amounts of data and storage is becoming a big issue for photographers. The use of RAID (Redundant Array of Independent Discs) storage systems was discussed, these spread the data over a number of linked hard drives which are viewed as a single drive. Loss of any drive within the group does not result in loss of data and in some systems each drive has enough information to set up any failed drive which has had to be replaced within the group. The price of these systems has reduced from £20,000 to £500 in the last few years to make them accessible to consumers.

Over the last few meetings Dr. Graeme Awcock of the University of Brighton has discussed the Landsat series of 7 US satellites, which have been continually mapping the Earth’s surface since 1972. This year’s talk entitled “A New Take on Atmospheric Pictures” was a live, online demonstration of how the freely available data can be accessed and manipulated to produce stunning images of the Earth’s surface. Galleries of images produced from the satellite data are available at a website called “Earth As Art”( http://eros.usgs.gov/imagegallery/ ), he selected one of these and showed how the data could be downloaded, decompressed then manipulated using the readily available free software and online advice. He showed how the selected image had been constructed from the satellite data, a process that could be done for any of the enormous number of images that have been produced over the years by this highly successful satellite series.

Dr. Alan Hodgson of 3M Security and Systems Ltd. discussed some important issues of printing in a talk entitled “Into the 3rd Dimension – How Deep Are Your Images”. Through the use of microscopic examination of cross-sections of materials from various types of printing methods he discussed how deep the dye is incorporated and why this is an important issue. The four characteristics which are most sensitive to the depth on the dye are drying characteristics, colour gamut of the dyes, gloss and image permanence. High quality ink-jet media is a fairly complex construction with a number of layers coated on either side of a plastic coated paper. The multiple layers control where the dye is located, the structure of the dye blob after it is absorbed into the surface and how the water in the ink is dealt with. These systems can be engineered to leave the dye closer to the surface with the water sinking into lower layers, therefore it is possible to handle the print more quickly. The use of different materials in these layers also has a strong effect on the hue of the absorbed dye and therefore the final colour gamut and the sensitivity of the dye to fading reactions, particularly through the action of light. These papers also have very thin top layers which do not absorb dye but act as protection from abrasion and produce high gloss. He also discussed the problems associated with pigment dyes which generally are not absorbed into the surface of the media; this results in images which are easily scratched and a surface gloss which varies with density as the pigments have higher gloss than the paper when dry. These problems are usually overcome by incorporating a resin with the pigment to produce improved physical properties.

Dr. Philip Green of the London School of Communications discussed the need for processes to calibrate displays and printers which are able to produce larger colour gamuts on their outputs. A Colour Gamut is defined as all the colours that can be represented by a particular medium and is the result of the representation provided by the colorants (dyes, phosphors etc.) and the range of the medium (TV, print paper etc.). Actual scenes have a much larger luminance dynamic range than media, such as TV or print paper, and the latter are very dependant on viewing conditions. He discussed how these gamuts may be visualised on CIE x,y co-ordinate diagrams and how white point normalisation can be used to compare various media in Lab 3D colour diagrams. When extended gamuts need to be accommodated it is necessary to render the camera colour data to the larger gamut of the medium, which can be done during RAW conversion. There are various sets of standard primaries which can be used for this conversion, e.g. opRGB (Adobe ’98) if often used for TV displays. Others are available such as PRMG, ROMM and ISO 15339 RPC7.

In a talk entitled “3D Imaging – A Review of Current Trends and Techniques” Dr Peter Plassmann of the University of Glamorgan reviewed the use of various methods to obtain accurate and quantifiable 3D data for objects. This talk did not cover such areas as 3D cinema, TV and still photography. He began by discussing various types of scanners which build up the image using light or laser beams. Triangulation laser scanning is a rapid process using a hand held device for observing small objects which have been tagged with reference points. There is no colour data just a 3D data cloud recorded on a computer. Larger objects, e.g. buildings and open spaces, are analysed using Time-of Flight laser scanners where the time is measured for the beam to travel to and from the device, after being reflected from the object, to produce an accurate distance and the laser is scanned over the whole field of view. In Structured Light scanning the device projects stripes of coloured light onto the object and the shapes taken up by the stripes produces the 3D data. This is accurate to sub-mm range and is used extensively in the medical field for wound analysis.
Finally he discussed the use of Steroscopy which simultaneously produces a pair of stereoscopic images which are then combined to produce a 3D image from which accurate measurements can be made using triangulation when the lens separation and object distance are known precisely. Although this was also developed for wound analysis it has had wider use and he showed an example of this type of analysis in preparation for cosmetic surgery.

The final talk was given by John Charnock of Print Research International Ltd. and was entitled “Photobooks and Printing Technology to Enable Photographers to Self Publish”. Historically publishing was an expensive process but this has very significantly reduced with a change in the role of publishers and the printing cycle. Previously a minimum print run required 5,000 copies, when this was added to the page preparation, finishing, advertising and marketing costs, self publishing was very expensive. Now it is possible to do most of this oneself. The use of software such as Photoshop allows the home user to prepare images and text etc. and design the pages to be printed. Modern digital printers such as HP Indigo, high speed laser and Ink-Jet can do small runs at relatively low cost and marketing and sales can now be carried out, again at home, on the internet by using for example social media and Amazon. Surveys have shown that up to 80% of people would like to self-publish e.g. to pass on images etc. to family. This has resulted in development of vanity publishing services such as LuLu, Blurb and createspace who will make the “books” for prices varying from £600 to £5000 depending on quality. There are other services that will print ones own images on to all sorts of material surfaces to produce personalised gifts. He presented a case study of self-publishing where a photographer had produced, a cost of about £1000, a short run of a high quality table-top book of nature photography on the Thames which he was able to sell in local bookshops. This book lead to more commissions from towns along the Thames in London.

To close the Symposium, the RPS Roy Robertson FRPS, warmly thanked all the speakers and the organisers for putting together such an interesting and stimulating set of topics.