Testing

The testing procedure is pretty straightforward. We used three test pieces: a page of black text and graphics and two photographs. One is a landscape scene with green trees, blue skies and red ironstone escarpments, taken in the Yosemite National Park. The other is a portrait shot of a model, illustrating flesh tones, a detailed textile pattern and a background of green foliage.

We printed samples on all four machines, with all five inks and on all six papers. There were a couple of exceptions to this, as InkTecShop couldn't provide ink for our Epson printer and StinkyInk couldn't cater for the Lexmark machine.

During testing, we encountered several problems with cartridges. The Colourpoint cartridge supplied by InkTecShop for the HP OfficeJet wouldn't print yellow and the replacement cartridge had to undergo third-level maintenance before it would produce satisfactory prints.

The unbranded Lexmark cartridge from InkTecShop also had blocked jets and took several cleaning cycles before it produced good prints. Ironically, once working, the prints it produced were highly rated by our panel of testers. Finally, the first JetTec cartridge for the HP OfficeJet was unserviceable and we had to obtain a replacement. All this puts a query on the reliability of some third-party cartridges.

In the end, we had some 230 different prints and we showed all these to a panel of typical printer users and asked them to rate each print out of 10. We did this in a small hall with dispersed overhead fluorescent light, so all the prints had very similar lighting.

We offered guidelines on rating, suggesting a score of nine or ten would be a print you'd be happy to have as a record of a wedding or important birthday, seven or eight would be a print you would be content to have as a record of a holiday and five or six would be a print good enough to capture a moment, but with minor flaws in reproduction. Scores of below five were reserved for prints you would be unlikely to keep because of their quality.

A comment that was made by several of our panellists was that many of the prints were of very similar quality and quite difficult to differentiate between. The prints were completely mixed in a double-blind trial so neither we, nor the panellists knew which print had been produced with which printer, ink and paper. They were identified purely by number.

Once the prints had been scored, we averaged the results for each one and arranged them in a table, so we could quickly see which prints were rated the highest.

comments powered by Disqus