08.Slow, slower, slowest
The time it takes to scan a slide is largely dependant on the quality settings you choose. The lower the quality, the faster the scan. However, scanning times differ significantly when set at the top optical resolution and maximum bit depth. If you've lots of slides that need cleaning, you could be looking at weeks, rather than hours, of work.
The Konica Minolta Dual IV was the fastest. It shares equal lowest resolution (3200 dpi) with the Canon CanoScan 9900F, but took about 32 seconds to scan a slide as against 171 seconds for the Canon 9900F. The slowest was the Canon Canoscan FS 4000 US at about 393 seconds. The Nikon V ED was also quite fast at about 49 seconds - the others took between 102 and 178 seconds.
However, these timeframes greatly increase if you turn on the slide cleaning systems (see Cleaning up below). We didn't score the timeframes using the cleaners because we thought it was unlikely you would use them all the time, even though they are a useful tool. The Nikon V ED scanned slides fastest with the cleaning tools on (about 105 seconds), while the slowest was the Konica Minolta Dimage Scan Elite 5400 at about 541 seconds. The others took between 360 and 455 seconds.
Digital Ice and FARE is technology designed to remove the effects of dust and scratches from slides or negatives. The former is a Kodak product and is available on several brands of scanner (Epson, Konica Minolta and Nikon in this test), The latter is a Canon technology.
Both work in a similar way, using a combination of hardware and software. Scanners measure the level of red, green and blue in an image, dot by dot. These are called the red, green and blue channels. A scanner with FARE or Digital Ice introduces an extra infrared or (D) channel. Most slides and colour negatives are transparent to infrared light, so if the slide is clean this extra channel will be clear. Dust, mould and scratches are not transparent and appear in the channel as dark areas. The software uses this channel in combination with information from the areas close to the fault to fix the image.
We used a slide that had been very badly affected by mould, as well as some with small amounts of dust and minor scratches. The results across all brands are impressive. There was some evidence of interpolation (see Dictionary) around the worst affected areas, but these could be fixed. Two spots on our mouldy slide were too degraded to be cleaned by either system, but fixing them manually would only take a few minutes - whereas fixing the original slide manually would be nearly impossible.
Traditional silver-halide black and white film and some specialty slide films cannot be used with this technology. An additional drawback is time, but it could be the difference between a lost image and one recovered forever.