High-resolution drum scanning

The ne plus ultra of film scanning methods is the drum scan. Photo Multiplier Tube (PMT) technology in the Howtek 4500 and Dainippon Screen 1045ai scanners provide a density range of up to 3.9, extremely low noise levels, and true optical resolutions up to 8000 dpi across the entire film area, resolving grain-level detail. Safe liquid mounting in Kami fluid eliminates scratches and reduces grain. This is no-compromise scanning, recommended for all serious printing and archiving.

We are different!

At MSP, we regard scanning as a craft, not a mere mechanical exercise. Great scanning requires not only great equipment but also a thorough knowledge of color reproduction, fanatical attention to detail, and a passion to pull the very best possible image from the film. This takes time, and it doesn't come cheap. Many labs offer drum scans at amazingly low prices, and in truth sometimes these scans are adequate. But there can be hidden costs: Images with exposure or color errors and small defects, all present to some degree in every photograph, go uncorrected, passed on to the customer as "accurate" or "straight" scans, but in fact they are still unusable without a great deal of further work, which the customer has neither the time nor skill to perform. One well known scan provider told a customer that correcting the color in his slide would cost "hundreds of dollars." We consider this totally unacceptable. Every scan from MSP is finished—fully color- and contrast-corrected and optimized for intended output. All press scans are converted to the correct color space and expertly sharpened.

Pricing

We are committed to provding the best overall value for our customers, which is not always the same thing as the lowest possible price. We do try to find a solution for each customer that will best suit his or her needs and budget. Please contact us and we'll be happy to furnish a quote for your job.

For the Technically Curious: Drum Scanning FAQ

What makes a drum scan superior?

There are really two big reasons to have your film drum scanned. The first is resolution. Drum scanners, which use photomultiplier tubes (PMT) as light sensors, simply record finer detail from your film than more ordinary scanners, which use charge-coupled devices (CCD). CCDs have improved greatly in recent years, but they still have inherent limitations owing to their size and tendency for flare and electronic noise, greatly aggravated as the CCD elements are made smaller and arranged in denser arrays. Although CCD scanner manufacturers now claim up to 8,000 dpi resolution, as high, in theory, as the best drum scanners, these resolution figures are based on file size and the number of image pixels created, not the amount of actual image detail recorded. The only way to determine the latter is to scan resolution targets (photographs of extremely tiny line patterns) and examine the results. Such comparisons invariably show that drum scanners have a very clear advantage over all CCD scanners at the same rated resolution. Is this important? It is if you plan to enlarge your work more than 4X, which means virtually all enlargements from 35 mm films and most from medium-format films, whether intended for offset press, inkjet, Lightjet, or film output.

The second reason is dynamic range, that is, the ability of the scanner to faithfully record a wide range of density values in a film from the clearest to the densest. A color transparency presents the greatest challenge in this area, often having density ranges of 3.5 or even higher (for the uninitiated, these numbers represent base-10 logarithms, or powers of 10, so a density range of 3 represents 1000:1, for example). Since many CCD manufacturers claim dynamic ranges over 4, why bother with the expense and bother of drum scanning? Here again, the manufacturers mislead. Some CCD scanners can actually record some information in the highest densities, but most of what is captured is simply noise from the device's electronic circuitry as the gain, or signal amplification, is turned up in an attempt to compensate for the relatively poor CCD sensitivity at low light levels (Those familiar with video cameras—CCD devices, in fact—have experienced the grainy images that result from shooting in low light). The drum scanner's photo multiplier tubes are thousands of times more sensitive than CCDs and do not require this ruinous signal amplification and therefore have extremely clean images right down to the darkest image areas. This can be seen in side-by-side comparisons. A good drum scan can actually retrieve more detail from the film than can be seen on a light box.

Why do you use the Howtek and Screen scanners and not, for example, the Heidelberg models that many other labs use?

The scanners we use have certain advantages over most other drum scanners. These models have a variable sampling aperture, the opening through which they "see" the film detail. These apertures can be as small as 3 microns, which allows a sample as small as 1/8000 of an inch to be taken, smaller than nearly any possible image detail. Furthermore, this variable optical resolution can be extremely useful when scanning grainy films such as color negative emulsions as it allows us great flexibility in finding a good balance between sharpness and smoothness. The Heidelberg Tango and its successor, the Primescan, are beautiful pieces of engineering, but their minimum aperture of 11 microns, or about 1/2300 of an inch, in our view makes them a bit less adaptable than they should be.

What about "Virtual Drum" Scanners?

One manufacturer, Imacon, makes CCD scanners featuring a "virtual drum." This description derives from its ability to curve the film so that the scanning axis is always at a right angle to the film. This is the only respect in which these scanners can be associated with true drum scanners. The Imacon is a CCD scanner, and while it is a very fine one, it still suffers from the drawbacks of its class, that is, lower resolving ability, dynamic range, and signal-to-noise ratios than true PMT drum scanners.

Copyright 2015 Michael S. Strickler