There are major trade-offs between linear and area sensors, and in this application an area sensor ultimately made far more sense due to the constraints linear sensors impose on the optics and backlight system.
A linear sensor operates at extremely high line rates, meaning the integration time is incredibly short - often in the microsecond range. This requires a very high-power backlight to achieve proper exposure. Creating a light source that is both powerful and perfectly uniform is difficult, requires substantial thermal dissipation, and limits LED selection significantly.
These constraints then affect the lens design. Because exposure times are so short, the lens needs to be fast, which makes consistent focusing much more difficult, especially once thermal expansion from the backlight and the realities of manufacturing and the postal system are considered.
Initially, I used this area sensor in a line-scan style mode by reducing the number of lines read out. Although it worked technically, it did not make for a viable commercial product. For this reason, I moved to using the same sensor as an area scanner and compositing multiple images, which has proven to be extremely effective. While the sensor itself is physically small, its optical and electrical performance is high.
Sensor size itself is a huge topic that is often difficult to communicate when selling a product like this. People naturally apply the logic of “bigger sensor = better,” but a film scanner is fundamentally different from a handheld camera. The requirements and operating conditions are completely different.
Unlike a camera, a scanner operates in a fully controlled environment. We control the light source completely rather than relying on natural lighting, meaning the sensor can operate continuously in its ideal range at base ISO, where performance is excellent. Because of this, a small sensor with small pixels can deliver extremely high image quality in this application.
Using a much larger sensor, such as full frame, would massively increase cost and complexity while providing little practical benefit. The sensor alone would currently cost more than the entire scanner. A larger sensor would also require a much larger and more complex lens, a larger enclosure, more demanding mechanics, and significantly higher overall system cost.
The important point is that scanner performance is determined by the performance of the complete optical system, not sensor size alone.
.
