pieusb

Python package to enable interfacing with USB-connected Reflecta (a.k.a. PacificImage) film and dia scanners.

Prerequisites

Windows

You need to swap whatever driver Windows auto-assigned to your scanner to WinUSB

Linux

You need to add a udev rule

sudo tee /etc/udev/rules.d/60-pieusb.rules >/dev/null <<'EOF'
SUBSYSTEM=="usb", ATTR{idVendor}=="05e3", ATTR{idProduct}=="0142", TAG+="uaccess"
SUBSYSTEM=="usb", ATTR{idVendor}=="05e3", ATTR{idProduct}=="0144", TAG+="uaccess"
SUBSYSTEM=="usb", ATTR{idVendor}=="05e3", ATTR{idProduct}=="0145", TAG+="uaccess"
EOF
sudo udevadm control --reload-rules && sudo udevadm trigger

If you only care about operating one scanner, you can skip the lines with irrelevant ATTR{idProduct}. Here is a summary of which scanners have which

Model	ATTR{idProduct}
DigitDia 6000 Multiple Slide Scanner	0142
CrystalScan 7200	0145
ProScan 7200	0145
ProScan 10T	0145
CrystalScan 3600	0145
DigitDia 4000	0142
RPS 10M (aka Pacific Image PrimeFilm XAs)	0144

macOS

It should just work™

Install

pip install pieusb

Usage example

import numpy
import pieusb
from pieusb.scanner import Scanner

for info in pieusb.get_devices():
    print(info.vendor, info.model)

def on_update(data):
    # progress runs 0.0 -> 1.0 within each phase, not across the scan
    print(f"{data.phase}: {data.progress * 100:.1f}%")

def on_complete(result):
    if result.error:
        raise result.error
    numpy.save("scan.npy", result.rgb)

with Scanner(pieusb.get_devices()[0]) as scanner:
    scanner.mode = "rgbi"
    scanner.resolution = 5000
    scanner.color_depth = 16
    scanner.tl_x, scanner.tl_y = 0, 0
    scanner.br_x, scanner.br_y = 14200, 9600

    scanner.scan(on_update, on_complete)   # returns immediately
    scanner.wait()                         # optional

Scanning more than one frame

Keep one Scanner open for the whole batch rather than opening one per frame. It is faster, and the reason is reuse_calibration:

with Scanner(pieusb.get_devices()[0]) as scanner:
    scanner.resolution = 5000
    scanner.reuse_calibration = True

    for _ in range(50):
        scanner.scan(on_update, on_complete)
        scanner.wait()

Shading (flat-field) correction is applied on the host from a reference the scanner produces during a calibration pass, so the reference has to be read at least once before it can be reused. The first scan above therefore calibrates whatever the option says; the rest skip it and correct from the cached reference, which is what saves the time. The scanner keeps the last word: when it decides its calibration has gone stale -- after a power cycle, or on its own drift checks -- it refuses the skip and that scan calibrates again.

Every scan is shading-corrected either way. reuse_calibration only ever determines whether a pass pays for its own calibration, never whether the image gets corrected. Since the cache lives on the Scanner, a fresh one starts cold: a loop that opens a Scanner per frame is correct but calibrates every time.

Status of verification

Verified on hardware

• ProScan 10T

That's it for now

WARNING

No guarantees are made as far as hardware safety goes.

It may be possible to set certain parameters, such as scan width or height, such that the physical limits of the scanner are exceeded. This may damage your hardware.

Proceed at your own risk.