Parts and parameters

Information-oriented. Look things up here; don't read it cover to cover.

What's in the HoloBox

Base Set — interferometry & wave optics

Component	Purpose
openUC2 injection-moulded cubes	Modular optical mounts
50 mm optical rails & mounts	Hold and position the cubes
Front-surface mirrors	Steer and reflect beams
Beam splitter (50:50)	Split / recombine a beam
Low-cost laser module (532 nm)	Coherent light source
Screens	Observe fringe patterns
Apertures, slits, gratings	Diffraction experiments
Converging lens	Expand / condition the beam
Base plates & puzzle pieces	Mechanical stability
Printed worksheets / build guides	Classroom material

Enables: Michelson interferometer, Mach–Zehnder interferometer, diffraction experiments.

Digital / holography add-on

Component	Purpose
Raspberry Pi camera (smart camera)	Records holograms; runs the reconstruction UI
LED + holder	Illumination for inline holography
Gel colour filters (red / green)	Quasi-monochromatic light
Aluminium foil + needle	DIY pinhole
Electronics module	LED control, phase shifting, triggering

Enables: inline holography, lensless microscopy, computational imaging.

🖼️ Image placeholder — box-contents-labelled.jpg

Show: A flat-lay photo of the full HoloBox contents with each part numbered to match these tables.

Inline-holography reconstruction parameters

These are the settings in the ImSwitch inline-holography widget. Defaults below are the software's built-in starting values.

Parameter	Default	Typical range	What it does
Wavelength	488 nm	400–650 nm	Colour of your light. Match it to your filter (green ≈ 532, blue ≈ 450, red ≈ 650).
Pixel size	3.45 µm	sensor-specific	Physical size of one camera pixel. Change only if you change camera/binning.
Distance dz	0	0–30 mm (1 µm steps)	How far to "rewind" the wave. This is your focus dial.
Numerical aperture (NA)	0.3	—	Describes the light-collection angle; affects resolution.
Colour channel	red	red / green / blue / white	Which channel to reconstruct. Match your filter colour. white = average of all channels.
ROI size	256 px	—	Side length of the square crop that gets reconstructed.
ROI centre	image centre	—	Where the crop is taken from.
Binning	1	1, 2, 4…	Combine pixels to speed up / widen field at lower resolution.
Flip X / Flip Y	off	—	Mirror the image horizontally / vertically.
Rotation	0°	0/90/180/270°	Rotate the image.
Update frequency	10 Hz	—	Live reconstruction frame rate.
Show raw	off	—	Show the un-propagated hologram (dz = 0) regardless of the slider.
Full frame	off	—	Ignore the ROI crop and reconstruct the whole sensor.

The three you'll actually touch

For a first hologram you only need wavelength, colour channel, and dz. Leave the rest at defaults until you have a reason to change them.

How the reconstruction propagates the wave (for reference)

The widget reconstructs by Fresnel free-space propagation. In plain steps:

1. Take the cropped, single-channel hologram (an intensity image).
2. Treat it as a light field by taking the square root of the intensity (E₀ = √intensity) — i.e. assume flat phase at the sensor.
3. Fourier-transform it into spatial frequencies (FFT).
4. Multiply by the Fresnel phase factor for distance dz: each frequency is multiplied by exp(iπ·λ·dz·(fx² + fy²)).
5. Inverse-transform back to an image and display its intensity.

Dragging dz changes that phase factor, which is what brings different planes into focus. The conceptual version of this is in How reconstruction works.

For advanced readers

A more general method is angular-spectrum propagation, which uses the factor exp(−i·2π·dz·√(λ⁻² − fx² − fy²)). Fresnel propagation is its small-angle (paraxial) approximation. The Münster thesis derives both. For school use the difference is invisible; both bring the sample into focus.

Offline reconstruction in Python

If you'd rather capture a still and reconstruct it yourself (a nice coding-class bridge): capture at full resolution, save as PNG, then in a Jupyter notebook:

import numpy as np
import matplotlib.pyplot as plt

def reconstruct_inline_hologram(hologram, wavelength, pixel_size, distance):
    """Reconstruct an inline hologram by Fresnel propagation.

    hologram    : 2-D intensity image (one colour channel)
    wavelength  : light wavelength in metres   (e.g. 532e-9 for green)
    pixel_size  : sensor pixel size in metres  (e.g. 3.45e-6)
    distance    : propagation distance in metres (your 'dz' focus dial)
    """
    ny, nx = hologram.shape

    # spatial-frequency grids
    fx = np.linspace(-(nx - 1) / 2 / (nx * pixel_size),
                      (nx - 1) / 2 / (nx * pixel_size), nx)
    fy = np.linspace(-(ny - 1) / 2 / (ny * pixel_size),
                      (ny - 1) / 2 / (ny * pixel_size), ny)
    FX, FY = np.meshgrid(fx, fy)

    # intensity -> field (assume flat phase at the sensor)
    E0 = np.sqrt(hologram.astype(float))

    # Fresnel phase factor for this distance
    kernel = np.exp(1j * np.pi * wavelength * distance * (FX**2 + FY**2))

    spectrum = np.fft.fftshift(np.fft.fft2(E0))
    field    = np.fft.ifft2(np.fft.ifftshift(kernel * spectrum))
    return np.abs(field)          # reconstructed amplitude image

# --- usage ---
img = plt.imread("hologram.png")          # load your capture
channel = img[:, :, 0]                     # red channel (match your filter)
recon = reconstruct_inline_hologram(channel,
                                    wavelength=532e-9,
                                    pixel_size=3.45e-6,
                                    distance=0.007)   # 7 mm — tune this
plt.imshow(recon, cmap="gray"); plt.title("Reconstruction"); plt.show()

Re-run with different distance values (a few mm up to ~30 mm) to refocus — that loop is the offline equivalent of dragging the dz slider.

Connection details

Item	Value
Camera Wi-Fi password (HoloBox image)	holobox123
Web interface address	http://192.168.4.1

Related

• Your first hologram
• Troubleshoot holograms
• Glossary