🔭 astrophysics

Characterization of the commercial spectrograph system for astronomical observations: PIXIS 1300BX Camera and IsoPlane 320A Spectrograph

This paper presents a comprehensive laboratory and on-sky characterization of the PIXIS 1300BX camera and IsoPlane 320A spectrograph system, validating its high quantum efficiency, moderate spectral resolution ( $R \approx 600 - 2600$ ), and suitability for multi-object spectroscopy as part of the upcoming A-SPEC survey.

Original authors: Jiwon Jang, Changsu Choi, Ho Seong Hwang, Haeun Chung, Hyeonguk Bahk, Dongkok Kim, Jae-Woo Kim

Published 2026-03-10

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Jiwon Jang, Changsu Choi, Ho Seong Hwang, Haeun Chung, Hyeonguk Bahk, Dongkok Kim, Jae-Woo Kim

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine you are a chef preparing a grand feast for the entire universe. You have a new, high-tech kitchen (a telescope) and a brand-new, incredibly sensitive camera (the detector) that you plan to use to take "photos" of distant galaxies. But before you start cooking, you need to make sure your new tools work perfectly. If your camera is blurry, or if it adds weird static to your photos, your scientific discoveries could be ruined.

This paper is essentially a detailed "test drive" and "user manual" for two specific pieces of equipment: a PIXIS 1300BX camera and an IsoPlane 320A spectrograph. The authors are preparing these tools for a massive project called A-SPEC, which aims to take a "census" of nearby galaxies by analyzing their light.

Here is a breakdown of what they did, using everyday analogies:

1. The Problem: The Camera Without a Shutter

Most cameras have a mechanical shutter (like a door) that opens and closes to let light in for a specific amount of time. The PIXIS camera they tested is special: it doesn't have a mechanical shutter. Instead, it uses an electronic "shutter" that reads the image row by row, like a scanner.

The Analogy: Imagine trying to take a photo of a room with a flashlight, but you have to scan the room from top to bottom with your eyes instead of snapping a picture instantly. The top of the room gets lit for a split second longer than the bottom because you scan it first.
The Issue: In the lab, this caused a "gradient" or a slope in the brightness of the image. The top rows looked brighter than the bottom rows, not because the light was different, but because the camera was "reading" them for a tiny bit longer.
The Fix: The authors invented a clever math trick (a "gradient correction") to subtract this extra brightness, ensuring their measurements were accurate. It's like realizing your scale is slightly off because you stepped on it slowly, so you adjust the math to get the true weight.

2. Testing the Camera (The "Eye")

They put the camera through a gauntlet of tests in a dark, temperature-controlled room:

Darkness Test (Dark Current): They turned off the lights and waited. They wanted to see if the camera "glowed" on its own due to heat (thermal noise).
- Result: When cooled down to -55°C (very cold!), the camera was almost perfectly silent. It was so quiet that any "noise" it made was less than 1% of the signal they actually wanted to measure.
Sensitivity Test (Quantum Efficiency): They shone different colors of light on the camera to see how many photons it could catch.
- Result: It's a superstar! It catches over 80% of the light from blue to red (400–800 nm). Think of it as a net that catches almost every fish swimming by, whereas older cameras might let half the fish slip through. This is crucial for seeing faint, distant galaxies.
Persistence Test: They flashed a bright light and then immediately looked at a dark scene. They wanted to see if the bright light "stuck" to the sensor like a ghost image.
- Result: The ghost vanished in less than 10 seconds. It's like a camera that forgets a flash immediately, so it doesn't ruin the next photo.

3. Testing the Spectrograph (The "Prism")

A spectrograph doesn't just take a picture; it breaks light apart into a rainbow (a spectrum) to see what chemicals are in a star or galaxy. They tested this with three different "prisms" (diffraction gratings) that act like different levels of magnifying glasses.

Separating the Light: They used a bundle of 7 optical fibers (like tiny glass straws) to send light into the spectrograph. They needed to make sure the light from each straw stayed separate and didn't mix with its neighbors.
- Result: The system kept the "straws" of light perfectly distinct. No cross-talk.
Sharpness (Resolution): They measured how sharp the rainbow lines were.
- Result: Depending on which "prism" they used, they could resolve details from 600 to 2,600 times the width of the light wave. This is like being able to read the fine print on a contract from a mile away. The higher the number, the more detail they can see.

4. The Real-World Test (On-Sky)

Finally, they took the whole setup to the Seoul National University 1-meter telescope and pointed it at a famous, bright star called Vega.

The Goal: To see if the lab tests matched reality.
The Result: It worked perfectly. The system successfully captured the star's light, broke it into a rainbow, and measured its brightness accurately. They even corrected for the Earth's atmosphere (which acts like a dirty window) to get a clean reading.

Why Does This Matter?

The authors are essentially saying: "We have built a reliable, high-performance tool."

For the A-SPEC Survey: This system is now ready to be used for the main mission of mapping nearby galaxies.
For the Future: They didn't just test the gear; they wrote a recipe book (Python code) and shared it with the world. Now, any other astronomer can use their methods to test their own cameras and spectrographs, ensuring that future discoveries are based on rock-solid data.

In a nutshell: They took a new, shutter-less camera and a prism, figured out how to fix its weird "reading" quirks, proved it's incredibly sensitive and sharp, and showed that it's ready to help us understand the universe better.

1. Problem Statement

The authors are preparing for the A-SPEC (All-sky Spectroscopic Survey of nearby galaxies), a project requiring a high-performance, multi-object spectrograph (MOS) system. While modern CMOS sensors are gaining traction, Charge-Coupled Devices (CCDs) remain superior for faint-object spectroscopy due to their high quantum efficiency (QE) across broad wavelength ranges and stability.

The study addresses the need to rigorously characterize a specific commercial combination—the PIXIS 1300BX CCD camera and the IsoPlane 320A spectrograph—before their deployment for commissioning observations. A specific technical challenge arises because the PIXIS 1300BX lacks an internal mechanical shutter. When tested standalone in a laboratory (without the spectrograph's external shutter), the sequential readout of the CCD causes a row-dependent exposure gradient, where pixels at the bottom of the sensor accumulate more light than those at the top during the readout process. This effect introduces systematic errors in standard characterization metrics like linearity, gain, and quantum efficiency if not corrected.

2. Methodology

The authors conducted a comprehensive characterization in two phases: laboratory testing and on-sky validation.

A. Laboratory Setup

Optical Bench: Utilized an integrating sphere coupled with a continuum Xenon lamp and a monochromator to provide uniform, tunable illumination (300–950 nm).
Gradient Correction Technique: To address the shutter-less readout issue, the authors developed a novel gradient correction method. They acquired "illuminated bias" frames (zero-second exposure under uniform light) and subtracted them from flat frames. This isolated the row-dependent signal gradient caused by the sequential readout time ( $\Delta t_{excess}$ ), allowing for accurate subtraction of this artifact from subsequent measurements.
Characterization Metrics:
- CCD: Measured readout noise (RDN), dark current, gain, full-well capacity (FWC), linearity, photon transfer curve (PTC), quantum efficiency (QE), and charge persistence.
- Spectrograph: Evaluated spatial separability of multi-fiber spectra and spectral resolution using three diffraction gratings (150, 300, and 600 gr mm $^{-1}$ ).
- Resolution Analysis: Employed two methods for spectral resolution: (1) a Fourier-based stacking method of Th-Ar lamp lines to minimize pixel sampling effects, and (2) a direct 2D-Gaussian fitting of the fiber beam image combined with datasheet dispersion values.

B. On-Sky Validation

Observations: Conducted at the Seoul National University Astronomical Observatory (SAO) 1m telescope.
Targets: Used the standard star Vega ( $\alpha$ Lyr) for flux calibration and wavelength calibration using Th-Ar arc lamps.
Calibration: Performed a two-step wavelength calibration (stellar lines + Th-Ar) and derived an instrumental response function by comparing observed counts to the CALSPEC reference spectrum.

3. Key Contributions

Gradient Correction Algorithm: The development and validation of a technique to correct for row-dependent exposure gradients in shutter-less CCDs. This allows for accurate characterization of commercial cameras that lack mechanical shutters, a common issue in laboratory testing of astronomical instruments.
Open-Source Software: The authors released Python packages (EvalCCD and EvalSpec) containing all analysis routines and recipes used for the characterization, facilitating reproducibility and future instrument evaluations.
Comprehensive Performance Benchmarking: Provided a complete dataset for the PIXIS 1300BX/IsoPlane 320A system, including noise modeling that separates gradient noise from readout noise, and a detailed comparison of spectral resolution measurement techniques.

4. Key Results

CCD Performance (PIXIS 1300BX)

Gradient Effect: The uncorrected readout process introduced a vertical gradient of ~20,000 ADU. The correction technique successfully reproduced manufacturer specifications.
Noise & Dark Current: At the operating temperature of -55°C, the dark current is negligible (< 0.01 e $^{-}$ pixel $^{-1}$ s $^{-1}$ ), remaining an order of magnitude below the readout noise for typical 15-minute exposures.
Readout Noise & Gain: Measured readout noise ranged from ~12 to 41 e $^{-}$ rms depending on the gain mode. The Photon Transfer Curve (PTC) analysis confirmed the gain and Fixed Pattern Noise (FPN) factors.
Linearity: The camera exhibits excellent linearity (>99% accuracy) up to near-saturation. High-capacity modes showed piecewise linear behavior, while low-noise/high-gain modes were optimal for faint targets.
Quantum Efficiency (QE): The system demonstrated QE > 80% from 400–800 nm and >90% between 450–750 nm. This performance is superior to many commercial CMOS sensors in the red/NIR range and comparable to specialized astronomical CCDs.
Persistence: Charge persistence was found to be negligible, decaying to sub-readout noise levels within <10 seconds.

Spectrograph Performance (IsoPlane 320A)

Spatial Separability: Successfully demonstrated the ability to resolve multiple fiber spectra with negligible cross-talk using a 7-core fiber bundle.
Spectral Resolution ( $R$ ):
- 150 gr mm $^{-1}$ : $R \approx 640$
- 300 gr mm $^{-1}$ : $R \approx 1270$
- 600 gr mm $^{-1}$ : $R \approx 2580$
- The Fourier-stacking method provided more robust resolution estimates at high dispersion compared to the PSF-based method, which tended to underestimate $R$ due to pixel sampling limitations.
On-Sky Validation: Flux calibration using Vega successfully reproduced the continuum shape and major absorption features, confirming the system's reliability for scientific observations. The reliable spectral range was determined to be 390–900 nm.

5. Significance

This study validates the PIXIS 1300BX and IsoPlane 320A as a robust, high-performance platform for moderate-resolution spectroscopy ( $R \approx 600 - 2600$ ). The system is particularly well-suited for the A-SPEC survey and similar extragalactic projects due to:

High and Stable QE: Critical for detecting faint galaxies and accurately determining stellar populations in the red/NIR range.
Low Noise: The combination of low dark current and manageable readout noise ensures high signal-to-noise ratios for long exposures.
Methodological Rigor: The proposed gradient correction technique solves a critical bottleneck in testing shutter-less CCDs, ensuring that future instrument characterizations are not compromised by readout artifacts.

The authors conclude that this commercial system, when operated with optimal gain and fiber configurations, provides a cost-effective and scientifically capable solution for upcoming spectroscopic surveys.