The Next Generation of Astro Cameras: Testing the QHYCCD QHY268M

ads-medium
The QHY268M and its sensor window, and the author’s image of the Pacman Nebula taken with it. Credit: Rouzbeh Bidshahri

After testing over a dozen cameras for astrophotography, none ticked all the boxes for what I was looking for (sensor size, price, read noise, quantum efficiency, FWC, etc.). All seemed to compromise something important to me.

But when I saw the numbers claimed by QHYCCD for their new QH268M ($2,399), I thought I might have found what I was looking for, so I put in an order. This camera and the ZWO ASI2600MM (which uses the same sensor) are currently among the most popular cameras on the market.

Here we’ll look at the camera’s characteristics and my analysis of its sensor. I’ll also recommend settings that work well for me, and compare it to a few other cameras.

Overview

The latest generation of CMOS sensors, as used in the QHY268M, appears to have solved a lot of issues older CMOS cameras suffered, such as amp glow, 12-bit analog-to-digital converters (ADC), microlensing effects, and mediocre quantum efficiency.

The QHY268M is light and compact, weighing 855 grams (1.9lbs), and requiring only 12.5mm of precious back focus with a 28.3mm sensor. Unfortunately, it lacks auxiliary USB ports.

Most telescopes work well with this size sensor. I’m using 36mm filters, thus avoiding the step up to the next size (50mm filters, typically seven of them), which means a few thousand dollars less expense! The camera’s compact size causes less strain on the focuser as well.

Package Integration – QHY Camera Eco System

The QHY268M, off-axis guider, and spacer combinations: Credit: QHY CCD

Filter wheels – QHY offers several size filter wheels, including both 36mm and 50mm, that work with this camera. I prefer these wheels, which bolt on (6 x M3 screws) rather than threading on, and connect directly to the camera via a special cable without taking up a USB port.

Off-Axis Guider (OAG) – You can integrate the QHY OAG by bolting it to the filter wheel. My medium-sized OAG costs $280. I like that this bolt-on setup is solid, uses just 10mm of space, has a helical focuser. The prism can be moved as needed. My only complaint is that the prism could be larger than 8x8mm, but the smaller size does help keep the system slimmer.

Spacers – The camera comes with six bolt-on spacers ranging from 1 to 10mm with M54 and M48 telescope-side thread. More spacers are available, if needed.

The bolt-on parts are more solid than conventional threaded connections, and there’s no concern about stuck threaded adapters. The orientation of the camera, filter wheel, and OAG can be locked in, and the spacing can modified (think camera Legos).

The off-axis guider (OAG) on the right is the QHY with its smaller prism, a trade-off that keeps the OAG slim but can make finding guide stars a bit harder. Credit: Rouzbeh Bidshahri

Performance

The Sensor: The APS-C chip measures 28.3mm on the diagonal and has 26 megapixels, with resulting images of 6,252 x 4,176 pixels.  It’s a good-sized sensor for most uses. Very large sensors can’t be properly illuminated by some optical systems. The camera uses the new generation Backlit Sony IMX571 mono sensor with a native 16-bit analog-to-digital (ADC). The 65,535 values makes for smoother transitions with less quantization error.

Download Speed: The 50MB files download in three seconds over a USB3 connection, up to 1/10 the time older CCDs would take. The camera also has 1 gigabyte of DDR3 memory (the ZWO ASI2600 has 256Megabytes by comparison). That saves a lot of time, especially when taking many short subs or going through autofocus routines and model-building.

Pixel Size: The camera has relatively small 3.76-micron pixels, though you can simulate larger pixels by “down-sampling” or binning.  This gives me more flexibility; small pixels are a better match with my shorter focal length telescopes, and I bin 2x to simulate larger pixels for my longer 1700mm telescope.

Quantum Efficiency (QE): While Sony didn’t release official numbers, QHYCCD tested the sensor and measured it at about 90% peak efficiency. This is one the highest numbers available and one of the characteristics that attracted me to this camera.

This means the camera will detect almost all the incoming light, so it’s much more sensitive and will require less integration time compared to older, less efficient sensors (older generation CMOS and CCDs were in the 60% range).

Read Noise (RN): The QHY268M has exceptionally low read noise; with my gain set to 56, read noise is only 1.55 (e-) / pixel, which is especially useful for narrowband imaging.

The image thus contains more signal and less noise (Signal to Noise Ratio, or SNR). Subs look cleaner and less grainy, and I don’t need very long subs to get good results (see my recommended settings below).

Full Well Capacity (FWC): FWC per square micron of this sensor is exceptionally high. Most importantly, that holds true even with high gains (i.e., low read noise).

For example, at gain 56 the FWC is 21k (e-)/pixel with a small dose of 1.55 (e-) of read noise. To compare, the ZWO ASI1600 at that same read noise level, would only have 2k worth of FWC, just 1/10 of the QHY268M!

This means I can expose long enough to get dim regions without clipping and oversaturating the brighter areas. The stars look much better without appearing bloated.

QHY268M Read Noise and Full Well Capacity vs. Gain: Credit: QHYCCD

 

Side-by-side comparison of Full Well Capacity of the QHY268m (left) and the ZWO ASI1600 (right), previously the most popular CMOS camera. Credit: QHYCCD, ZWO

Thermal Current and Control:

Dark Current: The camera’s dark current is negligible. For reference, the QHY268M has less than 1/10 the dark current of the ASI1600. It doesn’t really need to be cooled more than  -5 or -10 degrees C.

Less cooling means less power usage in the field and less chance of condensation or frost in the chamber. The cooler holds my set temperature to within 0.1 degrees C, and it includes a heated window (where light enters from the telescope) that prevents frost.

A feature I really like is that the software shows the humidity inside the sensor chamber so you can see if there is any risk of sensor frost. If I see high humidity, I can recharge the chamber with the external desiccant plug without opening the camera or sending it in for a $500 service.

The refillable desiccant and other features keep the camera free of humidity, preventing frost condensation inside the camera. Credit: Rouzbeh Bidshahri

Amp Glow: QHYCCD claimed he QHY268M has no amp glow. I tested this with actual dark frames downloaded from the camera and verified their claim.

This results in better data and simpler processing since dark-flat calibration frames are not essential as with older CMOS cameras with considerable amp glow.

Author’s test dark frames at 300 seconds and -5 degrees C, highly stretched to show any glow. Credit: Rouzbeh Bidshahri

Testing

Given how critical these numbers are, I decided to analyze the camera myself using Pixinsight and Sharpcap Pro. The measured values were similar to those claimed by QHYCCD.

Camera Properties test by Pixinsight. Credit: Rouzbeh Bidshahri

 

Sensor Analysis with Sharpcap Pro. Note the very good of 99.8% as well. Credit: Rouzbeh Bidshahri

My Settings

QHYCCD provides four readout modes to select from based on your priorities, including Read Noise and Full Well Capacity. I usually use Mode #1, the “High Gain” mode. CMOS cameras are a bit tricker than CCDs since you can adjust multiple settings such as mode, gain, offset, etc. I won’t get into the science behind it here, but the settings below are what I use. Note that these settings depend on many factors but they provide a good starting point.

Recommended settings. Credit: Rouzbeh Bidshahri

Quick Comparison to Other Cameras

With so many cameras and so many numbers, I use this method to get what I call the performance “Indicator Number” that I use to compare cameras.

First, consider these factors:

  • The size of the sensor required is mostly dictated by the optics and the desired field of view. Budget often determines the size as well since larger sensors and the larger filters they require cost more.
  • Pixel size can be manipulated by re-sampling to match the optics and sky conditions.

I then pick the top 3 equally important performance parameters:

  1. Full Well Capacity: Larger is better.
  2. Read Noise: Lower is better.
  3. Quantum Efficiency: Higher is better.

Since the ratio of FWC to RN is essentially the Dynamic Range, I normalize them by dividing each by the area of a single pixel (i.e., per micron). I can then compare sensors side-by-side using for a square micron of each chip. I then multiply by QE to account for uncounted photons.

Comparing these three key parameters and the resulting performance Indicator Number (a higher Number indicates better performance) between the QHY268M and some other workhorse cameras below, the QHY268M would be expected to greatly outperform the rest.

Side-by-side comparison of popular cameras. Credit: Rouzbeh Bidshahri

Final Thoughts

The QHY268M has worked very well for the past 10 months, with no issues. I have not detected any sensor tilt, microlensing artifacts, or frost issues. The drivers have worked and integrated well with my imaging software, N.I.N.A. Overall, the camera’s sensitivity and performance are excellent, as predicted by the numbers and analysis, and it has produced great results. The only thing I could ask for is it to have a larger sensor, which is why QHYCCD offers the QHY600m with almost identical characteristics. However, a larger full frame sensor doubles the cost of the camera, filter wheel and filters. For me, the QHY268M is in the sweet spot with regards to price, size, performance, and usability.

MSRP: $2,399

Website: www.qhyccd.com

Below are some images I’ve taken with the QHY268M. You can see more in my image gallery: https://www.astrobin.com/users/Rouzbeh/

The Wizard Nebula – NGC7380 – SHO – 10 hrs 15 mins of integration. Credit: Rouzbeh Bidshahri
The Crescent Nebula – NGC6888 – HOO – 4 hrs 45 mins of integration. Credit: Rouzbeh Bidshahri
The Bubble Nebula – NGC7635 – SHO – 22 hrs 35 mins of integration. Credit: Rouzbeh Bidshahri

 

About Rouzbeh Bidshahri

Rouzbeh Bidshahri is a mechanical engineer with a lifelong passion for astrophotography. He has tested dozens of telescopes ranging from 3 to 20 inches in aperture and has spent several years optimizing systems for very high-resolution planetary imaging in the sub 0.1 arcsecond/pixel range. He has contributed to several institutions such as ALPO (The Association of Lunar and Planetary Observers). His main area of interest has been designing and operating larger setups, and he is currently focusing on high resolution, long exposure photography for both broadband and narrowband deep sky imaging.

Related posts