CMOS OR CCD MATRIX

The company ZWO was the first to introduce a series of cooled CMOS matrix cameras for the deep sky on the market. The firms Atik and QHY also offer cooled CMOS in camera models also dedicated to the deep sky. The technical specifications are similar.

ZWO also offers a new camera with a Sony monochrome CMOS matrix (IMX571) cooled for the deep sky, which performs better than the Panasonic monochrome CMOS matrix.

In order to compare the performance of these affordable cameras, here is a comparative analysis of CMOS and CCD matrix cameras.

This update of this benchmarking is based on a better understanding of how CMOS arrays work which is more complex to grasp than CCD arrays due to camera gain setting options. It is therefore not based on the specifications of the manufacturers of the CMOS matrices which tend to present the best performance without taking into account the gain adjustment options necessary to obtain the details of the object from the deep sky.

All the cameras analyzed have a wide field matrix. There will be a comparison between the cameras with color matrix and the monochrome one. Considering that there are constantly technological innovations in these types of cameras, I will regularly update this benchmarking.

All monochrome CCD cameras feature the popular Kodak KAF-8300 matrix. Considering the large sales volume of these models, they are offered at a much lower price than other wide field monochrome CCD arrays. It is therefore interesting to compare their performance with those with a CMOS matrix, the price of which is very competitive.

Here is a table to compare these different types of cameras:

This rating is for the use of a single camera for the wide field of view and the deep sky. So for use under all circumstances. It therefore favors cameras with smaller pixels with a large matrix allowing use for the large field of view (short focal length). It will allow better sampling. For the deep sky (long focal length), we will have to adjust the sampling using the Bin 2 × 2 mode. As the image will be reduced by a factor of 4 in this mode, it is also necessary to favor a greater number of pixels to maintain a large image for the deep sky. For example, a camera of 8 Mega pixels (Mp) in bin 1 × 1 will be 2 Mp in bin 2 × 2. This will avoid buying two cameras, one for the large field of view (short focal length which is the focal length most used for nebulae) and the other for the deep sky (long focal length which is mainly used for galaxies and planetary nebulae). It must be remembered that it is not the majority of amateurs who can afford to invest in two cameras which represent an investment of more than $ 2 per camera! For more details and understanding of this statement, see the section Astronomical calculations.

Here is the description of some elements of the table:

  • Price in Canadian Money (CAD) (February 2021): For monochrome cameras, I added to the price of the camera, the price of the motorized filter wheel as well as the LRGB filters. For other currencies, use the daily exchange rate to get an estimate of the costs for the currency of your country.
  • Format - bits: This is the conversion of the analog signal to digital format. To learn more about the different image formats, click on this link.
  • Load capacity (full weel capacity): This is the load capacity of the light signals captured by a pixel before its saturation (Blooming). The larger the number, the better the dynamic range (greater signal-to-noise ratio). We can compare the load capacity of a pixel to a bucket of water, the deeper it is, the more it can be filled before it overflows, thus allowing to acquire a greater amount of information on the object. to image without saturating the stars.
  • Bin 2 × 2: click on this link
  • Camera cooling: At each cooling of 6o Celsius, the thermal noise decreases by 50%. It is therefore an important element to consider when choosing a camera.
  • Cooling regulation: This feature allows the cooling to be fixed at a fixed temperature, for example -20Celsius. This greatly simplifies the production of Blacks (Darks).
  • Read noise: Camera manufacturers indicate read noise (Read Noice - RMS) in electron (e-). The larger the number, the more noise is added to the image.
  • Quantum efficiency (RQ): Click on this link

The rating for each camera:

  • For each element, the best performing instrument receives the maximum rating of 10; the others are prorated. For the 12-bit, 14-bit and 16-bit formats, the calculation is performed on the ability of the matrix to distinguish different shades in the image: 12 bits = 4096 shades, 14 bits = 16384 shades and 16 bits = 65536 shades.
  • The Global Score: this is the average of the individual scores. The maximum overall score is therefore 10.

Note 1: Load capacity of CMOS matrices

For CMOS arrays, the load capacity varies according to the increase in gain (sensitivity) of the camera. For example, for the monochrome Panasonic matrix, at zero dB gain it is 20000e-. At a gain of 30dB, it decreases to around 500th - only! The load capacity of the board is with an average gain of 15 dB (or 150 in units of 0,1 dB) which gives a load capacity of about 3500e- for the Panasonic monochrome matrix. Here, we must consider, in the reality on the ground, that it is always necessary to increase the gain to get details in the object of the deep sky. By using an average gain of 15dB, this makes it possible to better assess the actual performance of CMOS cameras and not to use the manufacturers' specifications which mention a load capacity of 20000e- at 0dB for the Panasonic monochrome matrix, which is practically impossible. to use in reality.

Note 2: Bin 2 × 2


Monochrome CMOS matrices do not offer the hardware 2 × 2 bin which achieves 4x more signal than the 1 × 1 bin. The 2 × 2 Bin of CMOS matrices achieves more frames per second than the 1 × 1 bin. This characteristic is therefore of no use for deep sky imagery. It can be interesting for the imaging of planets by offering more images per second. 

Conclusion of this comparative analysis

The color CMOS matrix has an overall higher rating than the color CCD sensor (5,3 compared to 4,6). If your choice is a color camera, then the ASI 294MC Pro camera offers the best performance overall. Its price is also unbeatable ($ 1294 versus $ 3603 for the CCD camera).

For monochrome cameras, the new ASI 2600MM Pro camera with CMOS matrix has the best overall performance (7,9 compared to the other three which have scores between 6,5 and 6,6). We are therefore currently witnessing a victory for CMOS matrices over CCDs for affordable cameras. It is also the first CMOS camera in 16-bit format that can reproduce 65536 shades like CCD cameras, which is a technological feat.

Note that for monochrome CMOS matrices, the Bin 2 × 2 mode cannot be used. I remind you that the Bin 2 × 2 allows to acquire 4 times more signal than the Bin 1 × 1 for the same exposure time. In addition, this mode allows you to adjust the sampling of the camera according to the object to be imaged (as mentioned in the above statement). Monochrome CMOS matrices are therefore less versatile than monochrome CCD matrices and are therefore more dedicated to short focal lengths (large field of view). For me, this remains an important advantage, as I always use the 2 × 2 bin mode for all my images acquired with my Atik 383L monochrome camera, which allows me to get 4 times more signal.

 

Richard Beauregard
Sky Astro - CCD

Revised 2022/08/14

 

References  

According to manufacturers' specifications. For CMOS matrices, I adjusted manufacturer ZWO specs using an average gain of 15dB. To get the information, I used the graphics of it which are shown on their website.

For the RQ, the values ​​used are for red at 650 nanometers, green at 550 nm and blue at 450 nm thus allowing a comparison between the cameras. For color cameras, the RQ is adjusted as follows: RQ * 25% for red and blue and RQ * 50% for green.