You might ask, "What is the difference between a camera for EAA and one for astrophotography?". The answer is, nothing really. There is no hard and fast line between the two activities. Any camera that will work for astrophotography will work for EAA and vice-a-versa including DSLRs. Now, there are cameras that are more commonly used for EAA and these are the cameras with CMOS sensors selling for a few hundred dollars to no more than ~$1500. Based upon posts on multiple forums, I would estimate that the most common EAA cameras fall into the price range of $250 to $1300. Again, there is no hard and fast rule here. But you will find that the cameras costing more than ~$1500 are generally used by astrophotographers and not those doing EAA. Astrophotographers have traditionally used cameras with CCD sensors because of their higher sensitivity compared to CMOS sensors. But that is changing as the sensitivity of CMOS sensors is approaching that of CCD sensors while the CMOS sensors have the advantage of much lower read noise. Lower read noise makes live stacking of short exposures extremely practical for EAA. In addition, the two major suppliers of CCDs are Sony and On Semiconductor. In 2015 Sony announced that it would stop development of new CCDs by 2017 and On Semiconductor announced it would stop production of all CCDs in 2020. So as supplies of CCDs on hand dwindle, new astronomy cameras will eventually only use CMOS sensors. Therefore, in this blog we will concentrate on those with CMOS sensors costing no more than ~$1500 and only discuss a few using CCDs. That will leave out many CCD cameras from companies like SBIG, QSI, FLI, Apogee, QHY, Starlight Xpress, and Atik which are more often used by astrophotographers.
Another important point that needs to be made is the distinction between cameras for Deep Sky Object (DSO) viewing and cameras for planetary viewing. Once again there is no hard and fast line between the two, but typically cameras for planetary work have smaller sensors, fewer pixels and higher frame rates. This is because the planets are small bright objects and lucky imaging is employed to capture thousands of sub-second frames from which a few hundred of the best images are used to create a final image. In contrast, DSOs are much larger and fainter and exposures are several seconds to tens of seconds in length. Having said this, both types of cameras can produce pleasing images of both types of objects.
For most of us, the cost of a camera for EAA is the primary driver of what we ultimately buy. Fortunately, it is not necessary to spend a lot of money for a camera for EAA. An entry level camera like the 1.2 Mega Pixel (MP) Rising Tech IMX224 camera with the 6.1mm diagonal Sony IMX224 color sensor is available for $165. If you already have a telescope on a motorized mount the cost of entry is very minimal. While the IMX224 sensor is small, it is quite capable and provides a cost effective option for the EAA beginner. In fact, when first introduced around 2015, the ASI224MC using this sensor and costing $300 was widely popular as one of the earliest CMOS cameras to be employed by the EAA community. As larger CMOS sensors became available, cameras with increasing pixel count and sensor size have been widely adopted for EAA. It is now possible to find CMOS based astronomy cameras prices from the $165 camera mentioned above to many thousands of dollars. The most commonly used EAA cameras today have color sensors with pixel counts of ~9MP to ~20MP, and sensor diagonals of ~15mm to ~23mm. Obviously as the sensor size increases, the cost goes up as well. Other drivers of cost include a Peltier cooler to minimize thermal noise , an internal memory buffer to prevent lost frames during image download and a USB hub for connection to a focuser and guide camera. We will review a comprehensive list of available cameras and cost once we discuss the key camera attributes we need to understand before choosing which camera is best for our needs.
Color or Mono (B&W)
Perhaps the most important consideration when choosing a camera for EAA is whether to choose a color or monochrome camera. Many cameras are available with either color or mono (black and white) sensors. Color sensors are just mono sensors with a red, green and blue filter matrix on top of the individual pixels. This matrix is called a Bayer matrix after the Kodak scientist who invited it in 1976 to turn a mono camera into a color camera. Because the human eye is most sensitive to green light, the Bayer matrix is typically arranged as a 2 x 2 matrix of pixels with 2 green and 1 each red and blue filters. These days, the filters also act as micro lenses to focus off axis photons onto the pixel thereby maximizing the light collection sensitivity of the sensor. Sometimes cyan, magenta and yellow filters are used instead of red, green and blue but the result is the same.
To realize the full real time viewing experience only a color camera can show the rich colors present in nebulae, star forming clusters in distant galaxies and the different colors of stars at various stages of their lives. Being able to view objects in color is one of the major advantages of EAA compared to viewing with an eyepiece (EP). Because of the filter matrix associated with a color camera, some sensitivity is lost compared to a mono camera as a trade off for the simplicity of a one shot color camera (OSC). While there is nothing preventing the use of a mono camera for EAA, they are more often used for astrophotography combined with external filters to capture images at each color which are combined later to form a full color image. Some EAA'rs use a mono camera to view DSOs in black and white in real time to take advantage of the added sensitivity of a mono camera, or even use a mono camera with one of the possible narrow band filters available (hydrogen, oxygen, sulfur) to view specific detail in deep sky objects. This may be a good approach if you ultimately want to pursue astrophotography as well. We will see below that cameras with the mono version of a particular sensor cost significantly more than the color version which is another reason for the greater popularity of a OSC camera for EAA.
The size of the sensor chip will determine many important attributes of the camera. Most important, the larger the size of the chip, the bigger the field of view (FOV). In the Blog "Choosing a Telescope for EAA" we made the point that the FOV for an optical system consisting of a telescope and a camera is given by the equation:
FOV ~ 57.3 L / F
where L is the length of one side or diagonal of the sensor chip in mm, F is the telescope focal length in mm and FOV is in degrees. So the larger the sensor chip the larger the FOV. For instance, let's take the case of an 8" SCT with a Celestron f/6.3 focal reducer to achieve a focal length of 1260 mm (2000 mm x 0.63). Assume we are using one of the cameras like the Rising Tech 224 or the ASI224MC, with the 6.46mm diagonal Sony IMX224 chip. The FOV will be 0.29 degrees along the diagonal, slightly larger along the long axis and slightly smaller along the shorter axis of the chip. Now, if instead we use a camera with the Sony IMX294 chip with a diagonal of 23.2 mm the FOV will be 1.05 degrees along the diagonal which is 3.6 times larger.
The larger chip size has several advantages. First, it allows us to capture much larger DSOs in a single image frame. An excellent tool for calculating FOV and image scale for different optical configurations is Ron Wodaski's CCDCalc. The side by side images below were generated by CCDCalc using a telescope with a focal length of 1260mm (our 8" SCT with f/6.3 focal reducer or any other telescope combination to achieve the same focal length) and the IMX224 and IMX294 camera dimensions. Clearly, something like the Triffid Nebula is too large to fit into a single frame with the much smaller chip of the IMX224, but fits nicely into the chip of the IMX294.
In addition to enabling larger DSOs to fit inside a single frame, the larger FOV can make it much easier to align the telescope and to find objects because of the larger FOV. In effect, the sensor diagonal acts like the focal length of an EP in determining the FOV. So, the IMX224 with a diagonal of ~6.5mm gives a similar FOV as a 6 to 7 mm focal length EP, while the IMX294 gives a similar FOV as a 23 mm focal length EP.
On the other hand, the smaller sensor, like the smaller focal length EP, provides higher magnification of an object compared to the larger sensor. Compare the image frames for M61 below between the IMX224 and IMX294 sensors. Obviously, the image will appear larger on the computer screen with the IMX224 sensor and will fill the frame while the image appears much smaller with the IMX294 sensor. Now there is a trick here. The IMX224 has only 1.2MP while the IMX294 with 11.3MP has nearly 10X the number of pixel. The image from the IMX294 can be zoomed in or viewed on a much larger screen like a 60" TV without producing a blocky or pixelated image whereas the IMX224 image doesn't have enough pixels to do that. The IMX224 image will support a 720p video display format which is just below full HD or 1080p, while the IMX294 has more resolution than a 4k video display. Having said that, the IMX224 resolution is more than 2X that of the analog video cameras with 0.4MP which were the only options for EAA up until ~ 2015.
Another thing to be aware of as the sensor size increases, vignetting of the image will become more obvious. As the light travels through the optical path any narrowing of that path due to baffling inside the tube, the diameter of the opening at the back end of the telescope, the focuser, any adapters, focal reducers, filters, etc. can block some of that light toward the outer radius of the FOV. This will show up as a halo toward the outer edge of the image. This is typically not a problem in the very small sensors less than 10mm in diagonal. Focal reducers will make vignetting worse as they push the light cone further back from the exit of the telescope. Vignetting may or may not be an issue for the individual viewer depending upon how severe. This is why APS-C (28mm) and full frame cameras (43mm) can be challenging. However, vignetting can be corrected with flat frames applied to the image on the fly with live stacking software.
It should also be noted that as the sensor size increases, so does the number of pixels which means that file sizes get larger fast. This directly impacts the the amount of computer storage necessary if one wants to save images. It also can impact the ability to stack images live in software if the computer used doesn't have sufficient CPU capability.
The biggest driver of the move from CCD cameras to CMOS cameras after cost has been the fact that CMOS cameras have amazingly low read noise. What is read noise? It is the random noise due to the uncertainty in counting the number of electrons created in each pixel by the photons striking that pixel. Read noise is independent of the signal, or amount of photons collected, so it is independent of the exposure. This noise is introduced into the image data when the captured image frame is downloaded from the sensor to the output device, display and computer hard drive. Read noise for CCD cameras is typically greater than 5electrons (5e). The Atik Infinity camera with the Sony ICX825 CCD has a read noise of 6e while the Starlight Xpress UltraStar using the Sony ICX825 CCD has a much better read noise of 3.5e. CCD cameras used for astrophotography like the SBIG STF-8300 have as much as 9.3e of read noise. In contrast, CMOS cameras can have a read noise less than 2e depending upon the gain setting used. At this level, the read noise is less than the other sources of noise which impact the image and can effectively be ignored for EAA. The read noise is a strong function of gain used as shown in the plot from QHY below.
A lower read noise means that there is no penalty for taking many short exposures and stacking them live rather than taking a single long exposure. In fact, commensurate with the introduction of these low read noise CMOS cameras, live stacking software like SharpCap became widely available. Whereas in the bygone days of analog cameras with CCD sensors we used exposures of 30 seconds to several minutes to bring out detail in our images, EAA has moved more toward stacking many very short exposures, 5 and 10sec, to achieve the same total exposure time now that live stacking software is readily available.
The ability to use many short exposures has had multiple effects on EAA. First, instead of waiting a minute to see something on our display, we begin to see the faint evidence of a DSO in very short order. This gets better in real time right before our eyes with more detail and less noise with each additional frame added to the stack. Second, the use of short frames means that a less than perfect polar alignment of our equatorial (EQ) mount does not cause objectionable star trailing since the exposure is not long enough to show the effect of an imperfect alignment. Third, with very short exposures we can now use Alt-Azimuth mounts for EAA which are simply out of the question for traditional astrophotograpy. Alt-Az mounts have the advantages of being much easier to set up since a polar alignment is not needed nor is one possible. Also, Alt-Az mounts tend to be less expensive than EQ mounts. Fourth, shorter exposures can also help to avoid saturation of bright stars in the FOV, thus improving the dynamic range of the viewed image.
So, low read noise can be a big plus for EAA enabling short exposure live stacking, the use of less expensive mounts, and a simpler setup routine. Fortunately, most CMOS cameras have exceptionally low read noise and one can concentrate on other features in deciding which camera is best suited to their EAA needs.
Yet another choice one must make in selecting a camera is whether to purchase a cooled or uncooled camera. Many models are available with Peltier or Thermo Electric (TEC) cooling of the camera sensor to minimize thermal noise. Thermal noise is the result of heat buildup inside the camera from the camera electronics and from the background air temperature. Thermal nose results in a dark current within the camera sensor which shows up as random noise in the background of an image frame. Thermal noise is fairly predictable at a given temperature and can be controlled with a well regulated TEC. A TEC typically allow for temperatures to be maintained ~ -35 to -40 degC below ambient.
Cooling is more important for long exposures as the heat built up in the camera during these long exposures is greater than in short exposures. Therefore, if the strategy is to stack many very short exposures, say 10sec or shorter, the advantage of TEC cooling may not be worth the added cost and complexity. Also, the background noise from the dark current can also be very effectively removed by using dark frame subtraction. A set of dark frames, typically 16, can be collected and averaged at the exposure planned for live viewing. The dark frame average is used as a master dark frame which can then be subtracted on the fly with live stacking software like SharpCap to remove the dark noise from each frame. For this to be effective, the dark frames must be collected at the same exposure time, gain and temperature as the frames during live viewing. Also, because ambient temperatures can drop significantly during the night, new dark frames may need to be taken and a new master dark frame used throughout the night. On the other hand, with TEC cooling, a library of dark frames can be made ahead of time at different exposure times and temperature offsets from ambient to be used as needed throughout the night. This library can be made during the day or during a cloudy night so that no time is wasted on nights with good visibility.
Several things need to be considered when using a camera with a TEC. First, the lower temperature at the sensor can result in dew buildup on the sensor and on the camera's glass window. Many cooled cameras come with a sealed chamber to minimize water vapor causing dew build up inside the chamber. This still leaves the outside glass window of the chamber exposed to dew buildup which is why many cooled cameras now come with a dew heater strip surrounding the chamber window. Second, while the camera itself requires very little power, typically less than 2.5W which can be supplied via the USB connection to the camera, cooling requires an additional 12V power supply capable of supplying ~20 - 35W and an additional cable to the camera. Consideration for extra battery capacity when visiting a remote dark site is also necessary with a cooled camera.
QE, Full Well Depth, Bit Depth, Frame Rate
With so many different specifications for the sensors inside astronomy cameras, there is not single ideal camera. Most likely, once you determine the price range you plan to spend, whether to go with color or mono, cooled or uncooled and the chip size that will work best for your application, the remaining specs will already be determined. Nonetheless, it is worthwhile to go over those additional specifications as they will impact the performance of the camera.
Quantum efficiency (QE) is a measure of how well the pixels in the sensor convert an incoming photon into electrons. The silicon which makes up the sensor is responsive to light over a range of wavelengths centered in the visible but extending into the near infrared and ultraviolet, typically from ~300nm to 1,000nm. The QE varies with wavelength and is usually expressed in a graph with separate curves for red, green and blue light. Quantum efficiencies can vary quite a bit with values ranging from 50% to 84% for typical EAA cameras. The lower the QE the longer the exposure needed to collect the same amount of photons compared to a higher QE.
Another important attribute of the sensor inside a camera is called the full well capacity. This is a measure of the maximum number of photons that a pixel can collect before it is full of electrons and cannot detect any additional photons. The full well capacity is proportional to the size of the pixels with larger pixels capable of capturing more photons before they become full. Full well capacities can vary significantly from camera to camera with a range of about 15K electrons for the IMX183 sensor with a 2.4 micron pixel size to 64K electrons for the IMX294 sensor with a 4.63 micron pixel size. Full well capacity will determine the range between the brightest object and the dimmest object. If the full well capacity is small, bright stars will saturate unless the exposure is shortened which will reduce the intensity of the dimmer objects in the FOV. All other things considered, a higher full well capacity is preferred.
It should be noted that the stated full well capacity is measured with zero gain. Gain is equivalent to the ISO setting on a DSLR camera which multiplies the number of electrons captured at each pixel thereby increasing the camera's sensitivity by allowing resolution of smaller differences in the number of electrons captured. Gain values can be anywhere from 0 to ~450 depending upon how the camera manufacturer sets it up. The trade-off with higher gain is reduced dynamic range. Dynamic range is the ratio of the largest signal (brightest object) to the smallest signal (dimmest object or background sky). A large dynamic range is desired to show the full range of objects without washing out the brightest ones in the image. For instance, a large dynamic range is needed to capture the dark dust lanes without blowing out the core of a galaxy. As gain is increased the full well capacity of the camera is reduced because the multiplication factor uses up more the the well capacity. This, in turn, causes the dynamic range to decrease. So gain helps with sensitivity but hurts with dynamic range.
Yet another specification that you will find with each camera is bit depth. Typical cameras used for EAA have bit depths of 12 or 14 but some of the newest higher prices cameras are 16 bit. What is bit depth? After the camera captures an image frame an analog to digital converter (ADC) converts the analog voltage associated with the number of electrons in a pixel into an integer or digital value. With a 12 bit device, 2^12 or 4096 discrete values are possible for each pixel. At 12 bits, the camera uses a bit value of 0 for no voltage detected and 4096 for the maximum voltage detected. A 14 bit device has 4 times as many possible values so it also uses 0 for no voltage but can now use 16,384 bits for the maximum voltage detected providing a finer scale with which to better differentiate details within an image. A bit depth of 14 is most common among the current batch of CMOS cameras, but 16 bits is starting to show up in the latest high end CMOS cameras. A 16 bit device has 65,536 possible values. The more bits possible, the finer the resolution in the photon levels detected. All other things considered, a camera with a higher bit depth is desirable. Above is a comparison from ZWO showing the increased grey scale resolution with more bits
Cameras will also have a maximum frame rate specification which indicates how many full frame images can be downloaded from the camera per second. The download rate may be limited by external cabling, with USB3 supporting higher data rates than USB2. Frame rates are not super important for EAA since exposures are a few seconds or longer. Frame rates of 10 to 23 frames per second (FPS) are typical for cameras commonly used for DSOs. This is in contrast to planetary cameras with frame rates of 20 FPS to 170 FPS. Which is not to say that a camera used for DSOs cannot also be used for planetary imaging; just that it is not optimized in terms of pixel size and frame rate for planetary work. Another point to note, with binning the frame rate will increase since there are less individual pixels of data to transfer.
USB Hubs, Internal Memory, Binning, etc.
There are several other features to consider when choosing a camera which will be discussed here.
Most CMOS cameras today allow binning of the sensor pixels which simply merges the signal from adjacent pixels into one larger pixel. Typically binning is available in 2x2, 3x3 and even up to 4x4. Binning 2 x 2 means that four adjacent pixels are combined into one, which effectively increases the pixel size by a factor of 4, improves the sensitivity by a factor of 4, but reduces the resolution by the a factor of 2. Binning 3 x 3 combines a 3 pixel square of 9 pixels into one and 4 x 4 binning combines a 4 pixel square of 16 into one very large pixel. Because of the increased sensitivity, binning is helpful when searching for objects, doing an alignment and framing targets as it shortens the time needed to get a recognizable image. How does binning with a OSC camera keep the color since the adjacent pixels have different (R, G, B) color filters? The answer is that the binning is done in software which allows the camera to deBayer the image before binning so as to preserve the color information.
Now that we are discussing binning, we need to also visit the concept of image scale. Recall from the Blog " Choosing a Telescope for EAA" that image scale is determined by the ratio of the sensor pixel size to the telescope focal length.
Image Scale (arcsec/pixel) = 205 x Pixel Size (microns) / Focal Length (mm)
Now, it is widely accepted that typical seeing limits observable detail to ~ 2 arcsec per pixel. On nights of better seeing an image scale less than 2 arcseconds per pixel is possible. Consider an 8" SCT at f/10 with a focal length of 2000 mm. Let's use 4 microns as an approximate size of a pixel in a typical EAA camera. This gives an image scale of 0.41 arcsec/pixel which is much smaller than the typical seeing which means that we are oversampling relative to the sky conditions. Now, what if we use a Hyperstar lens to achieve a focal ratio of f/2 so that the telescope focal length is now 400mm. In this case we are operating at ~2.1 arcsec/pixel which is perfect for typical seeing conditions. Take another example of an 8" Newtonian at f/3.9, or 780mm which gives an image scale of ~1.1arcsec/pixel. And last, consider a 127mm refractor at f/5 with a focal length of 635mm which leads to an image scale of ~1.3 arcsec/pixel. The point is that with the small size of CMOS sensors, with most telescope setups we will be oversampling relative to a 2arcsec/pixel seeing limit. There is no harm in this, but it does say that we can bin 2x2 and get the advantage of 4X the sensitivity without loosing significant resolution in all but the best seeing conditions. With 2x2 binning of a 4 micron pixel the pixel size becomes 8 microns and the image scale increases accordingly as shown in the table below. Because the image scale with typical CMOS pixel sizes is much smaller than 2arcsec/pixel, binning will not reduce the resolution unless the seeing conditions are sub 2arcsec/pixel or even sub 1arcsec/pixel.
Another concept we should discuss is amp-glow. In the days when we used analog cameras with CCD sensor all cameras had to deal with amp-glow. Amp-glow appears as a bright region at an edge or corner of the image which is caused by IR radiation from the read out amplifier. These IR photons are picked up by the nearby pixels and show up as a background glow, hence the name amp-glow. Now, CMOS sensors have completely different circuitry but can still suffer the effects of glow from the other on board circuitry and this varies from camera to camera in intensity, number and shapes of glowing regions. The pictures below show several different types of amp-glow possible with different sensors.
Amp-glow can be handled with dark frame subtraction which, for EAA, means live stacking software must be used. Some cameras advertise amp-glow control which can virtually eliminate the effect without dark frame subtraction. An internal memory buffer is used to increase the readout speed thereby reducing the time the readout circuits are active. Several camera makers also have "Anti Amp-Glow" hardware and software which reduces the power in the CMOS circuitry thereby minimizing amp-glow but no real details are given about how they do this.
Minimum and maximum exposures vary by camera but typically have a minimum of tens of micro seconds which is sufficient for bright planetary objects and maximums of 30 to 60 minutes which is much more than is needed for EAA. The Atik Infinity has the shortest maximum exposure of 120 seconds but because it uses the very sensitive Sony ICX825 CCD this is more than sufficient for most EAA applications but does limit the total time for live stacking to 2 minutes.
Most dedicated astronomy cameras these days come in a cylindrical shaped body which are designed to have a small footprint. This is important when using the Hyperstar adapter on an SCT as it minimizes the amount of incoming light blocked by the camera. The bodies are typically ~3" or smaller in diameter. The ATIK Infinity is an exception as it has a rectangular body 70 x 113mm in dimensions. On the other hand, cooled camera bodies are 4" to 5" long which means that they will run into the base of some Alt-Az mounts like the Celestron Nexstar mounts.
Camera Options for EAA
Like mounts and telescopes, there are far too many cameras available for EAA to cover them all. We will concentrate on CMOS cameras with two exceptions for popular CCD cameras. Also, it is important to note that the manufacturers of the cameras we will discuss use the same sensors from Sony and Panasonic so one can find very similar cameras from ZWO, QHY, Mallincam, Rising Tech, Altair Astro, etc. Below is a table of the key characteristics of the most commonly used sensors in EAA cameras. The table is arranged from smallest to largest diagonal. Many of these sensors come in both a color and mono version, although color is much more commonly used for EAA to get the full benefit of real time viewing. The advantage of the mono sensor is its higher sensitivity, especially when used with narrow band filters to cut through local light pollution. The smallest sensor is the IMX224 which was discussed above and can be found in the least expensive cameras. Most sensors have diagonals in the range of 16mm to 22mm which can result in obvious vignetting especially if significant focal reduction is used. Vignetting can be addressed with flat fields when using live stacking software or minimized with a telescope which has a large fully illuminated image circle. Vignetting is especially problematic when an APS-C format sensor like the IMX071 is used or a Full-Frame sensor like the IMX455. We will not discuss the Full-Frame format sensors and cameras here as they are much more suited to astrophotography even though certainly can be used for EAA. The values of the Read Noise given in the table are the minimums as Read Noise varies with the Gain used. Quantum Efficiency (QE) is not available for all sensors for some reason.
Cameras are available for most of the sensors shown with TEC cooling. Also, all but a very few of the cameras come with a USB3.0 connection to the camera which provides faster download speeds compared to USB2.0. Camera bodies are all cyclindrical with the exception of the Atik Infinity. TEC cooling requires a much larger camera body length which will not clear the base of an Alt-Az mount when pointing near the zenith unless it is one of the side mount Alt-Az telescopes.
When choosing among cameras with the same sensor but from different manufacturers here are some additional things to look for:
1. DDR Memory: Built in DDR memory comes in either 128MB or 256MB. Note that 1 MB is the same as 8Gb as some express the memory in Gb instead of MB. It takes 8 bits (b) to make 1 byte (B). This added memory prevents frames from being dropped when transferring the data from the camera to the computer due to the high pixel count of many cameras. It can have the added benefit of reducing amp-glow as discussed below. ZWO cameras typically uses 256MB as does Mallincam while QHY uses 128MB except in their highest end cameras. Rising Tech generally does not indicate whether or not any of their cameras have memory.
2. USB Hub: Typically, cameras with cooling come with a 2 port USB2.0 Hub which can be used to connect two other devices such as a filter wheel, focuser or guide camera. This simplifies cabling and reduces the number of cables which must hang from the telescope to a computer or a USB Hub below. ZWO, Rising Tech and Mallincam provide 2 port USB2.0 hubs on their cooled cameras but QHY does not.
3. Anti-Amp Glow: In addition to DDR memory, some cameras have additional methods to combat amp-glow which may include software and additional hardware tricks which they do not explain in any detail. QHY indicates Anti-Amp-Glow features on some of their cameras while Rising Tech and Mallincam indicate it on almost all of their cameras. ZWO claims Anti-Amp-Glow for their ASI224MC and on their newest full frame camera the ASI240MC Pro.
4. Anti-Dew: Cooled cameras will often cause water condensation on the sensor window which may even ice up completely bringing a viewing session to a grinding halt until the camera is warmed. All cameras come with a sealed window using a gasket to keep water vapor outside the sensor. QHY and Mallincam add a thin heater on the chamber window which keeps water from condensing on the window. QHY cameras and the Atik Horizon II have a removable and re-chargeable silica gel pack to absorb any water vapor inside the sensor chamber which must be removed and baked from time to time to remain absorbent. Mallincam vaccum seals the chamber on some of its cameras. Rising Tech does not indicate the presence of a heater. ZWO seems to only have the heater on its latest designs like the ASI240MC Pro, but does does offer an after market heater kit which can be easily attached to the camera window although the power connection is not integrated into the camera itself.
5. Fan: A few of the lower cost Rising Tech cameras use a fan instead of a TEC for cooling which will not produce the same low temperature noise reduction as a TEC. Some cameras with TEC also have a fan to assist with heat dissipation.
6. Accessories: Additional accessories vary greatly. ZWO includes USB cables for both the camera and the Hub (if present), as many as a half-dozen adapters/spacers which are helpful in setting the correct camera spacing. They also supply a camera cover and a soft carrying bag. Rising Tech includes a 1.25" adatper, one 2m USB3.0 cable, and a 2m guider cable with its non cooled cameras. With its cooled cameras Rising Tech supplies 1.25 and 2" adapters, a 1.5m USB3.0 cable, a power supply for the cooler and a hard plastic carrying case. Similarly Mallincam provides a 2" adapter, one 15ft USB3.0 cable, and a guider cable with its non-cooled cameras and adds a power supply for the cooler and a hard plastic carrying case for its cooled cameras. QHY supplies their cameras with a 2" adapter, 1.5m USB3.0 cable, power and guider cables, a car power adapter, a desiccant tube and desiccant.
7. Software: Camera control and live stacking software has become very common for EAA since Robin Glover introduced SharpCap in 2010. SharpCap works natively with both ZWO , QHY and Starlight Xpress cameras and will work with other cameras using an ASCOM driver. The free version of SharpCap has camera control and live stacking capability. ZWO also has their own proprietary software called AstroLive which only works with their cameras. Rising Tech cameras come with Rising Sky software and Mallincam cameras with Mallincamsky. Atik cameras have their proprietary softwaren called Infinity which is considered one of the easiest and most intuitive software applications to learn because it does not have all of extra features not absolutely needed for EAA as does SharpCap. Starlight Xpress's software is called Starlight Live. While all of these do live stacking and on the fly processing, SharpCap is arguably the most complete package, especially if one uses the subscription version which includes plate solving, polar alignment and many other useful functions which are not necessary for EAA but make the process of setting up, aligning, focusing, etc. much easier. But it can have a steep learning curve.
Certainly camera manufacturers try to distinguish themselves from their competitors so there are other differences among them including the use of AR coatings on chamber windows, DDR3 vs DDR2 memory, global vs rolling shutters, etc. Check competing manufacturer's sites for these additional details.
We will only consider cameras under $1500 even though there are a multitude of cameras beyond that price range because cameras need not cost so much to bee fully sufficient to meet every EAA need. More expensive cameras can be considered if one anticipates using a single camera for EAA and astrophotography. The table above lists the wide assortment of cameras within this price range from ZWO, QHY, RisingCam, Mallincam, Atik and Starlight Xpress (SX). For those in Europe, Altair Astro in the UK has a wide assortment of the same cameras. It is widely understood that cameras from RisingTech, Altair Astro and Mallincam originate from the camera manufacturer ToupTek in China. (Note that cameras from RisingCam are commonly referred on the astronomy forums as RisingTech as I think this was their original name). These re-branded cameras are able to select different options from the original equipment manufacturer and may even provide different specifications to distinguish their cameras from their competitors. For instance, Mallincam specifies ZWO, QHY, Atk and SX all manufacture their own cameras but, as mentioned above, choose from the same supply of sensors as everyone else.
Cameras using the Sony IMX224MC color sensor were the first CMOS cameras used for EAA. Since the sensor only has a 6mm diagonal and 1.2MP, cameras with the IMX224 are now more often used as planetary cameras but can still be a great entry level EAA camera due to their low price. Because of the very small diagonal of this sensor the FOV will be very narrow for any telescope at a focal length of 1000mm or longer. Therefore it is common to use a focal reducer to get closer to 500mm focal length which will provide a FOV close to 0.5deg. Otherwise, it can be difficult to find DSOs unless you are using a plate solving software. Cameras with the IMX224 sensor can be obtained from Rising Tech on AliExpress for $165 or from ZWO from most astronomy retailers for $249. Both are uncooled cameras with a small form factor body which will not have a problem working at the zenith in an Alt-Az mount like the Celestron Nexstar. QHY offers a cooled version with anti-amp glow and internal memory for $669 but it will not work near the zenith on mounts like the Nexstar because of its large body.
Soon after the IMX224 cameras were introduced, the Panasonic MN34230 sensor with 16MP became available in cameras from a number of different manufacturers. These immediately became popular for EAA due to their larger sensors and reasonable prices. Cameras with the Panasonic sensor come in cooled, uncooled, color and mono versions. Cameras with mono sensors are always significantly more expensive than their color cousins. Rising Tech offers an uncooled color camera with the Panasonic IMX1600 chip for $648 and a cooled version for $997 while QHY offers its cooled color version for $899. ZWO discontinued their color version of this camera a few years back and now only offers the mono cameras in uncooled and cooled versions for $999 and $1280, respectively while Risking Tech's uncooled mono camera sells for $990. Mallincam's cooled color camera is available for $999.
More recently, the Sony IMX294 sensor with 11.3MP started showing up in cameras from ZWO, QHY, Rising Tech and Mallincam. With a similar sensor size as the Panasonic MN34230, the Sony IMX294 provides a much deeper well depth, slightly larger pixels for greater sensitivity and 14 bits instead of 12 bits ADC. ZWO offers its ASI294MC uncooled color camera for $699, or its ASI294MC Pro cooled color camera for $999. QHY has a color cooled version, QHY294c, for $999. Rising Tech has an uncooled color version for $725 and a cooled color version for $960. Mallincam's DSC10 is an uncooled version for $949.
1" format sensors with ~16mm diagonals can be found in the 20MP IMX183 and the 9MP IMX533 Sony sensors. The least expensive camera in the 1" format is the uncooled ZWO ASI183 for $549. The QHY183c and the ZWO ASI183MC Pro cameras are both cooled color cameras and retail for $699 and $799, respectively. Cooled mono versions are $999 from both QHY and ZWO. The IMX533 sensor can be found in the cooled ZWO ASI533c and Rising Tech 533c for $899 and $900, respectively.
Just under $1000, Atik offers their Infinity camera with its signature rectangular shape and Sony ICX825 CCD with 1.2Mp on an 11.2mm diagonal. When first introduced about 5 years ago the Infinity was well received because of its performance and its Infinity live stacking software which is regarded as one of the easiest to master. But at $975 I believe it is now over priced given the fact that one can obtain a much larger sensor with cooling and internal memory for about the same price. If you have your heart set on one of these, look for a used camera for ~$450 on the Cloudy Nights classifieds. I have the same thought on the SX Ultrastar also with the ICX825 which sells for $1050.
The camera table above lists 9 cameras above $1000, 4 of which are with mono sensors which always sell for much more than their color cousins. Unless you want to do mono EAA for the higher sensitivity, or want to use narrow band Ha or OIII filters to bring out specific details in nebulae, these are overkill for EAA. The QHY and ZWO cameras with the Sony IMX071 sensor are listed here because these have an APS-C format sensor with the largest diagonal at 28.4mm of all the cameras listed here. This would be a good camera if one wants to do both EAA and traditional astrophotography, but with such a large sensor expect to deal with significant vignetting.
Clearly there are many possibilities when it comes to cameras for EAA. With so many choices it can be challenging to decide which camera is best suited to ones' needs. A good approach may be the following. First decide your price range. When doing so, consider waiting to to buy something with a much larger sensor than the IMX224 like the IMX183 cameras from ZWO or QHY rather than rushing into one of the low end cameras with the idea of buying a better camera later. While the IMX224 cameras are a great choice for someone on a tight budget, cameras like those with the IMX183 or even the IMX294 are a better choice for the long run if you can eventually work it into your budget by waiting. Or, pick up a higher end camera on the used market to move up in features right away. Cameras are pretty robust so there is not a great deal of risk to buy used. In that case it is a good idea to ask for a recent 60sec dark frame to make sure that the camera does not have too many hot pixels.
Cooling is nice to have and essential for astrophotography, but for the short exposures used in EAA exposures cooling is not absolutely necessary. This will keep the cost down. If you are planning to use an Alt-Az mount with the telescope centered on the mount a cooled camera will make it impossible to view objects around 80 to 90 degrees in altitude.
A sensor in the 15mm to 23mm range gives a good FOV for large objects without having to be too aggressive with focal reducers. Color cameras add a satisfying dimension to real time viewing which mono cameras cannot. When you do purchase a camera make sure that you have the necessary spacers to place the sensor at the correct distance from the back of the telescope to get an image without artifacts. Screw on spacers are better than slide on spacers since they will have less flex, but both will work.
You can look through more detailed summaries of most of the cameras discussed here on the Oceanside Photo & Telescope (OPT) web site.
OPT links are Affiliate links.