Lorenzo Comolli's Astronomy HomePage

Introduction

In the middle of the last decade, and thanks to the success of DSLR in astronomy, Canon suddenly surprised the astrophotographers with a specialized camera. The Canon EOS 20Da was introduced in February 2005, and this was the first time for Canon and all other manufacturers to produce such a version of a standard camera. The 20Da was mainly based on 20D, but with substantial modifications, both hardware and software, and immediately was identified as the reflex for the astronomical usage, but also traditional diurnal usage was still possible. The main hardware modification was the optical filter, more transparent to the H-alpha line in the red (such a modification was made previously by many astrophotographers using filters like the Baader or Hutech). Also the software modification was notable, introducing for the first time in the world the Live-View feature, a very useful function that afterwards will be adopted by all the DSLRs. 20Da was a cheaper alternative to the more expensive CCD cameras with the same format. The success of this camera was good worldwide (not so much in Italy), but unfortunately it was discontinued less than an year after the introduction. Now, after 7 years, Canon try again to get in the astrophotography world, introducing the new EOS 60Da on 2 April 2012.

This camera was born using the experience gained with the 20Da, and considering the needs of the users. Also this time the camera is based on a prosumer model (2 digits), named 60D, that was introduced on the market in 2010. Curiosity was hight and we've decided to test in detail this reflex, one of the first arrived in Italy, with serial number 0281200753.

The package

The package include only the body of 60Da, together with many accessories. No bundles are available with lenses. The package is well distinguishable thanks to the Orion nebula shot printed as a background. Here is the included parts (Fig. 1).

All the main features and characteristics of the main model 60D are retained in this version, including the possibility to make HD videos. But here we'll focus on the astronomical features, leaving out nearly completely all the other "terrestrial" features. The main difference between 60D and 60Da is the optical filter suited to cut the infrared light, and positioned in front of the CMOS sensor. This filter, in the standard photography, has the purpose to render the spectral response similar to the naked eye. But in astrophotography this is not good because the standard filter will cut the H-alpha light nearly completely. We recall that the H-alpha light is emitted by the ionized Hydrogen, and has a deep red color, with a 656.3 nm wavelength. In this camera a special filter was designed, with fluorine coating, so that the H-alpha light is transmitted nearly three times more respect to the standard filter. Canon declares that this filter allows the camera to be used both in "terrestrial" and "astronomical" fields. In this test we've tried to verify this declaration.

But which are the main differences between 60Da and 60D? Obviously the infrared filter, more transparent to the H-alpha light. But about the software we've not noticed any difference, in contrast with 20Da where the Live-View was a great addition. Nowadays this feature is standard on all cameras. Unfortunately a modification of the feature, with longer exposure times, was not implemented. So, both camera are nearly identical. A comparison between 20Da and 60Da in in Tab. 1, or in more detail on the DPreview website (with 60Da-60D-20Da-20D): http://www.dpreview.com/products/compare/side-by-side?products=canon_eos60da&products=canon_eos60d&products=canon_eos20da&products=canon_eos20d&sortDir=ascending

To find other differences we need to look at the accessories. The ACK-E6 AC power adapter is useful to power the camera for long periods and in the cold, where the standard battery would last only 2-3 hours or less, much less that a complete astronomical night. The power supply need an AC input, and this is not so convenient for traveling amateurs, using mainly DC power from lead acid 12V batteries. The need for an inverter is quite unconfortable, needing more cables and increasing power consumption. A 12 V power adapter would be highly appreciated; nevertheless this can be homebuilt or purchases from third party sources.

Another provided accessory is the RA-E3 adapter cable, needed for using the TC-80N3 timer remote (with a N3 connector, standard on older xxD cameras) with the 60Da (adopting the simpler 2.5 mm mini-jack connector, such as in xxxD models). Unfortunately the timer remote is not included in the kit, even it's a fundamental accessory. It can be purchased separately, but the price is high (140€ in Italy), or a Chinese alternative can be found on Ebay at about 10€, just search "timer remote 60D" (these ones mount the proper connector, so the adapter cable provided by Canon is no more useful).

Our test where conducted during June and July 2012, in a period not so favorable to imaging due both to the temperature and to the short nights. Nevertheless we've conducted many tests both in the field and in lab. The tests on the sky was conducted both from light polluted locations and in the pristine sky of Colle del Nivolet, one of the best places in the Italian Alps. We've worked from our homes in the strongly light polluted Po Valley north of Milan, Italy, at an elevation of about 300 m. Fortunately we've got some dark and clear nights, so we've aimed to some faint object.

We've used two flat field apochromatic refractors, a TMB 80/500 and a TEC 140 (with field flattener), obtaining an imaging focal length of 500 mm and 1000 mm respectively. The first focal is great for the small pixels of 60Da, of 4.3 micrometers, and allowing to frame a while field of 2.5°x1.7°. The second focal length is allowing to frame smaller subjects, with a good sampling (0.88"/pix, with an Airy disk of 2" and a theoretical FWHM of 0.84"). To get a direct comparison, we've also tested together other Canon cameras, a standard 350D and a 40D, and more a 350D modified with a Baader ACF filter. This way we've compared the behavior both in "terrestrial" and "astronomical" scenarios. Even if we disposed of a 5D Baader modified, we've not compared because of the sensor size difference (full-frame vs. APS-C), making the comparison ambiguous.

And here are at the long awaited test, on the night sky. We've operated via PC using the EOS Utility software, saving the files on hard disk. This software allows to set all the parameters remotely, and also to take shots longer than 30 s, always using only the USB cable; so a RS-232 serial bulb cable is no more needed. The Live-View is possible also on the PC display, so the focusing is highly facilitated (Fig. 2), even if focusing on the 3" adjustable display is too very easy. Even if we've used the PC for many reasons, it can be avoided, thus the setup is much simplified.

In the short time of our tests, we've not tried to connect the camera to the most used astronomy software, such as Maxim DL or Astroart. They are useful to operate both the camera and the guider and to make dithering (we've done it manually, but that's not easy). Looking on the web, we've found that some US amateurs were able to connect the 60Da in Maxim DL; this is not surprising because the 60D (not-a) model is supported and the firmware seems to be the same in 60D and 60Da.

The supplied power supply was tested and we've found a small power consumption (Tab. 2). This is confirmed also by the long lasting of the standard battery: after using it for 1.5 hours at +18°C, for focusing and taking sequences of 5 minutes exposures, the 5 levels power indicator don't moved even one step! The weight of 60Da is contained (less than 700 g), similar to many beginners CCD, so that it can be mounted on any good telescope, without any problem of flexure or sliding of focuser.

As first object we've chosen the Butterfly Nebula (IC 1318), near Gamma Cyg, and we've used the TMB 80/500 refractor. We've imaged from Caronno Pertusella (VA, Italy), a heavily light polluted city. This region is rich in H-alpha emission and is a great benchmark for testing. After the first shot, only 3 s were needed to download the 27 MB raw file on the PC, and the Butterfly appeared immediately great. It was a single shot of 10 minutes at 800 ISO, without calibration and processing, a RAW file but showing great details. Then other shots were collected, together with darks, bias, flat and darkflat, so that a final image can be obtained in the better way. The same procedure was adopted, in the same night, with a 350D Baader modified, to get a comparison with a very common astronomical camera. The result is in Fig. 3.

The comparison shows the much better resolution of 60Da, thanks to the small pixels of 4.3 micrometers very well suited to exploit the resolving power of the telescope. The 80/500 has a resolving power of 1.5" and the sampling with 60Da is 1.76"/pix respect to 2.64"/pix of 350D. This resolution is surely an advantage in aestetical terms, but it's also an additional difficulty during imaging, because guiding needs to be better. The 60Da has a 14 bit A/D converter (compared to 12 bit of 350D) and so the dynamic range is wider. This translates into less noisy images, with more details.

Looking at the H-alpha behavior, the modified 350D is better respect to 60Da. Even if with the 60Da the nebula is well recorded, the 350D image shows much fainter parts, so that the black sky is nearly absent here, filled with faint nebulas. Looking carefully at full resolution, we've found that the 350D image shows fainter stars respect to 60Da, maybe thanks to larger pixels. This fact must not to be underestimated, and is a problem common to all the new sensors with small pixels.

The second test on the night sky was conducted in Tradate (VA, Italy), another heavily light polluted city. The test subject was the Veil nebula, near the star 52 Cyg (NGC 6960), gathered with a TEC 140 apo refractor and field corrector. This nebula has both red and green-blue regions. This time we've added another camera for comparison, a standard 40D, to get a reference with a non modified camera of the same prosumer level. The results are in Fig. 4, and they confirm the other test. The 60Da is surely much better respect to an unmodified 40D, that records nearly nothing of the H-alpha emission. n the other hand the 350D Baader modified is recording much more red light respect to 60Da.

At the end, a final test was conducted from the dark sky of Colle del Nivolet (AO, Italy), one of the best locations in the Alps, at 2530 m heigth. The adopted telescope is again the TEC 140 apo refractor, and the night of 21 July 2012 was moonless. The mount was a Gemini G41, with a 80/400 guidescope with Lodestar. The chosen subject was NGC 7380, an emission nebula in Cepheus, with mainly red light, but also with some blue hints. We've taken 27 exposures of 10 minutes each (total 4.5 hours), from the start to the end of the astronomical night. The dithering was automated with an uncommon method, because we've not connected the camera to PC.
The usage of the 60Da was found really comfortable: focusing a really easy task thanks to the Live-View at 10x on the large adjustable display. A few seconds were needed and this let you immediately forget all those complicated procedures at the PC, with FWHM and other methods. Moreover checking each shot was very easy, enlarging the image on the 1 megapixel display adjusted in a very comfortable position. After looking, the display can be directed toward the ground so that no disturbance is given to other photographers or observers in the field.

The result is in Fig. 5, and no comment should be added. Anyway, the nebula is greatly recorded and this testify the great potential of the 60Da for astroimaging. This time no comparison was made with other camera because we don't wanted to waste the precious time in this dark location. We've preferred to bet all on the 60Da and look at the result of 4.5 hours of total exposure.

The images shot at Nivolet were obtained with a quite low ambient temperature, 0°C. This temperature is good to get low noise, but to reduce even more we've added a fan applied on the back of the camera, so that the air flow cools even more the sensor. Even we've not measured the efficiency of this solution on 60Da, previous tests showed that the noise can nearly be halved. On the 60Da, the factor can be even higher than 2 because of the adjustable display, that can be moved on the outside and so a much smaller thickness is present between external air and the sensor. In the Fig. 6 the 60Da is shown mounted on the TEC 140, waiting for the night.

At the end of the astronomical test, we'd like to stress that to manage and process the huge 18 Mpix files of the 60Da a very powerful PC is necessary. We've used a i7 processor with 12 GB of RAM, SSD drive and Windows 7 64 bit operating system.

We would had liked to measure the quantum efficiency of 60Da and maybe also compare to a standard 60D and a Baader modified 60D. Unfortunately we don't had available those references and also the instruments useful to measure the quantum efficiency, or the transmission in H-alpha of the standard filter.

Nevertheless, from the performed test, we can without any doubt draw a classification of sensitivity to H-alpha light, between "standard", "Da" and "Baader modified" cameras. The improvement between each of these steps seems equal, and so, just to give some indicative numbers, we can say a 30-70-100% H-alpha transmission can be plausible. In other words, the "Da" is much better than "D", but is not at the level of a "Baader modification". The ratios seems to be similar to the older 20Da, even if Canon declared that 60Da has a transmission in H-alpha 1.5 times better respect to 20Da. However, the choiche of Canon not to declare the absolute transmission values is very regrettable, they would be much simpler to interpret respect to the imprecise and qualitative declarations.

Digital images are simply an array of numbers, and so interpreting these numbers is fundamental to enter into the intimate of photography, to better understand how a camera works. We've tried to interpret the data collected in laboratory, to get some more information respect to the ones that are declared.
Canon says the noise of 60Da is small: ok, we've seen, but how much? We've taken a series of 10 minute darks at 800 ISO, at different temperature, from +28°C down to -14°C. The graph at left of Fig. 7 (and the Tab. 3) shows the thermal noise as a function of the external temperature. The noise decreases with decreasing temperature, but the values are not in a straight line as expected from theory.

The Fig. 8 shows the central area of the dark exposures. At +8°C the noise is very small. The test showed the presence of many hot pixels, but this is normal for a sensor with so many small pixels. Fortunately there are no areas subject to the electroluminescence problem, so that the dark is very homogeneous, without brighter areas like in the 350D darks (comparison in Fig. 9).

From the analysis of Tab. 3, showing the bias values at 800 ISO at ambent temperature, a better read noise is noticed in 60Da respect to 350D, of about 30% (considering the bit depth difference).

Notice about gain: the short time we've granted from Canon Italia was not sufficient to get all the tests and information we'd liked, such as the delicate tests needed to measure the gain and other low level measurements. For the same reason, we've not tested the video recording functions, promising for the planetary imager, thanks to the full resolution mode of only the central part of the sensor (640x480 pix).

As we've seen, the night time tests confirms the value of 60Da for astronomical use. But a unique characteristic of 60Da is the possibility to be used also in the day time, for standard photography. Is it true that it can be used without problems? We've analyzed the behavior in the traditional usage, first imaging a panorama of countryside (where nearly all colors are present, from the brown of the ground to the green of plants, to the blue and white of the sky and clouds), and then for photographic hunt in the mountains, in the National Park of Gran Paradiso, at Colle del Nivolet.

Fig. 10 shows the results of the first test, as a comparison of 60Da with two 350D, one Baader modified and the other standard. Te result is quite interesting, showing that 60Da is behaving well also in this situation. The spectral response seems not to be exactly the same of a standard camera, but nevertheless is very similar, and -analyzed one by one- they do not show any problem or prominent color unbalance. Respect to the standard 350D, the 60Da get images only a little cooler, with the reds a little less saturated, and a small drift toward cyan. Maybe the internal settings tends to counterbalance the higher red sensitivity a little too much.

Also interesting is the comparison to the Baader modified 350D, where setting in both the Personal WB, the images are nearly identical! Both shows tones a little warmer respect reality, but whites remains white. So great images can be obtained with both.

About the second test in the mountains, one of us has used the 60Da together with a MTO 500 f/6.3 catadioptric tele-lens, taking some close shots expecially on the marmots (Fig. 11). In this field the 60Da showed to be very effective (such as his sister 60D). The Live-View is really useful, thanks also to the adjustable display, and the great optical viewfinder, really bright, so good that also manual focusing was possible with it (in comparison in the 350D viewfinder focusing is much more difficult, nearly impossible). Nothing to declare about the color behavior, being nearly perfect.

Another point is the grip, a little more uncomfortable respect to previous xxD versions. This new design seems more plastic-made and small, so that it resembles more to an entry level camera.

We'd like to stress that for all the tests we've adopted manual settings, excluding all autocatic corrections, acquiring in RAW format, so that the most raw images can be obtained.

After looking at the above results, we can ask if buying a 60Da is worthwhile, or maybe is better ot buy a standard camera and modify with a Baader filter (or Hutech). Let we compare advantages and disadvantages of these alternatives in the Tab. 4.

As shown, the total price is not so different, even if a 200€ advantage for the modified version is present (even increasable if the modification is made by a cheaper service, or homemade).

So, who will benefit of a 60Da? Surely the less "geek" amateur, that prefer a simple and safe solution, usable easily also during the day for traditional photography.

And, who will benefit of a Baader modified 60Da? To the amateur that do not accept any compromise, pretending the best from his images, and so accepting a little more complication during the buying and modification phase, and also for the daytime usage (or even better if no daytime usage is intended)

Defining the 60Da "The camera for astrophotography" is, in our opinion, a little too much, because even nowadays a modified camera will give better results on H-alpha subjects. Even if we recognize all the difficulties and limitations of the modified camera, if an astronomical use is needed, they are the best solution. We prefer to define the 60Da as "The factotum camera", because both "terrestrial" and "astronomical" imaging is possible with good results.

Another time Canon decided to produce an object well suited for many purposes, without unbalancing and so not precluding any users. This way, however, Canon needed to make a technical compromise and so we consider the 60Da in between a standard and modified camera, but with the advantage of a great flexibility, farther of the Canon warranty and support. We think that a 60Da is a great starting point for the novice astrophotographer, at a price that is a good balance of advantages and disadvantages

In the end we'd like to acknowledge Canon Italia for borrowing us the tested 60Da, and we make a call for the future, a dream for the astrophotographers: a truly astronomical camera, with a full-frame with 10-12 megapixel (more are not needed), with 16 bit dynamic range, without IR-cut filter, with an internal timer, and moreover with a black and white CMOS sensor!

	Canon EOS 20Da	Canon EOS 60Da
Introduction	1 June 2005	2 April 2012
Sensor	CMOS 22,5 x 15,0 mm	CMOS 22,3 x 14,9 mm
Total pixel	Nearly 8,2 megapixel	Nearly 18 megapixel
Dimension of pixel	6,4 micrometers	4,3 micrometers
Aspect ratio	3:2	3:2
RAW file size in pixel	3504x2336	5184x3456
Infrared filter	Integrated/fixed with a modified IR-cut (transmision H-alfa 2,5 times more than the standard filter)	Integrated/fixed with a modified IR-cut with fluorine cover (transmision H-alfa 3 times more than the standard filter)
Dust removal system	no	Ultrasound, integrated
Processor	Digic 2	Digic 4
A/D converter	12 (4096 levels)	14 (16384 levels)
Available exposures	30 s - 1/8000 s	30 s - 1/8000 s
White balance	Automatic, Daylight, Shadows, Clouded, Tungsten, Fluorescent, Flash, Custom, Color temperature	Automatic, Daylight, Shadows, Clouded, Tungsten, Fluorescent, Flash, Custom, Color temperature
ISO sensitivity	100-1600 (3200 with boost)	100-6400 (12800 with boost)
Live View	Yes, with 2X or 10X (first DSLR to implement this feature)	Yes, at 30 fps with 5X or 10X; can add a reference reticule
Display	TFT 1,8" (4,6 cm) 118000 pixel	TFT Clear View 3" (7,7 cm) with adjustable position, 1,04 megapixel, 3:2 aspect ratio
Focusing screen	Fixed	Interchangeable (3 optional types)
Lens adapter	Canon EF / EF-S	Canon EF / EF-S
Memory type	CF I/II and Microdrive	SD, SDHC and SDXC
Body material	Magnesium alloy and resin	Aluminium and resin (glass fiber reinforced)
Mass	685 g	642 g 755 g (including battery and SD card)
Connection to PC	USB 2.0 (Mini-B)	USB 2.0 (Mini-B)
Battery duration	700 shots at 20°C / 550 shots at 0°C ; battery code BP-511	1100 shots at 23°C / 1000 shots at 0°C ; battery code LP-E6
Included power supply	AC power supply ACK-E2	AC power supply AC-E6 (output 8 V 3 A DC) with DR-E6 false battery (kit ACK-E6)
Software	EOS Capture, EOS Viewer Utility, Digital Photo Professional	EOS Utility, ZoomBrowser EX, Digital Photo Professional 3.11.31
Supported operating systems	Win 98 (including SE), Win 2000, Win XP	PC and Mac - Win XP (SP2/SP3), Vista (SP1, excl. Starter Edition), Win 7 (excl. Starter Edition), OS X (v10.6-10.7)

Power on, doing nothing	40 mA with peaks to 90 mA
LCD on, looking at menu	140 mA with peaks to 160 mA
LCD on, looking at images	140 mA with peaks to 180 mA
While shooting a long exposure	200 mA (representative of night time use)
HD video recording	210 mA
Downloading images to PC	140 mA

Compariso of Bias	60 Da	350D Baader
Noise of single bias (pattern noise subtracted, @800 ISO, +22°C) [ADU]	6.42	2.30 (9.19 @14 bit equiv.)
Noise of the median of 9 bias (@800 ISO, +22°C) [ADU]	2.80	1.02 (4.08 @14 bit equiv.)
Temperature	Noise in 60Da 10 min dark @800 ISO [ADU]	Noise of the difference of two 60Da 10 min darks @800 ISO [ADU]
28°C	177.3	62.0
22°C	103.0	43.7
8°C	37.3	22.9
-14°C	17.9	12.5

COMPARISON	60Da	60D modificata Baader
Easiness in buying	The camera is ready “out of the box”	A good technician has to be found to get the camera modified; a Baader filter must be purchased; both has to be shipped to the technician, hoping that no problems will occur (very rare, but possible)
Warranty	2 years (in Italy)	The modification invalidate the original Canon warranty; the modification service can give you a warranty, but the repair can be quite slow
H-alpha performance	Surely much better respect to a standard DSLR, but not at the level of a specialized filter (like Baader or Hutech)	Excellent, maximum possible transmission
Diurnal usage	Perfectly possible also with the auto white balance	Possible, but with the need to use a personal white balance, to be redone each time the ambient light changes (Sun, shadows, lights in a building, flash,...)
Included accessories	Included are the AC power adapter ACK-E6 and the remote cable adapter RA-E3; unfortunately the AC adapter is uncomfortable to be used on the field, and a 12 V DC adapter should be evaluated	Nothing, but a 12 V DC power supply can be homemade with few euros and some soldering; these adapters are also available on the market.
Accessories to be added	Timer remote TC-80N3 (price ~140€) or as an alternative the Chinese equivalent (price ~10 €)	Same as 60Da
Prices (as of end of July 2012, for the Italian market)	60Da+sh: 1219€ (Amazon.it) Chinese timer remote: 10€ (Ebay.it) 12 V DC adapter: 55€ (AstroHobby.it) TOTAL: 1284€	60D+sh: 836€ (Amazon.it) Modification+sh: 120+10€ (AstroHobby.it) Filtro Baader+sh: 65€ (Unitronitalia.it) Chinese timer remote: 10€ (Ebay.it) 12 V DC adapter: 55€ (AstroHobby.it) TOTAL: 1096€