M33 - Triangulum Galaxy
M33 - Triangulum Galaxy. Click here for full resolution image.
M33, also known as the Triangulum Galaxy is a spiral galaxy located in the constellation Triangulum, about 2,73 million lightyears from Earth. It is a member of the Local Group of galaxies, to which also our own Milky Way and the Andromeda Galaxy belong. At only 60% of the size of the Milky Way, it is the third largest galaxy of the Local Group.
The relative close proximity makes it a rewarding target to photograph, revealing a lot of detail with modest magnification. There are many areas of active star formation, indicated by the purple/pink spots in the image.
M33 is sometimes referred to as Pinwheel Galaxy. However, this is not correct. That name is reserved for M101, a somewhat similar looking galaxy.
Planning
Object
Visual Magnitude: 5.79
Apparent size: 62 x 37 arcmin
R.A.: 1h 34m 59.97s
Dec: +30º 45’ 50.8”
Conditions
Astr. night: 18:37
Astr. dawn: 05:58
Moon: 58.8%
Moon set: 00:30
Humidity: 85-95%
Pressure: 1032-1036 hPa
M33 is well visible throughout most of the wintertime. After March it gets more difficult to observe. Imaging sessions were held on two consecutive days, 16 and 17 November 2018. The equipment could stay outside between the days, reducing the amount of calibration to be done on the second day. A half moon was out at the beginning of the night. At a roughly 80 degree angle from M33, and fairly low on the horizon setting around midnight it did not give too much problems. On the first evening, humidity was very high (95%) and there were quite some problems with dew on the lens. Wiping the lens clean and with dew-heater on full power, the lens could eventually be kept clean. On the second night, humidity had lowered to 85% and no problems with dew at all.
Capturing
The image is captured with 180s exposures at gain 0 using LRGB broadband filters. In addition, 300s exposures at gain 300 using the H-alpha (5nm) is added to highlight the regions of intense star formation.
With the TOA-130 and ASI1600 combination, the galaxy fills the complete frame diagonally. The diagonal orientation was achieved by rotating the camera to -108 degrees. Given the size of the object, proper rotation was important, as there would not have been enough field of view to play with and rotate in post-processing.
A total of 50% more RGB than L data was collected. In more recent pictures, the tendency is to make the total luminance exposure somewhat similar to the total colour (combined RGB) exposure, putting a little bit more emphasis on sharpness with less noise rather than colour.
Technical details
Telescope
Mount
Camera
Sensor Temp.
Takahashi TOA-130 + 35 flattener
10Micron GM1000HPS
ZWO ASI1600MM Pro
-30 ºC
Exposures
Luminance
Red
Green
Blue
Ha (5nm)
Ha(5nm)
Total Exposure
72 x 180s @ Gain 0/10
36 x 180s @ Gain 0/10
36 x 180s @ Gain 0/10
36 x 180s @ Gain 0/10
10 x 300s @ Gain 139/21
7 x 300s @ Gain 300/50
10.4h
Processing
All frames were calibrated with Bias (100), Dark (50) and Flat (25) frames, registered and stacked using the BatchPreprocessing script. The H-alpha images were captured using different gain settings between the two days. In the stack they were all put together, keeping normalisation on and weights based on noise evaluation.
In each of the five channels the noise was reduced using MureDenoise. This is a very powerful noise-reduction tool, but finding the right parameters is quite tricky. First there are the three detector settings that need to be determined, which was quite a complex process at the time. These days, there is an extra script added to PI, called MureDenoiseDetectorSettings. Feeding this script with two flat frames and two bias or dark frames returns the three detector settings. This is a big step forward. But still finding the proper variance scale is difficult. One way is to make several previews and treat them all with a different variance scale setting. Blinking through the previews can then help selecting the right value.
The Luminance was subjected to regular deconvolution and stretched. The R, G and B channels were combined. The RGB image came out with a diagonal colourcast, magenta top-right to green bottom-left. This was well corrected by DynamicBackgroundExtraction. The colour-balance was quite good already, but still a photometric colour correction was applied. Especially for galaxies, the photometric correction usually gives good results. After stretching the RGB image, it was combined with the Luminance, which gave a significant brightness boost to the image, perhaps even a bit too much. In an attempt to get maximum detail out of the image, tools were applied like LocalHistogramEqualization, DarkStructureEnhance and UnsharpMask. Also a bit more noise reduction was applied. All in all it gave a bright image, with plenty of detail.
The H-alpha channel was prepared using a fairly standard combination of background extraction, deconvolution, stretching and noise reduction. The resulting image had a higher signal with a mean brightness of around 8000 ADU, whereas the LRGB image had around 7000 ADU. This was corrected with a linear fit. Since the image was already stretched, a histogram transformation adjustment would have done the trick here as well.
The H-alpha channel was then added to the LRGB image, really bringing out the purple star-vorming regions like little knots of wool. A few small further final touches to increase contrast and saturation completed the process in PixInsight.
Just as a last touch, a little bit more vibrance was dialed into the image in Affinity Photo
This image is published on Astrobin
