and after weeks of hard labour under poor skies :-)
Posted: Wed Mar 12, 2025 12:43 pm
Large aperture telescopes are harder work (in every sense including physical exercise) but what are they actually good for in astroimaging? Well obviously in theory they can deliver higher resolution because of the larger aperture size D. Indeed they can deliver images of planets and the moon close to their theoretical resolution via lucky imaging. However - for deep sky imaging and at sizes above 200 mm or so - even under the best conditions (and certainly here in Britain) - resolution is constrained by atmospheric conditions.
Nevertheless - I think that a long focus telescope that is also fast - offers significant advantages for observing smaller deep sky objects as well as planets. Long focus makes the image larger and while not exactly true 'lucky imaging', a lower F number does permit the use of shorter exposures (5-10s) which provides the possibility of selecting from a larger number to build a sharper image. At the same time by using a higher gain and low F number such short frames can be still be bright enough that they combine into an image with sufficient SNR and information content that the image can be sampled sufficiently (0.41 arcsec. pixel in this case) for deconvolution - and particularly Russell Cronan Blur Exterminator software - to provide images with an apparent resolution less than 1.5 arcsec or so and significantly better than the sky limit --of usually ca 2.8 and no better than ca 2.4 arcsec for conventional > 60s exposure images.
So here the targets were three galaxies in Ursa Major -- M109, M108 and NGC2841 at contrasting distances and brightness. All three were near overhead and in a part of the sky unobscured for me and imaged over a series of nights from the 27 Feb until 09 Mar with varying degrees of haziness, frost, cloud etc and eventually a moon waxing up to first quarter.
The set up was a VX12 OO (UK) 300 mm, F4 Newtonian on a CEM70G mount with an 80mm, f = 400 mm guide scope and ASI 120 MM guide camera. For short 10 s subframe mono luminance images, a ZWO UV/IR cut off filter and ASI 294 MM camera with 2.315 u pixels (0.405 arcsec/ pixel) at gain 280 was used and for colour images, the same filter, longer 66 s subframes and an ASI 294 MC camera with 4.63 u pixels (0.81 arcsec/ pixel) at gain 124.
All three galaxies were imaged and the images processed in essentially the same way. For each galaxy a luminance image was produced - after frame preselection using the sharpcap Brightness and FWHM filters and then further selection using PixInsight subframe selector - to end up with about 1200 x 10s ASI 294 MM frames and processed in PixInsight. After gradient correction and background subtraction integrated images were sharpened using Blur Exterminator using the default settings (correction and 0.5X stellar and nonstellar) - the starting images were each of good quality in terms of estimated FWHM (ca 2.2) and Eccentricity (<0.4). The mono images were then each stretched and calibrated to match the extracted luminance channel of the OSC colour image to which they would eventually be added.
The colour OSC images of M108 and NGC2841 comprised 122 and 170 x 66s images respectively - again preselected from many more in Sharpcap and then using PixInsight subframe selector - were CFA drizzle integrated (1X) - background subtracted - colour corrected (using SPCC) and then aligned and registered to their corresponding luminance image (above). The higher resolution mono luminance was then transferred to the colour images using the LRGB tool in Pixinsight and then the images adjusted using Arcsinh, curves, Noise Xt and eventually Affinity. In the case of M109 I used an old colour image comprising several hundred OSC frames collected in previous years.
Overall the approach of separately collecting higher resolution luminance images seemed to work to deliver large images of these small-angular-sized deep sky objects with more detail in than would have been delivered simply by using an OSC camera to image for longer at the lower resolution (i.e. applying Blur Exterminator directly to these OSC images at 0.81 arcsec/ pixel delivered less detail than here). Perhaps unnecessary where skies are better?
M108 is a barred spiral galaxy -- 11.5' x 4.5 ' - (although not much evidence of this from our viewpoint) at a distance now re-estimated following the Spitzer photometric study down to 28 Mly. It is the closest and brightest of the three galaxies and perhaps the best coloured and most interestingly detailed image? At ~ 46 Mly NGC2841 - the tiger's eye - is the next furthest out - and a huge flocculent tightly-wound galaxy like M63 but larger at 150,000 ly across. M109 - 7.6' x 4.7' is the smallest angular sized and thought to be the furthest out of the the three galaxies at 67 +/- 23 ly . For M109 the colour noise is really high and the detail less compelling -- which is consistent with it being fainter and further out - really needs darker skies (rather than B6) and more time.
Nevertheless - I think that a long focus telescope that is also fast - offers significant advantages for observing smaller deep sky objects as well as planets. Long focus makes the image larger and while not exactly true 'lucky imaging', a lower F number does permit the use of shorter exposures (5-10s) which provides the possibility of selecting from a larger number to build a sharper image. At the same time by using a higher gain and low F number such short frames can be still be bright enough that they combine into an image with sufficient SNR and information content that the image can be sampled sufficiently (0.41 arcsec. pixel in this case) for deconvolution - and particularly Russell Cronan Blur Exterminator software - to provide images with an apparent resolution less than 1.5 arcsec or so and significantly better than the sky limit --of usually ca 2.8 and no better than ca 2.4 arcsec for conventional > 60s exposure images.
So here the targets were three galaxies in Ursa Major -- M109, M108 and NGC2841 at contrasting distances and brightness. All three were near overhead and in a part of the sky unobscured for me and imaged over a series of nights from the 27 Feb until 09 Mar with varying degrees of haziness, frost, cloud etc and eventually a moon waxing up to first quarter.
The set up was a VX12 OO (UK) 300 mm, F4 Newtonian on a CEM70G mount with an 80mm, f = 400 mm guide scope and ASI 120 MM guide camera. For short 10 s subframe mono luminance images, a ZWO UV/IR cut off filter and ASI 294 MM camera with 2.315 u pixels (0.405 arcsec/ pixel) at gain 280 was used and for colour images, the same filter, longer 66 s subframes and an ASI 294 MC camera with 4.63 u pixels (0.81 arcsec/ pixel) at gain 124.
All three galaxies were imaged and the images processed in essentially the same way. For each galaxy a luminance image was produced - after frame preselection using the sharpcap Brightness and FWHM filters and then further selection using PixInsight subframe selector - to end up with about 1200 x 10s ASI 294 MM frames and processed in PixInsight. After gradient correction and background subtraction integrated images were sharpened using Blur Exterminator using the default settings (correction and 0.5X stellar and nonstellar) - the starting images were each of good quality in terms of estimated FWHM (ca 2.2) and Eccentricity (<0.4). The mono images were then each stretched and calibrated to match the extracted luminance channel of the OSC colour image to which they would eventually be added.
The colour OSC images of M108 and NGC2841 comprised 122 and 170 x 66s images respectively - again preselected from many more in Sharpcap and then using PixInsight subframe selector - were CFA drizzle integrated (1X) - background subtracted - colour corrected (using SPCC) and then aligned and registered to their corresponding luminance image (above). The higher resolution mono luminance was then transferred to the colour images using the LRGB tool in Pixinsight and then the images adjusted using Arcsinh, curves, Noise Xt and eventually Affinity. In the case of M109 I used an old colour image comprising several hundred OSC frames collected in previous years.
Overall the approach of separately collecting higher resolution luminance images seemed to work to deliver large images of these small-angular-sized deep sky objects with more detail in than would have been delivered simply by using an OSC camera to image for longer at the lower resolution (i.e. applying Blur Exterminator directly to these OSC images at 0.81 arcsec/ pixel delivered less detail than here). Perhaps unnecessary where skies are better?
M108 is a barred spiral galaxy -- 11.5' x 4.5 ' - (although not much evidence of this from our viewpoint) at a distance now re-estimated following the Spitzer photometric study down to 28 Mly. It is the closest and brightest of the three galaxies and perhaps the best coloured and most interestingly detailed image? At ~ 46 Mly NGC2841 - the tiger's eye - is the next furthest out - and a huge flocculent tightly-wound galaxy like M63 but larger at 150,000 ly across. M109 - 7.6' x 4.7' is the smallest angular sized and thought to be the furthest out of the the three galaxies at 67 +/- 23 ly . For M109 the colour noise is really high and the detail less compelling -- which is consistent with it being fainter and further out - really needs darker skies (rather than B6) and more time.