Upvote
0
Deep Sky Astrophotography
- Thread starter jrista
- Start date
Click said:
This is not mine click, I wish it was ;D, this was taken by the Very Large Telescope in chili, and remastered by an Astro photographer, you can see the details in the link below the image
Upvote
0
How much of a difference do you think the 5D-III's AA-filter is having?jrista said:I have not had much clear sky time this year. Been extra heavy on the clouds. Originally that was because of El Ninio, which was hot and heavy until the end of spring this year. But, the pacific conditions are returning to La Ninia status...so, I honestly don't know why the clouds are still perpetual.
Anyway...I dropped $3700 on some new camera gear in April. I've only been able to use it once...on the moon!Just to give you guys an idea of how much better dedicated astro gear can be...this is from an ASI1600MM-Cool camera, running at -15°C. I used very high quality AstroDon LRGB filters (these alone cost $500), since the camera is monochrome. The camera has only 1.5e- read noise, and 0.008e-/s dark current @ -15C (compared to my 5D III, which has a monstrous 4e-/s @ 28C!!!) So, this is an ultra low noise camera.
I acquired high speed video sequences for four filters, LRGB, between 30-50fps. Graded, culled, aligned and stacked the top 15% of those in a program called AutoStakkert!2 to get individual Luminance (L), Red, Green and Blue channel images. I then processed those in PixInsight to combine it all into a deconvolved, high detail full-color image:
You can see a larger version here:
http://www.astrobin.com/full/260298/B/
And, for those who are willing to download the full size, you can see that here:
http://www.astrobin.com/full/260298/B/?real=&mod=
To contrast the quality of this camera with what I've been able to get with my 5D III. This is the largest lunar image I've had with the 5D III:
See full size here: http://i.imgur.com/NiWu0FV.jpg
I used video capture with this one, but the 5D III can barely sustain about 20fps. On top of that, the low pass filter and interpolation to convert the bayer CFA to a full color image really softens things up. The next highest resolution single-image of the moon I've got from the 5D III is this guy:
See full size here: http://i.imgur.com/r7DlVOz.jpg
This was a while ago, but it was on a night of very good seeing, so the detail was pretty darn good. But it only holds up at this size...again, the low pass filter and interpolation soften things up. The ASI1600 has neither a low pass filter, nor does it need to be demosaiced...you get a 100% fill factor on every channel.
I've also been wondering lately... does a lack of AA-filter results in better SNR?
Upvote
0
telemaq76 said:
Yeah, good quality narrow band filters are not cheap. Earlier this year, I was considering getting one of the new KAF-16200 cameras, with APS-H size sensors. The necessary filters for that were the 2" or 50mm square. Those suckers cost $1200, EACH!
Just to get a full set of narrow band filters, I would have had to spend $3600, and that was on top of the camera price of $7000, and the filter wheel price of about $2800. The insane cost of all of that was one of the reasons I went with the ASI1600 instead. It is a smaller sensor, but it was far cheaper overall. I managed to get a full set of NB filters, and the LRGB filters, and the filter wheel, camera, and all the necessary adapters for a little over $3700.
Upvote
0
StudentOfLight said:
The AA filter has a huge impact. With the AA filter, I can't really get stars less than about 4.2" in size (FWHM, full width half maximum, as we call it) most of the time. In the absolute best of conditions, my stars get down to about 3" with the 5D III...but, that is very rarely, and I don't think it can really get much better than that.
The diffraction limited performance of my 600mm lens is 0.76", and the diffraction limited performance of my AT8RC is 0.48". So, even a 3" star is 4x larger than my 600mm lens is capable of, and over 6x larger than the RC is capable of. However, with the ASI1600, I've had my stars down around 1.5" pretty often. Smaller pixels, but also no low pass filter. The low pass filters really blur the crap out of things, no question.
For astrophotography, you don't have any patterns that can create moire. Even if you image undersampled, you can recover detail and smoothness by drizzling, which will recover lost resolution. So, cameras that don't have a low pass filter will usually produce better results.
As for SNR. Yes, sharper details will have better SNR. If a filamentary structure is really 5" in size, but is being blurred to 10" in size, then all the signal information that should have been concentrated in half the area, is being spread out over twice the area. And, for any given unit area, the SNR will be lower when the information is blurred. So a low pass filter can definitely hurt your SNR. Same thing goes for not having ideal focus. If your focus is off, your blurring information, spreading the photons out over a greater area. The total signal coming from the object at large is the same, but the focused signal for specific parts of the object can be reduced (because they are being redistributed over a larger area.)
Upvote
0
jrista said:
Hi Jon. Thanks for the Flickr favourite on my cuckoo (I guess that was you!). I just got a modest telescope (4SE) and have been fiddling around with my DSLR as a webcam for planetary imaging using the telescope and using the tracking mount with a lens for deep sky stuff, but of course have been looking at how to improve things. So two questions...
With regards to planetary things, is a basic astro webcam better than the 5Ds? I've rigged it up using the Canon software and QuickTime to record the magnified LiveView feed directly on my computer - which works but is awkward. As for deep sky, is a dedicated astro cam better? Given it's such a cheap telescope and mount, would I be better off investing in filters? (I live in a moderately light polluted place). I saw it's possible to mount a deep sky CCD directly to Canon lenses...
Upvote
0
If you want to get good results with planetary, you need very high magnification (an SCT with a 2-3x barlow, for around f/20-f/30!!), and you need very high speed video. The idea with planetary is that you use very short exposures with very sensitive cameras (we are talking way more sensitive than any DSLR on the market, 75-85% quantum efficiency) and very low noise (2e- or less, many planetary cams have less than 1e- read noise!) so that you can get exposures that are on the time-order of scintillation. Scintillation is the word used to describe the main form of high-frequency high altitude turbulence or "seeing" that affects planetary imaging. Scintillation is caused by the jetstream.
At a frame rate of 100fps or more, your exposure lengths are 10ms or shorter, which is around the time-order of scintillation. There are larger-scale effects from seeing that occur over longer time periods...around a second or so, and larger around 10-30 seconds. With a DSLR, especially a high resolution one, in a high quality video mode, you are usually lucky to get 30fps...with the 5D III I can only get about 20fps sustained at full resolution. You also cannot choose an ROI (region of interest) with a DSLR, so your stuck reading the entire sensor out at full resolution, when most of the time for planetary you just need a tiny 400x400 pixel or smaller ROI.
DSLR video also entails all the read noise the camera can pump out, which can be rather extreme when in high speed video, because the sensor heats up, which jacks the dark current level up, which introduces even more noise on top of the base read noise (which, BTW, is higher with a higher speed video readout than a standard single frame readout.) Finally, the video coming out of a Canon DSLR is usually compressed, which is less than ideal for planetary imaging.
A dedicated planetary camera (I'd look at ZWO, their ASI camera line, particularly the new USB 3 products like the 224MC or 178MM) is going to be extremely low noise, extremely high sensitivity, capable of delivering hundreds of frames per second with ROI's down to 100x100 pixels, and with hardware binning on a mono camera, your frame rates can get upwards of 300fps to nearly 800fps (good for solar and lunar imaging). There is little comparison between a dedicated planetary camera, and a DSLR, these days.
The ZWO ASI178MM for example, is an excellent planetary cam. It has as little as 1.4e- read noise, uses a Sony BSI sensor, quantum efficiency is up around 80%, frame rates up to 240fps unbinned (480fps binned!), and it's 14-bit so the IQ is very high quality. It's a monochrome camera (although i think there is also a color version), so you can use LRGB filters with it to get the maximum quality (and ZWO just released a really nice mini filter wheel that supports 1.25" filters, and they sell an LRGB filter set themselves as well).
https://astronomy-imaging-camera.com/products/usb-3-0/asi178mm-mono/
https://astronomy-imaging-camera.com/products/accessories/efw-mini/
https://astronomy-imaging-camera.com/products/accessories/zwo-new-rgbl-filters-optimised-asi1600/
https://astronomy-imaging-camera.com/products/usb-3-0/178mm-kit/ (NOTE: This kit uses the manual filter wheel, not the new EFW-mini linked above! An electronic filter wheel is very handy...but, it does require a laptop or other computer to be connected, and software that supports filter wheel control to be used.)
The difference in planetary, lunar and solar imaging with the ASI178MM would be orders of magnitude better than imaging with a 5Ds. Absolutely no comparison. These cameras are also very small, very light weight, low power, so they are highly portable and easy to haul around to dark sites if you want. While the sensors are small, they can also be used to do longer exposure astrophotography, and with a mono camera with LRGB sensors, you'ed smoke the 5Ds for small FoV imaging (i.e. galaxies, or close up imaging of nebula details).
The 5DsR would make a great wide field camera, though. The 178 cant come close to delivering the kind of huge full frame field of view that the 5DsR could, and the R, lacking a low pass filter, would deliver great image quality. The difficulty there, actually, is getting a large enough image circle. Canon lenses certainly have them, but few are actually good for astrophotography. Most regular refracting telescoeps, and for that matter some of the lower end reflecting telescopes, just don't have a large enough image circle for a full frame sensor. That is one of the reasons I use my 600mm lens as a telescope...it's got a very large image circle.
As for light pollution. Imaging with a monochrome camera is far more efficient than imaging with a color DSLR. Note how I've said LRGB. That stands for Luminance, Red, Green, Blue. If you just used RGB filters, while you would be more efficient at gathering light than a color DSLR, it is really the L filter that sucks down the photons. An L filter is a full spectrum filter, square cutoff, blocking UV and IR, but otherwise passing the entire visible spectrum with over 95% transmission. The red, green and blue color filters in a bayer sensor are not even square bandpasses, they are more along the lines of a gaussian bandpass, and their transmission levels are usually not even 90%. So, combine L imaging, with RGB imaging. You can acquire photons far, far faster by getting lots and lots of L subs, and just enough RGB subs to avoid issues with noise and stack properly (usually about an hour to an hour and a half for each RGB channel, then all the rest of your time on the L channel).
So if you want to do full color imaging in the city, LRGB with a mono camera is still more efficient than a DSLR (barring FoV differences). The real benefits with a mono camera come from the ability to use narrow band filters, though. With narrow band imaging, you can block out 90% or more of the visible spectrum, and just image one narrow emission band at a time. Hydrogen alpha, Oxygen three, Sulfur two. You need longer exposures, and a many hours of data at least on each channel, but you don't really have to worry about light pollution.
At a frame rate of 100fps or more, your exposure lengths are 10ms or shorter, which is around the time-order of scintillation. There are larger-scale effects from seeing that occur over longer time periods...around a second or so, and larger around 10-30 seconds. With a DSLR, especially a high resolution one, in a high quality video mode, you are usually lucky to get 30fps...with the 5D III I can only get about 20fps sustained at full resolution. You also cannot choose an ROI (region of interest) with a DSLR, so your stuck reading the entire sensor out at full resolution, when most of the time for planetary you just need a tiny 400x400 pixel or smaller ROI.
DSLR video also entails all the read noise the camera can pump out, which can be rather extreme when in high speed video, because the sensor heats up, which jacks the dark current level up, which introduces even more noise on top of the base read noise (which, BTW, is higher with a higher speed video readout than a standard single frame readout.) Finally, the video coming out of a Canon DSLR is usually compressed, which is less than ideal for planetary imaging.
A dedicated planetary camera (I'd look at ZWO, their ASI camera line, particularly the new USB 3 products like the 224MC or 178MM) is going to be extremely low noise, extremely high sensitivity, capable of delivering hundreds of frames per second with ROI's down to 100x100 pixels, and with hardware binning on a mono camera, your frame rates can get upwards of 300fps to nearly 800fps (good for solar and lunar imaging). There is little comparison between a dedicated planetary camera, and a DSLR, these days.
The ZWO ASI178MM for example, is an excellent planetary cam. It has as little as 1.4e- read noise, uses a Sony BSI sensor, quantum efficiency is up around 80%, frame rates up to 240fps unbinned (480fps binned!), and it's 14-bit so the IQ is very high quality. It's a monochrome camera (although i think there is also a color version), so you can use LRGB filters with it to get the maximum quality (and ZWO just released a really nice mini filter wheel that supports 1.25" filters, and they sell an LRGB filter set themselves as well).
https://astronomy-imaging-camera.com/products/usb-3-0/asi178mm-mono/
https://astronomy-imaging-camera.com/products/accessories/efw-mini/
https://astronomy-imaging-camera.com/products/accessories/zwo-new-rgbl-filters-optimised-asi1600/
https://astronomy-imaging-camera.com/products/usb-3-0/178mm-kit/ (NOTE: This kit uses the manual filter wheel, not the new EFW-mini linked above! An electronic filter wheel is very handy...but, it does require a laptop or other computer to be connected, and software that supports filter wheel control to be used.)
The difference in planetary, lunar and solar imaging with the ASI178MM would be orders of magnitude better than imaging with a 5Ds. Absolutely no comparison. These cameras are also very small, very light weight, low power, so they are highly portable and easy to haul around to dark sites if you want. While the sensors are small, they can also be used to do longer exposure astrophotography, and with a mono camera with LRGB sensors, you'ed smoke the 5Ds for small FoV imaging (i.e. galaxies, or close up imaging of nebula details).
The 5DsR would make a great wide field camera, though. The 178 cant come close to delivering the kind of huge full frame field of view that the 5DsR could, and the R, lacking a low pass filter, would deliver great image quality. The difficulty there, actually, is getting a large enough image circle. Canon lenses certainly have them, but few are actually good for astrophotography. Most regular refracting telescoeps, and for that matter some of the lower end reflecting telescopes, just don't have a large enough image circle for a full frame sensor. That is one of the reasons I use my 600mm lens as a telescope...it's got a very large image circle.
As for light pollution. Imaging with a monochrome camera is far more efficient than imaging with a color DSLR. Note how I've said LRGB. That stands for Luminance, Red, Green, Blue. If you just used RGB filters, while you would be more efficient at gathering light than a color DSLR, it is really the L filter that sucks down the photons. An L filter is a full spectrum filter, square cutoff, blocking UV and IR, but otherwise passing the entire visible spectrum with over 95% transmission. The red, green and blue color filters in a bayer sensor are not even square bandpasses, they are more along the lines of a gaussian bandpass, and their transmission levels are usually not even 90%. So, combine L imaging, with RGB imaging. You can acquire photons far, far faster by getting lots and lots of L subs, and just enough RGB subs to avoid issues with noise and stack properly (usually about an hour to an hour and a half for each RGB channel, then all the rest of your time on the L channel).
So if you want to do full color imaging in the city, LRGB with a mono camera is still more efficient than a DSLR (barring FoV differences). The real benefits with a mono camera come from the ability to use narrow band filters, though. With narrow band imaging, you can block out 90% or more of the visible spectrum, and just image one narrow emission band at a time. Hydrogen alpha, Oxygen three, Sulfur two. You need longer exposures, and a many hours of data at least on each channel, but you don't really have to worry about light pollution.
Upvote
0
my very first h-alpha color image. i mixed an h-alpha picture i just made with a standard picture i ve done weeks ago.
.
It s great i can now shoot with full moon. That s why i bought that h-alpha 6nm filter. Most of the time full moon days are the clearest days of the month, no clouds
.It s great i can now shoot with full moon. That s why i bought that h-alpha 6nm filter. Most of the time full moon days are the clearest days of the month, no clouds
Upvote
0
jrista said:
That's fantastic, thanks. Really detailed and concise.
It was wishful thinking of me to hope for a one size fits all approach. I guess it all ultimately boils down to how much money each person thinks it's worth spending... it's so rarely good weather here, I've hesitated to buy much dedicated astro gear. Decisions, decisions...
(At least planetary imaging has the advantage that you don't need such long clear periods, I might start with that, although it's hard to beat the glory of good deep sky images).
Upvote
0
Yeah, you do need to factor in cost recovery time on your equipment. I purchased the ASI1600, and have only used it twice...once for the lunar image, once already a week ago (more, even) to get some Ha (hydrogen alpha) subs on Cave nebula (and they aren't even keepers due to a tilt issue). I won't be covering much of the cost of the new kit in usage much this year, it seems. However, over the long term...I definitely think it was worth it. It's a pretty amazing camera...incredibly low noise.
If you only have a handful of days a year to do any astro, then you would want to very carefully weight the options. That said, you can get a full monochrome CMOS kit with LRGB filters with the ASI178MM for about $800. The cost of getting into the mono game has never been so good.
If you only have a handful of days a year to do any astro, then you would want to very carefully weight the options. That said, you can get a full monochrome CMOS kit with LRGB filters with the ASI178MM for about $800. The cost of getting into the mono game has never been so good.
Upvote
0
telemaq76 said:
Beautiful telemaq76, not an overprocess, but maybe reduce the tint, more green and less purple.
Upvote
0