Hi Robin,
I found a video online of a presentation you made on 9 March 2019 at the Practical Astronomy show — it was great! It also occurred to me that you may well be the ideal source for whom I’ve searched to ask a question that has frustrated me for quite some time; to wit:
As I get more into astro photography, I’m extremely frustrated with the “fuzzy” suggestions proffered on the Web as to nominal exposures for celestial objects, such as, “take at least 3 or 4 hours of exposure” for a target rather than adjusting the exposure for the apparent magnitude of the target and my location’s Bortle class, which seems inherently more logical to me. A review of images and the equipment used in posts on Astrobin has been no more helpful because the integration times for any specific object are often unrelated to the posted equipment used. I’m trying to determine a proper overall exposure for a celestial target based upon its magnitude.
I realize there’s no true “standard” as there are too many variables, such as that particular evening ski’s seeing, transparency, etc. But it would be so very helpful, especially when beginning in AP, to have a reasonable, nominal starting point based on apparent magnitude and Bortle class and then, of course, your equipment. I’d like to create such a list, say from magnitude -4 to +14 that I can use for my refractor, and then adapt it correspondingly for my larger aperture SCT which also has a longer focal length. I can easily compensate mathematically for the size differences of the objective, subtraction of the secondary mirror area, the different focal lengths of the scopes, etc., and am aware of the logarithmic nature of the celestial magnitude scale. The difficulty is having a reasonable starting point.
For example, how do I come up with a ballpark integration exposure for a magnitude 4 object using my 100mm refractor with a focal length of 550mm in my Bortle 4 yard? After that, I can determine a correspondingly appropriate exposure for my 203mm objective SCT native, with a 0.7 focal reducer, with a HyperStar, etc. But determining an exposure starting point has eluded me. I could find no authoritative papers regarding ball-parking an exposure per magnitude on any academic site nor professional astronomy website.
Of course, I realize that I don’t know what I don’t know, and there are questions, such as, is the apparent magnitude of the surface brightness of a nebula to be treated differently than the magnitude of a star, or a cluster, or a galaxy? Watching your aforementioned presentation implies to me that you could provide valuable insight to my quest.
Finally, I do appreciate this has little to do with SharpCap and the great support you provide. But I’m desperate and want to be more efficient when shooting images, So please, “Help me Obi wan Kenobi. You’re my only hope.”
Thanks much in advance!
Jerry
Exposure Starting Point
-
admin
- Site Admin
- Posts: 13534
- Joined: Sat Feb 11, 2017 3:52 pm
- Location: Vale of the White Horse, UK
- Contact:
Re: Exposure Starting Point
Hi Jerry,
I'm not aware of any standard table or info that suggests the total integration time needed for various magnitudes of objects - there are a few problems with this idea I suppose
1) The magnitude usually indicates the brightest parts of the object, but often we are interested in seeing the fainter parts, and those will require much longer total imaging time that the bright parts
2) The answer depends on how much noise you find acceptable in an image - quadruple the total imaging time and you will halve the noise in the final image. Some people might find the original level of noise OK, some might like the halved level, some might want to go lower still
3) The equipment plays a part in the answer - f-ratio and pixel size are the main contributors (double the f-ratio - four times longer as four times less light on each pixel; double the pixel size - four times shorter as the bigger pixels collect four times more light)
4) The noise level in the final image isn't just determined by the noise of the target, but also the potentially much larger noise coming from the light pollution background, so the times needed in a highly light polluted area are much longer than in a dark sky area for the same results
5) The almost universal answer is going to be 'more time than you have available before the clouds/dawn/time to pack up arrive'
The approach that I think most people use is to start with a short time on the target and then add more if the noise level in the resulting image is unacceptable. Of course, that soon adds up, since every halving of noise requires four times more imaging time! There is also an impact from how you process the image - some tools now provide very capable noise reduction features which can help make great images from shorter integration times.
Sorry not to be able to give you an easy answer to the problem,
Robin
I'm not aware of any standard table or info that suggests the total integration time needed for various magnitudes of objects - there are a few problems with this idea I suppose
1) The magnitude usually indicates the brightest parts of the object, but often we are interested in seeing the fainter parts, and those will require much longer total imaging time that the bright parts
2) The answer depends on how much noise you find acceptable in an image - quadruple the total imaging time and you will halve the noise in the final image. Some people might find the original level of noise OK, some might like the halved level, some might want to go lower still
3) The equipment plays a part in the answer - f-ratio and pixel size are the main contributors (double the f-ratio - four times longer as four times less light on each pixel; double the pixel size - four times shorter as the bigger pixels collect four times more light)
4) The noise level in the final image isn't just determined by the noise of the target, but also the potentially much larger noise coming from the light pollution background, so the times needed in a highly light polluted area are much longer than in a dark sky area for the same results
5) The almost universal answer is going to be 'more time than you have available before the clouds/dawn/time to pack up arrive'
The approach that I think most people use is to start with a short time on the target and then add more if the noise level in the resulting image is unacceptable. Of course, that soon adds up, since every halving of noise requires four times more imaging time! There is also an impact from how you process the image - some tools now provide very capable noise reduction features which can help make great images from shorter integration times.
Sorry not to be able to give you an easy answer to the problem,
Robin
Re: Exposure Starting Point
The magnitude of a DSO, for example a galaxy, is the average of the bright and faint parts integrated over an ellipse. Who determines the size of the ellipse? Some governing body. Therefore, if today Galaxy X is said to be 20th magnitude, and then tomorrow the governing body says the ellipse is larger, Galaxy X will become fainter on the magnitude scale.
But Robin is right, you always want to integrate for the faintest parts, if possible. One time I spent 11 hours over 3 nights, and I still would have liked to have gone longer. Best advice I can give is to buy a cooled camera with low read noise. How long you spend integrating is a personal choice that can only be learned by doing. Soon, you'll get a feel for what it takes. I've turned down many objects due to the limitations of my equipment and skies.
And then there are times when you just want to capture something to prove a point. A few years ago I wanted to prove that I could capture the Integrated Flux Nebula in the region of M81 with my low-cost refractor and high-noise CCD. I did it but it was not publication quality. I just wanted to see if I could do it. In that case I accepted a higher than normal level of noise.
Brian
But Robin is right, you always want to integrate for the faintest parts, if possible. One time I spent 11 hours over 3 nights, and I still would have liked to have gone longer. Best advice I can give is to buy a cooled camera with low read noise. How long you spend integrating is a personal choice that can only be learned by doing. Soon, you'll get a feel for what it takes. I've turned down many objects due to the limitations of my equipment and skies.
And then there are times when you just want to capture something to prove a point. A few years ago I wanted to prove that I could capture the Integrated Flux Nebula in the region of M81 with my low-cost refractor and high-noise CCD. I did it but it was not publication quality. I just wanted to see if I could do it. In that case I accepted a higher than normal level of noise.
Brian
Re: Exposure Starting Point
Robin and Brian,
Thanks both for your responses! At the moment, I'm already doing what Brian has advised: I'm getting several hours of data on a target each evening as the weather allows. As a part-time nature and landscape photographer for many years, I'm used to measuring or calculating an exposure for my intended effect. With AP, it's been more of a crap-shoot with so much conflicting advice. So my desire has been to find a way to derive a logical starting point that's likely to result in at least an acceptable image. I've already learned that spending several hours recording in an evening is no guarantee I'll be satisfied with the results and will often need to collect still more data on the same target one or more additional evenings. If it's something like M42 or M45, they're so bright that even that much time is not necessary and still results in fine detail and the collection of wispy nebulae clouds. unfortunately, there aren't a plethora of such targets.
My biggest frustration is that, these days, we're only getting perhaps, 4 - 6 days a month with clear nightly skies lasting several hours and one good image can therefore take several weeks, an entire month, or even more depending on its magnitude and the weather conditions. This has often proven exasperating to say the least.
My typical subframe exposure lengths these days are down to 30 - 60 seconds. There are now so many StarLink satellites in orbit and aircraft aloft that I'm losing one subframe for every 10 or 15 that I shoot. I started at 10 minute subframes, dropped to 5 minutes, and was still losing too many to stay at those levels. At least Noise Exterminator and GraxPert's AI Noise Reduction do provide some solace to rescue us from the insane requirement of interminably long integration times otherwise necessary for a clear image.
Thank you both! It sounds as if how I've forced to adapt and am now doing is what everyone is doing, at least unless they live on a mountain above the clouds or in the rare locations where clear skies are the norm rather than the exception.
Take care,
Jerry
Thanks both for your responses! At the moment, I'm already doing what Brian has advised: I'm getting several hours of data on a target each evening as the weather allows. As a part-time nature and landscape photographer for many years, I'm used to measuring or calculating an exposure for my intended effect. With AP, it's been more of a crap-shoot with so much conflicting advice. So my desire has been to find a way to derive a logical starting point that's likely to result in at least an acceptable image. I've already learned that spending several hours recording in an evening is no guarantee I'll be satisfied with the results and will often need to collect still more data on the same target one or more additional evenings. If it's something like M42 or M45, they're so bright that even that much time is not necessary and still results in fine detail and the collection of wispy nebulae clouds. unfortunately, there aren't a plethora of such targets.
My biggest frustration is that, these days, we're only getting perhaps, 4 - 6 days a month with clear nightly skies lasting several hours and one good image can therefore take several weeks, an entire month, or even more depending on its magnitude and the weather conditions. This has often proven exasperating to say the least.
My typical subframe exposure lengths these days are down to 30 - 60 seconds. There are now so many StarLink satellites in orbit and aircraft aloft that I'm losing one subframe for every 10 or 15 that I shoot. I started at 10 minute subframes, dropped to 5 minutes, and was still losing too many to stay at those levels. At least Noise Exterminator and GraxPert's AI Noise Reduction do provide some solace to rescue us from the insane requirement of interminably long integration times otherwise necessary for a clear image.
Thank you both! It sounds as if how I've forced to adapt and am now doing is what everyone is doing, at least unless they live on a mountain above the clouds or in the rare locations where clear skies are the norm rather than the exception.
Take care,
Jerry
Re: Exposure Starting Point
Hi Jerry,
5 to 10-minute subs are hurting you more than helping. Stars blow out. Sky pollution takes up a lot of your dynamic range. Satellites and aircraft damage a large percentage of frames. It's not worth it but there is a better way. Reduce your exposure. SharpCap's "Brain" helps with that. Many here call it the "Optimal Exposure Calculator". I call it the "Minimum Exposure Calculator" mostly because you can expose longer but the rewards diminish and the risks increase from satellites, aircraft, and clouds. From my experience avoid going under the minimum exposure. Depending on a number of factors you could become the victim of "raining noise".
The difference between astrophotography and landscape photography is night and day (excuse the pun.) In conventional photography you enjoy a generous supply of photons to work with. It is the opposite in astrophotography where you never seem to have enough photons so you have to resort to long integration times. The role of exposure is to allow just enough photons to fall on the sensor so that it overcomes noise. (Noise comes from sky pollution, sensor temperature, and sensor "read noise".) If you don't expose long enough then you miss out on all of that faint delicate nebulae. However, like I said you could expose longer but now you risk spoiling frames from satellites, aircraft, and clouds. For me that exposure is 2 minutes using my noisy CCD and Bortle 5 skies. However, using my CMOS camera 30 seconds works fine. We can discuss Gain but I'd have to locate some graphs to help with that.
Brian
5 to 10-minute subs are hurting you more than helping. Stars blow out. Sky pollution takes up a lot of your dynamic range. Satellites and aircraft damage a large percentage of frames. It's not worth it but there is a better way. Reduce your exposure. SharpCap's "Brain" helps with that. Many here call it the "Optimal Exposure Calculator". I call it the "Minimum Exposure Calculator" mostly because you can expose longer but the rewards diminish and the risks increase from satellites, aircraft, and clouds. From my experience avoid going under the minimum exposure. Depending on a number of factors you could become the victim of "raining noise".
The difference between astrophotography and landscape photography is night and day (excuse the pun.) In conventional photography you enjoy a generous supply of photons to work with. It is the opposite in astrophotography where you never seem to have enough photons so you have to resort to long integration times. The role of exposure is to allow just enough photons to fall on the sensor so that it overcomes noise. (Noise comes from sky pollution, sensor temperature, and sensor "read noise".) If you don't expose long enough then you miss out on all of that faint delicate nebulae. However, like I said you could expose longer but now you risk spoiling frames from satellites, aircraft, and clouds. For me that exposure is 2 minutes using my noisy CCD and Bortle 5 skies. However, using my CMOS camera 30 seconds works fine. We can discuss Gain but I'd have to locate some graphs to help with that.
Brian
-
admin
- Site Admin
- Posts: 13534
- Joined: Sat Feb 11, 2017 3:52 pm
- Location: Vale of the White Horse, UK
- Contact:
Re: Exposure Starting Point
Hi,
if you really want to attack it from a scientific viewpoint, it's possible to do, but would rely on data that probably isn't available without trying to measure it yourself. Let me explain...
In astro imaging you get three components to the pixel value that comes out in your final image
1) Photons from your target that create electrons in the sensor pixel
2) Photons from light pollution that create electrons
3) Electrons created by random thermal noise in the sensor
For this calculation we will assume that you have a cooled camera and the thermal noise level is small comapred to the other two, just to simplify things, but it could be included.
Light pollution - let's assume that the rate of electrons accumulating from light pollution is 1e/pixel/s (you can work out a value for your setup and observing conditions using our calculator tool here - https://tools.sharpcap.co.uk/)
Target photons - we will start by assuming that the target is giving us 0.1 e/pix/s
We'll also assume that you are taking sub-exposures at least as long as recommended by the SharpCap 'brain' function, so that camera read noise can be ignored.
So, if you image for an hour, whether it is 1 hour long exposure or 180 x 20s ones, you will collect on average 3600 electrons per pixel light pollution and 360 from the target, to a total of 3960. The random noise that pixel values will show in that total is given by the square root of the number of electrons (Poisson distribution), so is about 62. We are going to subtract out the light pollution background in processing, leaving a signal of 360e from the target and a noise level of 62, so a S/N ratio of about 6 - not bad... You will be able to see the target and fluctuations in brightness in it, but the noise will be very noticeable.
Now consider a fainter part of the target - 0.01 e/pix/s from this part. Now our hour of imaging gives us only 36 electrons from the target, but the light pollution is still 3600, so a total of 3636 and a noise about 60. Now our S/N ratio is just 0.6 - you will probably be able to make out that *something* is there in this region, as long as it extends over a good number of pixels, but no real details
Fainter still at 0.001 e/pix/s and our hour brings just 3.6 electrons and an S/N of 0.06 - you will not be seeing this with just one hour of imaging.
So, what to do to improve matters? Well, two things really
1) Image for longer - quadrupling the imaging time multiplies the signal by 4, but the noise by only 2, so it doubles the S/N ratio. This will help our medium brightness area, raising the S/N to 1.2, but not much help for the faintest area - a S/N of 0.12 is still pretty rubbish
2) Reduce the light pollution level - darker skies have a big effect, so if you magically move to a world class dark sky site, with an average f/5 telescope and 3.75 micron pixel camera, your light pollution rate might be about 0.1 to 0.2 e/pix/s - suddenly you only get maybe 500 electrons of light pollution per hour, so the typical noise level drops from ~60 to ~22, giving a 3 times boost in S/N. This is equivalent to taking 9 times longer total exposure at the more light polluted site
3) Reduce the light pollution level with filters - if the light from your target is concentrated in particular spectral bands and the light pollution light *isn't* particularly in those bands (or if the light pollution is in a particular band), you can use filters that will reduce the light pollution more than they reduce the light from the target. In this case you need to do the maths related to how many light pollution photons get through the filter and how many target photons get through.
This is all very well in theory, but it all relies on knowing the brightness of the target, and not just the total brightness or the brightness of the brightest area, but the surface brightness (in magnitudes per square arc second) of the fainter bits that we want to resolve well in our image. I'm not aware of any database of this sort of information though
cheers,
Robin
if you really want to attack it from a scientific viewpoint, it's possible to do, but would rely on data that probably isn't available without trying to measure it yourself. Let me explain...
In astro imaging you get three components to the pixel value that comes out in your final image
1) Photons from your target that create electrons in the sensor pixel
2) Photons from light pollution that create electrons
3) Electrons created by random thermal noise in the sensor
For this calculation we will assume that you have a cooled camera and the thermal noise level is small comapred to the other two, just to simplify things, but it could be included.
Light pollution - let's assume that the rate of electrons accumulating from light pollution is 1e/pixel/s (you can work out a value for your setup and observing conditions using our calculator tool here - https://tools.sharpcap.co.uk/)
Target photons - we will start by assuming that the target is giving us 0.1 e/pix/s
We'll also assume that you are taking sub-exposures at least as long as recommended by the SharpCap 'brain' function, so that camera read noise can be ignored.
So, if you image for an hour, whether it is 1 hour long exposure or 180 x 20s ones, you will collect on average 3600 electrons per pixel light pollution and 360 from the target, to a total of 3960. The random noise that pixel values will show in that total is given by the square root of the number of electrons (Poisson distribution), so is about 62. We are going to subtract out the light pollution background in processing, leaving a signal of 360e from the target and a noise level of 62, so a S/N ratio of about 6 - not bad... You will be able to see the target and fluctuations in brightness in it, but the noise will be very noticeable.
Now consider a fainter part of the target - 0.01 e/pix/s from this part. Now our hour of imaging gives us only 36 electrons from the target, but the light pollution is still 3600, so a total of 3636 and a noise about 60. Now our S/N ratio is just 0.6 - you will probably be able to make out that *something* is there in this region, as long as it extends over a good number of pixels, but no real details
Fainter still at 0.001 e/pix/s and our hour brings just 3.6 electrons and an S/N of 0.06 - you will not be seeing this with just one hour of imaging.
So, what to do to improve matters? Well, two things really
1) Image for longer - quadrupling the imaging time multiplies the signal by 4, but the noise by only 2, so it doubles the S/N ratio. This will help our medium brightness area, raising the S/N to 1.2, but not much help for the faintest area - a S/N of 0.12 is still pretty rubbish
2) Reduce the light pollution level - darker skies have a big effect, so if you magically move to a world class dark sky site, with an average f/5 telescope and 3.75 micron pixel camera, your light pollution rate might be about 0.1 to 0.2 e/pix/s - suddenly you only get maybe 500 electrons of light pollution per hour, so the typical noise level drops from ~60 to ~22, giving a 3 times boost in S/N. This is equivalent to taking 9 times longer total exposure at the more light polluted site
3) Reduce the light pollution level with filters - if the light from your target is concentrated in particular spectral bands and the light pollution light *isn't* particularly in those bands (or if the light pollution is in a particular band), you can use filters that will reduce the light pollution more than they reduce the light from the target. In this case you need to do the maths related to how many light pollution photons get through the filter and how many target photons get through.
This is all very well in theory, but it all relies on knowing the brightness of the target, and not just the total brightness or the brightness of the brightest area, but the surface brightness (in magnitudes per square arc second) of the fainter bits that we want to resolve well in our image. I'm not aware of any database of this sort of information though
cheers,
Robin
Re: Exposure Starting Point
Robin,
Thank you for that great explanation! I’m lucky that my home is Bortle 4, but it is severely limited in sky access by a small opening overhead between the 90 - 110 foot trees surrounding my property. Essentially, I live within a small clearing in a forest – a double-edged sword for AP. So right now, objects such as M81, M82, and M101 are available to me. If I’m willing to drive I can get to large fields with views almost down to the horizon. That said, your description of comparing the electron accumulation at rates from 0.1 to 0.001 per pixel from our target was interesting. I know that signal data adds linearly and noise adds as the sq. rt., but your explanation brings home the magnitude of the challenge with such low signal levels. It also makes obvious the futility of doubling and quadrupling subsequent data gathering evenings to acquire less and less improvement due to diminishing returns, exacerbated by so few clear evenings per month.
One potential solution that comes to mind (though limited by the size of one’s pocketbook), is how large is your objective, or one you can purchase to collect more data per unit time, presuming you can and will carry it. Other options available to the well-healed, such as adaptive optics are not truly available to us yet, though may be in the future.
So, a larger scope (e.g. a C11 or some such), using a filter in light polluted areas (I have an Optolong L-Extreme that I especially like for nebulae, but take a big hit in additional exposure time), and going to a darker sky is about all the options I’m realistically afforded. After that, it’s more and more integration time (or a move to a drier clime in the Rocky Mountains; I’m in the U.S., so it’s less ridiculous than it sounds… well, still somewhat ridiculous, perhaps.)
One of the most salient points both you and Brian make is what magnitude truly encompasses. The fact that it’s the mean of the different brightnesses of an object is something I’d not previously thought enough about. I’m not interested in just capturing the highlights of an object – I’ve been doing that. I now want to go deeper and capture the filamentations and wispiness of nebulae and galaxies. That’s going to cost me in time, money, or both.
I still believe that I can assemble a reasonable starting point from an object’s apparent magnitude, but it’s now obvious it must be more conservative than what I’ve already created and can only be a starting point and guide, not a precise, trustable goal in and of itself as it is with terrestrial photography.
I want to thank you and Brian again for your generous insights and time! It’s been helpful and putting things into a more reasoned perspective for me (though a little frustrating
; but alas, the actual, sobering facts often are.).
Jerry
Thank you for that great explanation! I’m lucky that my home is Bortle 4, but it is severely limited in sky access by a small opening overhead between the 90 - 110 foot trees surrounding my property. Essentially, I live within a small clearing in a forest – a double-edged sword for AP. So right now, objects such as M81, M82, and M101 are available to me. If I’m willing to drive I can get to large fields with views almost down to the horizon. That said, your description of comparing the electron accumulation at rates from 0.1 to 0.001 per pixel from our target was interesting. I know that signal data adds linearly and noise adds as the sq. rt., but your explanation brings home the magnitude of the challenge with such low signal levels. It also makes obvious the futility of doubling and quadrupling subsequent data gathering evenings to acquire less and less improvement due to diminishing returns, exacerbated by so few clear evenings per month.
One potential solution that comes to mind (though limited by the size of one’s pocketbook), is how large is your objective, or one you can purchase to collect more data per unit time, presuming you can and will carry it. Other options available to the well-healed, such as adaptive optics are not truly available to us yet, though may be in the future.
So, a larger scope (e.g. a C11 or some such), using a filter in light polluted areas (I have an Optolong L-Extreme that I especially like for nebulae, but take a big hit in additional exposure time), and going to a darker sky is about all the options I’m realistically afforded. After that, it’s more and more integration time (or a move to a drier clime in the Rocky Mountains; I’m in the U.S., so it’s less ridiculous than it sounds… well, still somewhat ridiculous, perhaps.)
One of the most salient points both you and Brian make is what magnitude truly encompasses. The fact that it’s the mean of the different brightnesses of an object is something I’d not previously thought enough about. I’m not interested in just capturing the highlights of an object – I’ve been doing that. I now want to go deeper and capture the filamentations and wispiness of nebulae and galaxies. That’s going to cost me in time, money, or both.
I still believe that I can assemble a reasonable starting point from an object’s apparent magnitude, but it’s now obvious it must be more conservative than what I’ve already created and can only be a starting point and guide, not a precise, trustable goal in and of itself as it is with terrestrial photography.
I want to thank you and Brian again for your generous insights and time! It’s been helpful and putting things into a more reasoned perspective for me (though a little frustrating
; but alas, the actual, sobering facts often are.).Jerry
Re: Exposure Starting Point
Jerry
Don't overlook the simplicity of test shots and a correctly shaped histogram - to assess the exposure for an object under given sky conditions. These log histograms displayed in FITS Liberator are test shots of M42. Aim for A to be greater than zero and B to be less than 65535. If A is zero, then the histogram is touching the left hand axis. In this case, the object is under-exposed (clipped) and faint data is lost which cannot be recovered by processing. If B is 65535, then the histogram is touching the right hand axis. In this case, the object is over-exposed (clipped) and bright data is lost which cannot be recovered by processing.
[EDIT] if A=0 then look at increasing the offset (QHY)/Black Level (Altair)/Brightness (ZWO). This will move the histogram to the right.
For any object/sky conditions these simple tests will point to suitable exposures. Better to spend 5 minutes evaluating some test shots than spending all night gathering a load of data which next day turns out to be dud.
The data displayed in the histograms was actually used in an M42 HDR project which resulted in this image.
Dave
Don't overlook the simplicity of test shots and a correctly shaped histogram - to assess the exposure for an object under given sky conditions. These log histograms displayed in FITS Liberator are test shots of M42. Aim for A to be greater than zero and B to be less than 65535. If A is zero, then the histogram is touching the left hand axis. In this case, the object is under-exposed (clipped) and faint data is lost which cannot be recovered by processing. If B is 65535, then the histogram is touching the right hand axis. In this case, the object is over-exposed (clipped) and bright data is lost which cannot be recovered by processing.
[EDIT] if A=0 then look at increasing the offset (QHY)/Black Level (Altair)/Brightness (ZWO). This will move the histogram to the right.
For any object/sky conditions these simple tests will point to suitable exposures. Better to spend 5 minutes evaluating some test shots than spending all night gathering a load of data which next day turns out to be dud.
- M42-exposure-histograms.JPG (59.7 KiB) Viewed 300 times
The data displayed in the histograms was actually used in an M42 HDR project which resulted in this image.
- M42_HDR_50x10s_30x30s_30x60s_15x120s.PNG (671.29 KiB) Viewed 300 times
Dave
-
Jean-Francois
- Posts: 418
- Joined: Sun Oct 13, 2019 10:52 am
- Location: Germany
Re: Exposure Starting Point
Hello Jerry,
Concerning using a larger scope ...
If you double the diameter with the same FD ratio, then you double the focal length.
With a double focal length, you will have an image with double dimension in X and in Y ... what does a factor 4 in surface.
The double diameter produces a factor 4 in surface at the entrance ...
and you can imagine what happen ... 4x more light over a 4x larger surface ... you have win nothing.
The quick calculation is valid for sky object with a surface. That is a little bit different with the stars itself.
So, if you wish to win something, you need to change the FD ratio.
If you change FD=10 to FD=8 or 6 then you will win something.
The maximum you can win is to use the RASA telescope.
With FD=2 or 2.2, then you will improve the surface intensity on your images.
But ... it is clear, that with the same diameter, a RASA will have a short focal.
But today with the small pixels, it is fine for the resolution.
The advantage you will have more field of view without buying a larger camera.
I have a RASA8" with a QHY268M. I can not use exposure time longer than 2 minutes ... the sky background is then overexposed.
Regards,
Jean-Francois
Concerning using a larger scope ...
If you double the diameter with the same FD ratio, then you double the focal length.
With a double focal length, you will have an image with double dimension in X and in Y ... what does a factor 4 in surface.
The double diameter produces a factor 4 in surface at the entrance ...
and you can imagine what happen ... 4x more light over a 4x larger surface ... you have win nothing.
The quick calculation is valid for sky object with a surface. That is a little bit different with the stars itself.
So, if you wish to win something, you need to change the FD ratio.
If you change FD=10 to FD=8 or 6 then you will win something.
The maximum you can win is to use the RASA telescope.
With FD=2 or 2.2, then you will improve the surface intensity on your images.
But ... it is clear, that with the same diameter, a RASA will have a short focal.
But today with the small pixels, it is fine for the resolution.
The advantage you will have more field of view without buying a larger camera.
I have a RASA8" with a QHY268M. I can not use exposure time longer than 2 minutes ... the sky background is then overexposed.
Regards,
Jean-Francois
Re: Exposure Starting Point
Jerry
Having slept on last night's post, the histograms also give other information. Looking at the 120s exposure, the histogram tells me that for M42 with my equipment and sky conditions (Bortle 6) I will not gain anything by increasing the exposure. Increasing the number of frames will help (up to a point). If I were to use 120s with M13, the result would be a blown out blob. For this object 15s - 30s, perhaps less, would be more appropriate.
Objects like M42 with a high dynamic range (bright core and faint nebulosity) may require exposures of different values to achieve the best image.
The object, sky conditions, equipment and the moon will all affect what can be considered optimal exposures. With this, there are no rules - only guidelines. Look at viewtopic.php?t=2715 for the result of stacking more frames.
Dave
Having slept on last night's post, the histograms also give other information. Looking at the 120s exposure, the histogram tells me that for M42 with my equipment and sky conditions (Bortle 6) I will not gain anything by increasing the exposure. Increasing the number of frames will help (up to a point). If I were to use 120s with M13, the result would be a blown out blob. For this object 15s - 30s, perhaps less, would be more appropriate.
Objects like M42 with a high dynamic range (bright core and faint nebulosity) may require exposures of different values to achieve the best image.
The object, sky conditions, equipment and the moon will all affect what can be considered optimal exposures. With this, there are no rules - only guidelines. Look at viewtopic.php?t=2715 for the result of stacking more frames.
Dave