Planetary camera

[Ken82]

Whilst I sort out a scope for my widefield setup I’m going to put my c11 on and do a bit on mars. 
 

initially  I’ll use my asi174 but in the longer run I want to use this for guiding my permanent mount. So I’ll need a new planetary cam for the c11 and possibly also a mount when I get my widefield setup working. I’ve just bought a crayford focuser which will limit mirror flop. 
 

So in terms of sampling rate the pixel size of the asi 174 with c11 needs a Barlow for optimal ratio I believe ?  I think x2 Barlow was best when I started out with this some years ago but I’ve forgotten the math. 
 

Perhaps a smaller pixel camera is best as I wouldn’t need a Barlow ? 
 

So what alternatives are available that would give good results with c11 ? 
 

Thanks ken 

[Owmuchonomy]

ASI290.  Good ratio at native f/l.  Extremely good near IR sensitivity too.

Edited September 28, 2020 by Owmuchonomy

[vlaiv]

Will you be going mono or OSC for your planetary camera?

My personal view is that OSC is far easier way to do it than all those electronic filter wheels and derotations and ....

I would say - look at ASI224 / ASI385 for OSC planetary.

Look at ASI290 or ASI178 for mono.

Depending on what type of camera you get and what pixel size it has - you'll need to use barlow to get to optimum sampling rate. Best is to get only barlow element rather than barlow lens. This way you can dial in the distance needed to achieve wanted F/ratio. Probably the best choice there is Baader VIP barlow (one of the best barlows there is).

In any case, if you go with ASI290 or 2.9um pixel size - mono option, F/ratio that you need will be

F/11.375

For ASI178 and 2.4um pixel size - F/9.4

Both of these F/ratios are very close to native F/10 and I would not bother with barlow in either of these two cases.

ASI224 and ASI385 (which is just a bit larger sensor and worth getting if you want to do Lunar) have 3.75um pixel size, but are also OSC sensors, which means they require a bit larger F/ratio.

F/29.4

and special "processing" - you either need to do super pixel debayering or downsize final result x2 if you want very sharp image. In any case, I would aim for x3 barlow with these cameras.

ASI174 is not very suited for planetary imaging (it is good for solar ha though and lunar use) due to high read noise, but if you want to give it a spin before you purchase something else, then best F/ratio to do that would be:

F/23 in case of mono and double that F/46 in case of OSC version of ASI174 (for OSC version, you'll need to downsample it by x2 or use super pixel mode - otherwise it might look too soft).

[Owmuchonomy]

Excellent advice above.  I have used both the 290 and 174 mono for Solar and Lunar (examples in the link in my signature).  In fact I did a comparison on here some time ago.  If that is one of your priorities then strongly consider these options.  The IR sensitivity of the 290 produces excellent lunar images if a little bit softer than the 174.  Or other options for OSC as @vlaiv says.

[michael.h.f.wilkinson]

On 28/09/2020 at 23:41, vlaiv said:

Will you be going mono or OSC for your planetary camera?

My personal view is that OSC is far easier way to do it than all those electronic filter wheels and derotations and ....

I would say - look at ASI224 / ASI385 for OSC planetary.

Look at ASI290 or ASI178 for mono.

Depending on what type of camera you get and what pixel size it has - you'll need to use barlow to get to optimum sampling rate. Best is to get only barlow element rather than barlow lens. This way you can dial in the distance needed to achieve wanted F/ratio. Probably the best choice there is Baader VIP barlow (one of the best barlows there is).

In any case, if you go with ASI290 or 2.9um pixel size - mono option, F/ratio that you need will be

F/11.375

For ASI178 and 2.4um pixel size - F/9.4

Both of these F/ratios are very close to native F/10 and I would not bother with barlow in either of these two cases.

ASI224 and ASI385 (which is just a bit larger sensor and worth getting if you want to do Lunar) have 3.75um pixel size, but are also OSC sensors, which means they require a bit larger F/ratio.

F/29.4

and special "processing" - you either need to do super pixel debayering or downsize final result x2 if you want very sharp image. In any case, I would aim for x3 barlow with these cameras.

ASI174 is not very suited for planetary imaging (it is good for solar ha though and lunar use) due to high read noise, but if you want to give it a spin before you purchase something else, then best F/ratio to do that would be:

F/23 in case of mono and double that F/46 in case of OSC version of ASI174 (for OSC version, you'll need to downsample it by x2 or use super pixel mode - otherwise it might look too soft).

I actually used the ASI224MC on a C8, and got much better results with the 2x Meade TeleXtender (tele-centric Barlow), or perhaps the 2.5x TV PowerMate than I could ever get with the 3x TeleXtender. I only ever used the 3x on older chips with much bigger pixels. I now use an ASI183MC (same pixel size as ASI178) for planetary and lunar (and even deep sky), but for planetary I only need a tiny region of interest. Here too, I tried the 2x TeleXtender and got much poorer results than without (but with flip mirror and R&P focuser in place, in which case the C8 is working roughly at F/12).

The ASI178MC can also be used for planetary as OSC option.

 

[vlaiv]

5 minutes ago, michael.h.f.wilkinson said:

I actually used the ASI224MC on a C8, and got much better results with the 2x Meade TeleXtender (tele-centric Barlow), or perhaps the 2.5x TV PowerMate than I could ever get with the 3x TeleXtender. I only ever used the 3x on older chips with much bigger pixels. I now use an ASI183MC (same pixel size as ASI178) for planetary and lunar (and even deep sky), but for planetary I only need a tiny region of interest. Here too, I tried the 2x TeleXtender and got much poorer results than without (but with flip mirror and R&P focuser in place, in which case the C8 is working roughly at F/12).

The ASI178MC can also be used for planetary as OSC option.

 

Yes, because with OSC camera - sampling rate and presentation rate should not be the same since there is bayer matrix.

You got better results with ASI224 closer to F/15 than those at F/30 because you should sample at F/30 but work with data at F/15.

I know this is somewhat confusing to most people - but reason for this is bayer matrix.

As far as pixel size is concerned ASI224 has 3.75um pixels and for those pixels proper F/ratio for critical sampling rate is ~15. However, blue, red and each component of green are not sampled at each pixel - but at every other pixel (look at blue and red in the image. Green in the image can be viewed as "two separate" greens - again having same spacing). This means that for OSC sensor, actual sampling rate is twice lower than pixel size would suggest.

For this reason you need to sample at F/30 and then to reduce size of the image by factor of two - or preferably use another method to achieve what is needed - split debayer or super pixel debayer (first is better then second).

That way you will end up with F/15 data although you used F/30 optics.

Same thing goes for ASI178 - pixel size suggests F/9.4 is needed but if you use OSC camera, you need to sample at twice that and then use appropriate debayer method to bring it back to F/9.4

What you did was to sample at "9.4" but sample every other pixel and then use interpolation to fill in missing data.

[Ken82]

Just having a look at the recommendations. It’s a bit overwhelming the amount of planetary cameras now available!

when I started with the asi174 some years ago I only had a choice between the 120 and 174. I started to use the 174 for guiding so never fully explored planetary. 
 

I’m thinking to stay with a colour cam to get started so the 385 sounds like an interesting prospect as I’d like to do some lunar. 
 

Also what about the asi462 ? 
 

I’ll have to do a side by side comparison on my computer later. 
 

ken 

[michael.h.f.wilkinson]

5 minutes ago, vlaiv said:

Yes, because with OSC camera - sampling rate and presentation rate should not be the same since there is bayer matrix.

You got better results with ASI224 closer to F/15 than those at F/30 because you should sample at F/30 but work with data at F/15.

I know this is somewhat confusing to most people - but reason for this is bayer matrix.

As far as pixel size is concerned ASI224 has 3.75um pixels and for those pixels proper F/ratio for critical sampling rate is ~15. However, blue, red and each component of green are not sampled at each pixel - but at every other pixel (look at blue and red in the image. Green in the image can be viewed as "two separate" greens - again having same spacing). This means that for OSC sensor, actual sampling rate is twice lower than pixel size would suggest.

For this reason you need to sample at F/30 and then to reduce size of the image by factor of two - or preferably use another method to achieve what is needed - split debayer or super pixel debayer (first is better then second).

That way you will end up with F/15 data although you used F/30 optics.

Same thing goes for ASI178 - pixel size suggests F/9.4 is needed but if you use OSC camera, you need to sample at twice that and then use appropriate debayer method to bring it back to F/9.4

What you did was to sample at "9.4" but sample every other pixel and then use interpolation to fill in missing data.

Yet in practice, a factor of two is not really needed, or is even detrimental, especially because of the effective dithering that takes place (unless your tracking and seeing are both perfect).

[vlaiv]

4 minutes ago, michael.h.f.wilkinson said:

Yet in practice, a factor of two is not really needed, or is even detrimental, especially because of the effective dithering that takes place (unless your tracking and seeing are both perfect).

I agree, but bayer drizzle seems to be removed from latest AS!3 incarnation for some reason?

There are additional issues with bayer drizzle - fact that you have sparse data means that resampling used needs to be very tricky in order to work properly - if one gets that wrong in implementation it will negate benefits of drizzle.

[vlaiv]

15 minutes ago, Ken82 said:

Also what about the asi462 ? 

That one should be better version of ASI290 with additional benefit - IR performance.

With IR 820nm and above filter, this camera behaves almost as mono sensor - all 3 elements of bayer matrix have more or less the same response curve.

It has lower read noise which is excellent. Only drawback is that we don't yet have QE figures on this camera. According to ZWO, it is supposed to be higher than ASI290, but I'm somewhat skeptical of that:

[Ken82]

I’ve narrowed it down to either the 385mc or 462mc. 
 

With the 462 I’m assuming I’ll need an IR cut filter when using an SCT ? 
 

in terms of Barlow would x2 be sufficient?

ken 
 

 

[nfotis]

Waiting for my 462 as well for my C9.25. IR/UV cut is needed for focusing, according to FLO (I bought also a IR-permitting filter at 850mm, which should be useful)

I guess that a 2x Barlow is enough for it and the C11 (that's f//20, a bit oversampled I suppose, but not as big  a deal as a clear atmosphere)

If you check the ZWO FB page or Astrobin, you should see many people shooting planetary photos with a similar setup (ASI290 , 2x Barlow and a medium to large SCT).

It seems that an ADC is a good tool to have if the planets are quite low on the horizon, too.

N.F.

 

Edited October 9, 2020 by nfotis

[Ken82]

On 28/09/2020 at 22:41, vlaiv said:

Will you be going mono or OSC for your planetary camera?

My personal view is that OSC is far easier way to do it than all those electronic filter wheels and derotations and ....

I would say - look at ASI224 / ASI385 for OSC planetary.

Look at ASI290 or ASI178 for mono.

Depending on what type of camera you get and what pixel size it has - you'll need to use barlow to get to optimum sampling rate. Best is to get only barlow element rather than barlow lens. This way you can dial in the distance needed to achieve wanted F/ratio. Probably the best choice there is Baader VIP barlow (one of the best barlows there is).

In any case, if you go with ASI290 or 2.9um pixel size - mono option, F/ratio that you need will be

F/11.375

For ASI178 and 2.4um pixel size - F/9.4

Both of these F/ratios are very close to native F/10 and I would not bother with barlow in either of these two cases.

ASI224 and ASI385 (which is just a bit larger sensor and worth getting if you want to do Lunar) have 3.75um pixel size, but are also OSC sensors, which means they require a bit larger F/ratio.

F/29.4

and special "processing" - you either need to do super pixel debayering or downsize final result x2 if you want very sharp image. In any case, I would aim for x3 barlow with these cameras.

ASI174 is not very suited for planetary imaging (it is good for solar ha though and lunar use) due to high read noise, but if you want to give it a spin before you purchase something else, then best F/ratio to do that would be:

F/23 in case of mono and double that F/46 in case of OSC version of ASI174 (for OSC version, you'll need to downsample it by x2 or use super pixel mode - otherwise it might look too soft).

Ok im going to get the asi 462. What was the calculation for optimal F ratio? I was working to x5 pixel size but your F ratios appear to be slightly different ? Also why much more F ratio for a colour cam ? 

Thanks Ken 

[vlaiv]

11 hours ago, Ken82 said:

Ok im going to get the asi 462. What was the calculation for optimal F ratio? I was working to x5 pixel size but your F ratios appear to be slightly different ? Also why much more F ratio for a colour cam ? 

Thanks Ken 

First thing to understand is that sampling rate is really pixel size - it is distance between pixels. It may seem strange to think that way, but that is correct way of thinking about it and it is helpful to see why colour cam has different optimum sampling rate than mono camera.

With mono camera distance between pixels is the same as pixel size (mathematically speaking - same X values only shifted half a pixel - pixel size is measured from start of one pixel to end of it / start of next, while distance between pixels is measured between their centers).

With colour cameras that is not the case as sensor has what is called Bayer matrix - alternating colour filters on adjacent pixels, and it looks like this:

On the left side is Bayer matrix itself and on the right side is how you should "view" each color. Let's take red for example: what is distance between red pixels? It is now not equal to pixel size, it is equal to two pixel sizes as we have red pixel, green pixel, red pixel, green pixel, and so on ...

Same is true for blue and also for green pixels (we even have Gr and Gb notation for two different "grids" of green pixels).

This means that distance between same pixels in Bayer matrix is twice as big compared to actual pixel size and in turn this means that sampling rate is twice as low. For this reason we have to use twice as long focal length with colour camera in order to get proper sampling rate (if we talk about critical sampling rate).

For 2.9um camera pixel size and it being OSC camera you need F/22.75. Since you'll be using C11 - you only need x2.2 barlow (which would be x2 barlow with tuned distance to sensor to give x2.2 magnification).

 

 

[inFINNity Deck]

Hi Ken,

I can confirm Vladimir's advice to increase by a factor 2 the focal ration when imaging in colour. You may be interested in how we arrived at these factors, which you can find in an article that I wrote on determining the f-number. In it I dedicated a section on the colour camera where I too arrive at a f-number twice the f-number for a monochrome camera. For a monochrome camera the f-number should be at least 3 times the pixel size in microns, for a colour camera this is 6 times. The article is in Dutch, but if you open it in Chrome it should translate reasonably well.

Strictly spoken the image of a colour camera should be reduced in size by 50% as half of the pixels that are produced by a colour camera are the result of interpolation.

I do planetary imaging myself using a C11 EdgeHD and ZWO ASI174MM (pixel size 5.86 micron) or ZWO ASI290MC (pixel size 2.9 micron). For both I image at f/20 (2 x PowerMate) and that seems to work well:

ASI174: 5.86 x 3 = 17.58
ASI290MC: 2.9 x 3 x 2 = 17.40

C11 with 2 x Barlow = f/20

Here is Mars by C11/2x PowerMate/ASI174MM in LRGB:

 

After the processing I have enlarged the image by 200% using bicubic interpolation and added another sharpening. You could go to f/30 or even f/40, but that will only enlarge the image (like the Mars image shown here), hardly add more detail (although some gain may be achieved from stacking at that f-number).

Nicolàs

Edited October 12, 2020 by inFINNity Deck

[CraigT82]

I've got a comment on oversampling that I hope someone cleverer than me might have some thoughts on...

I think the reason why you wouldn't want to oversample is that your image would be dimmer that at critical sampling and you would need to compensate by increasing exposure or gain which could be detrimental to the image.

However if you can oversample whilst keeping the exposures still short enough to 'freeze' the seeing and keeping the gain at a reasonably low level, then by all means do it. Modern planetary cameras are so sensitive that this is entirely feasible in practice. 

As an example, my most recent imaging run on Mars, using a 290mono with  2.5x barlow (f/19 = 6.5x pixel size = significantly oversampled) I was using 5ms exposures and only 9% gain on all three channels. If I had sampled 'correctly' at f/9, the increased brightness meant I would have needed to either run the gain at near zero and/or or use very short exposures, which as I learned from Vlaiv not too long ago, is bad in terms of read noise.

Any thoughts?!

[vlaiv]

2 minutes ago, CraigT82 said:

I've got a comment on oversampling that I hope someone cleverer than me might have some thoughts on...

I think the reason why you wouldn't want to oversample is that your image would be dimmer that at critical sampling and you would need to compensate by increasing exposure or gain which could be detrimental to the image.

However if you can oversample whilst keeping the exposures still short enough to 'freeze' the seeing and keeping the gain at a reasonably low level, then by all means do it. Modern planetary cameras are so sensitive that this is entirely feasible in practice. 

As an example, my most recent imaging run on Mars, using a 290mono with  2.5x barlow (f/19 = 6.5x pixel size = significantly oversampled) I was using 5ms exposures and only 9% gain on all three channels. If I had sampled 'correctly' at f/9, the increased brightness meant I would have needed to either run the gain at near zero and/or or use very short exposures, which as I learned from Vlaiv not too long ago, is bad in terms of read noise.

Any thoughts?!

I agree with most you said except for some things about the gain.

Gain in planetary imaging is only really used to control read noise - it does nothing to actual signal as it is multiplication factor. I find it hard to believe you have saturation at critical sampling rate on Mars. We can do the math, but I think that it turns out to be something like 50-100e per pixel per exposure of signal.

Cracking up the gain to get lower read noise is of course a good thing, but I don't think that you can easily saturate 12bit ADC with high gain. In fact you would need e/ADU of about 0.025 to do that.

In fact, we can check what sort of e/ADU you might have been using with high gain on this camera. ASI290 has unity gain at 110 and best read noise at gain 350:

This is 240 x 0.1dB difference and e/ADU doubles at every 61 x 0.1dB so we have roughly 2^4 times difference in e/ADU or gain 350 is roughly 1/16 e/ADU = 0.0625 or about 3 times lower than saturation value (for 100e per exposure).

We can also see that going from F/19 to F/11 will change intensity by factor of ~ x2.97 = x3.

Recorded signal will be x3 stronger - again, I doubt that it would saturate in 12 bit. It would likely saturate in 8bit mode though on high gain.

In another words - don't worry about saturating exposure when using critical sampling rate, even on high gain settings. If you do saturate, make sure you are using 12 bit (16 bit) mode and if that does not help, then yes, back away a bit with gain.

[inFINNity Deck]

Best is of course to keep an eye on the histogram, keeping it around 80%-90% filled.

As long as it does not significantly increases exposure times it does not harm to oversample. Better to over- than to undersample as the latter does reduces detail.

Regarding noise due to higher gain: even if the signal gets noisy by it, then the stacking will take care of it, provided we have enough data:

 

Seeing was not particular brilliant that evening, but I think the idea is clear... 😉

Nicolàs

Edited October 12, 2020 by inFINNity Deck

[Ken82]

Thanks for the comments all 👍

So asi 462mc and x2 powermate is the way to go. 
 

Nicolas how do you connect this all at the back of the scope ? Just 2” push fit or other ? Care to share any images of the connections at the back ? 
 

Thanks ken 

[vlaiv]

45 minutes ago, inFINNity Deck said:

Best is of course to keep an eye on the histogram, keeping it around 80%-90% filled.

Please don't do that. Histogram is simply useless when determining exposure length.

First of all - it is not absolute measure - it is relative measure. Absolutely same signal will produce two very different looking histograms if you do 8bit imaging vs 16 bit imaging. Signal that is at 80% of 8bit range (0.8 * 256 = 204.8) is going to be at 0.3125% of 16 bit range.

It is the same signal but using 8bit mode would put it in your recommended range, while using 16bit mode would make it "very very dim - almost nothing can be seen - very poor signal" range. But in reality it is the same signal and hence same SNR.

In lucky imaging exposure length is determined by coherence length and coherence time.

Even small change in exposure above coherence time will create more blur but hardly any improvement in SNR. For example if we are in 6ms exposure range for our scope and sky conditions, going with 10ms exposure will only improve our SNR per exposure for less than 20%.

However blur introduced by motion of atmosphere (lack of ability to freeze the seeing) will be much worse on final image. This is because we are trying to do frequency restoration and our frequency spectrum looks a bit like this:

This is how much certain frequencies get attenuated. Level of blur determines how "narrow" this curve is - it shifts it towards origin. Now look at slope of the curve around 0.05-0.1 range in X axis - only small shift in frequency causes attenuation to double.

Above coherence time blur becomes much more damaging than noise - you simply can't restore image if you need to multiply certain frequency with factor of x2, because blur attenuated it twice as much and you only improved your SNR by max 20%.

Coherence time is ultimate exposure limit in lucky imaging - going above it will only damage your ability to sharpen more than it will improve SNR of stack.

55 minutes ago, inFINNity Deck said:

As long as it does not significantly increases exposure times it does not harm to oversample. Better to over- than to undersample as the latter does reduces detail.

Actually it does. This would be true if we had zero read noise cameras and we don't.

I'll do a quick math to show above claim that I made that planets produce about max 100e per exposure on critical sampling. This will be very important for remainder of this argument.

Mars magnitude now is about -2.5 in best conditions (good transparency, high altitude and Mars highest towards zenith).

We will use 200mm scope (unobstructed, obstructed scopes just reduce signal, so do mirrors and so on). Mag 0 star produces about 880,000 photons on top of the atmosphere per cm squared per second.

Mars is x10 brighter than this (mag -2.5) so it will produce 8,800,000 photons per cm^2 per second.

200mm of aperture has 10^2 * PI = ~314 cm^ of surface, so it receives 2763200000 photons per second from Mars.

We sample at F/11 so our focal length is 2200mm and we use ASI290 so pixel size is 2.9um. Sampling rate is ~ 0.27"/px.

Mars is currently ~23" in diameter so it covers angular surface of roughly 415.5 arc second squares. Single pixel covers 0.27 x 0.27 = 0.0729 arc seconds squared. This means that image of Mars sampled at 0.27"/px will cover roughly 415.5 / 0.0729 = ~5700 pixels

Light captured by aperture is spread over that many pixels. This means that each pixel gets hit by 484772 photons per second. Average QE over visible spectrum is about 60%, so we get ~ 290863 electrons for visible spectrum, or about one third per color - ~ 96954 electrons.

Remember, this is per second. Let's say we use 6ms exposure and that is one 166.666...th of a second, so we need to divide this number with 166.6666 to get number of electrons on average per pixel per exposure, and that is 581.724.

Under best conditions, in greatest transparency with Mars at opposition with very high QE camera and unobstructed scope without any losses by mirrors and all at best we get around 580e of signal.

Ok, so I overestimated number by factor of x2-x3 since I've not done this calculation on Mars before, but I have for Jupiter and there signal per exposure is less than above (as it is both less bright and bigger in size).

In any case, let's get back to point - read noise. Read noise is very important thing to consider when doing planetary imaging as it right on the edge of becoming major factor.

Let's say that our system is capturing 256e per exposure (mirrors, central obstruction, filter QE, transparency issues). Let's also say that read noise is about 2e.

Read noise becomes an issue when it is higher than about 1/5th of dominant noise source. With 256e of signal we have 16e of shot noise. Read noise here is 1/8th of shot noise - so not really a concern. But let's over sample by using x2 barlow.

This will split same signal over 4 adjacent pixels. We have 64e per each of these pixels of signal and 8e of shot noise.

Now ratio of shot noise to read noise is no longer in "safe zone" - we now have that to be 1/4 - higher than ~1/5.

 

[inFINNity Deck]

Hi Vladimir,

thanks for your elaborate and interesting answer. I have to say that I image using FireCapture and it seems to handle histograms very well. As soon as I go over 100% the image clearly shows overexposed areas, so it seems to do its calculations in respect to the bit-depth of the image, but I could be wrong there.

Mars I never tried at f-ratios higher than f/20, but I will try next time at f20 and f/40, see how that relates to each other and to the theory.

Nicolàs

 

[vlaiv]

1 minute ago, inFINNity Deck said:

Hi Vladimir,

thanks for your elaborate and interesting answer. I have to say that I image using FireCapture and it seems to handle histograms very well. As soon as I go over 100% the image clearly shows overexposed areas, so it seems to do its calculations in respect to the bit-depth of the image, but I could be wrong there.

Mars I never tried at f-ratios higher than f/20, but I will try next time at f20 and f/40, see how that relates to each other and to the theory.

Nicolàs

 

Most software indeed does this - stretches histogram to full range of data - hence my example with 8bit vs 16bit. You do have a point here - histogram is somewhat useful for checking of the clipping - either to the left (too small offset) or to the right (over exposure), but stats window can do the same (min max value).

You should determine suitable F/ratio with respect to pixel size. ASI178 and ASI174 have very different pixel sizes - in fact 5.86 / 2.4 = about x2.4 of factor between them - meaning that for same sampling rate one would require F/ratio to be larger by factor of 2.4 over the other.

If you choose F/18 for ASI178 (color version for example) - you should use ~ F/43 for ASI174 (again color version).

[inFINNity Deck]

53 minutes ago, Ken82 said:

Nicolas how do you connect this all at the back of the scope ? Just 2” push fit or other ? Care to share any images of the connections at the back ?

Hi Ken,

like this:

From from focuser down to the camera I use:

- 2" TeleVue 2x PowerMate
- ZWO EFW mini filter wheel with LRGB filters
- a 1.25" clamp that allows me to rotate the ADC (although I could do the same by turning the filter wheel in the PowerMate or by turning the PowerMate in the focuser, but in this way I find it easier)
- a 1.25" ZWO ADC
- another 1.25" clamp that allows me to rotate the camera in respect to the ADC (although I could do that in the T2 to 2" adapter as well)
- a T2 to 2" adapter that allows me to swap the ASI174MM for the ASI290MC without the need to refocus
- the ASI290MC or ASI174MM camera

The T2 to 2" adapter is very useful for adjusting the ADC when you want to image in mono as even within the bandwidth of the monochromatic filters dispersion will occur. So when I image in mono I first attach the ASI290MC, adjust the ADC using ASICAP (please note that the FireCapture ADC tool is not functioning properly), and then swap the ASI290MC for the ASI174MM. Using this adapter causes the two cameras to have the same focus position. Before I used a T2 to 1.25 adapter that screws into the camera, but when doing so both cameras have different focus points and that affects the setting of the ADC again.

The combination of the 1.25" clamp and T2 to 2" adapter creates more distance between ADC and camera, which makes it easier to adjust the ADC due to the longer path.

Prior to imaging I slew the C11 to a close by star and collimate it.

Nicolàs

Edited October 12, 2020 by inFINNity Deck

[symmetal]

6 hours ago, vlaiv said:

Please don't do that. Histogram is simply useless when determining exposure length.

First of all - it is not absolute measure - it is relative measure. Absolutely same signal will produce two very different looking histograms if you do 8bit imaging vs 16 bit imaging. Signal that is at 80% of 8bit range (0.8 * 256 = 204.8) is going to be at 0.3125% of 16 bit range.

It is the same signal but using 8bit mode would put it in your recommended range, while using 16bit mode would make it "very very dim - almost nothing can be seen - very poor signal" range. But in reality it is the same signal and hence same SNR.

An 8 bit image discards the 8 least significant bits of the 16 bit image so the overall image brightness or histogram won't change. The histogram is very useful to ensure you don't clip the exposure. 🙂

Thanks for the detailed explanations though vlaiv. Very useful, and also to InFINNity Deck's article he linked to in his post, which helped explain the method used in planetary imaging to determine optimum image scale.

Alan

[vlaiv]

1 hour ago, symmetal said:

An 8 bit image discards the 8 least significant bits of the 16 bit image so the overall image brightness or histogram won't change.

 

You are in fact right. I just checked this with ASI185 and sharpcap.

I was under impression that following happens:

12 bit mode is recorded in 16bit mode by padding LSBs with zeros (4 LSB)

10 bit mode is recorded in 8bit mode by discarding 2LSBs

Since ASI says on their website that cameras operate in 12bit / 10bit mode: