Which base camera

[sinbad40]

Hi all,

 

So I have been using my old dslr attached to my 80ED Pro, but would like to get an astro camera . I've been looking at the zwo585 and the Altair hypercam 183c pro. Not much difference in price, but 1 does have a cooling fan and 4GB buffer, the other is a new sensor so could be better in the long run. Both can be used for guiding as well.

I want to do dso mainly but will also try planets. Don't want to go higher in price while I get back into it. I will read up on adapting my d80 for more light at some point and will use that for wide field.

All comments and advice more than welcome.

 

Thanks

 

Glenn

[The Lazy Astronomer]

If you're making your foray into dedicated astro cams, then definitely look for one with set point cooling - it is one of the main advantages of an astro cam, so I think it would be unwise to choose one without (assuming you're going down the long exposure DSO route, and not planetary or something more akin to EEVA).

[ONIKKINEN]

Both cameras you have in mind have tiny pixels, which you probably wont be able to make use of with your scope in DSO imaging (without binning), and for planetary imaging there are better and cheaper options available. You can of course bin to reach better sampling, but it would be better to have bigger pixels to begin with. The 183 on paper seems like a bit of an old tech camera in terms of its technical specs, so that's a hard sell in my opinion. The 2 cameras you have in mind also have much smaller sensors compared to any DSLR, so you would lose a lot of field of view to that. I dont think either camera would give you the improvements you are looking for when compared to DSLRs, other than much higher QE, but that coupled with smaller pixels could actually take you to the opposite way in terms of imaging speed.

At around the budget you have you could look into modifying your existing DSLR instead, or looking for a premodified one from the used market?

Edited June 22, 2022 by ONIKKINEN

[sinbad40]

Thanks both.

I have had my d80 around 15 years, and doesn't look to hard to remove the filters, so that should be good to do and not a major loss if it fails. I just wanted something that could do a bit of everything without breaking the bank, and be used as a guidescope when using dslr.  Will look for bigger pixels or binning may be the easy way even on a cheaper camera.

[scotty38]

I know you said you don't want to "spend more while you get into it" but that suggests, well we all know what it suggests so my 2p is that if it's remotely possible I'd save your money until you have enough for a 533MC. For all intents and purposes it could be all the camera anyone ever needs 🙂

[AstroMuni]

40 minutes ago, scotty38 said:

I know you said you don't want to "spend more while you get into it" but that suggests, well we all know what it suggests so my 2p is that if it's remotely possible I'd save your money until you have enough for a 533MC. For all intents and purposes it could be all the camera anyone ever needs 🙂

Would you rate the 533MC better than the 294MC ? I know it has no amp glow.

[Rossco72]

After a long absence, my 2p worth may actually only be worth 1/2p but I have just moved from a modified Canon DSL to a 533MM Pro.

I haven't had a clear sky since 🤨 but after reading up as much as I could, this seemed the next logical step from a DSLR to me.

[scotty38]

22 minutes ago, AstroMuni said:

Would you rate the 533MC better than the 294MC ? I know it has no amp glow.

Well IMHO amp glow is a non starter as it calibrates out BUT the 294 can be tricky to calibrate in general hence why I didn't mention it. I had one and and had a few troubles, most of my own making, and would have one again but for an easy life the newer 533 has to be the easy, fuss-free option that's the main reason I suggested it.

But, if you want a bigger sensor and have no problem dealing with it then I'd have no qualms with a 294 either.

Edited June 23, 2022 by scotty38

[Adam J]

11 hours ago, sinbad40 said:

Hi all,

 

So I have been using my old dslr attached to my 80ED Pro, but would like to get an astro camera . I've been looking at the zwo585 and the Altair hypercam 183c pro. Not much difference in price, but 1 does have a cooling fan and 4GB buffer, the other is a new sensor so could be better in the long run. Both can be used for guiding as well.

I want to do dso mainly but will also try planets. Don't want to go higher in price while I get back into it. I will read up on adapting my d80 for more light at some point and will use that for wide field.

All comments and advice more than welcome.

 

Thanks

 

Glenn

If you are planning on getting one without cooling then both the 533 and 585 have zero amp glow and very low dark current. I would not get a 183 it has amp glow and is difficult to calibrate without set point cooling. 

FOV will be limited with a 80ED on the 585 but there are plenty of smaller brighter objects you can have a go at. So if the 533 is out of budget certainly get the 585 over the 183. 

Adam 

 

 

[Adam J]

32 minutes ago, AstroMuni said:

Would you rate the 533MC better than the 294MC ? I know it has no amp glow.

Better is the wrong word. The 294MC is a pain to calibrate and really for that reason most people shout stay clear in my opinion, but if you willing to work with it then it is a bigger sensor. 

Adam 

[Adam J]

11 hours ago, ONIKKINEN said:

Both cameras you have in mind have tiny pixels, which you probably wont be able to make use of with your scope in DSO imaging (without binning), and for planetary imaging there are better and cheaper options available. You can of course bin to reach better sampling, but it would be better to have bigger pixels to begin with. The 183 on paper seems like a bit of an old tech camera in terms of its technical specs, so that's a hard sell in my opinion. The 2 cameras you have in mind also have much smaller sensors compared to any DSLR, so you would lose a lot of field of view to that. I dont think either camera would give you the improvements you are looking for when compared to DSLRs, other than much higher QE, but that coupled with smaller pixels could actually take you to the opposite way in terms of imaging speed.

At around the budget you have you could look into modifying your existing DSLR instead, or looking for a premodified one from the used market?

Remember that the resolution of a OSC camera is 1.5X less than a Mono of the same pixel size. 

Adam 

[Adam J]

11 hours ago, The Lazy Astronomer said:

If you're making your foray into dedicated astro cams, then definitely look for one with set point cooling - it is one of the main advantages of an astro cam, so I think it would be unwise to choose one without (assuming you're going down the long exposure DSO route, and not planetary or something more akin to EEVA).

Not 100% true now as dark current and amp glow on the most recent cameras is so low that set point cooling is no longer 100% required. The 533mm for example will work well at high gain and 60s exposures. With not much additional benefit to using cooling unless your ambient temperatures are very high at night so 20c+. Of course its always best to get one with cooling if you can but the performance delta from uncooled is just not what it used to be. 

Adam

Edited June 23, 2022 by Adam J

[vlaiv]

8 minutes ago, Adam J said:

Remember that the resolution of a OSC camera is 1.5X less than a Mono of the same pixel size. 

Adam 

How did you arrive at that figure - x1.5?

[Adam J]

1 hour ago, vlaiv said:

How did you arrive at that figure - x1.5?

Most quote half the resolution of a mono camera for OSC but this is wrong. 

Ill explain with a diagram at lunch.

Adam

[ONIKKINEN]

1 hour ago, Adam J said:

Most quote half the resolution of a mono camera for OSC but this is wrong. 

Ill explain with a diagram at lunch.

Adam

Something to do with higher resolution in green compared to red and blue no doubt? I was under the impression that the resolution of an OSC image is still only the resolution of a single channel (unless dithered every frame and bayer drizzled), but it just so happens that there are 2 green channels which just increases green SNR, but not really resolution. Anyway, very interested in seeing a technical explanation for this.

[Adam J]

2 hours ago, vlaiv said:

How did you arrive at that figure - x1.5?

 

[Adam J]

26 minutes ago, ONIKKINEN said:

Something to do with higher resolution in green compared to red and blue no doubt? I was under the impression that the resolution of an OSC image is still only the resolution of a single channel (unless dithered every frame and bayer drizzled), but it just so happens that there are 2 green channels which just increases green SNR, but not really resolution. Anyway, very interested in seeing a technical explanation for this.

Yes its to do with green pixels and the location of the interpolated pixel which is calculated for the centre of each four pixel RGGB group, incidentally also one of the reasons why sensors tend to have additional rows / cols of pixels over and above those that are declared / output. But also the shifting of the location of the virtual RGB pixel also leads to resolution being calculated on the diagonal separation and not the X/Y pixel separation. At least that is how I understand it. 

Adam 

[vlaiv]

@Adam J

Very interesting point - if you rotate grid by 45 degrees then green pixels are indeed spaced at 1.414 pixel sizes apart, however, no debayering algorithm exploits this.

In above image - "spatial position" of debayered pixel is marked wrong.

If interpolation debayering is used (like linear or VNG or other) then every pixel of the grid is position of debayered pixel - but that does not mean that frequency of samples is altered.

If super pixel / split debayering is used - then pixel position can be anywhere you like in 2x2 group - it does not matter as separation between resulting pixels is x2 pixel size.

No debayering algorithm rotates green and exploits the fact that they are spaced like that.

[Adam J]

24 minutes ago, vlaiv said:

@Adam J

Very interesting point - if you rotate grid by 45 degrees then green pixels are indeed spaced at 1.414 pixel sizes apart, however, no debayering algorithm exploits this.

In above image - "spatial position" of debayered pixel is marked wrong.

If interpolation debayering is used (like linear or VNG or other) then every pixel of the grid is position of debayered pixel - but that does not mean that frequency of samples is altered.

If super pixel / split debayering is used - then pixel position can be anywhere you like in 2x2 group - it does not matter as separation between resulting pixels is x2 pixel size.

No debayering algorithm rotates green and exploits the fact that they are spaced like that.

I think its to do with Nyquist sampling theory. I thought that this is exactly what happens when pixel luminace is calculated and that the luminace component to the pixel is calculated separately from RGB balance. 

I will look into it in more depth. 

Something interesting on Adaptive Homogeneity-Directed that apparently adjusts the plain of interpolation. In essence I believe this is doing as I have described? Seems to be used by Astropixel processor. In general its orientated to take advantage of the Green pixel grid orientation but adapts itself when edges are detected to reduce colour artefacts. 

All in all though we are getting off topic now. 

Adam 

Edited June 23, 2022 by Adam J

[vlaiv]

Just now, Adam J said:

I think its to do with Nyquist sampling theory. I though that this is exactly what happens when pixel luminace is calculated and that the luminace component to the pixel is calculated separately from RGB balance. 

I will look into it in more depth. 

Adam 

I think you are right.

Sampling itself does not depend on orientation and resulting image, interpolated or not. Interpolation itself won't increase detail beyond sampling rate, but in this particular case - no resampling happens, so green pixels that are captured remain the same.

As far as luminance goes - luminance is some linear combination of raw components (coefficients depend on actual sensor response) and as such will have high frequency components of green which will be missing in blue and red - so this arrangement acts as sort of low pass filter (actually aliasing artifacts will occur if there are high frequency components in red and blue as these will be under sampled).

 

 

[Adam J]

6 minutes ago, vlaiv said:

I think you are right.

Sampling itself does not depend on orientation and resulting image, interpolated or not. Interpolation itself won't increase detail beyond sampling rate, but in this particular case - no resampling happens, so green pixels that are captured remain the same.

As far as luminance goes - luminance is some linear combination of raw components (coefficients depend on actual sensor response) and as such will have high frequency components of green which will be missing in blue and red - so this arrangement acts as sort of low pass filter (actually aliasing artifacts will occur if there are high frequency components in red and blue as these will be under sampled).

 

 

See what I added about about AHD but also Green is weighted over red and blue in the case of the luminescence calculation, but this will no doubt be adaptive with relative RGB pixel intensity (cant interpolate a red image with a non existent green channel) and to an extend assumes that green with be the brightest RGB component (not always the case with astronomy), should work well for galaxy imaging mind you. Most modern algorithms are adaptive for that matter. 

I would be happy to modify my previous statement (and almost did originally) to say that resolvable detail in mono will be between 1.414 and 2.0 OSC resolution though. 

Adam

Edited June 23, 2022 by Adam J

[vlaiv]

In any case - OSC can be made to operate on "full" resolution implied by pixel size - in case of astronomical imaging where images are stacked if one uses Bayer drizzle instead of other debayering methods (provided of course one is not over sampling and that Bayer drizzle is implemented properly).

[Adam J]

2 minutes ago, vlaiv said:

In any case - OSC can be made to operate on "full" resolution implied by pixel size - in case of astronomical imaging where images are stacked if one uses Bayer drizzle instead of other debayering methods (provided of course one is not over sampling and that Bayer drizzle is implemented properly).

Yes of course but it reduces SNR and to be honest I have never seen it work convincingly, most likely because most people are oversampling relative to seeing / optics in any case. 

Adam

[vlaiv]

2 minutes ago, Adam J said:

See what I added about about AHD but also Green is weighted over red and blue in the case of the luminescence calculation, but this will no doubt be adaptive with relative RGB pixel intensity (cant interpolate a red image with a non existent green channel) and to an extend assumes that green with be the brightest RGB component (not always the case with astronomy), should work well for galaxy imaging mind you. Most modern algorithms are adaptive for that matter. 

Adam

Not sure how AHD works - but if we observe naive linear interpolation method - then colors don't mix in interpolation stage.

Missing red pixels are calculated by neighboring present red pixels - same for blue and same for green (with exception for green being that different pattern of surrounding pixels is used to calculate missing values).

[vlaiv]

1 minute ago, Adam J said:

Yes of course but it reduces SNR and to be honest I have never seen it work convincingly, most likely because most people are oversampling relative to seeing / optics in any case. 

Adam

I think that most convincing implementation is in AS!3. There is little if any difference in level of detail of planets in OSC vs mono if both are properly sampled and conditions are good.