Please help me choose a camera!

[Thalestris24]

I've been going around in circles trying to decide on a better camera for eaa/live stacking with either a 200mm lens or 80/384mm. I've been using basic guide cams which are ok in some respects but I feel I could do better . Anyway, I think I've narrowed it down to either a 290 mono or a 178 mono. The 2MP 290 has high qe (80%?) but has a 16:9 aspect ratio which I'm not so keen on. On the other hand, it's only £219.5 as a GPCAM 2 and could make a good guide cam. The ~6MP 178 looks another, possibly better option. It's 14 bit and has some roi options but slightly smaller pixels. I don't have an absolute qe figure, though. If anyone's used one, perhaps they could tell me what it's like in terms of sensitivity? Maybe compared to a typical mono guide cam. Thanks . The basic ZWO178mm is a bit more expensive at £342 but it's not a huge amount in the whole scheme of things. 
I'm open to suggestions but don't really want to spend more than the cost of the 178mm, unless it's for something absolutely wonderful! Of course, I've looked at the SX Ultrastar but I think it's too expensive for what it is!
Thanks for any thoughts/suggestions.

Louise

[Martin Meredith]

Hi Louise

According to this site, https://www.highpointscientific.com/telescope-accessories/astro-photography/ccd-cameras/zwo-cmos-cameras/zwo-asi178mm-monochrome-imaging-camera

the 178mm has a QE of 81%. It does look to be in the sweet spot for EAA especially with ROI and the option to bin those smallish pixels, but the FOV is still quite small (more like the Lodestar than the Ultrastar). I imagine its a good galaxy camera, esp. on nights of good seeing.

What are your main requirements besides sensitivity? Do you have a range of FOVs that you want to cover?

cheers

Martin

 

 

[Thalestris24]

1 minute ago, Martin Meredith said:

Hi Louise

According to this site, https://www.highpointscientific.com/telescope-accessories/astro-photography/ccd-cameras/zwo-cmos-cameras/zwo-asi178mm-monochrome-imaging-camera

the 178mm has a QE of 81%. It does look to be in the sweet spot for EAA especially with ROI and the option to bin those smallish pixels, but the FOV is still quite small (more like the Lodestar than the Ultrastar). I imagine its a good galaxy camera, esp. on nights of good seeing.

What are your main requirements besides sensitivity? Do you have a range of FOVs that you want to cover?

cheers

Martin

 

 

Hi Martin

Thanks for your reply! Um, I was thinking more in terms of focal length i.e. 200mm @ F4 and 384mm @ F4.8 and the two give me a decent range of fov with both cameras (roughly 1-2 deg). I doubt I'll be able to pick much galactic stuff up with my murky Bortle 9 skies, though I did get some stacks of the Needle last year.

ZWO also quote a qe of 81% for the 178mm, so probably not much in it compared to the 290. Looks like the 178mm has the edge... I'd love to hear from someone who's used an uncooled one for eaa/live stacking .

Thanks

Louise

[vlaiv]

Not easy one. I'm inclined to say 178, but for EAA applications 290 has slight advantage in lower read noise.

If we for example take "10bit ADC" (or shell I better put it - 1024 FW capacity rather than bit count for easier understanding) for short exposures and we compare read noise of two cameras

ASI178 will get there with gain setting of about 233 - read noise from published graph is about 1.4e

ASI290 will have that much FW with gain of about 232 (strange coincidence that these two number are so close, don't read much into it) - read noise is about 1.1e

It's clear that 290 has slight advantage in this regard - lower read noise allows you to get same SNR with shorter exposures (for same total integration time) - and that is very positive thing for EAA.

On the other hand, 178 has three times as many pixels - this allows for binning - either "on the fly" or afterwards to further boost SNR. It gives good resolution with your proposed setup and in general I find it quite ok sensor.

I don't have any examples of 290 for DSO, but here are two images taken with 178 - a bit different model, but comparable setup. Mine is color cooled model, and these were shot at 384mm F/4.8 (80mm F/6 with x0.79 FF/FR):

Both were "binned" (superpixel mode) and second one cropped because it was from a two night session (not perfect alignment). Sky was border of white / red zone - around sqm 18.5. First one is 4h of exposure under good conditions, second is 8h over two night and slightly worse conditions (a bit of fog/haze - transparency issues).

[Thalestris24]

Thanks Vlaiv

Yeah I think the 178 has the overall edge though I don't want to spend on a cooled one. I'm thinking in terms of 30s subs so I won't be live stacking any galaxies like yours, unfortunately. I might just get a vague hint of fuzziness... So I mostly go for star clusters. Atmosphere here and Bortle 9 light pollution are terrible for astronomy

Louise

[Thalestris24]

On the other hand, the 294c is (a more expensive!) option but the sensor spec is way better than all the others. But it's colour only and I've always had probs with colour sensors and light pollution... I dunno…

Louise

[vlaiv]

23 minutes ago, Thalestris24 said:

On the other hand, the 294c is (a more expensive!) option but the sensor spec is way better than all the others. But it's colour only and I've always had probs with colour sensors and light pollution... I dunno…

Louise

Such a large sensor is suited to larger scopes.

There is really good mono alternative that is between 178/290 and 294 in both size and price. What do you think about 183 mono sensor?

[Thalestris24]

1 minute ago, vlaiv said:

Such a large sensor is suited to larger scopes.

There is really good mono alternative that is between 178/290 and 294 in both size and price. What do you think about 183 mono sensor?

Yeah, the 294 is a biggish sensor but nice biggish pixels too - at 4.6um similar to my canon 1100d which has 5.2um pixels and 550d - 4.3um. I'll have to have a closer look at its various fov's and roi's and binning possibilities. I'm having so much trouble trying to make a decision! We've discussed the 183 before - and I've been mulling over it again today... Hopefully, by the end of the 21st century, I'll have made a decision ha ha. Yeah have been looking at the 178/290 as well. The good thing about the 294 is the high qe which is still high in the red end of the spectrum. QE drops off markedly at 650nm with the 178 and to some extent with the 183 also. The 183 could be binned but a simple roi using the 294 looks like it would be good. Decisions... Decisions...

Louise

[vlaiv]

I would not consider ROI to be particularly interesting feature for EAA, at least not with fast CMOS sensors.

By fast I mean - fast download times - USB 3.0 devices. Even USB 2.0 devices have much faster download times compared to CCDs where download time depends on pixel read clock (higher clock speeds increase read noise - so its useful for framing/focusing and such, but you want slow/steady readout for actual subs).

With CMOS sensors you can just read out full frame and crop it in software if you want ROI type of functionality. ROI is useful if you go for ultra fast readout needed for planetary imaging to achieve high FPS, so you need to download frame in 5ms or something like that. For EAA where you take each sub in dozens of second exposures (even for 1-2s exposures), downloading full frame that takes couple of hundred of milliseconds is no biggie (most sensors can achieve at least 10-15FPS on full frame).

Here is alternative way of looking at color sensors - which can help when deciding if it's worth using them over mono (and only benefit would be color recording / display in "single go").

Look at color sensors as being 4 different sensors "overlapping" - that is basically what you get with bayer matrix. You have one sensor with red filter, one with blue filter and two with green filters. Each of those sensors has (and this needs a bit of mental work):

1. twice lower sampling rate than a mono sensor with such pixel size would have

2. less pixel blur because of smaller pixels

(with mono sensors, sampling rate and pixel size is tied together, but with above view - sampling rate is as would be with twice larger pixels but pixel size remains the same)

3. Because of the above, each sensor has 1/4 of QE that sensor with larger pixels would have.

To put above in perspective, imagine you want to use OSC sensor in place of mono sensor. You have certain resolution for mono sensor set - let's say you want to go for 1.6"/px. What would be appropriate OSC sensor to achieve same sampling rate and how would it behave? It would be the one having twice the smaller pixels - that would make it "sample at 0.8arcsec/px". But in reality every color would be still sampling at 1.6"/px. If you need help seeing this - just think of sampling resolution not in terms of pixel size, but in terms of how much you need to move to get to next sample. First red pixel would be at 0", then comes next pixel at 0.8" but that is not red, so you skip it, then you move again (we are moving in X direction) and you get to next red pixel - that one is at 1.6", .... so first red pixel is at 0", next red pixel is at 1.6", next one is at 3.2", .... Actual sampling rate in red is 1.6"/px. Same goes for other channels, with green being considered G1 and G2 (it's the same response but we look at it as two different sensors).

Now look at pixel size. Mono sensor would have whole pixel surface 1.6" x 1.6" collect the light. But with above sensor - only 0.8" x 0.8" will collect "red" light, same for "blue" light - so it can be thought of each channel having 1/4 of QE of mono pixel.

I think this is better way to think about OSC sensors, particularly for EAA applications.

[Thalestris24]

15 minutes ago, vlaiv said:

I would not consider ROI to be particularly interesting feature for EAA, at least not with fast CMOS sensors.

By fast I mean - fast download times - USB 3.0 devices. Even USB 2.0 devices have much faster download times compared to CCDs where download time depends on pixel read clock (higher clock speeds increase read noise - so its useful for framing/focusing and such, but you want slow/steady readout for actual subs).

With CMOS sensors you can just read out full frame and crop it in software if you want ROI type of functionality. ROI is useful if you go for ultra fast readout needed for planetary imaging to achieve high FPS, so you need to download frame in 5ms or something like that. For EAA where you take each sub in dozens of second exposures (even for 1-2s exposures), downloading full frame that takes couple of hundred of milliseconds is no biggie (most sensors can achieve at least 10-15FPS on full frame).

Here is alternative way of looking at color sensors - which can help when deciding if it's worth using them over mono (and only benefit would be color recording / display in "single go").

Look at color sensors as being 4 different sensors "overlapping" - that is basically what you get with bayer matrix. You have one sensor with red filter, one with blue filter and two with green filters. Each of those sensors has (and this needs a bit of mental work):

1. twice lower sampling rate than a mono sensor with such pixel size would have

2. less pixel blur because of smaller pixels

(with mono sensors, sampling rate and pixel size is tied together, but with above view - sampling rate is as would be with twice larger pixels but pixel size remains the same)

3. Because of the above, each sensor has 1/4 of QE that sensor with larger pixels would have.

To put above in perspective, imagine you want to use OSC sensor in place of mono sensor. You have certain resolution for mono sensor set - let's say you want to go for 1.6"/px. What would be appropriate OSC sensor to achieve same sampling rate and how would it behave? It would be the one having twice the smaller pixels - that would make it "sample at 0.8arcsec/px". But in reality every color would be still sampling at 1.6"/px. If you need help seeing this - just think of sampling resolution not in terms of pixel size, but in terms of how much you need to move to get to next sample. First red pixel would be at 0", then comes next pixel at 0.8" but that is not red, so you skip it, then you move again (we are moving in X direction) and you get to next red pixel - that one is at 1.6", .... so first red pixel is at 0", next red pixel is at 1.6", next one is at 3.2", .... Actual sampling rate in red is 1.6"/px. Same goes for other channels, with green being considered G1 and G2 (it's the same response but we look at it as two different sensors).

Now look at pixel size. Mono sensor would have whole pixel surface 1.6" x 1.6" collect the light. But with above sensor - only 0.8" x 0.8" will collect "red" light, same for "blue" light - so it can be thought of each channel having 1/4 of QE of mono pixel.

I think this is better way to think about OSC sensors, particularly for EAA applications.

Thanks, Vlaiv, though you've gone over my head a bit there! For me, mono sensors are better since they aren't affected by lp in the way that colour ones are. My live stacking consists of maybe 5 or 10 minutes of 30s or 1min luminance subs (as I've posted here). That's it! Whatever I get with that is what it is. I've not done any live stacking with a colour camera as I don't have a suitable one (I only have dslrs and a qhy8L). If I can get a reasonable image of a cluster or other bright object using a sensitive colour camera then that will do! If I can somehow do that without a colour cast caused by lp, that would be better but I usually get lp artefacts with any long exposure attempts, even with narrowband. ROI seems to be useful for focusing and framing. Software binning is maybe also an option with a bigger sensor. I try to imagine what the outcome would be like with 10 or 20 binned 2x2, live stacked subs but I think I'll just have to try it and see . My current live stacking cams are only 1.3MP so too small for meaningful binning, I think. At the end of the day, I think I just want to live stack bigger pictures and without too much noise! If I can pick up some Ha nebulosity that would be good. Likewise with faint galaxies and stuff.

Louise

[Thalestris24]

The ZWO site pointed me to this review of the 183 mono - it's practically a book on it! Despite its shortcomings it might be the one for me after all. I'm not a perfectionist, I can live with a little noise, plus the qe at 650nm is better than my Atik 383l+. With my poor seeing I think it will only be useable at 384mm with binning 2x2 but that's ok. I'll, um, see what I still think in the morning...

Louise

[pete_l]

11 hours ago, vlaiv said:

I think this is better way to think about OSC sensors, particularly for EAA applications.

I don't disagree with any of your analysis. But a big attraction of EAA is to see the colours!
Also I believe that a lot of EAA is done for outreach, so the target would be the large, bright, attractive, astronomical targets.

When we choose to dive deep into more "interesting" objects astronomically speaking, we would use a different setup that is more capable of capturing the photons and that doesn't require short, real-time duration subs.

[vlaiv]

2 hours ago, pete_l said:

I don't disagree with any of your analysis. But a big attraction of EAA is to see the colours!
Also I believe that a lot of EAA is done for outreach, so the target would be the large, bright, attractive, astronomical targets.

When we choose to dive deep into more "interesting" objects astronomically speaking, we would use a different setup that is more capable of capturing the photons and that doesn't require short, real-time duration subs.

There is a nice way to go about this, although it involves larger budget - dual setup. One to capture mono data at full resolution, and one to capture color data at half the resolution. This effectively means same pixel size in both mono and OSC - same cameras one color, one mono. This way you get the speed of mono setup, and still provide color. Human eye is less sensitive to noise in color, so images will still look pleasing.

Other way one might approach outreach is to do seriously aggressive F/ratio reduction. This will distort image somewhat on most systems, so stars will be less than perfect, but for outreach and quick look-see that shouldn't really matter as much.

For this latter case, I think that eyepiece projection and small sensors can be good approach. I'm going to try to do something like that myself. Turn 8" F/8 scope into ~F/2.5 system and see what the image looks like.

[Martin Meredith]

Perhaps getting off Louise's original topic, but a further alternative to the mono/OSC outreach situation is to use mono + filters and rely on automation (live scripts) to do the right thing. One approach would be to present the images coming from coloured filters as monochrome at first (building up the stack as usual), then 'reveal' the effect of colour. I've done this myself and it is a bit of a 'wow' moment to see colour appear in this way. It works really well with open clusters. There's usually plenty to say about the mono image building up anyway and it benefits from not having the distraction of the inevitable colour noise that one gets in short exposures.

All that is lacking at present is software to make this process easy (a single click should be enough). One current solution that I've been looking at is to use Nebulosity in scripting mode (i.e. driving Neb from another program).

Martin

[London_David]

I agree with everything above, but if you're looking to break the indecision... buy the ASI290mono.

I have an ASI290 mono and the ASI294Pro OSC and I like them both a lot. I bought the 290 first because of the cost, then the 294 because I thought I was missing out on the colour.

However, I've discovered... I wasn't. While colour is great, I find the mono is actually more exciting for EAA because the camera is more sensitive. You see more, faster. So if I want a quick look at the sky I grab the 290. If I want to see deeper things I grab the 290. If I want a look at a nebula or extended objects and I want to see colour -- I set up the 294 and prepare for more computer fiddling than with the mono which is annoying at 2 am. That said I'm most interested in galaxies and deeper objects -- so your interests may differ.

Overall I find EAA in mono more interesting because I get better results faster and easier -- and at 2am that has actually become more important for me than I thought it would. As @vlaiv said, mono noise is a lot less distracting and intrusive than colour noise. Colour can be a pain to make look good. I'm also tending to do more very short exposure EAA (0.1s - 10s exposures, usually around 2s) and live stack on galaxies since it can fight bad seeing. I live in Zone 2 London, so my LP is terrible too and I have no issues with short exposures. Of course I'd rather be at a dark site where I'd get far better results, but I can still observe -- and the mono gives me the easier option of going the narrowband route to cut LP with filters.

I don't want to oversell mono particularly. But having both a mono and a colour, and originally thinking I would prefer colour... I actually use my more sensitive mono camera far more because its faster and more fun.

Technically for me (and this does repeat more detailed info above) -- I haven't used the specific cameras your looking at so a broader answer to your original question: Check out your field of view on http://astronomy.tools/calculators/field_of_view/ with cameras and objects your interested in to see what you want to look at and how they might work with your telescope, don't worry about pixel size with CMOS because of binning, go for the camera with the lowest read noise, and highest QE you can get that fits what you want to look at in terms of sensor size.

I'm not really up to date on the exact cameras... but between a 290 sensor and a 178, I'd go for the 290, partly due to the relative per micron noise on the 178 (from a quick reading of the Jon Rista CN review).

Also -- I've been super happy with my ASI290mono purchase.

[Thalestris24]

Hi Guys

Thanks for your input  However, I'm not into outreach - I just tinker around for my own interest and from within my flat via open windows. As such, I have a very limited view of the sky and a very limited time on a limited number of targets in the east. Nevertheless, I have multiple setups but I've gravitated towards live stacking (rather than eaa as such) over the last 6 months. I live near a city centre and suffer from dreadful light pollution and in a place where the atmosphere and weather are both terrible! Still, I've found that live stacking with SharpCap Pro is a good compromise for me and mono/luminance imaging helps hide the lp. To date, I've been live stacking using a qhy Minicam 5s (cooled qhy5l-ii mono) and an uncooled gpcam v1 mono i.e. normally just a guide cam. I've had some reasonable results using SharpCap but would just like a better camera with more scope, so to speak. As I've mentioned, I'm using a 200mm Takumar lens (on an eq3 pro) and a TS Photoline f6 (=f4.8) 80mm at 384mm (on a Heq5). Both are good for what I'm trying to do. I'm not aiming to produce works of art, just some basic images. It's just a hobby... I also have a 115mm photoline triplet with a mono Atik 383l+ on an AVX. I've not used it in a while but it's there to do long exposure (has to be with the slow Atik!) either just luminance or H-alpha. So that's where I am and where I'm coming from. The cooled 183m seems to be a reasonably good camera albeit with some flaws. I've read Christian Buil's analysis (he gives it a lower qe at Halpha than zwo or qhy which is a concern but not the end of the world) as well as Jon Rista's long review. I'd considered the 294mc which has a very good spec but is colour and I know from experience that results with colour will not be good here. Mono is simple, clean - and quick and gives the best resolution and best images! I've looked at the 290 and 178 but, on reflection, I think the 183 will give me versatility and (software) binning options. It's pretty much like a bigger 178.  So there you go! I've not pressed any buttons yet just in case I change my mind! The qhy183 might be the one - I'm not 100% sure how they compare in practice but I've got the impression that the qhy has better amp glo suppression. With short exposures it shouldn't be an issue but maybe I'll do long exposures sometimes
 

*Just saw your post, David - thanks!

Louise

[Thalestris24]

Oh well, I've done it now! Life's too short to waste time being indecisive! Have ordered a qhy183m today. Fingers crossed it isn't an altogether bad decision! I think it should be fine for short exposure imaging/live stacking especially with binning and/or ROI. Well, time will tell, lol.

Louise

[vlaiv]

45 minutes ago, Thalestris24 said:

Oh well, I've done it now! Life's too short to waste time being indecisive! Have ordered a qhy183m today. Fingers crossed it isn't an altogether bad decision! I think it should be fine for short exposure imaging/live stacking especially with binning and/or ROI. Well, time will tell, lol.

Louise

I think it's very good choice. I also view it as 178 with larger chip size.

[Thalestris24]

56 minutes ago, vlaiv said:

I think it's very good choice. I also view it as 178 with larger chip size.

Thanks for all your input and analysis, Vlaiv! Yeah, it's similar to the 178 but may have higher qe at 656nm though only 12bits, of course. I'll find the practical truth of it soon enough! I wonder how difficult it is to measure effective qe/sensitivity  with lasers - I have 3, violet (405nm), green (532nm) and red (650nm). Obviously measuring absolute qe is out of the question but I ought to be able to do an effective relative sensitivity measurement at the different wavelengths just by measuring brightness of a number of exposures. I just need a way of dispersing the laser light - small lens, maybe. I'd be assuming the light intensity of the 3 lasers was the same, and that's probably not true, but they might be close enough to give a meaningful result. On the other hand, maybe it doesn't really matter ha ha

The 183 should be shipped tomorrow so will have it by the end of the week. I'd better check if I need and fittings/extension tubes etc.

Louise

[vlaiv]

34 minutes ago, Thalestris24 said:

Thanks for all your input and analysis, Vlaiv! Yeah, it's similar to the 178 but may have higher qe at 656nm though only 12bits, of course. I'll find the practical truth of it soon enough! I wonder how difficult it is to measure effective qe/sensitivity  with lasers - I have 3, violet (405nm), green (532nm) and red (650nm). Obviously measuring absolute qe is out of the question but I ought to be able to do an effective relative sensitivity measurement at the different wavelengths just by measuring brightness of a number of exposures. I just need a way of dispersing the laser light - small lens, maybe. I'd be assuming the light intensity of the 3 lasers was the same, and that's probably not true, but they might be close enough to give a meaningful result. On the other hand, maybe it doesn't really matter ha ha

The 183 should be shipped tomorrow so will have it by the end of the week. I'd better check if I need and fittings/extension tubes etc.

Louise

If you want to measure relative sensitivity, maybe easiest way to do it would be with StarAnalyzer?

Just take camera, do a spectrum of a known star - one that you have reference spectrum of and do instrument response - this will give you relative sensitivity curve of your setup - camera + scope (which is what you want really).

[Thalestris24]

1 minute ago, vlaiv said:

If you want to measure relative sensitivity, maybe easiest way to do it would be with StarAnalyzer?

Just take camera, do a spectrum of a known star - one that you have reference spectrum of and do instrument response - this will give you relative sensitivity curve of your setup - camera + scope (which is what you want really).

Yeah, maybe could do that but I'd need a clear sky first and I haven't had one for the best part of two months...

[Fraunhoffer]

On 26/03/2019 at 17:24, Thalestris24 said:

Thanks for all your input and analysis, Vlaiv! Yeah, it's similar to the 178 but may have higher qe at 656nm though only 12bits, of course. I'll find the practical truth of it soon enough! I wonder how difficult it is to measure effective qe/sensitivity  with lasers - I have 3, violet (405nm), green (532nm) and red (650nm). Obviously measuring absolute qe is out of the question but I ought to be able to do an effective relative sensitivity measurement at the different wavelengths just by measuring brightness of a number of exposures. I just need a way of dispersing the laser light - small lens, maybe. I'd be assuming the light intensity of the 3 lasers was the same, and that's probably not true, but they might be close enough to give a meaningful result. On the other hand, maybe it doesn't really matter ha ha

The 183 should be shipped tomorrow so will have it by the end of the week. I'd better check if I need and fittings/extension tubes etc.

Louise

Hello, Thanks for posting this and also all the folks that have replied.
Ive been looking at the mono 178and 183 based cameras and been in orbit trying to decide. I also have some quite bad skies.

(I like the idea of a larger fov on the 183 that would help me locate things with my non-goto.)

Has it arrived ? How have you got on with the 183?

Regards

 

[Thalestris24]

2 minutes ago, Fraunhoffer said:

Hello, Thanks for posting this and also all the folks that have replied.
Ive been looking at the mono 178and 183 based cameras and been in orbit trying to decide. I also have some quite bad skies.

(I like the idea of a larger fov on the 183 that would help me locate things with my non-goto.)

Has it arrived ? How have you got on with the 183?

Regards

 

Hi

Yes, I've been using the qhy183m - only a few times so far, cos of the weather... I like it very much  and it has a high QE. FOV depends on your focal length together with size of sensor. The 183M sensor is quite small. I've been using it on a reduced f6 80mm APO - at 384mm fl. I've also been using it bin 2x2 which gives me 2772 x 1842 px. (5Mpx) It has small pixels (2.4um) so bin 2 effectively doubles them up (4.8 x 4.8um) and overall noise is still low. I have problems with LP and gradients but here's a cropped M3 I did last week captured with SharpCapPro live stacking (Bortle 9 , -10 deg C, unity gain):

It is usb 3.0 so fast image downloads.

Hth

Louise