Mini review of ASI183mm Pro

[gregbradley]

I have been using astro CCD cameras now for about 12-14 years. Starting with an SBIG ST2000XM and then getting ripped off about $4,500 on Astromart on an ST10XME that never arrived. To the person's credit they did eventually repay me over several years (I think it took 3 years).

Then an STL11000XCM one shot colour which was lovely but a tad noisy on the dim areas so I got the STL11000m which was a fabulous camera and used by many in making images that are still top shelf by today's standards.

Then the largeKAF16803 cameras like Apogee U16M - a good camera with a ridiculously slow cooldown time (it would take 30 minutes to cool down and if the power supply got interrupted which it would very occasionally (I use a generator at my dark site) it took even longer as it would e slow to warm up (10-15 minutes) and then an additional 30 minutes to cool back down).

I upgraded to a FLI Proline 16803. My mainstay camera for the last 9 years or so. Hard to fault. Its big and heavy though. The large sensor gives a very widefield which creates a great look in images. No FLI Adaptive optics though and no FLI off axis guider either.

Then a FLI Microline 8300. The sensor was perfect, no flaws but I found I did not get very interesting images out of it. I see many excellent images from the QSI683 so QSI would be my choice now for a KAF8300 camera.

I got a Starlight Express 694 Trius which I used for several years. Very sensitive to narrowband and high QE plus low read noise. But I had trouble with non round stars with a bump in them and excessive fuzz around stars which in part was my processing flow but not something I encountered with these other SBIG and FLI cameras. I found out later the bump in the stars was corrected with a firmware update from Starlight Express.

Then a FLI microline 16 (KAF16200) which has been a very good camera overall as well.
I briefly had an SBIG ST8300 which was great for its cost but again the 8300 did not suit my scopes well.

Now recently I got an ASI183mm Pro to see what all the fuss about with these CMOS cameras. After studying the various models for a long time (there are so many models) I chose the 20mp, backside illuminated, 84%QE, 60%+ QE in Ha, low read noise ASI183mm Pro, $1750 at Testar.

First off this camera is tiny and weighs almost nothing. The Trius is small and compact but heavier (still light). Mine cools down 47C below ambient which is almost FLI like performance (50 to 55C). The images are very clean and the sensitivity is very high. Amp glow is there on exposures longer than about 2 minutes but so long as you have accurate darks this callibrates out well. Having gain and offset as 2 additional variables is a bit of a pain. But my FLI cameras have 2 download speeds, the slower one has the lowest read noise and its easy to forget to switch from the faster download speed when focusing to the slower download speed for imaging. Fortunately the download speed is able to specified in the Sky X imaging table.

QE is very high in narrowband and seems similar to the Trius 694 (Sony ICX694 CCD), perhaps even a bit higher. Read noise is about half of the Trius and the FLI ML16 (KAF16200), and nearly a quarter of the 16803.

The small pixels will test your optics, tilt and collimation as its very resolving but its quite a small camera so tilt is less of an issue than it is usually. The sensor seemed to be level as the tilt correction was the same for my other cameras which must be from my adapters.

One interesting aspect of the smaller pixels is that they create rounder stars as there are more pixels in every star. Quite noticeable.

The software supplied with the camera is very good and worked straight out of the box which is great. Most of the other cameras have had their little troubles at various times.

After watching a longish video about the differences of CMOS and CCD and exposure times it does seem shorter exposures are ideal for CMOS more so than CCD where longer exposures seem to rule.

I have been using 300 second gain 53 offset 8 for LRGB and either 300 or 600 seconds gain 111 offset 8 for narrowband and a conservative -10C which the camera achieves very quickly and easily and should all year round.
I could probably go -20 in the cooler months. Despite not getting strong benefits from cooling below a certain point (0 degrees perhaps) if it can cool lower then you may as well take it.

I don't see vertical lines in the sensor like you get in Kodak CCD sensors either. The sensor seems flawless in terms of defects.

Highest dynamic range is at gain 0 which I have yet to try out. Offset - all that does is shift the histogram to the right and prevents black clipping of data.

84% QE and perhaps higher is the most impressive thing about this sensor. It gets the image faster. So low read noise, super high QE, great narrowband sensitivity (over 60% QE for Ha) smallish but decent well depth (15,000 electrons is pretty small). Low cost - if this were a CCD with that performance (none of the usual CCDs have that) it would be many times more expensive. These ZWO cameras cost less than a DSLR.

File sizes are large as its a 20mp sensor and 39mb versus 32mb for the 16803 and similar for the 16200 cameras.

The 6200mm file sizes are 122mb -ouch. CCDstack always moans about not enough memory so its going to be a test.

Binning does not work on CMOS hardware but does have it with software binning. Its not as effective though and it 2X (or is it 4X?) the read noise.
It does improve the SNR though. I have not experimented with 2x2 much so far but don't visually see too much difference.

So overall. CMOS have much lower read noise, the backside illuminated ones (ASI183, 533,2600, 6200 sensors) have 84% QE for mono and still very high in colour. But there are only 3 mono versions - the 1600 (most popular but has microlense reflections), the 183mm (with small pixels)
the 6200 (the most expensive at around AUD$7500). The rest are all one shot colour which have their own advantages and disadvantages.

As far as ZWO versus QHY cameras go I am not in a position to comment but they seem similar. The main comment I read about that is in the past QHY had terrible software driver issues. As I said the ZWO software was the most seamless and easy implementation of any camera maker so far. The Sky X driver was a little hidden on their site but it worked well once found!

The 6200m and the APS one shot colour 2600 have zero amp glow, high QE are 16bit and low read noise but are out of stock everywhere.

Anyway these are my observations so far for those navigating these cameras.

Greg.

[JA]

Wow - what an incredibly detailed history Seems like the 183mm has something of a real niche and is something of a giant killer. You mentioned that exposures anything above 2 minutes need accurate darks, that begs the question :If you are lazy can you do without darks for less than say 1-2minutes?

Best
JA

[gregbradley]

Originally Posted by JA

Wow - what an incredibly detailed history Seems like the 183mm has something of a real niche and is something of a giant killer. You mentioned that exposures anything above 2 minutes need accurate darks, that begs the question :If you are lazy can you do without darks for less than say 1-2minutes?

Best
JA

I would have to experiment. But if you take an image and its a fairly decent exposure it seems to overwhelm the amp glow.
I just checked some 60 second exposures and there is no amp glow even when stretched except on the red subs but its the tiniest mount of amp glow and only when super stretched so really zero.

Exposures of 3 minutes would have only very slight amp glow but at the point you probably would want it removed. 2 minutes probably would still have no real amp glow.

I should add to my review a bit about the pluses and minuses of such small pixels.

2.4 micron pixels are so small that really to achieve the goal of 1 arc/sec per pixel, which is often considered an ideal image scale for most locations' seeing then that gives you something like an FSQ106ED 106mm aperture F5 and .99 arc sec per pixel.

In other words a shorter focal length widefield system.

On my AP Riccardi Honders 305mm F3.8 I am getting .43 arc seconds/pixel which is very low and would be considered oversampling. Oversampling consequences really are only a loss of some sensitivity. The plus is the potential for higher resolution if the seeing is good.

This is exactly what I have found. Firstly I find it quite hard to focus compared to my other cameras. The differences between shifted focus seem a lot less and it can be hard to tell if you improved the focus or not so I take bigger shifts in focus and compare those to narrow the in focus point more precisely.

But when the seeing is good I am finding it takes advantage of the good optics more than the other cameras and I am getting a bit more resolution on galaxies. But nights of poor seeing would be a waste of time as the stars would be too large.

Such a camera would be a waste on a large scope like the CDK17. It would be so far past being limited by the seeing that you would be wasting sensitivity too much.

In short it suits a shorter focal length setup best. 250mm F4 Newt images show that it shines with those.

Most APOs should do really well with it. Lens imaging would be very good, Redcat 51 should work really well with it.

The sensor is small so I get virtually no vignetting. I still use flats but you could get away with not using them.

These CMOS sensors do not like bias subtraction so no need for them either. Darks though do get rid of the amp glow when it does show itself so above 2-3 minutes darks are going to be needed.

So not a camera for everyone but in the right setup it would be ideal.

Greg.

[Camelopardalis (Dunk)]

Nice one Greg, glad you're liking it

Everytime I've tried raising the logic of high QE and low read noise, I've been virtually slapped down like the heretic I probably am

The 183 has very low dark current, so typically the shot noise is higher than the dark noise.

So far, I've resisted the temptation to go below gain 50, as below that quantisation noise becomes an issue...but short subs are the order of the day with my f/4 newt.

[Merlin66 (Ken)]

Interesting review.
I went for the ASI 183MM for solar imaging (SM60DS/ ED80/ x2.5 PM) and found "major issues" - ZWO accepted it back and I eventually went for the ASI 1600MM which has worked well for me.
From the ZWO site:
High Speed

Fast FPS can be used in solar(white light) and lunar imaging, as well as for live viewing/EAA.The high speed readout may also be used for real-time focusing, true lucky imaging of double stars and other small objects, planetary imaging of the major planets in the solar system, and much more.
Notice: depends on user’s feedback, this IMX183 sensor has grid pattern noise when do Ha solar imaging, please take flat frame for calibration(out of focus or add a barlow lens to take flat frames). We recommend our ASI1600/ASI174 mono camera for Ha solar imaging.

[gregbradley]

Originally Posted by Camelopardalis

Nice one Greg, glad you're liking it

Everytime I've tried raising the logic of high QE and low read noise, I've been virtually slapped down like the heretic I probably am

The 183 has very low dark current, so typically the shot noise is higher than the dark noise.

So far, I've resisted the temptation to go below gain 50, as below that quantisation noise becomes an issue...but short subs are the order of the day with my f/4 newt.

Thanks Dunk.

I haven't shot with the gain below 50 so I should try it as an experiment on a high dynamic range object and see how it goes. A lot of my uses of astro gear is from use rather than theory. The results are easy to see usually.

Greg.

[gregbradley]

Originally Posted by Camelopardalis

Nice one Greg, glad you're liking it

Everytime I've tried raising the logic of high QE and low read noise, I've been virtually slapped down like the heretic I probably am

The 183 has very low dark current, so typically the shot noise is higher than the dark noise.

So far, I've resisted the temptation to go below gain 50, as below that quantisation noise becomes an issue...but short subs are the order of the day with my f/4 newt.

Originally Posted by Merlin66

Interesting review.
I went for the ASI 183MM for solar imaging (SM60DS/ ED80/ x2.5 PM) and found "major issues" - ZWO accepted it back and I eventually went for the ASI 1600MM which has worked well for me.
From the ZWO site:
High Speed

Fast FPS can be used in solar(white light) and lunar imaging, as well as for live viewing/EAA.The high speed readout may also be used for real-time focusing, true lucky imaging of double stars and other small objects, planetary imaging of the major planets in the solar system, and much more.
Notice: depends on user’s feedback, this IMX183 sensor has grid pattern noise when do Ha solar imaging, please take flat frame for calibration(out of focus or add a barlow lens to take flat frames). We recommend our ASI1600/ASI174 mono camera for Ha solar imaging.

Interesting. I suppose that would be the same for similar architecture sensors like the 6200, the 2600 and the 533 cameras. All BSI small pixels.

Greg.

[Camelopardalis (Dunk)]

Hmmm that's interesting, wonder why that would be?

I've used my 183 for lunar imaging (at gain 50) and it works great.

Absolutely agree with your spirit of experimentation, Greg