Jump to content

  •  

CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.

Photo

Sony IMX183 mono test thread - ASI, QHY, etc.

  • Please log in to reply
997 replies to this topic

#101 dvalid

dvalid

    Ranger 4

  • -----
  • Posts: 322
  • Joined: 02 Apr 2014
  • Loc: Georgia

Posted 07 December 2017 - 06:35 PM

 

Jon, I know, this question has been asked many times, opinions are divided though. Your vision would be very valuable to me. 

 

Does the pixel size affect total exposure time? Does the sensor with smaller pixels need longer exposure to collect the same amount of signal (presume all other properties are equal)? 

 

It seems, the IMX183 is a very sensitive sensor and its tiny pixels are great for resolution, but how about exposure times? This question emerged once again to me after seeing a single 300 sec frame of ZWO294 camera. 

 

Thanks in advance.

 

David

On a non-normalized basis, yes. Absolutely, the pixel size affects exposure time. If you are going for resolution, then it will be a factor. This is actually a part of the reason why I am using 10 minute subs, vs. the 3 minute subs I used with the ASI1600. The smaller pixel size collects fewer photons per unit time, so to get the same pixel SNR you need longer subs. The other part of the reason is the higher read noise of the ASI183. It's 2.4e- vs. 1.3e- (for my narrow band work), so that, too, required me to increase exposure times.

 

Note that on a normalized basis, the differences in Q.E. are not as large (or rather, they remain unchanged as it's an instantaneous factor). They mostly boil down to the read noise difference. Also note that Q.E. affects exposure time, but the Q.E. difference isn't even going to account for a single stop in this case...84% vs. 60% (~1.4x), whereas the strait up read noise difference is a stop...2.4e- vs. 1.3e- (~1.9x), and the area-normalized read noise difference is over a stop...3.8e- vs. 1.3e- (~3x). The read noise, on a non-normalized and normalized basis, is the more significant factor.

 

Factoring in the 1.4x BENEFIT to the ASI183 due to Q.E. against the 3x DETRACTOR from the ASI183 due to read noise, plus accounting for glow and dark current, I came up with 7 minutes as the necessary exposures to get the same kind of SNR per sub as I was with the ASI1600 at 3 minutes.

 

I ended up shifting from a moderately high gain of 150-200, down to gain 53, in order to balance out my dynamic range as well. At gain 200, the ASI183 had less DR and more read noise (by a small amount, ~1.5e- vs. 1.3e-) than the ASI1600. While at a higher gain, I needed maybe 5 minute subs, I was clipping stars a bit too much. Moving to gain 53, I used 10 minute subs, but did not clip stars as much. It's always a balancing act. 

 

I shared a single 10 minute Ha sub earlier in the thread, in a comparison of calibrated vs. non-calibrated to show amp glow correction. Note that even though it is narrow band, I have extremely high LP, plus moonlight, plus haze...so it isn't the best of subs (few of my subs are any better these days....it's amazing how bad the LP has gotten around here, even 3nm filters suffer too much background contamination IMO). I did see the ASI294 single exposure, and I am pretty sure that guy is at a fairly dark site, deep yellow zone, maybe green zone. So take that into consideration. I have many 5, 7 and 10 minute single sub images from my green zone dark site that look just as good. wink.gif The power of a dark site is superior to just about any camera technology, by a long shot. 

 

Jon, thank you for the detailed answer. I do not want to hijack this thread, but this question is directly related to the ASI183 too.

 

To compare apples to apples, let's consider we have two hypothetical cameras with the equal sensor size, QE and the area-normalized read noise, but with different pixel size and therefore full well capacity. The pair of CCD's - ICX695 and ICX814 are in good approximation to this criteria. 

 

  3e RN per 4.54 µm pixel for ICX695 http://www.flicamera...eets/MLx695.pdf

  2e RN per 3.69 µm pixel for ICX814 http://www.flicamera...eets/MLx814.pdf

 

 

So, on one hand, the camera with the smaller pixels will collect less photons per time interval. On the other hand, the camera has less full well capacity and different gain setting, therefore we should get almost the same ADU distribution histogram and SNR (less RN!) at the end. Does this mean that the both cameras will reach SN ratio in equal time, but the one with smaller pixels would have higher resolution as a bonus?

 

I know, something is wrong with this conclusion, there are a lot of examples that demonstrate advantage of the larger pixels for collecting the signal faster. But what? 



#102 Jon Rista

Jon Rista

    ISS

  • *****
  • topic starter
  • Posts: 22916
  • Joined: 10 Jan 2014
  • Loc: Colorado

Posted 07 December 2017 - 09:06 PM

FWC only comes into play when we start talking about DR. Dynamic range is a ratio, FWC over RN. When discussing background SNR, we can effectively ignore FWC. So the only things that really matter are signal over noise. 

 

We also need to be clear about what SNR we are talking about. There is pixel SNR and object SNR. Assuming read noise scales with pixel size, the object SNR will be the same regardless of the pixel size, assuming the whole object fits within the sensor frame. Literal pixel SNR will be impacted by pixel size, however if read noise does indeed scale, we can normalize the image scale and that will eliminate any SNR differences. 

 

The key is the NON-NORMALIZED context, where you don't normalize image scale. In that case, even if read noise scales, the smaller pixel is going to acquire less signal, so pixel SNR will be lower. If we assume we have 2.5 micron and 5 micron pixels. The 2.5 micron pixels have 1.5e- read noise, the 5 micron pixels have 3e- read noise. We expose both sensors for the same time, the big pixel gets 27e- signal and the small pixels get 6.75e- signal:

 

SNR5 = 27/SQRT(27 + 3^2) = 27/6 = 4.5:1

SNR25 = 6.75/SQRT(6.75 + 1.5^2) = 6.75/3 = 2.25:1

 

The pixel SNR of the smaller pixels for a given exposure time is indeed smaller. But that isn't surprising, since it represents 1/4 the area of the object that the larger pixels see (assuming the same scope). If you combine four small pixels together:

 

SNR25 = (6.75 * 4)/SQRT(4 * (6.75 + 1.5^2)) = 27/SQRT(27 + 3^2) = 27/6 = 4.5:1

 

You still acquired the same amount of light on the object as a whole, and your SNR on an area-normalized basis is the same. 

 

There IS a tradeoff for higher resolution. You either divide the light reaching the sensor into smaller buckets, and have lower SNR in a given exposure time but higher resolution, or you combine the light into larger buckets that have higher SNR in a given exposure time but lower resolution. A bigger scope with bigger pixels doesn't necessarily buy you anything either, though. The tradeoff with a bigger scope is f-ratio...you get a longer focal length, but you usually increase f-ratio, so it's not really any different than using smaller pixels on the same smaller scope. This is basically what the etendue formula is saying...note that image scale involves focal length and pixel size, so if you factor image scale back into focal ratio and pixel size in the etendue formula, you end up with f-ratio and pixel size. The thing about etendue is, it doesn't account for read noise.

 

A^2 * s^2 * q = A^2 * (206.3 * p / f)^2 * q = (206.3 * p / F * SQRT(q))^2

 

The cost of resolution is usually lower pixel SNR unless you compensate for the lower flux per pixel.  With larger scopes, this often requires a significant increase in cost, especially for the camera/sensor to get large enough pixels. (Key exception is stars...a point source is defined as a source of light that is 1/4 the airy disc in size, meaning all the light from that point is focused onto the sensor...so star intensity grows with aperture as well as f-ratio. This differs from extended objects, which are only affected by f-ratio.)


Edited by Jon Rista, 07 December 2017 - 09:18 PM.

  • dvalid likes this

#103 dvalid

dvalid

    Ranger 4

  • -----
  • Posts: 322
  • Joined: 02 Apr 2014
  • Loc: Georgia

Posted 08 December 2017 - 02:59 AM

Ah yes, I took the noise from the signal itself out of the equation. Got it, thank you!


Edited by dvalid, 08 December 2017 - 03:46 PM.


#104 Jon Rista

Jon Rista

    ISS

  • *****
  • topic starter
  • Posts: 22916
  • Joined: 10 Jan 2014
  • Loc: Colorado

Posted 09 December 2017 - 02:01 AM

Quick test of the updated FGPA firmware.

 

Old Dark, 900s:

4rJhad7.jpg

 

New Dark, 900s:

XSLz2Aq.jpg

 

Solid improvement on the corner glows, and also removed a slight glow across the top edge. Slight reduction in the starburst glow. My stretches are extreme, and the glows look significantly smaller in the default SGP autostretch (more like Paul's example from #57).


  • h2ologg and calypsob like this

#105 Jon Rista

Jon Rista

    ISS

  • *****
  • topic starter
  • Posts: 22916
  • Joined: 10 Jan 2014
  • Loc: Colorado

Posted 09 December 2017 - 11:58 AM

It also appears as though the baseline dark current has been reduced considerably with the latest FPGA firmware. Below is a comparison of 1 hours worth of 900s darks before and after the update.

 

BEFORE, 4x900s darks:

88N5quY.jpg

 

AFTER, 4x900s darks:

iY3zUvu.jpg

 

This is actually very welcome. I was getting inconsistent and high results for dark current in my original measurements. I wonder if the copy of the camera I have was missing something in the firmware that controlled the dark current. Based on the results I'm getting now, dark current is about 40% lower than it was before!!


Edited by Jon Rista, 09 December 2017 - 12:12 PM.

  • h2ologg and Pug like this

#106 premk19

premk19

    Viking 1

  • *****
  • Posts: 635
  • Joined: 19 Apr 2015

Posted 09 December 2017 - 12:19 PM

Thanks for the update, Jon! I'm considering this camera strongly for LRGB imaging with my 115mm f/7 refractor for nights when my seeing allows it. Do you think I can get good SNR with ~80 5-min or shorter subs? As much as the starburst pattern is perfectly calibrated out, I'd like to keep it as minimal as possible, hence my interest in shorter subs. 



#107 suvowner

suvowner

    Ranger 4

  • *****
  • Posts: 316
  • Joined: 22 May 2016
  • Loc: Arkansas

Posted 09 December 2017 - 12:49 PM

from zwo facebook page:

 

Just confirmed that IMX183 and IMX294 sensor used AR coated glass, so don't need to worry about the halo on bright star

 

the 183mm pro available end of this month and will take pre orders soon.....

 

for hyperstar imaging at 425mm/F2.1 this is probably the camera to have.....i think hyperstar is a bit more susceptible to the bright halo, so the ar coated sensor glass and the smaller pixels seems to make this a better choice over the asi 1600 ??   


  • Jon Rista and Pug like this

#108 Jon Rista

Jon Rista

    ISS

  • *****
  • topic starter
  • Posts: 22916
  • Joined: 10 Jan 2014
  • Loc: Colorado

Posted 09 December 2017 - 01:09 PM

Thanks for the update, Jon! I'm considering this camera strongly for LRGB imaging with my 115mm f/7 refractor for nights when my seeing allows it. Do you think I can get good SNR with ~80 5-min or shorter subs? As much as the starburst pattern is perfectly calibrated out, I'd like to keep it as minimal as possible, hence my interest in shorter subs. 

Sure, I think 80-100 5 minute subs would be fine. You might be able to use unity gain (Gain 111) at f/7 as well. Remember that I am at f/4 with a 150mm aperture, so that means my stars get brighter a lot faster. Unity gain is nearly devoid of banding as well, so I think 5m subs @ g111 would be quite ideal for your setup.

 

 

from zwo facebook page:

 

Just confirmed that IMX183 and IMX294 sensor used AR coated glass, so don't need to worry about the halo on bright star

 

the 183mm pro available end of this month and will take pre orders soon.....

 

for hyperstar imaging at 425mm/F2.1 this is probably the camera to have.....i think hyperstar is a bit more susceptible to the bright halo, so the ar coated sensor glass and the smaller pixels seems to make this a better choice over the asi 1600 ??   

Hmm, interesting! I have indeed noticed some diffraction patterns with very bright stars, but I have not noticed the strong microlens shape that the ASI1600 has. I'll see if I can get some examples of the diffraction pattern. It is more of a starburst, with thin diffraction spikes radially around the star. It does take a BRIGHT star to do it, though.


  • premk19, DSO_Viewer and Pug like this

#109 Jon Rista

Jon Rista

    ISS

  • *****
  • topic starter
  • Posts: 22916
  • Joined: 10 Jan 2014
  • Loc: Colorado

Posted 09 December 2017 - 01:19 PM

Swamping the Read & Quantization Noise

 

Ok, so I wanted to share a quick example of what it means to "swamp the read noise" in general, and specifically with this camera, what it takes for shot noise to overpower quantization error.

 

Below is a set of comparisons between twilight flat frames (i actually had some stars in them, but I cropped out just a center empty portion of the field). The series included 0.001s exposures which were read noise limited, basically a bias frame with so little photon signal that it was basically read noise limited. I then acquired successive exposures, each one 10x longer than the last. I ended up only needing three of them, 0.01s and 0.1s exposures, as by the time I hit 0.1s exposures I was swamping the read noise squared by 33x! As you can see from the chart below, swamping the read noise squared by 4x is mostly sufficient to swamp both the read noise and mos of the quantization error. A 10xRN^2 exposure would be plenty to swamp quantization error entirely. You can see by 33xRN^2 (a 100x longer exposure than the read noise limited one) that shot noise completely dominates and is significantly larger in magnitude.

 

lM8odrz.jpg

 

Note that in the above charts, I have converted the ADU back to electrons, and normalized the offsets. The exact offset does not matter, however note that the scale on the left edge is in electrons. Peak to peak error within the read noise is only about 6e-. Standard deviation is less than 2e-. Peak to peak error in the 4x swamped signal is about 12e-, and standard deviation is just over 2e-. Peak to peak error in the 33x swamped signal is about 35e-, and the standard deviation is about 7e-. Note the standard deviation is denoted by the greenish-blue bar, and the median is the yellow line, in each graph. To the right side of each graph is the histogram plot for each signal.

 

While in a perfectly ideal world, totally swamping the quantization error with shot noise is preferred, with 12-bit cameras the general usage pattern is to stack lots of frames. By the time you have swamped the read noise squared by 4x, you have mostly eliminated the quantization error. By the time you stack 16-25 frames, quantization error would be entirely swamped. By the time you stack 80+ frames, it would effectively be eliminated.


Edited by Jon Rista, 09 December 2017 - 01:47 PM.

  • Pug likes this

#110 Jon Rista

Jon Rista

    ISS

  • *****
  • topic starter
  • Posts: 22916
  • Joined: 10 Jan 2014
  • Loc: Colorado

Posted 09 December 2017 - 01:57 PM

DFPN in Dark Frames

 

Another thing I wanted to demonstrate is FPN, in this case DFPN (dark fixed pattern noise). This is the tendency of the fixed structure of the sensor and each pixels non-uniform response to dark signal (DSNU) to lead to a growing signal that is fixed in time, but mostly random in space. As you can see in the animation below, as I stack more and more frames, the random noise (which is read noise and dark current and the dominant noise term in the first frame) averages down, while the FPN gets stronger and stronger. By 40 frames, the signal has started to converge on the underlying FPN pattern, and by 60 frames the pattern is fairly stable. There is some small variation in the noise, however the more you stack beyond this point, the more the DFPN will assert itself. This pattern includes both hot pixels, as well as spatially random noise.

 

peZUrm9.gif

 

This is actually the reason we create master darks, and the reason we calibrate with a master dark. Stacking many frames into a dark averages out the read noise, and leaves you with the DFPN. Once you have the modeled the DFPN in a master dark, it can be removed from the light frames without adding any meaningful amount of additional random noise. It should be noted that EVERY camera, CCD or otherwise, even the cleanest CCDs on earth, still have DFPN. Stack enough CCD frames, and you will have hot pixels, as well as spatially random but fixed in time noise. For those acquiring deep LRGB signals, DFPN is usually not an issue. For those chasing faint signals in narrow band data, DFPN can play a role in the SNR of your background sky and fainter signals. So good calibration is key. Dithering is also important, as sometimes you have additional patterns that are not entirely fixed, or perhaps not fixed at all...variable banding, for example. Dithering helps average out those semi-fixed or random patterns. I highly recommend both proper calibration as well as dithering with CMOS cameras (and really, with any camera, CCD included, for best results.)


  • calypsob, rockstarbill, sergtert and 1 other like this

#111 rockstarbill

rockstarbill

    Fly Me to the Moon

  • *****
  • Posts: 5660
  • Joined: 16 Jul 2013
  • Loc: Snohomish, WA

Posted 09 December 2017 - 03:34 PM

Excellent improvement! 



#112 rockstarbill

rockstarbill

    Fly Me to the Moon

  • *****
  • Posts: 5660
  • Joined: 16 Jul 2013
  • Loc: Snohomish, WA

Posted 09 December 2017 - 03:39 PM

from zwo facebook page:

 

Just confirmed that IMX183 and IMX294 sensor used AR coated glass, so don't need to worry about the halo on bright star

 

the 183mm pro available end of this month and will take pre orders soon.....

 

for hyperstar imaging at 425mm/F2.1 this is probably the camera to have.....i think hyperstar is a bit more susceptible to the bright halo, so the ar coated sensor glass and the smaller pixels seems to make this a better choice over the asi 1600 ??   

If these cameras have AR coated sensors, that is a strong reason to purchase one! 


  • Pug likes this

#113 Stamos

Stamos

    Vostok 1

  • -----
  • Posts: 145
  • Joined: 30 Oct 2015
  • Loc: Athens, Greece

Posted 10 December 2017 - 12:12 PM

This is indeed a very big improvement, especially the dark current reduction.

Seems like they have found a way to eliminate the corner-side glows but the starburst glow still dominates half the frame (i hope they will find a way reducing this too).

Excellent job Jon! 



#114 Jon Rista

Jon Rista

    ISS

  • *****
  • topic starter
  • Posts: 22916
  • Joined: 10 Jan 2014
  • Loc: Colorado

Posted 11 December 2017 - 01:56 AM

This is indeed a very big improvement, especially the dark current reduction.

Seems like they have found a way to eliminate the corner-side glows but the starburst glow still dominates half the frame (i hope they will find a way reducing this too).

Excellent job Jon! 

I think some of the improvement may have been due to temperature. The original darks were -15C, the new ones were -17C. I had actually set -20C, however apparently the ambient was just a bit too warm to achieve that. I doubt that a 2C difference is going to make up much of a difference, but the improvement in dark current is probably less than 40%. By my calculations based on a dark current chart ZWO gave me, I'd say the improvement is around 30% or so. Still a very nice improvement, but my previous value may have been a bit exaggerated due to the temp discrepancy.

 

As for the 10 minute examples. I have been having problems with my mount again lately. I am still trying to acquire enough data on the targets I've been imaging for a while here, and it's been difficult getting good data. I've lost several whole nights, and too many subs from other nights. My DEC axis is now ending up completely unresponsive after dithers sometimes...seems to be completely random, it's not cable tug, nor does it seem to be belt slip. I'm at a total loss, so I think next time the weather hits I'll be taking the whole thing apart down to every last individual piece and rebuild the whole thing from the ground up, with high performance grease, polishing the gears, and probably some mount casing modifications so I can properly tune the belts. Tonight hasn't been all that bad. I've lost a bit over a dozen 10 minute subs, so over two hours, but things seem to have finally settled into place and I may be able to get enough data to finish at least one of these objects. Resolution probably won't be all that great, sadly, but SNR should be good once I finally integrate all the data.


  • Pug likes this

#115 suvowner

suvowner

    Ranger 4

  • *****
  • Posts: 316
  • Joined: 22 May 2016
  • Loc: Arkansas

Posted 12 December 2017 - 11:07 AM

Jon do you think the improvement in dark current would allow you too push the gain a bit higher and get good results more like the asi 1600 ??



#116 Jon Rista

Jon Rista

    ISS

  • *****
  • topic starter
  • Posts: 22916
  • Joined: 10 Jan 2014
  • Loc: Colorado

Posted 12 December 2017 - 04:52 PM

Jon do you think the improvement in dark current would allow you too push the gain a bit higher and get good results more like the asi 1600 ??

Probably not. The main issue at the higher gains is the read noise, which bottoms out at ~1.5e-. At gain 173 (iirc), which is like gain 200 on the ASI1600, ~0.5e-/ADU, the read noise is nearly 1.8e-, which is higher than unity on the ASI1600. So using very short NB exposures with this camera just isn't as viable as with the ASI1600.

 

It is actually surprising how quickly a rise in read noise starts affecting your swamp factor. At 1.3e-, you only need a 17e- background sky signal to swamp by 10x. At 1.5e-, you need 23e-, at 1.8e- you need 33e-, at 2e- you need 40e-, at 2.4e- you need 58e-. So, at Gain 173 on teh IMX183, you need twice the exposure as on an ASI1600. Since I use 3 minute subs with the 1600, I would need 6 minute subs with the 183 (ignoring any other factors and other noise terms). I am using 10 minute subs at Gain 53, which has 2.4e- read noise, which is pretty much where I need to be.

 

Interestingly, my skies have darkened a bit lately, I'm middle of a red zone, and late night/early morning (if there is no moon) I have seen my SQMs get closer to orange zone border territory. My true background skies are ending up as little as 2-3xRN^2 on those nights. Interestingly, that is actually exactly what you would expect with a narrow band filter. Less skyfog noise, higher contrast, and any area of the sky that has no on-band signal at all will ultimately become read noise limited as your skyfog gets darker. My object signal areas also drop, but I am usually still swamping by about 4-5xRN^2. I haven't images with the ASI1600 for a while, so I don't know what kind of swamp factors I would have with it on similar nights. I suspect, though, that I would be swamping by less than 10x. 


  • rafiosso likes this

#117 akulapanam

akulapanam

    Mercury-Atlas

  • *****
  • Posts: 2505
  • Joined: 26 Aug 2012

Posted 13 December 2017 - 10:18 AM


Jon do you think the improvement in dark current would allow you too push the gain a bit higher and get good results more like the asi 1600 ??

Probably not. The main issue at the higher gains is the read noise, which bottoms out at ~1.5e-. At gain 173 (iirc), which is like gain 200 on the ASI1600, ~0.5e-/ADU, the read noise is nearly 1.8e-, which is higher than unity on the ASI1600. So using very short NB exposures with this camera just isn't as viable as with the ASI1600.

It is actually surprising how quickly a rise in read noise starts affecting your swamp factor. At 1.3e-, you only need a 17e- background sky signal to swamp by 10x. At 1.5e-, you need 23e-, at 1.8e- you need 33e-, at 2e- you need 40e-, at 2.4e- you need 58e-. So, at Gain 173 on teh IMX183, you need twice the exposure as on an ASI1600. Since I use 3 minute subs with the 1600, I would need 6 minute subs with the 183 (ignoring any other factors and other noise terms). I am using 10 minute subs at Gain 53, which has 2.4e- read noise, which is pretty much where I need to be.

Interestingly, my skies have darkened a bit lately, I'm middle of a red zone, and late night/early morning (if there is no moon) I have seen my SQMs get closer to orange zone border territory. My true background skies are ending up as little as 2-3xRN^2 on those nights. Interestingly, that is actually exactly what you would expect with a narrow band filter. Less skyfog noise, higher contrast, and any area of the sky that has no on-band signal at all will ultimately become read noise limited as your skyfog gets darker. My object signal areas also drop, but I am usually still swamping by about 4-5xRN^2. I haven't images with the ASI1600 for a while, so I don't know what kind of swamp factors I would have with it on similar nights. I suspect, though, that I would be swamping by less than 10x.

I’m not sure I agree with you. I suspect a lot of what you are seeing is the impact of smaller pixels. If you normalized and shot at the same arc sec resolution ,the 183s QE which is almost double the 1600 at ha , would offset the increase in read noise.

#118 suvowner

suvowner

    Ranger 4

  • *****
  • Posts: 316
  • Joined: 22 May 2016
  • Loc: Arkansas

Posted 13 December 2017 - 01:15 PM

 

 

Jon do you think the improvement in dark current would allow you too push the gain a bit higher and get good results more like the asi 1600 ??

Probably not. The main issue at the higher gains is the read noise, which bottoms out at ~1.5e-. At gain 173 (iirc), which is like gain 200 on the ASI1600, ~0.5e-/ADU, the read noise is nearly 1.8e-, which is higher than unity on the ASI1600. So using very short NB exposures with this camera just isn't as viable as with the ASI1600.

It is actually surprising how quickly a rise in read noise starts affecting your swamp factor. At 1.3e-, you only need a 17e- background sky signal to swamp by 10x. At 1.5e-, you need 23e-, at 1.8e- you need 33e-, at 2e- you need 40e-, at 2.4e- you need 58e-. So, at Gain 173 on teh IMX183, you need twice the exposure as on an ASI1600. Since I use 3 minute subs with the 1600, I would need 6 minute subs with the 183 (ignoring any other factors and other noise terms). I am using 10 minute subs at Gain 53, which has 2.4e- read noise, which is pretty much where I need to be.

Interestingly, my skies have darkened a bit lately, I'm middle of a red zone, and late night/early morning (if there is no moon) I have seen my SQMs get closer to orange zone border territory. My true background skies are ending up as little as 2-3xRN^2 on those nights. Interestingly, that is actually exactly what you would expect with a narrow band filter. Less skyfog noise, higher contrast, and any area of the sky that has no on-band signal at all will ultimately become read noise limited as your skyfog gets darker. My object signal areas also drop, but I am usually still swamping by about 4-5xRN^2. I haven't images with the ASI1600 for a while, so I don't know what kind of swamp factors I would have with it on similar nights. I suspect, though, that I would be swamping by less than 10x.

I’m not sure I agree with you. I suspect a lot of what you are seeing is the impact of smaller pixels. If you normalized and shot at the same arc sec resolution ,the 183s QE which is almost double the 1600 at ha , would offset the increase in read noise.

 

interesting point......with my cpc 800 on a wedge it seems my keeper rate falls off pretty quick past 3-4 minute subs, the asi 1600 would be no problem to stay well below that, not sure if the 183 would allow the shorter subs, I would be imaging at F2.1 vs Jons F/4 ....



#119 Jon Rista

Jon Rista

    ISS

  • *****
  • topic starter
  • Posts: 22916
  • Joined: 10 Jan 2014
  • Loc: Colorado

Posted 13 December 2017 - 07:21 PM

 

 

Jon do you think the improvement in dark current would allow you too push the gain a bit higher and get good results more like the asi 1600 ??

Probably not. The main issue at the higher gains is the read noise, which bottoms out at ~1.5e-. At gain 173 (iirc), which is like gain 200 on the ASI1600, ~0.5e-/ADU, the read noise is nearly 1.8e-, which is higher than unity on the ASI1600. So using very short NB exposures with this camera just isn't as viable as with the ASI1600.

It is actually surprising how quickly a rise in read noise starts affecting your swamp factor. At 1.3e-, you only need a 17e- background sky signal to swamp by 10x. At 1.5e-, you need 23e-, at 1.8e- you need 33e-, at 2e- you need 40e-, at 2.4e- you need 58e-. So, at Gain 173 on teh IMX183, you need twice the exposure as on an ASI1600. Since I use 3 minute subs with the 1600, I would need 6 minute subs with the 183 (ignoring any other factors and other noise terms). I am using 10 minute subs at Gain 53, which has 2.4e- read noise, which is pretty much where I need to be.

Interestingly, my skies have darkened a bit lately, I'm middle of a red zone, and late night/early morning (if there is no moon) I have seen my SQMs get closer to orange zone border territory. My true background skies are ending up as little as 2-3xRN^2 on those nights. Interestingly, that is actually exactly what you would expect with a narrow band filter. Less skyfog noise, higher contrast, and any area of the sky that has no on-band signal at all will ultimately become read noise limited as your skyfog gets darker. My object signal areas also drop, but I am usually still swamping by about 4-5xRN^2. I haven't images with the ASI1600 for a while, so I don't know what kind of swamp factors I would have with it on similar nights. I suspect, though, that I would be swamping by less than 10x.

I’m not sure I agree with you. I suspect a lot of what you are seeing is the impact of smaller pixels. If you normalized and shot at the same arc sec resolution ,the 183s QE which is almost double the 1600 at ha , would offset the increase in read noise.

 

Oh sure, pixel size is playing a role here. If you normalized the image scale, I agree.



#120 Jon Rista

Jon Rista

    ISS

  • *****
  • topic starter
  • Posts: 22916
  • Joined: 10 Jan 2014
  • Loc: Colorado

Posted 17 December 2017 - 02:11 PM

Received an update from ZWO on the ASI183. They are confirming that the peak Q.E. is 84%, and that Q.E. at the Ha line is 50%. Pretty amazing. These IMX183 cameras should be pretty phenomenal at short to moderate focal lengths because of that. 

 

Sounds like ZWO is preparing to get the ASI183 cameras to market very soon here as well. It appears as though QHY has already released their QHY183M, too! So if you were hoping to get one of these cameras for Christmas, sounds like it should be a possibility. 


Edited by Jon Rista, 17 December 2017 - 02:43 PM.

  • Pug likes this

#121 andysea

andysea

    Gemini

  • *****
  • Posts: 3376
  • Joined: 03 Sep 2010
  • Loc: Seattle, WA

Posted 17 December 2017 - 02:41 PM

Oh dear, I just bought a 294... now I need to shell out another 1.5k for this camera and accoutrements :) I'm still ahead compared to a fancy CCd I guess.


  • Pug likes this

#122 NorthField

NorthField

    Viking 1

  • *****
  • Posts: 848
  • Joined: 01 Jun 2017
  • Loc: SW Missouri

Posted 17 December 2017 - 02:46 PM

Think I’ll have a convo with my favorite vendor ( the one that says they’ll have it first ) tomorrow.. this will be my Christmas, even if it’s Easter...
  • Pug likes this

#123 DSO_Viewer

DSO_Viewer

    Viking 1

  • -----
  • Posts: 826
  • Joined: 03 Jan 2016

Posted 17 December 2017 - 03:21 PM

Received an update from ZWO on the ASI183. They are confirming that the peak Q.E. is 84%, and that Q.E. at the Ha line is 50%. Pretty amazing. These IMX183 cameras should be pretty phenomenal at short to moderate focal lengths because of that. 

 

Sounds like ZWO is preparing to get the ASI183 cameras to market very soon here as well. It appears as though QHY has already released their QHY183M, too! So if you were hoping to get one of these cameras for Christmas, sounds like it should be a possibility. 

How would this camera work in 2 x 2  bin mode using longer focal length scopes like my GSO 10 inch RC reduced to f5 for a 1270 mm focal length and my C9.25 reduced to f7 for a 1645 mm focal length. I could get flat and coma-free field of views since the sensor is only 16 mm diagonal. Thank you!

 

Steve


  • Pug likes this

#124 Jon Rista

Jon Rista

    ISS

  • *****
  • topic starter
  • Posts: 22916
  • Joined: 10 Jan 2014
  • Loc: Colorado

Posted 17 December 2017 - 04:34 PM

 

Received an update from ZWO on the ASI183. They are confirming that the peak Q.E. is 84%, and that Q.E. at the Ha line is 50%. Pretty amazing. These IMX183 cameras should be pretty phenomenal at short to moderate focal lengths because of that. 

 

Sounds like ZWO is preparing to get the ASI183 cameras to market very soon here as well. It appears as though QHY has already released their QHY183M, too! So if you were hoping to get one of these cameras for Christmas, sounds like it should be a possibility. 

How would this camera work in 2 x 2  bin mode using longer focal length scopes like my GSO 10 inch RC reduced to f5 for a 1270 mm focal length and my C9.25 reduced to f7 for a 1645 mm focal length. I could get flat and coma-free field of views since the sensor is only 16 mm diagonal. Thank you!

 

Steve

 

It should work just fine. Note that CMOS cameras won't have the read-noise reducing benefits of a CCD, however at a high gain (unity or maybe 173), you should still have a read noise advantage (at a higher gain, ~3e- total for the 2x2 set of pixels). You would end up with 5 megapixel images. I will say, even at 1645mm, unless you have good seeing a lot, the image scale of ~0.6"/px is going to be pretty small even with 4.8 micron pixels at 1645mm.


  • DSO_Viewer and Pug like this

#125 rockstarbill

rockstarbill

    Fly Me to the Moon

  • *****
  • Posts: 5660
  • Joined: 16 Jul 2013
  • Loc: Snohomish, WA

Posted 17 December 2017 - 05:43 PM

The really small pixels are tempting.


CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.


Recent Topics






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics