Skip to main content

16-bit CMOS upgrade - best choice for photometry? (IMX 455)

11 posts / 0 new
Last post
Lbew
Lbew's picture
16-bit CMOS upgrade - best choice for photometry? (IMX 455)

Hello!

I've been looking to upgrade to a monochrome camera for astrophotography for a while now. I currently use a QHY-5 178 CMOS colour camera by QHY (IMX178 sensor), which does have low read noise, however the field of view is tiny (0.33° * 0.22°) and it oversamples. Excellent choice for lucky imaging of planetary objects or planetary nebulae, not so much for most deep sky objects.

The observations I've done in the last months with the colour CMOS camera show a result (most recent one being a TrES-1b transit), yet it's not of great use for photometry. My thinking is that if I'm looking to purchase a cooled monochrome camera anyways, and if doing so would greatly benefit photometry, I might as well purchase one now instead of waiting for two years or so.

An important factor for me is sustainability. I expect that I'll switch telescopes or components within the next 10 years, so a quality camera which will last for that period and still be viable with new equipment (and higher expectations) is a better investment now compared to a cheaper camera which just gets the job done with the current equipment. 

I've been looking at the ZWO ASI6200mm Pro (IMX455) as the camera of choice. To me it seems to be an excellent choice for astrophotography and importantly photometry in combination with a filter (huge field of view with 0.61" resolution!). The full specifications are available here. Important is a 9576*6388 resolution, 16-bit ADC, 3,76um pixel size, 80% QE, 1.2e(-) read noise, no amp glow, and 51400e(-) full well capacity. There's also a section on the website on "ASI6200 VS KAI1100", that has been my best comparison reference so far. 

Update: I decided to for the QHY600 PH over the ASI6200mm Pro in the end. 

Should this CMOS camera be treated like a CCD? The photometry guide speaks of "perculiarities" for each camera. What kind can I expect when using a CMOS compared to the more common CCD for photometry?

Currently this sensor looks like it offers the best price-to-performance especially when compared to the prices of CCDs with similar specifications. I currently use a 203.2mm (8 inch) Schmidt-Cassegrain telescope, with a 0.63x reducer by Starizona with field correction. My telescope's focal length is 1280mm. It sits on the EQ6-R mount (guided) which has no tracking issues, and has perfect Go-To with astrometric solving. 

Leon 

 

msheald
16-bit CMOS upgrade - best choice for photometry? (IMX 455)

Hello! One thing t keep in mind is that the 8" SCT won't be able to fully use the 42mm diagonal of the IMX455 chip.

    I have a meade 8"SCT that I run at 0.6X reduction, and even my SBIG ST-402 vignettes with its FOV of 14x20 arcminutes.

    If you plan to use the scope long term, then a smaller chip might be a better match for the scope and perform just as well as the IMX455 (and be less expensive!). However, if you want to purchase the IMX455 as the first step in a complete system upgrade in the coming years, go for it.

    I suspect that we'll se a number of CMOS cameras coming on-line in the near future, so photometry choices are likely to expland so that we can match our scopes and photometry interests. Best regards.

Mike

Bikeman
Bikeman's picture
Also the photometric filters

Also the photometric filters & filter changer  would have to be compatible with the size of the sensor. 

I wish there was a sensor that had all the goodies of the IMX455 (same bit depth, full well, pixel size...) and just ~1/4 the area, so around 15 Mpix  (would not mind if the price tag would scale as well :-)  ). That would be a great choice for a lot of us I guess, like a ca. "Four Thirds" format variant of this sensor perhaps.....they could even take partly damaged ones and mask the cropped-away part :-) no problem with that. 

There is already an APS-C format astro-camera fitting this wishlist based on the IMX571 sensor....but AFAIK only in color versions. Something like the ZWO ASI 2600MC pro, but in monochrome, would be a better match for most of us wrt our telescopes I guess.

Does anybody here have insights if/when this will happen?

 

Cheers

HB

kj4nus
CMOS upgrade

May I suggest you take the CCD1 course. There you will find the answers you seek.   You are considering a $5500.00 camera, what's $35.00 ?   OK, with that said lets discuss a few points.Your mount EQ6R cost $1400.00 and the OTA $1000.00. The saying goes: spend as much of the budget as possible on the mount. It will become the weak link in the upgrade ladder. If you're going into photometry and producing science research grade data then a different approach might be prudent.  Lets look at the ZWO 6200 :  small pixels ,that's a problem if you want to optimize the huge pixel array. It only takes 2-3 pixels per star to get the info for your selected star. You would waste that expensive resolution. Did you realize the filter wheel uses the much more expensive 2 inch filters? And as mentioned earlier the OTA back has to accomodate an unusually large sensor without obstructing/vinetting which drives the price way up. Also did you review the manufactures choice of optics to pair with the camera?  A fast modified newtonian.  F 2.8. $2000.00. No f10 SCT"s here. Therefore limited selection of optical choices. In the $5000 to $6000 range some APO's may work, but they are F4-5. doubling your exposure times. Did you see the exposres were a dozen 900 second expusures.Will the EQ6R handle that with the added weight of camera and filterwheel and extra length?  The 6200 is an amazing camera BUT it is only one of several components all of which must operate in harmony. That matched system yields glorious wde field images which unlike smaller arrays enables printed enlargements.  You are not alone we have all been there.    Optoins would be a camera with larger pixels upto 9um ,a more modest array size. Your EQ6R is limited to a 44lb load capacity so a larger heavier SCT won't match however an APO say 100mm aperature might work well for photometry. Plenty of choices here starting at $2000.     Perhaps select telescope first then the camera.You will likely have a far better match. In closing I feel compelled to mention the huge benefit derived from working with SBIG/DIFRACTION LIMITED They sell, service, support  provide documentation and tech support for science grade cameras. They include custom software integrating camera and computer. I have ZWO QHY and SBIG. There is a reason why so many photometrist, researchers and universities use SBIG..Did you know they have an adaptive optics upgrade umder $1000.They warrant a look.Clear skys keep us posted. steve

Lbew
Lbew's picture
16-bit CMOS upgrade - best choice for photometry? (IMX 455)

Thank you for the detailed replies in such a short time frame! I will respond here

Your mount EQ6R cost $1400.00 and the OTA $1000.00. The saying goes: spend as much of the budget as possible on the mount. It will become the weak link in the upgrade ladder

That is true, however the EQ6-R offered me the best payload to price ratio. I did not find a comparable mount (for a reasonable price) which would be capable of holding a 14" SCT or so. I currently cannot really predict how large of a telescope will be pratical in future - would it make sense to get a 14" telescope, or (provided there is space) would the money be saved for something larger later down the line? At that point I'd sell the EQ6-R; when buying it was an issue of practicability vs. future compatibility, and this is not easy to predict. I think the camera would be more applicable across telescopes, especially with binning, but still keep the excellent sensor specifications

 

Lets look at the ZWO 6200 :  small pixels ,that's a problem if you want to optimize the huge pixel array. It only takes 2-3 pixels per star to get the info for your selected star. You would waste that expensive resolution. Did you realize the filter wheel uses the much more expensive 2 inch filters?

The small pixels would mean excellent sampling for my telescope regarding photography/pretty pictures and astrometry (asteroid tracking for example), if seeing is good (1-2" FWHM).
I agree (only from what I've read, you speak from experience here) that it would not be the best choice for photometry in terms of pixel size, and binning is an option to save file space, but the small pixels are of great use in photography and astrometry. 

And as mentioned earlier the OTA back has to accommodate an unusually large sensor without obstructing/vinetting which drives the price way up. Also did you review the manufactures choice of optics to pair with the camera?  A fast modified newtonian.  F 2.8. $2000.00. No f10 SCT"s here. Therefore limited selection of optical choices. In the $5000 to $6000 range some APO's may work, but they are F4-5. doubling your exposure times.

Although the pixels are small, and it will oversample without binning on telescopes with a higher focal length, I have not found a CCD at a similar pricepoint with the same suitability for astrophotography/astrometry, and camera specifications, notably noise and full well. I am unsure how extreme the vignetting is, and how much less it would be with a larger SCT, but I have contacted the reducer manufacturer. I expect they have a good prediction on hand. Maybe it can be mostly fixed with flats, maybe there will be a large portion not for use. The manufacturer's reccomendation I understand - this camera would be terrific with a RASA 36 or newtonian!

I wish there was a sensor that had all the goodies of the IMX455 (same bit depth, full well, pixel size...) and just ~1/4 the area, so around 15 Mpix  (would not mind if the price tag would scale as well :-)  ). That would be a great choice for a lot of us

Bikeman mentioned that there is a lack of a sensor with the same bitdepth and full well but smaller size, I equally frown at this. The ZWO 1600mm for example has a smaller sensor, but is much worse in both other factors.With this camera, it is a matter of choosing a best fit for the telescope across several fields. The 6200mm CMOS gives me great signal, good sampling, and a high resolution of the sky across a relatively large FOV. Then there is future compatibility for wider field using the small pixels or 2x2 binning with a higher focal length, still retaining a good field of view. 

 However, if you want to purchase the IMX455 as the first step in a complete system upgrade in the coming years, go for it.

It is a camera not perfectly suited for the current telescope in terms of sensor size, but has good specifications otherwise, and looks to be a safe-bet as a camera for a future system, I agree with Mike. 

In closing I feel compelled to mention the huge benefit derived from working with SBIG/DIFRACTION LIMITED They sell, service, support  provide documentation and tech support for science grade cameras. They include custom software integrating camera and computer. I have ZWO QHY and SBIG. There is a reason why so many photometrist, researchers and universities use SBIG..Did you know they have an adaptive optics upgrade umder $1000. 

I looked at SBIG cameras however I did not find a comparable CCD at the same price point. The adaptive optics for 800.00$ are very intriguing, though a camera with 6-9um pixel size quickly makes my telescope undersampled. It does surely work for photometry, but is a limiting component for photography and astrometry. With the 0.63x reducer -> 1280mm FL, the ZWO camera sampling is good all-round according to my calculation. 

A question would be, is a CCD with an inherent pixel size of ~7um better than the ASI 6200mm Pro binned at 2x2 pixels? The CCDs I looked at within the same price range all have a smaller full well and more noise, according to the specifications. To me it looks that the ZWO camera offers the best price-to-performance , but I could be mistaken here.

Leon

CrossoverManiac
CrossoverManiac's picture
I've been looking to upgrade

I've been looking to upgrade to a monochrome camera for astrophotography for a while now. I currently use a QHY-5 178 CMOS colour camera by QHY (IMX178 sensor), which does have low read noise, however the field of view is tiny (0.33° * 0.22°) and it oversamples. Excellent choice for lucky imaging of planetary objects or planetary nebulae, not so much for most deep sky objects.

 

Optec does sell a 0.33x focal reducer.  According to the website, it will work with an SBIG ST-7, which measures 8mm across diagonally.

Lbew
Lbew's picture
SBIG ST-7 & Comparison Table

Optec does sell a 0.33x focal reducer.  According to the website, it will work with an SBIG ST-7, which measures 8mm across diagonally.

A full well of 200ke(-) at that price is impressive, however that camera is not as useful for astrometry and astrophotography. I would have to purchase another one, where I'd reach the price of the 6200mm again.

I find this table very interesting, especially regarding versatility - small pixels for astrometry and photography, with a large field of view, yet still excellent performance with 2x2 binning: 

 https://astronomy-imaging-camera.com/wp-content/uploads/ASI6200-bin1vsbin2vsCCD-1024x600.png

 

I've made some phone calls to distributors and had several conversations with users of the camera - there are some disagreements, and (almost) all agree it is not the perfect choice in terms of sensor size for the 200mm SCT, but in the 2/3 of the capture area without vignettting it is better in price-to-performance and quality of results across all fields in comparison to a ASI1600mm Pro, or dedicated photometry CCD. 

Nevertheless I will have a phone call with the ZWO technical department tomorrow for a final opinion on the matter - though if there are better cross-field options here, or if my reasoning is wrong / there is something I have not considered, please do let me know. 

Leon  

Bikeman
Bikeman's picture
I will have a phone call with

I will have a phone call with the ZWO technical department tomorrow

If you could somehow catch them off-guard talking about plans for future 16 bit mono CMOS sensor cameras with smaller chip sizes (APS-C or 4:3 ), that would be great :-). Let us know. Because I think you are right, even if you are literally wasting some of the off-center real estate of the sensor, it is still a good choice if the budget allows. But wouldn't it be a shame to see a comparable but smaller and less expensive sensor variant being offered soon after making that investment?

Cheers

HB

 

Lbew
Lbew's picture
Re: I will have a phone call with

The ZWO techncial departement unfortunately did not respond to the phone calls even within their official business hours and information provided on the forum is somewhat vague. 

The best I've heard for now also are rumours about the introduction of a APS-C or 4:3 chip. One developer on the ZWO ASI forum mentioned "it depends on Sony and the needs of our customers", that is all I could find apart from "knows a guy who knows a guy who says there will be a xyz.." 

I personally have opted for the QHY600PH with the same IMX455 sensor due to the better suitability for photometric application specifically (see QHY600 tests thread on this forum for a fantastic list of reasons and tests), and a terrific experience in customer support with my current QHY camera. Having sent ZWO an email on another issue, their response was rather short and vague. 

We'll see if smaller 16-bit sensors come along in 2021. I suspect they will gradually be introduced replacing the common QHY163/183 and ZWO ASI1600 cameras, and of course CCDs at a similar pricepoint and sensor size. 

Leon 

MMU
MMU's picture
Suitability of Sony IMX455

Leon, although my path has been different, I'm also arriving at many of the same conclusions as you regarding the suitability of the IMX455 for photometry. (I've been doing monochrome CCD photometry for about 20 years using an SBIG ST-9 camera with a 10" SCT on a wonderful mount: GM2000HPS by 10Micron.)

Your original post asked about "peculiarities"; let me offer some of the items I've identified as I look at the IMX455:

  • USB 3.0: Cameras using this sensor typically use USB 3.0 for download of images. Further, some users have encountered problems with the USB link hanging, particularly if the computer is busy doing other things. This is leading me to realization that replacement of my "observatory computer" would be a part of an upgrade to a IMX455 camera.
  • No mechanical shutter: Although it seems straightforward, I've gotten quite used to the mechanical shutter on my SBIG camera. I'm going to need some time to gain confidence that an opaque filter will do the same thing as a closed shutter.
  • File size: This is huge (in more ways than one). Image file size is more than 2 orders of magnitude larger with the IMX455 than with my ST-9 camera. This is triggering upgrades to my home network, a complete re-think of how I permanently archive photometry images, a shift to compressed FITS files, and careful consideration of where in the processing pipeline I perform binning.
  • Subframe imaging will be the norm: I'm going to stick with my current investment in 1.25" photometry filters and only use the center 18mm x 18mm of the IMX455 sensor (a 4608x4608 square). This will provide me with an image 23 arcminutes across, which is plenty. (My current 13 arcminute square almost always provides multiple candidate comparison and check stars.) I know that it seems insane to pay for a large sensor and only use the center of it, but the economics are pretty persuasive: I can't match the performance of the IMX455 in a smaller sensor without paying more, not less (e.g., GSense 2020 or GSense 400 -- both of which offer better performance (at a higher price tag), but under my local sky, the better performance is lost because of skyglow).
  • Binning will be the norm: I anticipate using a mix of 1x1, 2x2, and 3x3 binning, depending on the field. (Aperture photometry can use 3x3, while crowded fields where I use PSF-fitting will use 1x1 or 2x2). This interacts with the software of my processing pipeline (including image calibration), storage space requirements, and approach to permanent archival.
  • Mode/Gain Selection: Different mode/gain combinations seem appropriate depending on the type of photometry and the field/filter combination. This adds a (small) wrinkle to the session planning process.
  • Nonlinearity: I've had real hassles dealing with the nonlinearity of the ST-9 in the center of its dynamic range, have had mixed results doing linearity corrections, and have a good understanding of how it is degrading my photometry. The linearity curves I've seen for IMX455 also have issues. And no one has published any curves yet describing how IMX455 nonlinearity changes with temperature or with mode/gain settings. (I'm not talking about near-saturation nonlinearity -- this is nonlinearity in the middle of the transfer curve.) If I go ahead with an IMX455 camera, I'm expecting to invest quite a few nights in nonlinearity characterization and experiments.

Despite all that, I tend to agree with you that cameras based on the IMX455 seem a pretty good cost/performance balance.

- Mark

Eric Dose
Eric Dose's picture
Focal reducers, vignetting

Focal reducers are fine--I use a 0.67X. But a 0.33X focal reducer is extreme. Be careful of vignetting. Be especially careful of backfocus range which may be limited and depends on your telescope. I designed an adapter to get the backfocus exactly right for my system (C14 Edge > Atlas focuser > STXL 6303E), and it has worked very well for the 5 years since.

Vignetting of up to maybe 20-30% flux loss at the sensor corners shouldn't be a big concern, as it can be pretty well removed by proper flat-field calibration.

Log in to post comments
AAVSO 49 Bay State Rd. Cambridge, MA 02138 aavso@aavso.org 617-354-0484