CCD Camera Noise Tests and Comparison


Why all this? Marketing and advertisement sayings “our CCD cameras are low-noise” is nice, but what that exactly means? Some parameters of a new CCD camera are more important then others for a potential camera buyer (based on everyone’s personal preference), but in the end, all of us would like to have a CCD camera that works perfectly well. By this I mean that the camera gives consistent and repeatable results and that bias frames contains pure Gaussian (completely random) source of noise. This means that stacking more single frames actually lowers the noise and increases the SNR (Signal to Noise Ratio) of the resulting image. That’s what all of us wants (apart from the World peace).

The purpose of this camera testing and comparison is to:

  • check that the camera works well and doesn’t contain any non-Gaussian source of noise like e.g. pattern noise which can’t be removed by stacking more frames together and which in fact introduces an unwanted signal (which in the end worsens SNR)
  • verify the readout noise of your particular CCD camera (cooled down to a set-point stable temperature) in order to check if it meets manufacturer’s specification (and also the CCD chip specification) as camera makers can only make things worse by using poor electronics that introduces unwanted noise to every frame that is read out of the camera.
    • NOTE: even a high readout noise camera can produce nice results based that there is not any pattern noise (or other serious defect) in the image. Especially large format cameras (36x24mm or bigger) allows for use of a large aperture telescope (to capture more light) and let’s you down-size the final image resolution to some acceptable value (to hide the remaining noise in the low SNR background part of your image). I do myself own a CCD which has one of the worst CCD chips inside, but I use it for LRGB imaging only where the readout noise is not important.

      This test focuses on camera readout noise which is the most important parameter for narrow band imaging (which is my favorite discipline), where the narrow band filter limits the sky background flux to some marginal (zero) value (it passes wanted signal, e.g. H-alpha and cuts off all the rest light spectrum). In this case the readout noise plays significant role in how clean the final image is. The narrower the filter (while the best is to have also highest possible filter transmittance along to the narrow bandpass) the higher demand on low RN (readout noise) camera – the more important this parameter becomes. On a dark sky this could mean that a difference in image cleanliness of a camera (A) having RN of 4e- and camera (B) having RN of 10e- by a factor of 6x (you would need to image 6 times longer with camera (B) in order to have the same clean image as from camera (A) – comparing pixel to pixel signal). The reality is much more complicated than this simplified case as there’s something like e.g. optimal subexposition duration based on the sky flux, filter and camera used. To simplify things, again, with camera (B) would be optimal to shoot narrow band images of a subexposition duration like 90 minutes (in order to get the most out of the camera) while with camera (A) you may end up with only 15-20 minutes per subexposition.

      The last, but not least, purpose is to put more pressure on camera manufacturers in order to increase QC (quality control) and produce perfect-most cameras within the limits of every particular CCD device.

      Method of Testing

      To compare cameras against each other I have extended the “simplified” (but very precise) method closely described by Craig Stark in his document Signal to Noise Part 3: Measuring Your Camera. In brief, I use a set of 10 bias frames to create a master frame that is used for subtraction from single biases from which I compute StdDev value. Therefore I compute 10 numbers which are averaged so as the test result is precise. Then I use 5 flat field calibration frames (which I calibrate with the master bias) for gain computation. Because all ABG CCD cameras on the market today are linear at around 25000 to 30000 ADU levels an ordinary (correct) flat field frame is the only thing I need for this. Again, to increase the precision (and to check the consistency of results) I compute 4 values of gain from 5 flat field frames. Then the RN value is simply computed as gain multiplied by previously computed StdDev value. Last step is to compute TSN (Total System Noise) that is affected not only by readout noise (must be always higher) but also by thermal noise (and other sources). The TSN characterizes all noise sources in a raw bias frame together. If it is much larger than RN then there’s something wrong. Notice that with lower CCD temperature the TSN on the same camera is lowered. To make the calculations easy and straightforward I wrote a simple script in Octave (free version of Matlab) that runs on my Linux box.

      Camera Test Results

      Sorted by CCD detector and Readout Noise value in ascending order.

      Last update: 26th August 2013

      DOWNLOAD the results table!

      NOTE on FWC – Full Well Capacity: the table might be a bit misleading on this as the number computed (65535 levels of an 16-bit A/D converter multiplied by computed gain of particular camera) may be effectively higher (or lower) than what’s real (true) full well/pixel capacity (saturation signal) that is a „system“ property of every particular CCD device. This real number is usually specified by SONY/Kodak for every chip. e.g. Kodak KAF-8300 has Full-Well 25500 e- based on Kodak’s datasheet. Therefore „optimal“ value of gain in order to get most out of the chip with 16bit ADC would be 25500 / 65535 = 0.389 e- / ADU. If the camera’s gain is higher than this „optimal“ number then the stars become sooner saturated (burned out white). On the other hand, if the camera’s gain is lower than this value then the dynamics is sacrificed. Neither case is wanted :)

      Interpretation of Results

      The Observation column contain my personal opinion on the obtained results. If I get consistent results in the computation log and the TSN is only a hair above RN (provided that the CCD temperature is below zero degrees of Celsius) then the camera works well = doesn’t have any problem. Whether RN meets the manufacturer’s specification or not is the other side of story. If the TSN is lower than RN (happened with some cameras) then any calibration with bias frame effectively introduces more noise into the image then it removes – this is bad. Also, if the histogram shape is anyhow different from a standard Gaussian curve then there’s a suspect on pattern noise (that may come from e.g. power source or any other electronical device in the surroundings of the camera may interfere). In such a case a look at the master bias frame could reveal more secrets on how bad the issue really is.

      May you find that there are many cameras listed as having issues – the reason is simple – that’s exactly why I started to dig into noise measurement – because my first and second CCD camera I ever purchased had some kind of problem. Also, people who are willing to provide their calibration data for my measurement do so only in case they expect (and need to confirm) that the problem exists. Others probably rather do not want to hear the numbers as they are affraid that there could be anything wrong while they are happy with the camera they have. Though I would very much prefer the other approach: to test brand new cameras that you have just received. In case of a problem you could claim it back to manufacturer or dealer to resolve the problem. This may save you money and frustration.

      How to capture calibration data needed for analysis?

      If you find this article interesting and you own any kind of cooled astro CCD camera, please help with this research by providing your calibration frames. I’d be happy to do a test run for you. The only inconvenience that persist, is that you would have to upload the large archive file to some website and provide a download link to me – best contact is through the comments section below as comments needs to be approved prior being published.

      To measure the readout noise of your CCD camera in an uniform manner (so as the results are comparable against each other) please provide a set of 10 single bias frames and 5 flat field frames in native FITs format.

      Bias frames:
      - capture 10 consecutive single bias frames with cooling on, binned 1×1. If you have set-point cooling camera cool to at least -10 degrees Celsius for Sony chips and -20 for Kodak chips (the colder the better). If you can’t control the cooling, turn it on and wait for up to 30 minutes to stabilise. Make sure you shoot the frames in a dark room no matter if your camera has mechanical shutter or not in order to prevent any possible light to “sneak in”. Nosepiece is not light-proof unless you cover it with aluminum foil and fix with e.g. a rubber band.

      Flat field frames:
      - capture 5 consecutive flat field calibration frames (again, with cooling on) using the same binning (i.e. 1×1). A very good tool for this is an EL-panel or a light-box. Use sheets of white paper (or RED/H-alpha filter, if filter wheel is in place) to dim the light if needed. It doesn’t matter if you shoot real flats with optics attached to the camera or just plain flats without any optics involved, when the CCD is just lying on the surface of EL-panel. If you have a mechanical shutter be sure to keep the exposition of a single flat frame a bit longer (e.g. couple of seconds). Aim for around 25000 ADU levels (as almost every ABG camera stays in linear range at this level).

      NOTE: you may use real calibration data that you already have on your hard drive. Some vignetting or dust motes on flat fields doesn’t matter as these frames will be center-cropped.

      37 komentářů u CCD Camera Noise Tests and Comparison

      1. Graham Roberts napsal:

        You Sir deserve more recognition from your work preformed. I consider you a pioneer in leading these test results and a credit to our society of amateur astrophotographers. Thank you for your work.

      2. Hi Pavel,
        A user has pointed out that you have some H694 results on your read noise page and that they show a high pattern noise level. The test cameras must have been early ones with the old firmware version loaded and so have some cooling regulator noise. If you still have access to these cameras, I can mail an update to fix the problem.

        Best regards,

        • pavelpech napsal:

          Hi Terry,

          thank you for your comment, I very much appreciate it. I like when people care about what they do, why they do it, how they do it etc.
          I have heard that when you disabled TEC cooling/regulation (or somehow like that) in your cameras (updated firmware) then the visible (easy to see by a human eye) pattern disappears :) from my point of view it’s a great improvement because now you do not have to calibrate images at all (under certain circumstances when the user is doing dithering – that takes care of removing of hot pixels – we really do not even need to calibrate by bias). It’s rare but it’s good to have this option :)
          The other findings that I have on SX cameras (H694) is that (when comparing to Atik) if you crop 1 (in words „one“) pixel (line and column) from the final image then the StdDev (CTRL+I in MaxIm DL on whole Area) drops from insane big numbers to the same level as Atik camera (460EXM) has – therefore the TSN in my calculations would drop by a factor of five or so which looks much better then.
          Last test I performed was a comparison of noise drop after combining 10 single bias frames. If I crop the SX camera FITs frames then the drop level is the SAME as Atik has. This means that there is essentially no fundamental problem in SX cameras (otherwise I would flood you and the SX user’s yahoo group) and that the noise is lowered by combining more subs (that’s the rule number one).

          best regards

      3. Pingback: CCD Sensor QE (Quantum Efficiency) Comparison | Xpírijensis astrofotografa

      4. Volker Umland napsal:

        Hello Pavel,

        Recently i saw your very interessting site with Camera Noise Tests and Comparison.
        Now i got a Atik460exm, (Serial 0943533253, Firmware 3.27).
        I made some Bias and Flats after your terms of reference in your Blog.
        Please, can you check this data for me and tell me about the noise of my cam?
        The files can be downloaded here: *** (its about 100 MB)
        I also own a QSI583 and i captured some data for this cam too.
        Maybe you are also interessted on this? (yes, i am :-) )
        Download the Files here: *** (About 150 MB)

        Thanks a lot for your time.

        Greetings from Hamburg
        Volker Umland

        • pavelpech napsal:

          Hi Volker,

          here is the result for your Atik460EXM (your Atik is even better than mine):
          Gain = 0.25
          Readout Noise = 4.69e-
          Total System Noise = 5.03e-

          and here is the result for your QSI583:
          Gain = 0.47
          Readout Noise = 9.40e-
          Total System Noise = 10.47e-

          thank you for your data and have a clear skies!

      5. Daniel Amado napsal:

        Hi Pavel,

        I am about to decide between buying an Atik 383L+ Monochrome CCD Camera or a QSI 683WSG Monochrome CCD Camera.

        I want to perform a low read noise test of an Atik 383L+ Monochrome CCD Camera before buy it.

        In the QSI website, they show two distinct read noise frames sample comparison between one of its cameras model and other manufacturer cameras model. They specify detailed instructions about how to make suitable bias frames to do the test. I have found all this information very useful.

        To do this comparison, they perform a Fast Fourier Transform (FFT) algorithm read noise test with Imaje J program.

        The QSI team claims that The FFT test of their cameras Read noise frame, reveals no pattern noise, as demonstrated in his test.

        Instead, the FFT test of a Read Noise Frame sample from other manufacturer camera published in his website, introduces non-Gaussian periodic noise into the image. This produces vertical lines in the FFT test.

        To perform the test is necessary to have a total of 9 good bias frames without anomalies like cosmic rays neither bright pixels. it is important to stretch the black and white levels in the histogram so that you can see the difference between the lightest and darkest pixels when making the bias frames.

        The link of this instructions is:

        Could you please send me 9 good bias frames samples of an Atik 383L+ Monochrome CCD Camera or send me a link with the bias frames of this camera?

        How perform Atik 383L+ against QSI 683WSG?

        Thank You.

      6. Yves napsal:

        Hi Pavel,
        can pass your email address so I can pass dropbox link to H814 data …
        flat’s are really quick and dirty so only use small center crop, if needed I will provide new ones …
        (tried to PM you on CLN but your account does not accept PM’s anymore)

      7. Blonder napsal:

        Hello Pavel,

        I don’t know if you are still interested on camera data.

        Today I read your very interessting post „CCD Camera Noise Tests and Comparison“.
        I ownned an Atik420 Monochrome camera.
        I made some Bias and Flats following your instructions in your Blog.
        Please, can you check the data for me and tell me about the noise?
        Thanks a lot for your time.

        For the download link, please send me an E-Mail.

        Greetings from Aachen

        • pavelpech napsal:

          Hi, yes, I am still interested in flat and bias sets from various CCD cameras in order to keep my comparison of quality of cameras from different manufacturers up to date. I’ll look into your data ASAP…

          • pavelpech napsal:

            here is your Atik420 measurement:

            Gain = 0.12
            Readout Noise = 3.82e-
            Total System Noise = 4.18e- (cooled at -10deg C).

            looks very good to me,

      8. Jean-Marie Locci napsal:

        Pavel, I just ran into this blog page, excellent work and very helpful! I own a QSI583wsg and I suspect a minor problem. As I am not afraid of scary results I’ll run these tests.

        I have read Craig Stark page also, yours is more „understandable and pragmatic“, but thanks to both for your contribution to progresses of AP.

        Best wishes 2014.

        • pavelpech napsal:

          Hi Jean-Marie,

          here is the measurement result of your QSI583wsg camera when cooled down to -15:
          CCD Camera Measurement:
          Gain = 0.47
          Readout Noise = 9.09e-
          Total System Noise = 9.84e-

          thank you for your contribution :)

      9. pavelpech napsal:

        New data from FLI-PL16803 cooled down to -25deg C:

        CCD Camera Measurement:
        Gain = 1.18
        Readout Noise = 8.71e-
        Total System Noise = 9.94e-

        Remark: this is the FLI-PL CCD camera that ASA uses in Nerpio/Spain observatory. There were few strange things: 1. sky-taken flats with some dimm stars present (this affects precise measurement of Gain), 2. flats were taken in high speed download mode and biases in low noise readout mode which is not correct approach for scientific measurements on ADU level. At least the -25deg C cooling temp was same as well as bin 1×1 mode.

      10. Emil napsal:

        Hi Pavel,

        Excellent data! I have an ASI120MM (with an MT9M034 cmos) planetary camera that I’d like to test as well. If I have time this weekend, I’ll send you some data (let me know where to send it). I also have a Basler Ace acA640-100gm (with an ICX618) that I could test. Unfortunately, I don’t think the capture software allows for FITS output, but I need to check that.

        Both camera’s are uncooled by the way, so they probably won’t give the best results. But I was really pleased with what especially the ASI120MM camera can do with very short exposure times on relatively bright deepsky targets.

      11. pavelpech napsal:

        New data from Trius-SX694 (SONY ICX-694) camera cooled down to -20deg C:

        CCD Camera Measurement:
        Gain = 0.34
        Readout Noise = 6.13e-
        Total System Noise = 6.67e-

      12. pavelpech napsal:

        New data from QSI-690 (SONY ICX-814) camera cooled down to -20deg C:

        CCD Camera Measurement:

        High Gain mode:

        Gain = 0.09
        Readout Noise = 4.12e-
        Total System Noise = 4.42e-

        NOTE: this mode results in very low dynamic range(!)

        Low Gain mode:

        Gain = 0.19
        Readout Noise = 4.25e-
        Total System Noise = 4.57e-

        NOTE: perfect data from my point of view. Highly recommended to use the low-gain mode for all the time. Acknowledgment „Astrowood“ for the data capture.

      13. Peter Nagy napsal:

        Hi Pavel,

        After reading your recent CN posts regarding to read noise and gain measurement about QSI660 camera, here are my data:

        The gain was set to „high“. Sorry, I don’t yet have low gain images for binned 1×1. Kevin Nelson from QSI suggested to use high gain for 1×1 and low gain for binned 2×2 or greater. I will contact Kevin Nelson about it. I am concerned about using low gain for binned 1×1 because it may require longer sub-exposure times but hopefully not that much longer. I am also concerned about low well depth capacity for high gain at 1×1 because my first images had bloated stars at 2 minutes sub-exposure (FWHM of 5″ to 7″ at 0.96″/pixel using TEC 140 APO at F/7). At first I thought it was due to bad seeing or some wind (I have AP1100 mount so it should withstand minor wind) but after reading CN posts about two people getting similar results for well depth of 10,000 electrons for high gain at 1×1 using same camera as mine, I got a little concerned. I don’t want to jump to conclusions right now until the next clear and better seeing night.

        Please let me know your results. Feel free to post the results at CN.

        Peter (AKA Peter in Reno from CN)

        • pavelpech napsal:

          thank you Peter, but I would need in total 15 fits, 15 files, 10 biases and 5 flats. You have posted/uploaded a combined frames (I guess), but I need raw frames to make statistics to work (reject outliners etc.). Could you please repost the individual frames? many thanks!

          • pavelpech napsal:

            Peter, your QSI-660 camera’s results:

            High gain mode, 1×1, -10C
            Gain = 0.15
            Readout Noise = 3.94e-
            Total System Noise = 4.27e-

            Low gain mode, 1×1, -10C
            Gain = 0.36
            Readout Noise = 4.15e-
            Total System Noise = 4.48e-

            looks perfect to me :)

      14. adrian napsal:

        Could you test my QSI690 for me. Camera seems very noisey.

        • pavelpech napsal:

          Hi Adrian,
          your camera looks perfectly OK. Your data at -20deg C with HIGH gain settings give following results:

          CCD Camera Measurement:
          Gain = 0.09
          Readout Noise = 3.48e-
          Total System Noise = 3.73e-

          your gain is like 0.085 which is the same as the other user’s QSI690 in high gain mode…


      15. Nick napsal:

        Could you provide the power sources being used as part of the analysis for each entry?

        I know from experience with my ATIK 383L+ that the noise increases as my Bosch S5 12V battery exhausts itself and the voltage it supplies drops (usually the second day without a charge as the S5 is a 100Ah battery capable of delivering 800A). This means that if the power source cannot provide enough current (amps), then the voltage may drop over time.

        The same is true with PSUs where if I use a 2A 12V mains PSU I get more noise than if I use a 13.8V 7A PSU (the maximum voltage I use) the noise levels are considerably lower.

        • pavelpech napsal:

          Hi Nick, that’s not possible as I have no clue what kind of PS and actual Voltage of the PS was used when this or that data were taken. I ordinarily ask camera owners to provide the data to me without asking for voltage measurement of the power supply they use. Anyway, I am aware of the fact that Atik 383L+ is somehow ‘sensitive’ on actual Voltage. I measured more samples from one 383L+ owner using different PS and the results were different, but from my perspective, just a little bit, nothing dramatic. The results were not very impressive, I like Atik very much, but the 383 is „noisy“ from my point of view. At the moment the lowest 8300 noise cameras are Moravian G2-8300 and QSI683.

      16. SX-H694 napsal:

        Hi Pavel

        If you are still carrying out these tests, please let me know how best to provide you some data on my camera above. I can put them up in dropbox if that is useful.



        • pavelpech napsal:

          Yes I do, I send you my e-mail address now… the dropbox would work fine! just put them all into one folder so as I do not have to download file by file :-) thanks! Pavel

          • SX-H694 napsal:

            Thank you Pavel!


            CCD Camera Measurement:
            Gain = 0.37
            Readout Noise = 6.17e-
            Total System Noise = 6.71e-

      17. Tomasz napsal:

        I am going to buy Attik Titan mono. BUT before I would like see noise( histogram) from this camera .You published excellent table but may You send a link to raw data from Attik Titan (cooled, dark, ~100s)

      18. Peter Nagy napsal:

        Hi Pavel,

        QSI sent me an e-mail saying that they fixed the glows issues in their Sony ICX-XXX cameras and asked me if I wanted to return my camera for an upgrade. I returned the camera and they fixed the camera and returned it back to me. I took 30 minutes darks and the glows appear to be non-existent and possibly better than the original design. I also uploaded 10 Bias and 5 Flats (-10C, 1×1 and high gain) for you to measure the read noise, system noise and gain. I am providing you links not only for flats or bias but also darks (before and after camera fix) for you to see and evaluate. I greatly appreciate it and look forward seeing the results.

        Zip file of 10 Bias and 5 Flats (108 MBytes):

        Single 30 minutes dark before camera fix:

        Single 30 minutes dark after camera fix:

        Master dark of 30 subs of 30 minutes each:

        Master bias of 100 subs:

        The following is your original measurement of read noise, system noise and gain before the camera fix:

        High gain mode, 1×1, -10C
        Gain = 0.15
        Readout Noise = 3.94e-
        Total System Noise = 4.27e-

        Peter (AKA Peter in Reno from CN)

        • pavelpech napsal:

          CCD Camera Measurement:
          Gain = 0.18
          Readout Noise = 3.88e-
          Total System Noise = 4.18e-

      19. Nick Haigh napsal:

        Hi Pavel,

        Ive just obtained a 460ex mono, and would like you to run your tests! Please email me and I’ll reply with a dropbox link. Or whatever you think is best!



      20. Rixon Reed napsal:

        Hi Pavel, Thank you for all of the work you’ve done here. Do you have any thoughts about the Starlight Xpress Trius SX36. Seems like it has good potential on paper. I’m using a Tak FSQ-85 with reducer and this camera seems to offer a good tradeoff between pixel size and FOV. Thanks. Rixon

      21. pavelpech napsal:

        Hi Rixon,

        I had only one data sample (set of biases and flats) from SX36, but there’s some problem with the data as the flats do not match the X/Y dimensions of the biases :-( I didn’t have time to look into it.

        Anyway, what I know about that chip (KAI-16070). The specification is much better in terms of dark current noise than KAI-11002 which is, as I say, one of the worst chips ever – but I still own it because it is cheap and very large which counts for LRGB work. But there’s still major issue with every chip OnSemi/TrueSense/Kodak KAI-16000/16070 and so far (until now) there’s no solution other then ugly one (increase readout noise over at least 10e-).
        The problem is with some horizontal transfer of charge.
        And it is visible only when you have:
        - really low noise CCD camera electronics (I tested a G3-16000 with 7.9e- readout noise)
        - do stretch the image very deep (which I do) and it’s most visible in narrow band images

        The issue I describe can be seen here:
        horizontal lines left to stars (direction of frame readout).

        you know I am nit-picker who wants stuff only 100% perfect which doesn’t exist on the market :-)

        good luck

        • pavelpech napsal:

          I managed to resolve the issue with data. The flats were taken with different control software than the biases which resulted into different X/Y dimmenstions of the images. After cropping biases few pixels (48 and 40px) to match flats I got following measurement data out from SX-H36:

          CCD Camera Measurement:
          Gain = 0.59
          Readout Noise = 10.88e-
          Total System Noise = 14.16e-

      22. Peter napsal:

        Hello Pavel,

        I have a Atik One 9.0 less than one year and have noticed the master bias frames look unusual when compared with my other Sony sensor cooled cameras, if you have time would you analyse these for me?

        Could you also look at the two bais master files included in the file one of 200 frames @ -12C and another of 50 frames @-20C, these seem to have different noise patterns, do you think the CCD has good contact to the Cooler as the -12C master has the look of a 0C bias, what do you think?

        Kind regards


      23. Martin Weiss napsal:

        Hi Pavel,

        I finally found your page … Really hard to find!
        Would you be interested in some assistance in programing? I would like to write a software for doing astrocam test automated and – perhaps this might be helpful for you (e.g. aquire raw images in an automated way, run the processing, …).

        So the user could have a free-of-charge tool to check the camera quality and if something went wrong.

        And also thank you for your positive report on the 094MC PRO – I like it very much and use it on my CDK 12.5 at the moment :-)

        Best regards


      24. Robin Kalfat napsal:


        I noticed that you were considering this combination back in 2019 and looking for information. I now find myself in the same position :) . Did you ever end up buying the FSQ-85 and using it with a full frame sensor? How did it perform? I’m trying to decide if there is any sacrifice in performance compared to the fsq-106. I too want to take it to chile! FSQ-106 is too big!



      Napsat komentář

      Vaše emailová adresa nebude zveřejněna. Vyžadované informace jsou označeny *


      Můžete používat následující HTML značky a atributy: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>