How fast can I image?


Biologists that use, or are interested in using, microscope cameras and don’t speak engineering


Clarity on the relevance of camera specs to biological experimentation



Hamamatsu listing Synonyms used by other vendors
Readout time (sCMOS only)  
Readout speed Imaging frequency
Pixel clock rate Pixel readout rate, Pixel scan rate, Clock rate
Read-out modes Binning, Sub-array

Biological processes can be dynamic, with events happening over a multiple log range of time scales. To understand whether your camera can keep up with your experiments, look at readout time, readout speed, pixel clock rate and consider how much of the sensor you’ll be using to acauire your images.


Readout speed tells you how many frames-per-second the camera can capture an image, with conversion to 1/readout speed providing information on the speed of individual frames. For a camera with 100 fps, each frame is captured at 1/100 seconds or 10 ms.


The pixel clock rate tells you how fast individual pixels are read out from a CCD or EM-CCD camera, providing a lower bound for imaging speed—you won’t be able to image any faster than the camera can process, and you still need to add on exposure time. The pixel clock rate is often given in MHz, so 1/pixel clock rate will tell you how long it takes to read each individual pixel.


Pixel clock rate is an important specification for CCD and EM-CCD cameras, but not for CMOS cameras because of chip architecture (read here for more on the differences between CCD, EM-CCD, and CMOS). Unlike CCD and EM-CCD cameras where photoelectrons from each pixel are converted into voltage one-at-a-time via a single amplifier, in CMOS cameras each pixel has its own amplifier, so conversion of photoelectrons happens in parallel in all the pixels. This parallel processing of photoelectrons is what gives Gen II sCMOS cameras such fast readout speeds.


Readout time is a specification Hamamatsu supplies for its CMOS cameras, and provides the readout rate of the entire array—its 1/readout speed. So for the ORCA-Flash4.0, the readout speed at fast scan is 100 fps, giving a readout time of 1/100 = 10 ms.

I still need to image faster; what can I do?

For some experiments, capturing higher speeds is more important than high resolution. By binning pixels in a CCD camera, groups of pixels are combined on the sensor to increase readout speed, but at the cost of slightly lower spatial resolution.


CCD camera binning is a function that is intrinsic to the sensor and built into the camera, with typical capabilities of binning pixels by 2 × 2, 4 × 4, and 8 × 8.


Note that another advantage of binning—many may argue the main advantage—is an increase in the signal-to-noise ratio (SNR). The signal is multiplied by the amount of additional area, but the noise (read noise and dark noise) remains the same, and the loss of spatial resolution is linear. For example, in 2 × 2 binning, the signal increases by a factor of 4, spatial resolution decreases by 2×, but the noise is unchanged . For 4 × 4 binning, the signal increases by a factor of 16 with no increase in noise, and the spatial resolution decreases 4×.


Binning in sCMOS is possible, but due to the structure of the sensor it occurs after digitization of the signal, and thus does not increase readout speed. For sCMOS cameras in particular, the main advantage of binning is to increase SNR.


In addition to binning, most cameras can achieve faster frame rates by reducing the size of the frame. By capturing a subset or sub-array of the whole effective area, there are fewer pixels to readout. In a CCD camera, subarrray readout speeds are determined by the overall size of the subset of pixels but in sCMOS, only shortening the vertical dimension of the region increases speed. Thus, in the Flash4.0, a 2048 (h) × 128 (v) region has the same frame rate as a 500 (h) × 128 (v).


  • Readout speed provides the maximum frame rate in frames-per-second.
  • Pixel clock rate (CCD and EM-CCD only) describes how fast each pixel can be read and provides a lower bound for imaging speed.
  • Pixel clock rate are not as relevant in CMOS cameras because of differences in chip architecture from CCD/EM-CCD cameras.
  • Readout time for CMOS cameras describes how fast the entire array can be imaged, and is 1/readout speed.
  • Cameras with binning (CCD and EM-CCDs only) and subset/sub-array capabilities can increase readout speed, enabling faster frame rates.


Faster frame rates are possible but come at the price of either lower resolution—binning—or smaller field-of-view—subset/sub-array.


NEXT SECTION: Will I be able to differentiate 1500 molecules from 1475? What’s the dynamic range?

Contact us for more information.

  • Literature
  • Price
  • Delivery
  • Custom order
  • Demo
  • Support
  • Other

Contact us