We have noticed there is some confusion about the various Sony Pregius image sensors and their resolution. The resolution doesn’t necessarily increase with a higher number in the sensor’s name. Therefore we will give in this blog a brief overview of the resolution of the Sony IMX250, IMX252, IMX253 and the IMX255 sensors. We also address the question which of these IMX sensor should I choose?
The four mentioned sensors have a different pixel resolution -the number of horizontal and vertical pixels. As the pixel size is equal to 3.45 µm for all sensors, the diagonal and aspect ratio will be different as well. In terms of other characteristics, like noise performance, the sensors perform the same.
Sensor | Resolution H | Resolution V | Diagonal (mm) | Aspect ratio |
IMX250 | 2448 | 2048 | 11.1 | ~5:4 |
IMX252 | 2048 | 1536 | 8.9 | 4:3 |
IMX253 | 4096 | 3000 | 17.6 | ~4:3 |
IMX255 | 4096 | 2160 | 16.1 | ~16:9 |
For the resolution the number of pixels to be used as recommended by Sony is listed.
How to determine which sensor to use?
So if the only difference is the number of pixels and aspect ratio, which sensor should you choose? Like in most cases this depends completely on your application.
Optimize for Field of View
The desired field of view will determine if you need a sensor with a smaller or larger horizontal and vertical resolution at a fixed aspect ratio. At a fixed pixel size and a fixed aspect ratio your field of view will increase with increasing number of pixels. With this relation it is thus possible to decide between the IMX252 and IMX253.
To obtain the largest field of view, a square sensor aspect ratio is most efficient. A square sensor makes optimal use of the field of view that can be reached with the lens system in front of the camera.
Optimize for display on screen
So in general a square aspect ratio is desirable for machine vision applications in which the recorded image is not being displayed on a screen. However once you display recorded images on a screen a different sensor aspect ratio could be more ideal for you.
In global security applications, for instance, images are often displayed on screens. The aspect ratio of screens is mostly 4:3 or 16:9 therefore it would make sense to choose for these aspect ratios for your sensor as well. You will then be able to project the whole sensor image on your screen.
In the case an image is being displayed on a screen it is often a requirement to have a frame rate of about 30 fps. Video with a lower frame rate could exhibit delays. This frame rate requirement could also force a limitation to your field of view. If the horizontal and vertical resolution of the sensor becomes too large it could become impossible for the interface to reach 30 fps.
To give an example let’s assume you are using a CoaXPress camera with a CXP-6, 6.25 GBit/s data rate. In that case mono/bayer cameras with all IMX sensors can run at 30 fps. However if you desire 8 bit RGB or 10 bit YCbCr at the output, it is not possible to achieve a frame rate of 30 fps with the IMX253 and IMX255 in full resolution.
In summary: although the differences seem to be small between the various IMX sensors there are some choices you have to make. The most important question you have to ask yourself is if you want to optimize for a large field of view or for display on screen.