It's the effect of the wavelength of far infrared light being quite a bit longer.

Think of a reduction to extremes: if you have a sensor that is a centimeter square and you're trying to 'catch' a wave that is a meter long there is a fair chance the sensor will be bypassed entirely, but if you are trying to catch millimeter waves your sensor will be easily able to capture the photons.

The most practical example of this effect is the size of radio antennae, they get longer as the wavelength gets longer.

Wow you're right, huge difference in the sizes of the "full res" images:

> MIRI: Full Res, 1306 X 1133, TIF (1.78 MB) [1]

> NIR Cam: Full Res, 4833 X 4501, TIF (24.06 MB) [2]

Maybe it's a mistake, they suggest it should offer an "incredible amount of detail":

  This Mid-Infrared Instrument (MIRI) image also offers an _incredible amount of detail_, including a cache of distant galaxies in the background. 

[1] https://webbtelescope.org/contents/media/images/2022/033/01G...

[2] https://webbtelescope.org/contents/media/images/2022/033/01G...

MIRI works at longer wavelengths than NIRCam, so its angular resolution is lower (longer wavelengths mean more diffraction). It also has a smaller field of view.

Those two factors mean that it has fewer pixels per image.

Well, it's incredible in the sense that I can't believe it