Mobile phone sensor sizes vs digital cameras

Joined
10 May 2004
Posts
12,981
Location
Sunny Stafford
Megapixels is how cameras are marketed, but the camera's sensor (the "retina") is what determines how clear the image is. It is normally displayed as a fraction, less than 1. The closer to 1 it gets, the bigger it is, and SLRs can go above 1.

Rough examples:

1/10 - camera phones from 10-15 years ago
1/6 - compact cameras
1/1.7 - bridge cameras
4/3 - four thirds
1.5" (3/2) - Canon G1X
35mm (full frame) - SLRs

I've been using Google and https://www.dxomark.com to get sensor sizes for phones and their so-called DXOMARK scores. Think of it as being Passmark scores but for cameras instead of CPUs and GPUs.

Samsung S4: sensor 1/2.3, score 75
Samsung Note 4: sensor 1/2.6, score 83
Samsung S8: sensor 1/2.55, score 88
Samsung A50: sensor 1/2.5, score 85
Samsung A71: sensor 1/2.8, score 83
Samsung S20 Ultra: sensor 1/1.33, score 122

I currently own the S20 Ultra and the Canon G1X, which incidentally scores only 60, which is half of the Ultra. When I take photos of the same subject with both devices, the Ultra comes out tops every time. Is there any idea why this is, even though the Ultra's sensor is still below 1? Do the scores really mean much? The G1X (Mark I) is a 2012 model, but it has served me well and I want to hang onto it. Just a bit miffed as the Ultra is the first phone I've had that takes better pictures (out of the box) than an actual camera.
 
Technology moves on and things get smaller. It's likely the same technology and performance from full frame cameras in 2012 is now going in to camera phones now.

Same reason if you compared a full frame camera from 2012 to a modern day equivalent, the newer one is going to provide better results despite being the same sensor size.
 
So many things affect the final photo though especially on phones - despite on paper good specs for instance my Note 4 largely produces inferior images to my Note 1 especially in less ideal light conditions and that isn't entirely down to software processing.
 
Back
Top Bottom