OUCH! Digitalcameraworld: New cameras are not getting any better!

Brownie

Legendary Member
Site Supporter
Site Supporter
Followers
21
Following
1
Joined
Oct 22, 2021
Posts
4,944
Likes Received
3,821
Name
Tim
Country
United States
City/State
SE Michigan
From SAR.

I don't think any of these charts are earth shattering, these differences are pretty small. The point they're trying to make is that there haven't been any massive improvements. Link to original DCW article at the bottom.

Evidently improvements in sensor tech – at least in terms of core stills image quality – have been slowing in recent years. Instead, we now see the likes of Sony marketing other features, like the speed of a sensor’s readout, and its corresponding impact on burst shooting speed and video capture performance.

SNR has much to do with pixel size. The RIII is the winner based on larger pixels. Not surprising except when the age is considered. It does seem to indicate that they haven't improved sensor and processor technology as much as we might hope.
Screenshot-2023-08-01-at-11.13.55.png


The DR is a bit different. Looks like they have improved that to some degree, but really only between ISO 200 and 2400+/-. The RV is improved a bit over the RIV in the higher ISO range.
Screenshot-2023-08-01-at-11.14.00.png



This is all Pixels. What's surprising is it looks like they tweaked the RV for some reason, leading to lower resolution (very minor) across most of the scale. Maybe this was to improve something in another area, since most of this stuff requires a trade-off of some kind.
Screenshot-2023-08-01-at-11.13.50.png


SAR link:

DCW Link:
 
Interesting, but not surprising. The thing that has moved on is the in camera processing and post processing so we can get more out of it.
 
Interesting, but not surprising. The thing that has moved on is the in camera processing and post processing so we can get more out of it.
The next biggest will likely be computational NR, such as cellphones and what OM has employed. Instead of implementing NR in post, the camera will be programmed to provide the best NR possible for the image, based on settings and metering. By the time you upload it the image will be as noise-free as possible. I think OM's kicks in automatically at 3200 or so. Not surprising it would show up first on a smaller sensor camera.
 
Agreed, and actually, you have to admire what they are doing. It's pushing the boundaries, but whether it can be implemented on high MP full frame sensors is yet to be seen.
 
Agreed, and actually, you have to admire what they are doing. It's pushing the boundaries, but whether it can be implemented on high MP full frame sensors is yet to be seen.
From a resolution/pixel size standpoint, OM's 20MP M-4/3 sensor is identical to an 80MP FF sensor. Panasonic's 25MP is the same as 100MP in FF. Given that, I don't see where there'd be an issue with 61MP and beyond.
 
It seems that the further I go down my photography journey the less I care about the nth degree of sharpness or low noise. My "best" iq camera (GFX100S) is the one I use the least.

I definitely appreciate the advances in af/tracking and subject recognition. Those things make a real difference to me an as previously mentioned - software is going to play an even larger role in the final results regardless.
 
So basically the improvements are that we are getting the same/similar results but with more pixels to work with!
 
From a resolution/pixel size standpoint, OM's 20MP M-4/3 sensor is identical to an 80MP FF sensor. Panasonic's 25MP is the same as 100MP in FF. Given that, I don't see where there'd be an issue with 61MP and beyond.
Interesting stats for those sensors, because when I zoom in on some OM1 shots on forums etc. they don't hold up as good as they initially looked.
 
Interesting stats for those sensors, because when I zoom in on some OM1 shots on forums etc. they don't hold up as good as they initially looked.
No, because you're starting with fewer pixels.

The problem is the term 'resolution'. We use it to describe higher megapixel cameras, but that's not where the term comes from. What it really describes is the sensor's ability to resolve between contrasting points from pixel to pixel. Smaller pixels yield higher resolution than larger pixels. So, a camera with more megapixels has higher resolution because there are smaller pixels, not because there are more of them. But because there are more of them, you can pixel peep a lot more.

Since we know an M-4/3 sensor is roughly 25% the size of a FF sensor, we know that a M-4/3 sensor of a given number of pixels has the same 'resolution' as a FF sensor with 4X the number of pixels. So, a 16 MP 4/3 sensor can 'resolve between pixels' the same as a 64MP camera. It doesn't mean the image will be as large or as croppable because even though the pixel size is the same, you're starting with far fewer.
 
No, because you're starting with fewer pixels.

The problem is the term 'resolution'. We use it to describe higher megapixel cameras, but that's not where the term comes from. What it really describes is the sensor's ability to resolve between contrasting points from pixel to pixel. Smaller pixels yield higher resolution than larger pixels. So, a camera with more megapixels has higher resolution because there are smaller pixels, not because there are more of them. But because there are more of them, you can pixel peep a lot more.

Since we know an M-4/3 sensor is roughly 25% the size of a FF sensor, we know that a M-4/3 sensor of a given number of pixels has the same 'resolution' as a FF sensor with 4X the number of pixels. So, a 16 MP 4/3 sensor can 'resolve between pixels' the same as a 64MP camera. It doesn't mean the image will be as large or as croppable because even though the pixel size is the same, you're starting with far fewer.
So ultimately, FF 61mp is better.
 
So basically the improvements are that we are getting the same/similar results but with more pixels to work with!
Or they are just in different features of the sensor. This whole thing is kind of cherry picking what makes a sensor better.
 
So ultimately, FF 61mp is better.
Well, some think so. Some don't. I think it's a bit much myself. But your post was based on if they can implement NR on a larger sensor. The answer is it shouldn't be a problem based on pixel density, resolution, pitch, or whatever you want to call it.
 
Well, some think so. Some don't. I think it's a bit much myself. But your post was based on if they can implement NR on a larger sensor. The answer is it shouldn't be a problem based on pixel density, resolution, pitch, or whatever you want to call it.
I guess not, no. There already is some in the A7's anyway, I turn it off.
 
Back
Top