How many pixels

Replytoken

Mu-43 Hall of Famer
Joined
May 7, 2012
Messages
3,612
Location
Puget Sound
Real Name
Ken
If you want to compare with the same technology, then I do not see any reason why you cannot scale up. If you are considering a different sensor technology and/or pixel size, then would you not use that same technology on the size of the G9 sensor in your comparison?

--Ken
 

ScottinPollock

Mu-43 Veteran
Joined
Oct 26, 2017
Messages
444
A full frame sensor with the same pixel density as the G9 would have 4 times the number of pixels.
 

wjiang

Mu-43 Legend
Joined
Sep 7, 2013
Messages
7,330
Location
Christchurch, New Zealand
You can just scale up, however economics and other factors will likely make it impractical. Many more pixels requires a higher readout and media write time, more processing power (and more heat). IC fabrication yields will be lower, leading to significantly higher costs, etc.

Oh, and using the same pixel density means you get the same noise characteristics at a per pixel level, so unless you purposefully down sample in PP, there is no advantage to having a bigger sensor for noise at a pixel level.

That's why the S1r is only 47MP I suspect.
 
Last edited:

wjiang

Mu-43 Legend
Joined
Sep 7, 2013
Messages
7,330
Location
Christchurch, New Zealand
Well... FF sensors dimensions do vary. My initial math was based on the standard 2x crop. Comparing the dimensions and pixel density of the G9 to the sensor dimensions of a D850, it'd be 3.815.
Yeah it depends on the mount specifications, like how many pixels around the edges are reserved for guard and calibration/software correction. The modern mirrorless systems all use some form of software correction AFAIK.
 

Replytoken

Mu-43 Hall of Famer
Joined
May 7, 2012
Messages
3,612
Location
Puget Sound
Real Name
Ken
Well... FF sensors dimensions do vary. My initial math was based on the standard 2x crop. Comparing the dimensions and pixel density of the G9 to the sensor dimensions of a D850, it'd be 3.815.
Very true. I forgot that the dimensions do vary.

--Ken
 

ralf-11

Mu-43 All-Pro
Joined
Jan 16, 2017
Messages
1,290
...
Oh, and using the same pixel density means you get the same noise characteristics at a per pixel level, so unless you purposefully down sample in PP, there is no advantage to having a bigger sensor for noise at a pixel level.
...
unless you can simulate the spatial summation in the vertebrate eye...
 

ToxicTabasco

Mu-43 All-Pro
Joined
Jul 2, 2018
Messages
1,299
Location
South West USA
Real Name
ToxicTabasco
Some contend that the FF sensor is 4 times the size of the MFT. With the pixel density of the G9 that would make it a 80+mp sensor. However using the crop factor of 2x that would be a 40mp ff equivalent. Which makes me wonder if the higher density MFT sensor would produce sharper images without aliasing.
 

wjiang

Mu-43 Legend
Joined
Sep 7, 2013
Messages
7,330
Location
Christchurch, New Zealand
Some contend that the FF sensor is 4 times the size of the MFT. With the pixel density of the G9 that would make it a 80+mp sensor. However using the crop factor of 2x that would be a 40mp ff equivalent. Which makes me wonder if the higher density MFT sensor would produce sharper images without aliasing.
:confused-53: It's a 2x linear crop, which equates to 4x surface area (2^2=4). Different ways of saying the same thing.
 

Michael Meissner

Mu-43 Veteran
Joined
Sep 19, 2018
Messages
426
Location
Ayer, Massachusetts, USA
Some contend that the FF sensor is 4 times the size of the MFT. With the pixel density of the G9 that would make it a 80+mp sensor. However using the crop factor of 2x that would be a 40mp ff equivalent. Which makes me wonder if the higher density MFT sensor would produce sharper images without aliasing.
However, note as you increase the pixel density, diffraction becomes an issue sooner with larger f/stops. According to this calculator:
Diffraction can start to be noticed if you are pixel peeping on an image from a 4/3rds sensor at f/7.3 for a 16MP sensor and f/6.5 for a 20MP sensor. But it likely means for most people when you view an image reduced for web viewing or printed, f/8 should be fine, but you might start noticing things being less sharp at f/11.

<edit>
Now Sharp's new 8K micro 4/3rds video camera would start seeing diffraction at f/5.1 (due to the 33MP sensor needed for 8K). Now, I imagine if you are buying a $3k camera, you are not going to be putting a consumer lens like the 14-42mm variants on it. The f/2.8 pro zooms, the fast primes, and the forthcoming Panasonic 10-25 mm f/1.7 zoom will probably be ok, but you might want to watch out for lenses in the f/5.6 range.
 
Last edited:

ToxicTabasco

Mu-43 All-Pro
Joined
Jul 2, 2018
Messages
1,299
Location
South West USA
Real Name
ToxicTabasco
Interesting, based on the above info, FF still can go 2x more MP on a D850 before diffraction. And that m4/3 is already at the limit for diffraction in the amount of MP on the sensor. Thus, now I know why people are saying MFT is dead. No ability to increase resolution due to no room for MP unless they solve diffraction in the camera or lens at f/11.

Also, FF with a sensor 4x larger still has a lot of room for ISO performance. Where the m4/3 is maxed out. Perhaps the FF cameras will have 4 stops of light improvement over the m4/3 someday. Now I can see why Lumix did the S series. For those who need it.
 

Machi

Mu-43 Top Veteran
Joined
May 23, 2015
Messages
623
@ToxicTabasco
I not sure on what you based your conclusion. FF has in principle 2 stops advantage over M43 and that will stay the same if technology will be comparable. Effects of diffraction are matter of pixel size and f ratio to the aperture so FF with same pixel size as M43 would be similarly limited.
Also this "m4/3 is already at the limit for diffraction in the amount of MP on the sensor" is completely false. For retaining all details in the scene imaged by some lens one needs to go to f/16 with 16MPix M43 sensor, ~f/14 with 20MPix sensor, f/11 with 33MPix sensor and f/8 with 64MPix M43 sensor. That's why Olympus limits hi-res mode to f/8 as beyond that it cannot observe more details in the scene. At numbers f/4-f/16 diffraction can cause blurring details but that depends on how good optically is the used lens. In case of lens with well corrected optical aberrations blurring because of diffraction will be visible sooner than with not so well corrected lens.
That's why every lens has a sweet spot where it's sharpest. At wide apertures it's limited by aberrations and at low apertures it's more and more limited by diffraction.
Pixel count in general is matter of technology, convenience and cost effectiveness. Diffraction is not limiting factor for Pixel count in majority of cases with exception of ultrazooms (P1000 and his brothers).
 
Links on this page may be to our affiliates. Sales through affiliate links may benefit this site.
Mu-43 is a fan site and not associated with Olympus, Panasonic, or other manufacturers mentioned on this site.
Forum post reactions by Twemoji: https://github.com/twitter/twemoji
Copyright © 2009-2019 Amin Forums, LLC
Top Bottom