New Ricoh GR, Nikon A, Sony RX100 - what happened to the law of physics?

wolfie

Veteran
For years the catch cry for small sensor cameras was that if you wanted a bigger sensor in a compact you would lose out becasue a bigger lens was the inevitable outcome - but doesn't seem to be the case when the GR has a sensor 900% bigger than the 1.7 previous version and the lens is not even 100% bigger from what I see ... so what gives?:confused:
 
Magic. According to Arthur C Clarke, that is.

That new Sigma lens looks very interesting, indeed. I wonder if this design will influence a future m43 version? Note that is is 18-35mm, though. The Canon EF-S zoom with constant f2.8 is 17-55mm, and while their smaller 17-85mm is f3.5-4.5. If the Sigma f1.8 zoom was longer, it would be MUCH larger.
 
I think 2 important factors are:
-digital lens correction, which on some cameras probably happens in the raw files already, and
-microlenses on the sensors negating the need for light to hit the sensor at a 90-degree angle, meaning the light doesn't have to be "steered" as much as it used to in order to fall into the "light gathering bucket" that's a pixel on a sensor (film cameras often have smaller lenses than digital cameras; partially because those photos weren't viewed as large as digital photos often are, so IQ faults aren't noticed as easily, but also partially because for film, it doesn't really matter from which angle the light ray hits the film).
 
I think 2 important factors are:
-digital lens correction, which on some cameras probably happens in the raw files already, and
-microlenses on the sensors negating the need for light to hit the sensor at a 90-degree angle, meaning the light doesn't have to be "steered" as much as it used to in order to fall into the "light gathering bucket" that's a pixel on a sensor (film cameras often have smaller lenses than digital cameras; partially because those photos weren't viewed as large as digital photos often are, so IQ faults aren't noticed as easily, but also partially because for film, it doesn't really matter from which angle the light ray hits the film).

+1
 
Digital correction is the new norm. M43 lenses were corrected this way, eg Oly 12mm, Panasonic 12-35mm has ~ 5% distortion. Other manufacturers are copying that also. See the RX100 review at lenstip's polish site:

Test Sony DSC-RX100 - Optics - Camera Test - Optyczne.pl

It is clear that in the case of 10.4 mm we are dealing with a very large barrel distortion, which the program Imatest estimated at -11.2%.

LX7:
It is clear that in the case of 4.7 mm there is a very large barrel distortion, which the program Imatest estimated at -9.69%. It would seem that the value comparable with that obtained Sony RX100.

Fuji X10:
For wide-angle effect even throws in the eyes, and the measurements indicate that we are dealing with barrel distortion at -6.17%.

Canon S100:
The widest viewing angle is very large distortion rate of -10.6%.

Olympus XZ-2
In the case of RAW files, the measured values ​​are equal to -3.82% for wide-angle, by -0.39% area equivalent to 50 mm, and 0.55% for the telephoto position. These excellent results show that it is possible to create a clear lens with a well-corrected lens distortion. In this respect, Olympus deserve applause.

Other then photozone and lenstip, most of the review sites do not report this as adobe/cameras correct it as default. I guess we will see the distortion correction for GR & Nikon A when these cameras are reviewed. The only problem I see with digital correction is when you have people shots at the sides/edges of the frames, they are stretched.
 
To be sure, lens correction and microlenses are a big part of the change. So is the very latest in computerized lens design, which has changed the craft from art to science.

Five or more years ago, on another photo website, the so-called experts would shut you down if you dared predict that certain things would be possible with cameras and lenses. As Wolfie posted, the cry was the unbreakable laws of physics. Well, the laws of physics may be unbreakable, but apparently many people didn't understand those laws or didn't realize there were ways to use those laws to reach certain goals. Many others were simply guilty of conventional thinking.

Mind you, many of these fabulous new lenses come with a price - as in a high pricetag. But that is likely to change as well with time.
 
So, is there a theoretical limit as to how good this digital wizardry can get? Are we going to see a Fuji XE-1 sized camera with a 24-200/1.8 at some point? :D

All I really want is the X100 with the X10 lens. I think I'd be happy at that point. Cue the laughter...
 
So, is there a theoretical limit as to how good this digital wizardry can get? Are we going to see a Fuji XE-1 sized camera with a 24-200/1.8 at some point? :D

All I really want is the X100 with the X10 lens. I think I'd be happy at that point. Cue the laughter...

I haven't a clue what the limits are. So far, it seems the current ways of making lenses smaller haven't plateaued yet. Perhaps in the future, liquid lenses will take off in the future, perhaps sensors will bend a little to accomodate more extreme lens design, and new ways of using carbon might become a huge factor (diamond is made of carbon so it's obviously possible to give carbon-based structures some impressive optical qualities, and scientists all over the world are doing incredible things with placing carbon atoms right where they want them)...

So all in all, my guess is, we're in for a treat in terms of small, high quality optics. I can't wait to see the first smartphone-sized full frame fast zoom camera! I do hope that by that time, someone will have figured out a nicer user interface than a smartphone's touchscreen though!
 
For years the catch cry for small sensor cameras was that if you wanted a bigger sensor in a compact you would lose out becasue a bigger lens was the inevitable outcome - but doesn't seem to be the case when the GR has a sensor 900% bigger than the 1.7 previous version and the lens is not even 100% bigger from what I see ... so what gives?:confused:

The alleged "physicists" were wrong LOL. But do note that Ricoh had to slow down the lens from f1.9 to f2.8 to make it fit the package. That plus the digital correction is making this stuff possible.
 
We live in interesting times ...

It certainly would be interesting to know how much is purely firmware correction and if there are special microlenses etc involved, ceratinly as far as i know, neither Ricoh or Nikon are claiming any super-duper rare earth glass in the lens is responsible for the compact dimensions.
From a quick read on Ming Thien's site about the new GR he attributes some of it down to be a fixed lens verusus and interchangeable lens so that it can be pared down exactly to suit the sensor, which also means it can perform better optically as well.
 
It certainly would be interesting to know how much is purely firmware correction and if there are special microlenses etc involved, ceratinly as far as i know, neither Ricoh or Nikon are claiming any super-duper rare earth glass in the lens is responsible for the compact dimensions.
From a quick read on Ming Thien's site about the new GR he attributes some of it down to be a fixed lens verusus and interchangeable lens so that it can be pared down exactly to suit the sensor, which also means it can perform better optically as well.

When Fuji released the x100.. During one of the interviews they said they were able to make the 23f2 so short by designing it specifically for the sensor.. But also someone did some more detailed looking at an x100 which had been taken apart and the rear lens element to sensor distance is very close (far closer than the norm). It s a special design of the lens in more ways then one..

But other aspects such as lens correction firmware must come into play. An example look how huge the Zeiss 24 is for the Nex series versus the 23 on the x100. I have heard people say that the Zeiss lens is that way because it is designed the normal way to not require any lens firmware corrections..

Gary
 
Software can correct barrel distortion, CA, etc, but tend to make the corner softer as an expense (there is no free lunch)

If Ricoh and DPR's samples are to be trusted, the new GR has very good corner from wide open. I believe it is from their microlenses magic, which is not surprising given their experience with the M module.
 
I don't think they used any microlenses, otherwise it would have been in their marketing ads like they did with their M module. Since the sensor is from Sony, it should be the same sensor as nex-5n, which magicaly is better with wide angle lenses compared to other Sony models nex-3/5 and nex-7. Also M43 has bunch of lenses with good corners with ~5% distortion correction eg Oly 12mm f/2, Panasonic 20mm f/1.7 pancake lens.

Software can correct barrel distortion, CA, etc, but tend to make the corner softer as an expense (there is no free lunch)

If Ricoh and DPR's samples are to be trusted, the new GR has very good corner from wide open. I believe it is from their microlenses magic, which is not surprising given their experience with the M module.
 
Back
Top