Do the math - focal length and Format demystified

[Stefan Steib]

OK - this will probably astonish many, it is probably known in parts by most, but I haven´t seen this yet put into such a nice, stringent and logical explanation in one run. I am sure this should be posted to all the sections of GetDPI mostly for the smaller sensor cameras, but I post it to Sony now. Maybe the admins put it somewhere to keep as a basic point of start and information.

Watch it ! It is really worth doing. I definitely knew everything in this video, but even I was astonished as of the connection to the advertisements and the actual amount of cheating. I was not "aware" of this......

And finally: of course this works the same the other way upwards to medium format lenses with their bigger lens diameters. The exit pupil´s size defines the brightness of the image. And now you know why we at Hartblei use MF Zeiss glass for 35mm Lenses with very nice results. :angel:

Are Sony, Olympus and Panasonic Cheating Their Customers (crop factor with ISO & aperture)? - CanonWatch

 

[Ben Rubinstein]

I have known of this for a while. A couple of caveats. Although the math is linear, when doing the comparisons to quality for a give iso, you need equivelent sensor architecture. Trying to compare IQ from an original 1Ds to a D7000 and it will fall down. It is also dependant on needing to compare quality from bigger sensors to smaller, in the real world where iso is now so good, at least with say APC crop, until a certain point, yes there is less light hitting the sensor and yes the noise would be less from a FF sensor but it's more than good enough. What he says is true, but it is not anywhere near as relevant as it used to be with real world usage for many people. The DOF issue is a constant of course but I believe that it is understood by most these days.

In the world of GetDPI where we are so often striving for the cutting edge in the best of technology I think that video is relevant. For most buying a camera today whose images are rarely presented at anything more than screen size or A4 prints and whose cameras are not being used over iso 1600 where the difference becomes apparent and relevant for non pixel peepers, I'm just not sure that these fact matter any more. Of course it is true and yes the manufacturers are cheating with their iso values to give equivelent brightness images, but if the IQ is still good enough and to be frank these days for the majority it is, it just doesn't matter any more.

 

[mjr]

I have a question, when is this information actually important? Ok, you shoot m4/3 or whatever and pretty quickly you will know your field of view, after all a 10mm lens is always a 10mm lens, regardless of format, you only have to know what you're shooting with and the field of view.

ISO, what does it matter how ISO changes on different formats? Who cares if people think different sensors are more or less noisy at a given ISO? Surely, you are shooting your camera at its optimum setting, i.e lowest ISO for any given situation. Why do you need to know that ISO 400 on a full frame is different to ISO 400 on a different format? You're shooting with what you're shooting with. If you have all the settings on your full frame and your m4/3rds cameras exactly the same, the m4/3rd shot will be over exposed, but why would you, that seems crazy.

If you have 2 systems side by side, do people really make a shot with 1 format and then feel it's important to replicate it exactly with another format? I can't see why, surely you just use each system at its optimum settings? Is anyone really going to say, if I was shooting this on a full frame it would need 400 ISO so if I'm shooting my m4/3's camera I need to multiply that 400 by some equation to know what ISO I need to set on my m4/3 camera?

I don't get it. This video could be 5 mins long. Different formats have different crops, multiply the lens fl by the crop to know what its equivalent is on 35mm, done. ISO on different formats is equal when the amount of light hitting the sensor is equal.

What am I missing?

 

[Annna T]

OK - this will probably astonish many, it is probably known in parts by most, but I haven´t seen this yet put into such a nice, stringent and logical explanation in one run. I am sure this should be posted to all the sections of GetDPI mostly for the smaller sensor cameras, but I post it to Sony now. Maybe the admins put it somewhere to keep as a basic point of start and information.

Watch it ! It is really worth doing. I definitely knew everything in this video, but even I was astonished as of the connection to the advertisements and the actual amount of cheating. I was not "aware" of this......

And finally: of course this works the same the other way upwards to medium format lenses with their bigger lens diameters. The exit pupil´s size defines the brightness of the image. And now you know why we at Hartblei use MF Zeiss glass for 35mm Lenses with very nice results. :angel:

Are Sony, Olympus and Panasonic Cheating Their Customers (crop factor with ISO & aperture)? - CanonWatch

Well, I didn't like it. If this type of arguments are the only ones imagined by the traditional DSLRs manufacturers to counter the emergence of smaller mirror less cameras, then they won't do very well in the future.

Recent MFT bodies have come a long way : they counter the ISO/less total light problem with improved stabilization, improved sensors and faster lenses. My Olympus 12-40mm F2.8 is an extraordinary lens and way lighter than my Canon 24-105mm F4.
Also not everyone is looking for shallow DOF. Deeper DOF is an advantage in many situations.

But above all : people aren't cheated by Olympus and Panasonic as the video maintain : if they ever owned a DSLR, they have the means to understand what they will get. If they are upgrading from a smartphone or a point and shoot, they will get better IQ anyway.

I have both FF and MFT cameras and love MFT. Especially compared to my Canon FF, which aren't delivering much more than MFT.

Switching regularly between FF and MFT, I'm more amazed by the performance of MFT than by the one of FF (which we expect to be good anyway). The A7r is possibly a game changer in terms of weight and performance.. But we need to see what they will manage with lenses. Personally I prefer smaller and lighter lenses, but sharp.. I'm willing to exchange speed against diminished weight.

I didn't like the way the video insists heavily that you have to multiply the ISO BY four : it is just the equivalent of two stops, but I guess that would sound less dramatic.

Conclusion : we hare happy now, we can have FF offering performance equivalent to MF at incredible prices and MFT cameras at half the weight and size of the traditional FF DSLRs.

 

[Stefan Steib]

I don't get it. This video could be 5 mins long. Different formats have different crops, multiply the lens fl by the crop to know what its equivalent is on 35mm, done. ISO on different formats is equal when the amount of light hitting the sensor is equal.

What am I missing?

Well the simple fact that an advertised f2,8 lens on an MFT is f5,6 actually ?
And - Beni - the sensitivity of newer sensors may have been improved , but not by a square of 2 ? The number of Photons caught is still the measurement of quality. And THIS - and I fear MOSTLY this is the reason of the superiority of Medium Format. More Photons- more quality.

Solutions so far - lenses like the Voiglaender with 0,95 and the Metabones Speedbooster. AS mentioned.

I also have an Olympus e-PL3. Nice little camera. But - not a match for a larger format, equipped with wide open f-Stop lenses.

Simply a different purpose and no pun intented. Just different.
I think the Lens/Camera Makers should simply communicate that, the idea to spell focal lengths equivalents translated to 24x36 but leaving the f-stops on misleading absolute values is just...... fraud ?

And yes- all the makers are involved in that. Nikon and Canon as well.

Greetings from Germany
Stefan

 

[Annna T]

Well the simple fact that an advertised f2,8 lens on an MFT is f5,6 actually ?

I'm not sure to get it : from what I learned, F=focal length/diameter of the aperture of the lens, it is a physical dimension. I don't see how they could cheat on that.

I also have an Olympus e-PL3. Nice little camera. But - not a match for a larger format, equipped with wide open f-Stop lenses.

The E-Pl3 still has the crappy 12Meg Panasonic sensor you find on the E-P1 in 2009. You should rather compare with the E-M5, E-Pl5 etc.. Which have a state of art Sony sensor to be fair.

 

[Shashin]

Very good. There are only a few things the presenter does not understand:

Focal Length
ISO
Aperture
Depth of Field
Bokeh

It is amazing that one person can put so much nonsense in one video about photography. I stopped when he started bashing camera companies.

 

[Stefan Steib]

Therefore the explanation with the signal to noise level.
Did you watch the full 10 Minutes ?

Regards
Stefan

 

[Stefan Steib]

Very good. There are only a few things the presenter does not understand:

Focal Length
ISO
Aperture
Depth of Field
Bokeh

It is amazing that one person can put so much nonsense in one video about photography. I stopped when he started bashing camera companies.

Aha. This is of course a very detailed answer.
Could you evaluate your superior findings to enlighten the rest of the world ?

Regards
stefan

 

[Shashin]

Well the simple fact that an advertised f2,8 lens on an MFT is f5,6 actually ?
And - Beni - the sensitivity of newer sensors may have been improved , but not by a square of 2 ? The number of Photons caught is still the measurement of quality. And THIS - and I fear MOSTLY this is the reason of the superiority of Medium Format. More Photons- more quality.

Solutions so far - lenses like the Voiglaender with 0,95 and the Metabones Speedbooster. AS mentioned.

I also have an Olympus e-PL3. Nice little camera. But - not a match for a larger format, equipped with wide open f-Stop lenses.

Simply a different purpose and no pun intented. Just different.
I think the Lens/Camera Makers should simply communicate that, the idea to spell focal lengths equivalents translated to 24x36 but leaving the f-stops on misleading absolute values is just...... fraud ?

And yes- all the makers are involved in that. Nikon and Canon as well.

Greetings from Germany
Stefan

The f-number does not indicate depth of field, but luminance at the image plane.

The size of the sensor is irrelevant in terms of exposure (as this speaker claims)--do you increase exposure if you crop your image? Pixel size determines the number of pixels gathered. But it is not a measure of noise as there are more factors that contribute to noise. Fuji X-trans sensors have much better noise suppression than Bayer sensors.

 

[Shashin]

Aha. This is of course a very detailed answer.
Could you evaluate your superior findings to enlighten the rest of the world ?

Regards
stefan

Focal length is not field of view. It is the distance form s' to the image plane. Field of view is a product of focal length and the image area. But sensor size does not change focal length.

ISO (ey-es-oh) for digital cameras is not defined the same way for film. ISO has clearly specified what ISO for digital cameras is. And, surprise surprise, is linked to noise. Sensor size is irrelevant to exposure--it does not matter how many photons the entire area of the sensor intercepts. Pixel area is the important factor.

Aperture, f-number to be exact, is proportional to luminance at the image plane. It is not a description for depth of field. Also, f-number is based on the exit pupil size, not entrance pupil size as it correlated to the angular size and with complex optics, tho is important, but this is a nit pick over a nit.

Depth of field is not an f-number while that is a variable. It is not Bokeh, which is simply the rendering of the out of focus area, which is at all apertures.

Stefan, these are not my "superior" findings. Anyone with a basic foundation in photography should spot these.

 

[Stefan Steib]

The f-number does not indicate depth of field, but luminance at the image plane.

The size of the sensor is irrelevant in terms of exposure (as this speaker claims)--do you increase exposure if you crop your image? Pixel size determines the number of pixels gathered. But it is not a measure of noise as there are more factors that contribute to noise. Fuji X-trans sensors have much better noise suppression than Bayer sensors.

??? new laws of optics developed by you ?

"...The size of the stop is one factor that affects depth of field. Smaller stops (larger f numbers) produce a longer depth of field, allowing objects at a wide range of distances to all be in focus at the same time...."

Aperture - Wikipedia, the free encyclopedia

Regards
Stefan

 

[Annna T]

Therefore the explanation with the signal to noise level.
Did you watch the full 10 Minutes ?

Regards
Stefan

Is that addressed to me ?

I looked at the whole 39min and thought that it could be summarized in 5min and so I lost 34-35min.

I know very well what concerns the relationship between total light hitting the sensor and noise, but I still don't see how Olympus et al. are supposedly cheating when they say that their lens have a max aperture of F2.8, because this is just a physical relationship between the focal lens and the aperture diameter.

The total light reaching the sensor and the noise relationship is another story which you can bring out if you want to speak about "full"equivalence, but I still don't think that Panasonic or Olympus are cheating anyone.

I'm sorry, but I think that with the recent sensor developments, MFDB will slowly become obsolete, just like 4x5 and 8x11 view cameras. Offering a small portable view camera able to mount an A7r in the rear standard could be a hit however.. Especially if that allows architecture photographers to rise and fall the rear standard. That Sony sensor in the rear standard and some sharp lenses with enough coverage on the front standard with everything working electronically and seamlessly and not over 2kg for the whole rig.... That would be great.

Or may be we just need Zeiss to make a super TSE lens for the A7r.

 

[Shashin]

??? new laws of optics developed by you ?

"...The size of the stop is one factor that affects depth of field. Smaller stops (larger f numbers) produce a longer depth of field, allowing objects at a wide range of distances to all be in focus at the same time...."

Aperture - Wikipedia, the free encyclopedia

Regards
Stefan

It is a variable in depth of field, but f/5.6 does not give a specific DOF in and of itself nor describes it. The f-number system is so you can get equal exposure among different focal lengths and allowed exposure to be standardized simply because luminance at the image plane is directly correlated.

From your same Wiki article:

The aperture stop of a photographic lens can be adjusted to control the amount of light reaching the film or image sensor. In combination with variation of shutter speed, the aperture size will regulate the film's or image sensor's degree of exposure to light. Typically, a fast shutter will require a larger aperture to ensure sufficient light exposure, and a slow shutter will require a smaller aperture to avoid excessive exposure.

 

[Stefan Steib]

As the video said. If you corellate all factors - photon count per available area, f-stop and focal length you´ll end up with a reationship that shows it is NOT correct to just use the relative opening compared to focal lenght and exit pupil. The signal to noise ratio (collected Photons) comes into the calculation.
Nothing else is said in the video.
And yes, the sensors have become better, but so did the MF Sensors. I am pretty sure the new Sony CMOS Sensor in the Phase/Blad/Pentax is no worse than a modern MFT/APS-C or even smaller sensor.
There may be backside illumination and special schemes to improve light collection. But the relative factor of enlargement in the end answers also your question about cropping. If you crop you will even loose more light/information+resolution.

Regards
Stefan

 

[Shashin]

As the video said. If you corellate all factors - photon count per available area, f-stop and focal length you´ll end up with a reationship that shows it is NOT correct to just use the relative opening compared to focal lenght and exit pupil. The signal to noise ratio (collected Photons) comes into the calculation.

??? A pixel pitch of a given size collects the same number of photons at a given f-number. This has nothing to do with sensor size. Or are you saying cropping an image makes it noisier?

Since we like math, this is the way of calculating the number of photons hitting a pixel. From CCD Arrays, Cameras, and Displays, Second Edition by Gerald C. Holst published by SPIE Press and JCD Publishing, page 34:

n = (pi/4)(L*A/F^2(1+M)^2)To*Ta*t

Where

n is the number of photons striking the pixel
L is the spectral photon sterance
A is the pixel area
F is the f-number
M is the optical magnification
To is the spectrally averaged optical transmittance
Ta is the spectrally averaged atmospheric transmittance
t is the integration time--shutter speed

 

[Stefan Steib]

"...??? A pixel pitch of a given size collects the same number of photons at a given f-number. This has nothing to do with sensor size. Or are you saying cropping an image makes it noisier?..."

If you bring the cropped part of the image back to the same size - of course. The noise in this case will not be more per se, but as the percentage of needed enlargement will grow, the existing noise will be enlarged too ....

Which is btw clear from the definition of M in your shown formula for the needed Magnification.

Regards
Stefan

 

[Shashin]

Stefan, no, the magnification in the equation is independent of the format size, but simply the magnification at the image plane. You need to account for focus distance.

You are also very confused by noise. It is a property of the pixel, not the viewing distance, which is what you are saying if the making the cropped area larger or smaller changes that. Noise is not a subjective quality.

 

[Stefan Steib]

This all has been discussed in length - e.g. here:

Full Sized vs. Cropped Sensors
.....
1. The sensor size alone determines the maximum useful f-number (N); and, in fact, the maximum f-number for high-resolution photography is given by 0.5 times (sensor diagonal in mm).


2. We can take an essentially identical photograph with any sensor size by scaling the focal length, the f-number, and the ISO sensitivity.


3. The smallest sensor we considered (denoted 1/2.5”) achieves maximum useful DoF at an f-number of N=3.5 while the full frame 35 mm sensor requires N=21 for a factor of 36 difference in transmitted light. If the full frame sensor gives the same signal/noise ratio (S/N) at ISO sensitivity 1600 as the small sensor does at ISO 80, the small sensor can still use a higher shutter speed. A PS sensor that could give low noise at ISO 800 or 1600 would appear to have a real advantage over FF sensors.

4. If maximizing the DoF is not the aim, larger sensors clearly win because of their ISO sensitivity advantage. A fast lens (N=1.4) with a full frame detector is impossible to match with the small sensor. Probably N=1 is the maximum aperture we can expect with a small sensor, and no company at present even offers N=2. The take home lesson is that small sensors should be coupled with large aperture lenses, i.e. small N values. Also, small sensors that support large ISO sensitivities should be sought. The vendors are showing some interest in higher sensitivities, but larger lenses are in conflict with their drive to smaller cameras. Unfortunately, none of the available PS cameras offer very high quality lenses......
______________________________________________________

It is clear that a small sensor is not having the same signal to noise ratio like a larger one, both with the same resolution. the conclusions in the video are exactly stating that.

there is another very good article
Do Sensors ?Outresolve? Lenses?

see

Figure 2. Graphic representation of the MTFs of two hypothetical lenses and a sensor of 100 lp/mm (5 microns). Wavelength of the light is 0,000555mm.

and:
"...The signal-to-noise ratio, however, imposes an inflexible limit to the effective resolution of the whole system, mostly due to photon shot noise.
."

Regards
Stefan

 

[tsjanik]

I didn't make it past the first five minutes. He states focal length is no longer useful because of different sensor sizes; so why was it useful with film that comes in different sizes? His comment about ISO variation for different size sensors is disingenuous; yes a small sensor gets less total light than a larger sensor at a constant f value, but the light/unit area is the same. Smaller sensors often has smaller photo sites which is a factor in light collected/pixel. This is when I stopped watching.
I must admit I was already disposed not to like this presentation when I saw the author. He has a review of the 645Z up too and he clearly is unfamiliar with the Pentax 645 system, but that didn't stop him. I think he is trying to generate web traffic a la Ken Rockwell.

Tom