cyainparadise
Veteran Member
But, that doesn't make it right to spread false information. Many people want to understand what their equipment, and making things up, doesn't help that person to better understand their equipment.
-
Welcome to the new forums! Please read this first. For known issues we are working to resolve, click here.
If I could place the RX100 sensor in A500 for bird shooting, .........
- Thread starter Faith Yeung
- Start date
That's a little harsh, I clearly labelled my post as theorizing and did not dress it up as the absolute truth.
I also think your not giving it even a chance, ISO standards are great, but don't pretend they tell you everything about a sensor, otherwise DxO marks would be useless right?
I also think your not giving it even a chance, ISO standards are great, but don't pretend they tell you everything about a sensor, otherwise DxO marks would be useless right?
cyainparadise
Veteran Member
I'm not talking about sensors. The topic is, and was, 'apertures'. You, somehow turned the subject to sensors. The standard of how to determine the aperture size was set long before digital came about.
dlkeller
Veteran Member
The flaw or at least irrelevence of your argument is that total light has nothing to do with exposure, it is the light on a specific area that determines the exposure. I would guess, using your argument, that the ISO of the sensor/film would also vary depending on the size of the sensor/film format.
The topic is the effective aperture when using a different sensor size. Just as the way to determine focal length has been determined long before digital came about yet we can still talk about effective apertures.
I'm not trying to change the definition of aperture or how exposure works, if something that I said has broken an actual physical law, then please be specific and explain it. I would honestly love to hear it.
Saying that the accepted knowledge and standards support your views without explaining why isn't helping me or anyone else.
Draek
Senior Member
Nope, read the post again. At most I'm saying that *ISO* is something that came about with photography, but even that would be a stretch.
Draek
Senior Member
For the most part, no, it doesn't. Certainly not to that degree.
cyainparadise
Veteran Member
If the aperture were to change with the sensor size, then you would expect every camera/lens maker would be facing a class action lawsuit, as none of them make mention of it, when you buy a FF lens.
For example, have you seen Sony, Canon, Nikon, Sigma, Tamron, etc., put anywhere in their sales material that if you were to use the lens on a camera with a smaller than FF sensor, that the maximum aperture is reduced?
If I used an A57, and bought a 70-200mm f/2.8, that is made for a FF body, and I were to find out that on the A57, the maximum aperture is reduced by a factor of 1.5X, you can bet that I would be contacting an attorney to sue for false advertising.
You and what's-his-face don't know what you're talking about.
I have stated clearly that the physical aperture size has NOT changed. We are talking about the equivalent aperture that would give the same result on FF. No one is suggesting that the physical size of the glass changes depending on the sensor in the body attached.
You seem to be taking this the wrong way for some reason, there is no hostility intended in any of my posts, and this is starting to turn into the sort of conversations I hate to have.
So good luck with everything, I'll stay away from now on. Cheers.
You seem to be taking this the wrong way for some reason, there is no hostility intended in any of my posts, and this is starting to turn into the sort of conversations I hate to have.
So good luck with everything, I'll stay away from now on. Cheers.
It depends on what you wanna use the aperture for. In the end you can do everything without the FF equivalent aperture but you cannot without the real aperture.
If you care about the shutter speed, all that matters is the real aperture.
If you care about the DOF you do not need crop factor or relative aperture. Just use the focal length, the real aperture and the size of the sensor. No need for the fancy relative aperture.
If you care about composing the picture and the field of view you do not need the relative aperture either. Just look through the viewfinder or calculate the stuff (without the real aperture).
The relative aperture only matters when you want to compare FF aperture to different sensor sizes. And even there you can just calculate everything that matters without even knowing the crop factor or relative aperture.
Light gathering abilities do not matter in any way. They do not directly influence every day photography. Using this as an argument does not make sense.
Every sensor is acting different when getting exposed to the same number of photons. It makes more sense to compare the noise at certain ISO levels and comparing the dynamic range.
All you do is confuse people. A FF lens mounted to APS-C does not change in any way and the different sensor sizes have a certain effect on the resulting picture.
Most of the people never use FF cameras. Even in the rare case someone wanna compare the aperture on a non FF sensor to anything else calculating an aperture ratio using an APS-C sensor as the comparing element would make more sense. Almost everybody used a APS-C camera at some point.
Dave Oddie
Senior Member
But the exposure properties have not changed. Set ISO 100 on an A77 and it is as sensitive light as the A99's sensor when set to ISO 100. So what you do not have to do is open your lens up more on the A77 to get a correctly exposed shot when you set the camera to ISO 100 compared to when using an A99.
If there was a change in exposure properties then you would have to use different settings for aperture and/or shutter on an A77 and A99 when they were both set to ISO 100. That is clearly not the case.
Now I am sure someone will say because the A77's senor is nosier then the A99's then there is indeed a loss of sensitivity with the smaller sensor Even that isn't correct because whatever the physics of it as far as noise goes that does not define how sensitive the sensor is. This was also true in the days of film. Films of the same speed had different amounts of film grain.
I remember an using ISO 64 slide film made by Perutz. It was as grainy as hell but gave some fantastic colours (and it was cheap!). Compared to Kodachrome 64 which was virtually grain free it was far grainier (noisier in today's terms). It was however equally sensitive.
Bottom line is the "film" in a Nikon V1 is as sensitive as the "film" in A77 is as sensitive as the "film" in an A99. If it were not you would need to set wider apertures or slower shutter speeds when using the V1 and A77 to obtain a correct exposure and you do not.
Also consider external exposure meters (yes they still make them and very sophisticated they are too). These would be impossible to use and obsolete if sensitivity varied with sensor size and smaller sensors were indeed less sensitive.
It has not changed period. There is no equivalence issue at all.
Resolution is not the same as sensitivity and it is the sensitivity of the sensor that dictates how much light is required to obtain a correct exposure.
At ISO 100 you will set F5.6 and 1/1000 for EV15 on any camera with any sized sensor. Basically whatever size the sensor is then the "sunny 16" rule will always apply.
It doesn't matter so long people do not start thinking they can only use their small sensor cameras in blinding sunlight because they have been misled that they require much more light to operate in.
Allan Olesen
Senior Member
- Messages
- 3,391
- Solutions
- 7
- Reaction score
- 1,586
I can see that I have really stirred things up. Almost everybody thinks that I am wrong. I am not going to reply to every single post, so here are instead two replies to all of you.
It seems that my opponents in this thread can be divided into two groups:
1. Those who do not agree that a 2.7x crop sensor behind an f/5.6 lens will receive the same total amount of light as a full frame sensor behind an f/15 lens.
2. Those who understand that my claim about the total amount of light is correct, but who also thinks it is irrelevant.
In order to keep some structure in the discussion, I will write one separate reply to each of those two groups. If you are in group #2, please reply to my other post.
This is my reply to those in group #1:
I assume that we can all agree that a sensor behind an f/15 lens will require a 7.3x higher ISO than a sensor behind an f/5.6 lens if scene lighting and shutter speed are the same.
The different definitions of ISO are based on light quantity per unit of sensor area. There is a direct (but reciprocal) relation between ISO and exposure measured in seconds * lux or seconds * lumen / m².
This means that at a 7.3x higher ISO, the necessary amount of light per sensor area will be 7.3x smaller, no matter which sensor size you look at.
So, the full frame sensor behind the f/15 lens will...:
If you agree with me now on this point, but think it is irrelevant, please reply to my other post.
If you still do not agree with me on this point, please feel free to reply to this post.
It seems that my opponents in this thread can be divided into two groups:
1. Those who do not agree that a 2.7x crop sensor behind an f/5.6 lens will receive the same total amount of light as a full frame sensor behind an f/15 lens.
2. Those who understand that my claim about the total amount of light is correct, but who also thinks it is irrelevant.
In order to keep some structure in the discussion, I will write one separate reply to each of those two groups. If you are in group #2, please reply to my other post.
This is my reply to those in group #1:
I assume that we can all agree that a sensor behind an f/15 lens will require a 7.3x higher ISO than a sensor behind an f/5.6 lens if scene lighting and shutter speed are the same.
The different definitions of ISO are based on light quantity per unit of sensor area. There is a direct (but reciprocal) relation between ISO and exposure measured in seconds * lux or seconds * lumen / m².
This means that at a 7.3x higher ISO, the necessary amount of light per sensor area will be 7.3x smaller, no matter which sensor size you look at.
So, the full frame sensor behind the f/15 lens will...:
- have a 7.3x larger area
- receive a 7.3x smaller amount of light per unit of sensor area.
If you agree with me now on this point, but think it is irrelevant, please reply to my other post.
If you still do not agree with me on this point, please feel free to reply to this post.
Allan Olesen
Senior Member
- Messages
- 3,391
- Solutions
- 7
- Reaction score
- 1,586
I can see that I have really stirred things up. Almost everybody thinks that I am wrong. I am not going to reply to every single post, so here are instead two replies to all of you.
It seems that my opponents in this thread can be divided into two groups:
1. Those who do not agree that a 2.7x crop sensor behind an f/5.6 lens will receive the same total amount of light as a full frame sensor behind an f/15 lens.
2. Those who understand that my claim about the total amount of light is correct, but who also thinks it is irrelevant.
In order to keep some structure in the discussion, I will write one separate reply to each of those two groups. If you are in group #1, please reply to my other post.
This is my reply to those in group #2:
First, we have to ask the question:
Why is the OP of this thread interested in the aperture number of his lens?
The aperture will affect two properties of the final image:
Now, let us look at my claims in my first post in this thread:
"So if you have a crop factor of 2.7 and use a 100-300 f/5.6 lens, it will behave as a 270-810 f/15 lens would on full frame.
That is:
Depth of field will be the same.
Light gathering ability will be the same. (Which I later clarified to mean: Total amount of light gathered will be same.)
Noise in the full picture will be the same if both sensors are of equal technology."
Are these claims relevant if you want to compare DOF and noise?
Well, since claim #1 directly is a claim about DOF, it must be relevant when comparing DOF. (And so far nobody seems to have disputed this claim anyway.)
And since claim #3 directly is a claim about noise, it must be relevant when comparing noise. It is possible that it is wrong, but it cannot be irrelevant. (And when you have finished reading this post, you will probably also agree that the claim is correct).
So that leaves claim #2: That the total amount of light gathered will be the same. Why is that claim relevant?
It is relevant because the total amount of light gathered is the single property which explains almost all differences in noise in the full image when you compare modern cameras. So this is actually the reasoning behind my claim #3.
Oops. That was actually a totally new claim. And you are probably not going to believe this claim either unless I prove that I am right. So that is what I am going to do:
Looking at DxO's measurements for a camera is the easiest way to find out how much light was needed to get an image with a certain amount of noise in the full picture. Their "Sports (Low-Light ISO)" score on dxomark.com tells us the ISO which will produce a certain amount of noise for a given camera. I will refer to this score as the "ISO score".
ISO is proportional to the reciprocal value of the amount of light per sensor area. If Camera A has an ISO score of 1000, and Camera B has an ISO score of 2000, then Camera A will need twice the amount of light per sensor area to create an image with the same amount of noise.
If we draw the sensor area of these two cameras into the comparison, we can also find the difference in total amount of light which hits the sensor at these two ISO settings. The total amount of light is (sensor area) *(amount of light per sensor area). This value is proportional to (sensor area) / (ISO score).
For SLT cameras we will have to multiply the result by 0.7 since 30% of the light is lost before it hits the sensor. (Dxomark bases their score on the light which enters into the camera in direction of the sensor, not the light which actually hits the sensor after the mirror).
So let us look at some modern cameras:
So, there you have it:
Total amount of light explains dxomark's noise measurements within +/- 0.3 stops, even when comparing cameras with a 17x difference (more than 4 stops) in ISO score.
And that is why I find total amount of light very, very relevant when someone asks which result he should expect with a given combination of lens and camera.
It seems that my opponents in this thread can be divided into two groups:
1. Those who do not agree that a 2.7x crop sensor behind an f/5.6 lens will receive the same total amount of light as a full frame sensor behind an f/15 lens.
2. Those who understand that my claim about the total amount of light is correct, but who also thinks it is irrelevant.
In order to keep some structure in the discussion, I will write one separate reply to each of those two groups. If you are in group #1, please reply to my other post.
This is my reply to those in group #2:
First, we have to ask the question:
Why is the OP of this thread interested in the aperture number of his lens?
The aperture will affect two properties of the final image:
- In all cases: The DOF
- In case he is short of light: The noise
Now, let us look at my claims in my first post in this thread:
"So if you have a crop factor of 2.7 and use a 100-300 f/5.6 lens, it will behave as a 270-810 f/15 lens would on full frame.
That is:
Depth of field will be the same.
Light gathering ability will be the same. (Which I later clarified to mean: Total amount of light gathered will be same.)
Noise in the full picture will be the same if both sensors are of equal technology."
Are these claims relevant if you want to compare DOF and noise?
Well, since claim #1 directly is a claim about DOF, it must be relevant when comparing DOF. (And so far nobody seems to have disputed this claim anyway.)
And since claim #3 directly is a claim about noise, it must be relevant when comparing noise. It is possible that it is wrong, but it cannot be irrelevant. (And when you have finished reading this post, you will probably also agree that the claim is correct).
So that leaves claim #2: That the total amount of light gathered will be the same. Why is that claim relevant?
It is relevant because the total amount of light gathered is the single property which explains almost all differences in noise in the full image when you compare modern cameras. So this is actually the reasoning behind my claim #3.
Oops. That was actually a totally new claim. And you are probably not going to believe this claim either unless I prove that I am right. So that is what I am going to do:
Looking at DxO's measurements for a camera is the easiest way to find out how much light was needed to get an image with a certain amount of noise in the full picture. Their "Sports (Low-Light ISO)" score on dxomark.com tells us the ISO which will produce a certain amount of noise for a given camera. I will refer to this score as the "ISO score".
ISO is proportional to the reciprocal value of the amount of light per sensor area. If Camera A has an ISO score of 1000, and Camera B has an ISO score of 2000, then Camera A will need twice the amount of light per sensor area to create an image with the same amount of noise.
If we draw the sensor area of these two cameras into the comparison, we can also find the difference in total amount of light which hits the sensor at these two ISO settings. The total amount of light is (sensor area) *(amount of light per sensor area). This value is proportional to (sensor area) / (ISO score).
For SLT cameras we will have to multiply the result by 0.7 since 30% of the light is lost before it hits the sensor. (Dxomark bases their score on the light which enters into the camera in direction of the sensor, not the light which actually hits the sensor after the mirror).
So let us look at some modern cameras:
So, there you have it:
Total amount of light explains dxomark's noise measurements within +/- 0.3 stops, even when comparing cameras with a 17x difference (more than 4 stops) in ISO score.
And that is why I find total amount of light very, very relevant when someone asks which result he should expect with a given combination of lens and camera.
Allan, when you become so technically detailed to support a fallacious argument it is dangerous. It misinforms. It may be a language or terminology issue, but I fear not.
The simplest explanations put forward here as to why you are wrong state that what matters is the amount of light received on the focal plane PER UNIT OF AREA, not PER TOTAL AREA OF THE SENSOR.
How that unit of area is divided into sensels, and how efficient those sensels are, determines the DxO score (as a major factor). The size of the image/sensor does not.
This bad physics/math/science/commonsense is bothering because even Sony themselves - or their ad agency, probable staffed by execs with a partial understanding of the technology they are selling - has managed to fall into this trap. The current NEX-5R advert on UK TV tells the public that this camera captures more light because it has a larger sensor.
There's nothing 'political' about this argument, nor are there two alternative equally valid points of view. Of course, larger sensors combined with the right sensor technology and the right lenses can indeed tackle low light better than smaller sensors in general - but not for the simplistic reason suggested by the advert.
David
The simplest explanations put forward here as to why you are wrong state that what matters is the amount of light received on the focal plane PER UNIT OF AREA, not PER TOTAL AREA OF THE SENSOR.
How that unit of area is divided into sensels, and how efficient those sensels are, determines the DxO score (as a major factor). The size of the image/sensor does not.
This bad physics/math/science/commonsense is bothering because even Sony themselves - or their ad agency, probable staffed by execs with a partial understanding of the technology they are selling - has managed to fall into this trap. The current NEX-5R advert on UK TV tells the public that this camera captures more light because it has a larger sensor.
There's nothing 'political' about this argument, nor are there two alternative equally valid points of view. Of course, larger sensors combined with the right sensor technology and the right lenses can indeed tackle low light better than smaller sensors in general - but not for the simplistic reason suggested by the advert.
David
Last edited:
Cetonid
Senior Member
- Messages
- 2,588
- Solutions
- 1
- Reaction score
- 102
I was going to chime in, but you said it better than I could.
Bottom line... whether you use the entire area of a smaller sensor or a crop to the same physical size area on the larger sensor, f2.8 in either system delivers the same light to the same size area.
DOF is based on actual focal length, F stop, distance to subject regardless of sensor size.
With all that said, something like the the old Panasonic FZ50 with the RX100 sensor and a 15x zoom (28-420mm equiv.) could be a nice little walk around wildlife camera at a measly 2lbs.
Bottom line... whether you use the entire area of a smaller sensor or a crop to the same physical size area on the larger sensor, f2.8 in either system delivers the same light to the same size area.
DOF is based on actual focal length, F stop, distance to subject regardless of sensor size.
With all that said, something like the the old Panasonic FZ50 with the RX100 sensor and a 15x zoom (28-420mm equiv.) could be a nice little walk around wildlife camera at a measly 2lbs.
Draek
Senior Member
For a pretty damn narrow definition of "modern". For instance, switching to an improved sensor alone made Olympus jump forward in 0.65EV between the E-PL3 and the E-PL5, and even against contemporary Panasonic cameras there's an improvement of around half a stop.
Haven't run the numbers in other segments, but I suspect your data happened to be well correlated by pure chance and that we'd see similar variance in other sensor sizes given enough samples. Except for 1" of course since, you know, there ain't any.
cyainparadise
Veteran Member
You're confusing aperture, which is the subject at hand, and which doesn't change when you use the lens on a different sized sensor body, with DOF. Your misunderstanding, or misstatement, is confusing people.
The physics might be correct but that does not mean the argument makes any sense.
You are telling people that the FF aperture equivalent is a good way to compare DOF and noise.
That might be correct for the DOF.
But only if comparing more than two different sensor sizes at the same time and need a reference system (That's where DPR uses the FF equivalent aperture sometimes).
If you only have two sensor sizes or lens/sensor combinations you can simply calculate the DOF without caring about the equivalent and get much better information.
If you want to compare noise it makes no sense to use the equivalent aperture. Your claims are only correct if the technology the sensor are based on is very similar. Use different sensor generations and the equivalent aperture is way off.
A comparison tool that only works in this very special case makes no sense. Why not just use low-iso data? It is not only much better because not depending on anything but is giving you direct data about the noise and is not some claim.
It's like comparing a distance in throwing distance. If you use the same person and the same object it might be accurate.
One does not need the FF equivalent aperture at any time. There are almost ALWAYS ways to get better data when using different parameters.
You are telling people that the FF aperture equivalent is a good way to compare DOF and noise.
That might be correct for the DOF.
But only if comparing more than two different sensor sizes at the same time and need a reference system (That's where DPR uses the FF equivalent aperture sometimes).
If you only have two sensor sizes or lens/sensor combinations you can simply calculate the DOF without caring about the equivalent and get much better information.
If you want to compare noise it makes no sense to use the equivalent aperture. Your claims are only correct if the technology the sensor are based on is very similar. Use different sensor generations and the equivalent aperture is way off.
A comparison tool that only works in this very special case makes no sense. Why not just use low-iso data? It is not only much better because not depending on anything but is giving you direct data about the noise and is not some claim.
It's like comparing a distance in throwing distance. If you use the same person and the same object it might be accurate.
One does not need the FF equivalent aperture at any time. There are almost ALWAYS ways to get better data when using different parameters.
Bruce Oudekerk
Senior Member
Here is a thought experiment in a carefully controlled situation that I get to choose
Let’s say that I have an 8x10 view camera (basically a light tight box similar to all cameras) with a 300 mm lens 15 feet from an artificially illuminated neutral grey wall. I meter the solid grey wall and determine that f8.0, 1/1000 sec and ISO 400 is an appropriate exposure. It would seem that I could put a small piece of Tri X 35mm film or a 8x10 TriX sheet film centered on the film plane of that properly focused 300mm lens and, if developed appropriately, both would produce the same density negative suitable for printing. My assumption is that this scenario is directly analogous to various size digital sensors.
If this is true, I would think we now have a problem with semantics or nomenclature. Certainly in the example of the 8x10…more photons/energy fell on the total surface of that large 8x10 sheet (more light?) than on the 35mm film frame; whereas it is also fair to say the same energy (same light?) fell on both the 8x10 and the 35mm frame if measured in visible light energy per area. These two disparate concepts are certainly reconcilable, but not when using the same terminology. So is it the concept or the terminology that is the problem here?
Just for clarification, let’s look at this differently, and I don't know the answer. Would an engineer that designs photovoltaic cells, consider it the 'same' or 'less' amount of light falling on a postage stamp vs. the football field the stamp is on? It would seem that this is a similar situation. Although I suspect they would neatly sidestep the issue by using differnt terminology.
So does it boil down to the "amount of light' being measured in Luminous flux or Illuminace? Or am I missing something?
What I’m not missing is that, no matter the focal length of the lens, no matter the light tight box size, no matter the dimensions of an ISO 400 light sensitive sensor or film …if any lens is focused on this ‘film plane’, F8 and 1/1000 second will lead to a proper exposure of the area illuminated. At least in a perfect world.
Bruce
Last edited:
Keyboard shortcuts
- f
- Forum