Amateur Sony Shooter
Veteran Member
Ah, that silly "auto correction" on my iPad!
-
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
Steve West
Veteran Member
Allan is talking about imaging the *exact same scene* onto each detector which means you have to move the camera. In this case, the solid angle of light each pixel receives from the larger detector is much greater. If you simply crop the image to your smaller detector, you no longer have the same scene on each and advantage of the larger detector disappears except for any improvement in pixel noise.
I did the experiments. If I simply crop the FF scene to APS on my A99 and compare to the A77 with same lens with the camera at the same distance, I see about a stop of noise improvement. If I move the camera farther away to get the same scene on the A77 as I saw on the A99, the A99 (with the same lens) has about a 3 stop improvement.
When I went through this with Allan a few weeks ago, he doesn't talk about throughput and image noise in the way we as croppers see the situation. We can only get within a certain distance of the bird or scene, and we tend to compare a crop camera's performance with FF assuming that the distance to the scene is fixed at one location. In this case, the smaller detector is a simple crop of what the FF would see. But compared to moving closer and getting the same scene on the FF, the apparent speed (or f/#) of the APS detector is vastly reduced.
I did the experiments. If I simply crop the FF scene to APS on my A99 and compare to the A77 with same lens with the camera at the same distance, I see about a stop of noise improvement. If I move the camera farther away to get the same scene on the A77 as I saw on the A99, the A99 (with the same lens) has about a 3 stop improvement.
When I went through this with Allan a few weeks ago, he doesn't talk about throughput and image noise in the way we as croppers see the situation. We can only get within a certain distance of the bird or scene, and we tend to compare a crop camera's performance with FF assuming that the distance to the scene is fixed at one location. In this case, the smaller detector is a simple crop of what the FF would see. But compared to moving closer and getting the same scene on the FF, the apparent speed (or f/#) of the APS detector is vastly reduced.
Dave Oddie
Senior Member
It has nothing to do with sensor area. A lens speed is not, never has been and never will be based on the size of the sensor or film. It is rated on the size of the physical characteristics of the lens - that ratio of is maximum entry pupil size to its focal length.
I was saying F15 is nearly four stops slower than F5.6 is where "4" came from. I should have said nearly 3 but its not relevant anyway because the implication of your post is that the F5.6 lens loses 3 stops of light because you put in on a smaller sensor camera is nuts.
But that does not alter the characteristics of a 300mm F5.6 lens on any sized sensor.
No it won't. In order for the longer lens to gather the same amount of light as the shorter one it needs a larger entry pupil size.
For ISO 100 a light level of EV15 requires an exposure of 1/1000 set at F5.6. You do not have to modify that to be approx 1/125 sec because you put your F5.6 lens on a smaller sensor camera.
At ISO 100 a full frame user would have to set their 810mm F5.6 lens to F5.6 (and shutter to 1/1000) to get a correct exposure at EV15.
At ISO 100 a 1 inch sensor user would have to set their 300mm F5.6 lens to F5.6 and a shutter of 1/1000 to get a correctly exposed shot at EV15.
The implication of what you are saying as I read it is that you are suggesting that for a light level of EV15 an exposure of F15 and around 1/1000 sec will yield a correct exposure for the full frame user with their 810mm lens.
If so that is clearly wrong as that would give an under exposed shot. You would have set the exit pupil to 54mm and that isn't large enough on a 810mm lens to give a correct exposure at a light level of EV15 if you set the shutter to 1/1000.
Dave Oddie
Senior Member
I agree but that is not what you are doing. You are ignoring the focal lengths effect on a lens light gathering properties
The short hand of saying a 300mm lens on a 1 inch sensor is equivalent to an 810mm one might be lazy when in reality it is just a crop but it doesn't alter anything about a lens is light gathering capabilities. That is a function of the focal length and its maximum entry pupil size and nothing else.
With respect to what? Depth of Field I presume.
Draek
Senior Member
Will Sony make an SLR with a small, 1" sensor, and where the mirror assembly (and thus the viewfinder image) is significantly higher than what the camera will actually capture?
Just buy a Nikon V1 and stick a red-dot sight in there.
Just buy a Nikon V1 and stick a red-dot sight in there.
Draek
Senior Member
Wrong.
Suppose I'm out and shooting with two Sony cameras, an A900 and an A100. Suppose I'm using the aforementioned 100-300/5.6 Minolta. Suppose now I mount it on the A900, and shoot at ISO200 1/800 f/5.6 to get a 'proper' exposure. Suppose now I mount it on the A100, set it at ISO200 and f/5.6; what should my shutter speed be to get an exposure as similar as possible to that from the A900?
(hint: it's not 1/300).
That's what people mean by "light gathering", that's what it has meant for *decades*, and while I do understand why you got to the conclusion you did, for most people and most situations it's an entirely useless calculation to make. Photographers do not count photons.
Dave Oddie
Senior Member
cyainparadise
Veteran Member
that if I use a lens that was designed for a full-frame camera, like the A99, on an APS-C sized senor camera like the A57, that the aperture is smaller? Thus, a lens with a f/2.8, is af/4.2, and the 500mm f/8, which is a FF lens, would have an aperture of f/12? Funny, because the AF system won't work with such a small aperture.
If what you say is true, Sony can expect to be sued any day now.
If what you say is true, Sony can expect to be sued any day now.
cyainparadise
Veteran Member
you're confusing the smaller sensor size with using a tele-converter where the aperture is changed. For example if you use a 2X TC, and your lens has a maximum aperture of f/2.8, it then becomes a f/5.6, due to the TC.
cyainparadise
Veteran Member
if what you say is true, then putting a f/2.8 FF lens on an APS-C body, and then using a 2X TC, my aperture would then become an f/8.4.
Allan's post is an example of the kind of confident misinformation which pervades forums. I'd love just to be able to erase it permanently - because you never know who may read it, and have insufficient technical knowledge to realise it's a completely fallacious argument. Doesn't seem to matter how many other intelligent people point it out, this ridiculous position about 'less light' reaching smaller sensors (in overall dimensions) is a myth which seems to persist.
DPR's comparison refers only to equivalent depth of field, and even then this comparison only applies when the scene is scaled in all respects - that is, a 2.7m high subject photographed at 2.7m away (on full frame) compared to a 1m subject 1m away etc.
This is one of the things I love about small sensors - being able to shoot with 'miniature as full scale' effect, placing the camera on the ground, inside dolls-houses, within architectural models, inside a guitar, under the floorboards - or just to make an interesting plate of food into a full landscape foreground for a restaurant view.
There are specific cases where format change does affect relative lens aperture, and it's all do with subject scale. You can photograph an egg in an eggcup using a 2/3rds sensor camera, fill the frame, and your working aperture will be pretty much as marked; try it with 10 x 8 sheet film to fill the frame, and the required 2X magnification will turn your marked f/8 into a real f/32.
But at long distance, telephotography of birds etc, this is not significant even for extreme differences in film or sensor size. f/5.6 is f/5.6 is f/5.6 and never becomes f/15 - and the luminous flux, photons per square millimetre reaching the sensitive surface, remains identical.
David
DPR's comparison refers only to equivalent depth of field, and even then this comparison only applies when the scene is scaled in all respects - that is, a 2.7m high subject photographed at 2.7m away (on full frame) compared to a 1m subject 1m away etc.
This is one of the things I love about small sensors - being able to shoot with 'miniature as full scale' effect, placing the camera on the ground, inside dolls-houses, within architectural models, inside a guitar, under the floorboards - or just to make an interesting plate of food into a full landscape foreground for a restaurant view.
There are specific cases where format change does affect relative lens aperture, and it's all do with subject scale. You can photograph an egg in an eggcup using a 2/3rds sensor camera, fill the frame, and your working aperture will be pretty much as marked; try it with 10 x 8 sheet film to fill the frame, and the required 2X magnification will turn your marked f/8 into a real f/32.
But at long distance, telephotography of birds etc, this is not significant even for extreme differences in film or sensor size. f/5.6 is f/5.6 is f/5.6 and never becomes f/15 - and the luminous flux, photons per square millimetre reaching the sensitive surface, remains identical.
David
tbcass
Forum Pro
All he is doing is arguing semantics which is a bit useless IMO. If all that was necessary was to move closer all a wildlife photographer would have to do is put a 28mm lens on his camera and walk up to the Bear he is trying to photograph. I doubt you would live very long doing that. No, all Allan is doing is arguing for arguments sake.
Faith Yeung
Leading Member
New to this conversation, but it seems that everyone is talking about an equivalence without first laying out the underlying assumptions. I'm no expert photographer/doctorate in optics or anything like that, but I have a basic understanding of everything involved, so I'll give it a go.
THE FOLLOWING MAY BE A LOAD OF CODSWALLOP:
Assumption number 1: The pixels per unit area stays the same.
Assumption number 2: The total number of pixels (resolution) stays the same.
Under 1 we don't multiply the f number as the sensor exposes the same (a great analogy is simply chopping a 35mm film to the size you want, of course the remaining film will expose exactly the same as the entire film would have), with the corollary that you lose resolution (both in the film and sensor examples).
Under 2 we multiply the f number, the film analogy here being that you are not only cutting away the extra film, you are also changing the film material so that its exposure properties change (this is one effect of a change in pixel density). So by multiplying the number you take into account both DoF properties, but also the implicit exposure properties of the sensor. The corollary here is that you will also loose iso performance.
To be clear: physically the focal length and f number of the lens have not changed, this is purely for equivalence purposes.
Also it is unlikely that either assumption will hold when comparing a full frame camera to a 1inch camera of the same generation and technology. A full frame sensor will likely have both a lower pixel density and a higher resolution than its smaller counterparts.
I wouldn't spend too much time worrying about it either way, but perhaps a poor approximation would be to multiply by half the crop factor. You won't get an accurate answer without calibrated tests I imagine, and there are many traps in that process.
Also, who cares?
Draek
Senior Member
No, he's saying that if you use an f/2.8 lens on a 2.7x crop camera, the total number of photons captured by the sensor will be the same as those of a f/4.2 lens on a FF camera. Which makes sense, because out of all the photons being captured by the f/2.8 lens, you're throwing away any which doesn't hit the part the "cropped sensor" is "cropped from", to put it some way, and the ratios between the areas it's covering and the area you're actually capturing is the same 2.7x.
The problem, of course, is that this figure is only useful for calculating stuff such as the quantum efficiency of a given sensor technology, but for actual photography f/2.8 is f/2.8 is f/2.8 and ISO100 is ISO100 is ISO100; that consistency is in fact a prime reason behind why ISO was standardized as such, and if the total number of photons captured mattered, then it's likely the ISO figure would've had a factor integrated into it to compensate accordingly.
The problem, of course, is that this figure is only useful for calculating stuff such as the quantum efficiency of a given sensor technology, but for actual photography f/2.8 is f/2.8 is f/2.8 and ISO100 is ISO100 is ISO100; that consistency is in fact a prime reason behind why ISO was standardized as such, and if the total number of photons captured mattered, then it's likely the ISO figure would've had a factor integrated into it to compensate accordingly.
cyainparadise
Veteran Member
The facts haven't changed from the days of film to digital. There's no need for new tests, the facts are out there, and have been for decades. The camera companies know the facts.
You're just typing words without knowing the facts. As to "who cares?" Anyone who takes their photography seriously cares.
cyainparadise
Veteran Member
Keyboard shortcuts
- f
- Forum