Apeture affecte from fulframe too APSC ?

[ropo54]

Advertisement (gone after registration)

LCT:

 

Thank you.

 

So, using the calculator . . . a Noctilux with an APSC sensor at appx 6.3 feet distance to subject yields the equivalent DOF as using the Noctilux on a full frame sensor at 5 feet from subject. 

 

Rob

Edited January 28, 2018 by ropo54

[Advertisement]

The Leica Q3 is available for pre-order at B&H Photo Video and Adorama

[lct]

Well i'm aware of the term "equivalent DoF" but i don't know its precise meaning sorry. All i can say is DoF depends, among other factors, on the circle of confusion (CoC) mentioned in the Dof Master site quoted above and CoC values depend on the size of the sensor so there is no surprise if DoF is not the same when using the same lens on FF and APS-C cameras. Now contrary to what i read in this curious thread, DoF changes when the subject distance changes as well so what you've got here is a combination of different sensor sizes and subject distances in the calculation of the depth of view. But again i'm no techie at all so take what i say with a lump of salt .

[ropo54]

My inquiry concerned whether one can obtain the same 'look' wide open from the Noctilux on ASPc as one can by using it on a full-frame sensor. 

 

Secondarily, I also have wondered whether the APSc lenses when used with full frame sensors perform 'better'

because of the larger pixels, albeit fewer pixels.  That has been discussed some on other threads, with the prevailing thought being more pixels, even if smaller, is probably the better option. Though, some opinions have differed on that issue as well.

 

Thank you for your thoughts.

 

Rob

Edited January 28, 2018 by ropo54

[lct]

My inquiry concerned whether one can obtain the same 'look' wide open from the Noctilux on ASPc as one can by using it on a full-frame sensor.  [...]

 

If same look means same FoV, same DoF and same perspective you won't get it by moving your feet as perspective would change then. Only way i can think of is asking Leica for a Super-Noctilux 35/0.75 i'm afraid .

[lct]

[...] I also have wondered whether the APSc lenses when used with full frame sensors perform 'better' because of the larger pixels, albeit fewer pixels.  That has been discussed some on other threads, with the prevailing thought being more pixels, even if smaller, is probably the better option. Though, some opinions have differed on that issue as well. [...]

 

I have never used APS-C lenses on FF bodies so i can only compare my M lenses on cameras with small and larger pixels and the better IQ i've got there is with larger pixels certainly. As far as M lenses on FF bodies are concerned, i prefer the IQ of my 12MP Sony A7s mod to that of my 24MP M240 but it's perhaps just me.

[Jared]

As far as DoF is concerned we agree of course but DoF is not all photography. I've been using both FF and APS gear for 15+ years and when i had an f/1.4 lens to use it was a 35 on APS and 50 on FF. For both lenses, shutter speeds and apertures have always been the same because there is no significant difference between them besides DoF, sorry to disagree with you.

I’m not certain where we disagree. A 35mm on APS-C has the same field of view as a 50mm in full frame. I think everyone agrees with that. A 35mm on APS-C will have a bit more depth of field at a given aperture than a 50mm on full frame. For a given ISO and shutter speed, a 35mm at f/1.4 would generate the same irradiance—the same light per unit area—as any f/1.4 lens on any format. I never suggested otherwise. If ISO 100, 1/60s, f/1.4 is the “correct” exposure, for example, it is the correct exposure with any focal length on any format. That’s the whole reason photographers find focal ratio convenient to use in determining exposure. It is invariant. I don’t think I ever suggested otherwise.

 

However, from a signal-to-noise ratio perspective, larger sensors perform better at a given megapixel count because each individual pixel is larger (covers more area) and therefore collects more photons. That’s really one of the primary advantages to larger formats. Not the only one, but it’s a big one. It’s why, in particular, high ISO performance tends to be better on larger chips. The extreme example—compare an iPhone pic taken under less than ideal lighting to an M10 low light image. Much of the difference in image quality is from physically larger pixels.

[Jared]

Advertisement (gone after registration)

Now contrary to what i read in this curious thread, DoF changes when the subject distance changes as well so what you've got here is a combination of different sensor sizes and subject distances in the calculation of the depth of view. But again i'm no techie at all so take what i say with a lump of salt .

That’s not contrary to what has been said in this thread. The fact that depth of field changes with subject distance is the whole reason a 35mm at f/1.4 on APS-C has more depth of field than on full frame. You won’t shoot it at the same distance on the two formats since you wouldn’t get the same field of view. On APS-C you will “back up” to get the same field of view. That additional distance is why DOF increases with a smaller sensor.

 

The physics, obviously, is the same as it always was.

[Jared]

Secondarily, I also have wondered whether the APSc lenses when used with full frame sensors perform 'better'

because of the larger pixels, albeit fewer pixels. That has been discussed some on other threads, with the prevailing thought being more pixels, even if smaller, is probably the better option. Though, some opinions have differed on that issue as well.

 

Thank you for your thoughts.

 

Rob

You mean like putting a TL lens on an SL camera and letting it crop down to 10 megapixels? If so, then you will definitely get better signal-to-noise performance due to the larger pixels, but I wouldn’t consider that an advantage. You could achieve the exact same effect by using the lens on a CL and just down-sampling in Photoshop to the same ten megapixels. Basically, you can always trade away resolution for SNR if you want to. Generally better to start with the higher resolution (assuming it actually has truly higher resolution, not just more pixels) since you can trade resolution for SNR, but it is generally not possible to go the other way (ignoring Drizzle and some particular deconvolution algorithms which require either multiple exposures or a known point spread function).

[Jared]

I have never used APS-C lenses on FF bodies so i can only compare my M lenses on cameras with small and larger pixels and the better IQ i've got there is with larger pixels certainly. As far as M lenses on FF bodies are concerned, i prefer the IQ of my 12MP Sony A7s mod to that of my 24MP M240 but it's perhaps just me.

Sounds like you favor better SNR over higher resolution for your photography. Nothing wrong with that. Which is more critical will depend on subject and lighting. Were I shooting primarily nightscapes (TWAN) I’d take your A7s in a heartbeat over any Leica made. Vanishingly small amounts of dark current and read noise in that chip make it a superb choice for low light situations even if the megapixel count is a bit lower than many photographers look for these days.

[FlashGordonPhotography]

Smaller pixels do not always have worse SNR than larger ones. That would only apply if the larger sensor were a directly scaled up version of the smaller one and all other pieces in the processing pipeline were the same, which never happens in real life. The D850 and A7R3 have smaller pixels than their 24MP equivalents and still have less noise and more DR.

 

It's a generalisation that is often, but not always true.

 

Gordon

[lct]

Sounds like you favor better SNR over higher resolution for your photography. [...]

 

I favor resolution for my job, reason why i use an A7r2 then, but with the same lenses i prefer my Kolari mod A7s' IQ by far. Not sure if it comes from pixel size, Kolari mod or both though.

[Jared]

Smaller pixels do not always have worse SNR than larger ones. That would only apply if the larger sensor were a directly scaled up version of the smaller one and all other pieces in the processing pipeline were the same, which never happens in real life. The D850 and A7R3 have smaller pixels than their 24MP equivalents and still have less noise and more DR.

 

It's a generalisation that is often, but not always true.

 

Gordon

 

 

True.  I was assuming similar quantum efficiency, similar read noise, similar thermal noise, and scaled full well capacity (proportional to pixel pitch squared) which, as you pointed out, is often but not always the case.  Some sensors are just plane better than others.  Frankly, though, the differences have been getting smaller over the past couple or three years.  Most of them are close enough now that, within a given format, I'd generally take the camera with the better ergonomics (or the features that were a better match for my photographic needs) over the camera with the better sensor.

[lct]

I’m not certain where we disagree. A 35mm on APS-C has the same field of view as a 50mm in full frame. I think everyone agrees with that. A 35mm on APS-C will have a bit more depth of field at a given aperture than a 50mm on full frame. For a given ISO and shutter speed, a 35mm at f/1.4 would generate the same irradiance—the same light per unit area—as any f/1.4 lens on any format. I never suggested otherwise. If ISO 100, 1/60s, f/1.4 is the “correct” exposure, for example, it is the correct exposure with any focal length on any format. That’s the whole reason photographers find focal ratio convenient to use in determining exposure. It is invariant. I don’t think I ever suggested otherwise. [...] 

 

OK but why saying « you get essentially the same image in every single meaningful way out of a 50mm f/2 full frame lens/camera combo as you do out of a 35mm f/1.4 APS-C lens/camera combo. You can consider them equivalent » as you wrote above? A reader could believe that one can get the same performance in low light with a 50/2 on FF as with a 35/1.4 on APS-C, which is untrue obviously. When f/1.4 is needed, all one can get at f/2 is an underexposed pic like mine above and it won't always be possible, or desirable, to push it in PP.

[FlashGordonPhotography]

OK but why saying « you get essentially the same image in every single meaningful way out of a 50mm f/2 full frame lens/camera combo as you do out of a 35mm f/1.4 APS-C lens/camera combo. You can consider them equivalent » as you wrote above? A reader could believe that one can get the same performance in low light with a 50/2 on FF as with a 35/1.4 on APS-C, which is untrue obviously. When f/1.4 is needed, all one can get at f/2 is an underexposed pic like mine above and it won't always be possible, or desirable, to push it in PP.

 

Again. True but not always. I would choose an m4/3 camera over an APSC one in low light because the PenF has IBIS. Meaning that I can often shoot at a lower shutter speed or ISO than I might on a CL. Occasionally several stops slower, completely obliterating any minor sensor performance differences. Not to mention every incremental increase in resolution also requires incrementally improved technique to hide pixel blur.

 

Gordon

[Guest VVJ]

I would choose an m4/3 camera over an APSC one in low light because the PenF has IBIS. 

 

Only a matter of time I believe.  Sony has IBIS.  The in March to be released Fuji X-H1 will have IBIS.  I really don't believe Leica can afford to go another 3-4 years without IBIS.  

[jaapv]

Again. True but not always. I would choose an m4/3 camera over an APSC one in low light because the PenF has IBIS. Meaning that I can often shoot at a lower shutter speed or ISO than I might on a CL. Occasionally several stops slower, completely obliterating any minor sensor performance differences. Not to mention every incremental increase in resolution also requires incrementally improved technique to hide pixel blur.

 

Gordon

Not just for IBIS. A good panorama setting and 4K stills options like long 30 fps bursts with pre-release, and post-focus are really enhancements that in part  allow for images that cannot be captured otherwise and at the very least make photographic life a lot more pleasant. Not to mention a hi-res setting that comes with IBIS.

Those things are not gimmicks, I would like to see some of it on a Leica APS-C.

[Jared]

OK but why saying « you get essentially the same image in every single meaningful way out of a 50mm f/2 full frame lens/camera combo as you do out of a 35mm f/1.4 APS-C lens/camera combo. You can consider them equivalent » as you wrote above? A reader could believe that one can get the same performance in low light with a 50/2 on FF as with a 35/1.4 on APS-C, which is untrue obviously. When f/1.4 is needed, all one can get at f/2 is an underexposed pic like mine above and it won't always be possible, or desirable, to push it in PP.

 

 

What I said is accurate (ignoring for the time being Gordon's points about IBIS).  Assuming the chips in the two cameras are essentially "scaled" versions of each other aside form pixel pitch, at any rate.  

 

Assume that, for example, you can hand hold the 35mm/APS-C combo at 1/60s.  Likewise, you can handhold the 50mm/full frame combo at 1/60s.  This is basically the one-over-focal length rule of thumb, but adjusted for the crop factor in the APS-C camera as it should be--smaller pixels.  Now assume that the light is such that the 35mm at f/1.4 and 1/60s requires ISO 1600.  When you shoot with the 50mm at f/2 under the same light at the same 1/60s, you would need to push the ISO up to 3200 in order not to be underexposed.  A full frame sensor at ISO 3200 should yield the same signal-to-noise ratio as an APS-C sensor at ISO 1600 if the two chips are equal in quality.  That is, if they have the same number of megapixels, the same quantum efficiency, the same read noise, the same thermal noise, and full well capacities that are proportional to pixel area (basically, just scaled versions of each other), then you should get the same results out of the ISO 1600 with APS-C as you do at ISO 3200 with full frame.

 

As a result, you do, in fact, get the same low light performance out of each.  You'll have to bump the ISO up on the full frame camera, but that bumped up ISO will actually put them on equal footing.  Same signal-to-noise under low light conditions.  Same depth of field.  Same field of view.  Same shutter speed to minimize motion blur.  

 

Now back to Gordon's points... Yes, IBIS will let you hand hold with a slower shutter speed as long as your subjects are fairly static.  So would lens based image stabilization if you popped an SL lens onto your CL (thought that sort of defeats the point of a smaller camera body).  And yes, chips are not all created equal.  But in order to make an apples-to-apples comparison, it's reasonable to assume similar quality in the chips and similar IS/IBIS/VR availability or lack of ability in either format.  Under those circumstances, a 50mm f/2 lens on a full frame body really will behave like a 35mm f/1.4 lens on a CL.  Same field of view.  Same DOF.  Same image quality for a given shutter speed (though the ISO will necessarily be one stop higher on the full frame if you are shooting each lens wide open).  

[lct]

[...] Now assume that the light is such that the 35mm at f/1.4 and 1/60s requires ISO 1600.  When you shoot with the 50mm at f/2 under the same light at the same 1/60s, you would need to push the ISO up to 3200 in order not to be underexposed. [...]

 

Why would i have to shoot at f/2 and why would i have to push in PP? I have a f/1.4 lens so i will shoot at f/1.4 and i won't have tu push this way. Sorry Jared but this does not make any sense for me. For all purposes but DoF, at the same aperture, my 50/1.4 lens on FF behaves the same way as my 35/1.4 lens on APS-C. 

[Advertisement]

Hallo,

Look here: Leica L-Mount

[jaapv]

What I said is accurate (ignoring for the time being Gordon's points about IBIS).  Assuming the chips in the two cameras are essentially "scaled" versions of each other aside form pixel pitch, at any rate.  

 

Assume that, for example, you can hand hold the 35mm/APS-C combo at 1/60s.  Likewise, you can handhold the 50mm/full frame combo at 1/60s.  This is basically the one-over-focal length rule of thumb, but adjusted for the crop factor in the APS-C camera as it should be--smaller pixels.  Now assume that the light is such that the 35mm at f/1.4 and 1/60s requires ISO 1600.  When you shoot with the 50mm at f/2 under the same light at the same 1/60s, you would need to push the ISO up to 3200 in order not to be underexposed.  A full frame sensor at ISO 3200 should yield the same signal-to-noise ratio as an APS-C sensor at ISO 1600 if the two chips are equal in quality.  That is, if they have the same number of megapixels, the same quantum efficiency, the same read noise, the same thermal noise, and full well capacities that are proportional to pixel area (basically, just scaled versions of each other), then you should get the same results out of the ISO 1600 with APS-C as you do at ISO 3200 with full frame.

 

As a result, you do, in fact, get the same low light performance out of each.  You'll have to bump the ISO up on the full frame camera, but that bumped up ISO will actually put them on equal footing.  Same signal-to-noise under low light conditions.  Same depth of field.  Same field of view.  Same shutter speed to minimize motion blur.  

 

Now back to Gordon's points... Yes, IBIS will let you hand hold with a slower shutter speed as long as your subjects are fairly static.  So would lens based image stabilization if you popped an SL lens onto your CL (thought that sort of defeats the point of a smaller camera body).  And yes, chips are not all created equal.  But in order to make an apples-to-apples comparison, it's reasonable to assume similar quality in the chips and similar IS/IBIS/VR availability or lack of ability in either format.  Under those circumstances, a 50mm f/2 lens on a full frame body really will behave like a 35mm f/1.4 lens on a CL.  Same field of view.  Same DOF.  Same image quality for a given shutter speed (though the ISO will necessarily be one stop higher on the full frame if you are shooting each lens wide open).

Now, this is all nothing new. We had this whole discussion identically when the M9 came out; the sensor was identical to the M8 except for size, yet some people couldn't believe that, as noise performance appeared to be one EV better. 

At any rate, there is your apples-to-apples comparison, nine years ago.

 

As for the DOF comparison. The perception of DOF varies by more than one EV value between different persons because of eye acuity variations, wearing spectacles or not, etc.

Look at a print a bit closer and you'll halve your DOF. Or just print twice as large. Have a look at a 100% crop on your screen and it will have vanished altogether. So who cares?

 

Just learn to use your tools to take the photograph you want to take. And that is done in actual life, not in theoretical discussions.

[Starwolfy]

However, from a signal-to-noise ratio perspective, larger sensors perform better at a given megapixel count because each individual pixel is larger (covers more area) and therefore collects more photons. That’s really one of the primary advantages to larger formats. Not the only one, but it’s a big one. It’s why, in particular, high ISO performance tends to be better on larger chips.

 

Hello Jared !

 

I agree with you and I do realize, now, that FF sensors do not have better iso than APSC...they are simply larger, collect more photons, and so for a given exposure get a cleaner image than with a smaller APSC sensor with the same exposure.

I understand from your explanation and from the video that if you were to give as much photons to both FF and APSC sensors they would be indentical grain wise. His point is that it is a non-sense to compare both system performance based on the same "exposure" since even with the same "exposure" there is a difference of amount of photons collected simply due to sensor sizes. And the reason is that "iso" is not a correct measurement parameter as it seems.