 |
 | Home | Contact Us | FAQ | Search & Site Map | Link to Us |
 | Sign In | Join | Other 45 Sites in Network |
| PhotoKB HomeDiscussion Groups Digital Photography Digital PhotoDSLR CamerasZLR CamerasPoint & Shoot CamerasFilm Photography 35 mmLarge FormatMedium formatDarkroomFilm and LabsOther EquipmentPhoto Technique Nature PhotographyPeople PhotographyTechnique GeneralGeneral Photo Topics General TopicsAustralian PhotographyUK PhotographyDirectoryPhoto Clubs |
Photo Forum / Digital Photography / DSLR Cameras / August 2006Question about lens F-rating on digital slr
Take a full frame camera and a 1.6 crop digital. Put a F2.8L lens on both. Set both cameras to automatic and same ISO and take exactly the same shot. Will the shutter speed on the 1.6 crop camera be the same as the other, or do camera manufacturers "tune" the internal gain (effectively the ISO) to compensate for the different intensity of the light on the sensor? I've wondered about this since I read in an Australian lens review that putting a full frame lens (like the F2.8L) on a 1.6 crop digital can actually result in reduced performance over putting a lenses specifically designed for the smaller sensor. Any thoughts? > Take a full frame camera and a 1.6 crop digital. > Put a F2.8L lens on both. Set both cameras to automatic and same ISO and > take exactly the same shot. Will the shutter speed on the 1.6 crop camera > be the same as the other, Yes. > or do camera manufacturers "tune" the internal gain (effectively the > ISO) to compensate for the different intensity of the light on the > sensor? Not unless they want to break the entire notion of why some folks by an SLR they don't. Keep in mind that handheld external meters are sold and in wide use, and that external studio strobes are used with SLR's and if manufacturers start inventing their own scales and compensating in software for shutter speeds and such, they'd effectively break their cameras for use with all this fun external gear we SLR junkies like to us. > I've wondered about this since I read in an Australian lens review > that putting a full frame lens (like the F2.8L) on a 1.6 crop > digital can actually result in reduced performance over putting a > lenses specifically designed for the smaller sensor. I'm curious too what they meant by that. It would have to be if anything optical performance they're talking about in some respect, or the difference in depth of field in the two cases because of the perspective change. -- Todd H. http://www.toddh.net/ >> Take a full frame camera and a 1.6 crop digital. >> Put a F2.8L lens on both. Set both cameras to automatic and same ISO [quoted text clipped - 4 lines] > > Yes. I agree, but we should define some of the terms above more carefully. Two cameras on fully "automatic" settings might well choose different apertures, meaning they'd end up with different shutter speeds as well. Change "automatic" to "aperture priority" mode, using the same actual aperture setting on both (eg, 2.8), and then we're getting closer. But now we need to look at what we mean by taking "exactly the same shot". It is impossible to take the *same* shot with both cameras using the same lens. Presumably, we are actually talking about standing in the same place and aiming the camera at the same subject, but then, the 1.6 crop camera is going to be looking at a different version of that scene than the full frame camera, so it may well still meter differently. But if both are set to spot metering, using the same metering point, or if the scene is something flat like a grey card, then I think we can say the results would be the same (and would, as observed, also be the same as on a film camera or as reported by a handheld light meter). >> I've wondered about this since I read in an Australian lens review >> that putting a full frame lens (like the F2.8L) on a 1.6 crop >> digital can actually result in reduced performance over putting a >> lenses specifically designed for the smaller sensor. > > I'm curious too what they meant by that. I'm guessing it has to do with the issue of digital sensors being more sensitive to the angle of light striking them than film is. That's why you'll see "digitally optimized" lenses that supposedly try to overcome this, but that still work on 35mm film or full frame digital. A lens designed for film would in theory have more of these issues with digital, whether full frame or smaller. Although here, it would actually seem APS would be *better* than full frame. In any case, I would *never* assume "reduced performance" was meant to imply anything whatsoever about shutter speed. Definitely optical quality. --------------- Marc Sabatella marc@outsideshore.com Music, art, & educational materials Featuring "A Jazz Improvisation Primer" http://www.outsideshore.com/ > Take a full frame camera and a 1.6 crop digital. > Put a F2.8L lens on both. Set both cameras to automatic and same ISO and > take exactly the same shot. Will the shutter speed on the 1.6 crop camera > be the same as the other, or do camera manufacturers "tune" the internal > gain (effectively the ISO) to compensate for the different intensity of > the light on the sensor? I'm seeing the question you asked, and then the question you probably mean. The answer to the question as you asked it is, no, unless by coincidence. The automatic exposure algorithms may be different, and one camera may choose an entirely different shutter speed *and* aperture from the other. The answer to the question I think you probably really meant to ask is, yes, f/2.8 is f/2.8 regardless and independent of the sensor size or the size of the image circle the lens projects. 1/125 at f/2.8 with ISO 100 means the same thing no matter what camera you're using. > I've wondered about this since I read in an Australian lens review that > putting a full frame lens (like the F2.8L) on a 1.6 crop digital can > actually result in reduced performance over putting a lenses specifically > designed for the smaller sensor. There is a myth that lenses can be "optimized" for digital and then somehow offer improved performance, usually with an extra coating of unexplainium on the glass to reduce reflections off the sensor or somesuch bollocks. I've never seen any non-theoretical evidence that it's not just a way to convince you to buy more shiny new lenses.  SignatureJeremy | jeremy@exit109.com >Take a full frame camera and a 1.6 crop digital. >Put a F2.8L lens on both. Set both cameras to automatic and same ISO and >take exactly the same shot. Will the shutter speed on the 1.6 crop camera >be the same as the other, or do camera manufacturers "tune" the internal >gain (effectively the ISO) to compensate for the different intensity of >the light on the sensor? The shots will be the same, within the tolerances of the cameras. Canon cameras (and a few others) have higher sensitivities then rated, EG the 30D achieves almost ISO4000 when set to 3200... The ISO on a digital camera is "tuned" by the factory to match a standard, since it's just a gain setting, as opposed to a film emulsion. >I've wondered about this since I read in an Australian lens review that >putting a full frame lens (like the F2.8L) on a 1.6 crop digital can >actually result in reduced performance over putting a lenses specifically >designed for the smaller sensor. > >Any thoughts? That's bull, it would be the other way around... you would actually take advantage of the less distorted center of the lens as well. Don't believe all you read... <>>I've wondered about this since I read in an Australian lens review that >>putting a full frame lens (like the F2.8L) on a 1.6 crop digital can >>actually result in reduced performance over putting a lenses specifically [quoted text clipped - 6 lines] > all > you read... I have to agree. Optimizing a lens for digital would just mean not having to cover the size of a 35mm frame, which wouldn't make it any better or worse -- just cheaper. Just depends on the individual lens. A 35mm lens that gets good reviews should get better reviews (or at least the same) on a 1.6x digital. > I've wondered about this since I read in an Australian lens review that > putting a full frame lens (like the F2.8L) on a 1.6 crop digital can > actually result in reduced performance over putting a lenses specifically > designed for the smaller sensor. It is more difficult (expensive) to make a lens covering a larger frame. So a lens covering a smaller frame can be sharper. http://digitcamera.tripod.com/#slr >> I've wondered about this since I read in an Australian lens review that >> putting a full frame lens (like the F2.8L) on a 1.6 crop digital can >> actually result in reduced performance over putting a lenses specifically >> designed for the smaller sensor. > It is more difficult (expensive) to make a lens covering a larger > frame. So a lens covering a smaller frame can be sharper. The easiest (and best) part of a lens is the center. So a lens covering a larger frame can be sharper, as it avoids all the problematic border areas. See? It cuts both ways. -Wolfgang > >> I've wondered about this since I read in an Australian lens review that > >> putting a full frame lens (like the F2.8L) on a 1.6 crop digital can [quoted text clipped - 9 lines] > > See? It cuts both ways. Is it easier (cheaper) to make a lens with soft corners, or a lens that does not need to cover the corners of the larger frame at all? (All other being equal, i.e., same sharp center). http://digitcamera.tripod.com/#slr >> >> I've wondered about this since I read in an Australian lens review that >> >> putting a full frame lens (like the F2.8L) on a 1.6 crop digital can >> >> actually result in reduced performance over putting a lenses specifically >> >> designed for the smaller sensor. >> > It is more difficult (expensive) to make a lens covering a larger >> > frame. So a lens covering a smaller frame can be sharper. >> The easiest (and best) part of a lens is the center. So a lens >> covering a larger frame can be sharper, as it avoids all the >> problematic border areas. >> See? It cuts both ways. > Is it easier (cheaper) to make a lens with soft corners, or a lens that > does not need to cover the corners of the larger frame at all? (All > other being equal, i.e., same sharp center). Depends on the lens. Making a low end lens with vignetting and soft borders on a FF sensor (but a good center) is probably cheaper than a lens with a good center, good border and little vignetting, especially as the borders are _used_ by and the vignetting _matters_ on an crop sensor. Not to forget: a FF lens you an sell to everybody, a crop-lens only to those who have crop sensors and don't worry about upgrading. Unless it's an extreme wideangle lens (say 10mm) for said crop cameras, you'll have to compete with all the full frame lenses, too. Of course, if you already have a world class fast telezoom with stabilizer and all the fun, it's not going to be easy to construct a crop-frame lens with the same capability, though building it may be cheaper (less glass) and lighter. -Wolfgang >Take a full frame camera and a 1.6 crop digital. >Put a F2.8L lens on both. Set both cameras to automatic and same ISO and >take exactly the same shot. Will the shutter speed on the 1.6 crop camera >be the same as the other, or do camera manufacturers "tune" the internal >gain (effectively the ISO) to compensate for the different intensity of >the light on the sensor? This comes up every once in a while. As the others have mentioned, all things being equal they will produce the same results. The only difference between FF and cropped sensors is the field of view that you see and capture. >I've wondered about this since I read in an Australian lens review that Whoever posted that review should be taken out and beaten with a wet noodle until they submit to the error of their ways. >putting a full frame lens (like the F2.8L) on a 1.6 crop digital can >actually result in reduced performance over putting a lenses specifically >designed for the smaller sensor. There is no such thing as special "digital" glass, so no, the results will not be reduced at all. In fact, they got it backwards. Using a FF lense on a 1.6x cropped camera will use the "sweet spot" of the optics in the lense center, usually resulting in better sharpness and contrast across the 1.6x cropped frame. Lenses that were made 10+ years ago before DSLR cameras even existed will function perfectly on cropped sensor cameras and produce the same quality images. >>I've wondered about this since I read in an Australian lens review that >>putting a full frame lens (like the F2.8L) on a 1.6 crop digital can >>actually result in reduced performance over putting a lenses specifically >>designed for the smaller sensor. > There is no such thing as special "digital" glass, so no, the results > will not be reduced at all. Ah, but there is: - lenses that don't fill out full frame, only the crop frame. Cheaper, lighter and --- if build well --- just as good. - lenses that (e.g. by multicoating) deal "better" with light entering from the sensor's end.[1] Where "better" is probably when compared to that manufacturers normal methods or marketing. > Using a FF lense on a 1.6x cropped camera will use the "sweet spot" of > the optics in the lense center, usually resulting in better sharpness > and contrast across the 1.6x cropped frame. Compared to that FF lens on a FF sensor. A bad FF lens will not magically become good on a crop sensor. Good lenses can beat a 8MP 1.6x crop sensors resolution (see photozone.de->Reviews->Canon EF 35mm f/1.4 USM L) under good conditions. Not all lenses are good. Compare the 75-300mm f/4-5.6 IS lens at 300mm. (Of course, that lens probably would have even weaker borders on full frame.) > Lenses that were made 10+ years ago before DSLR cameras even existed > will function perfectly on cropped sensor cameras and produce the same > quality images. The 75-300mm is from 1995 ... and most tests do not test things like "shooting into the sun", flaring, ... -Wolfgang [1] digital sensors reflect much stronger than film (which rather scatters light than reflecting it, too). Which exacerbates the problem of in-lens reflections back to the sensor/film. That's the reason TTL flash metering is done with pre-flash(es), not with light reflected from the film/sensor. Of course, good lenses try hard to avoid reflections anyway --- you'll get bad results otherwise even with film when shooting e.g. into the sun. Thus these lens avoid having a flat, parallel pane of glass even as a protective first element, and for that reason 'protective' filters in front of a lens --- which are flat, parallel panes of glass perpendicular to the lens! --- can degrade the pictures with stray light and in-lens reflections. >>>I've wondered about this since I read in an Australian lens review that >>>putting a full frame lens (like the F2.8L) on a 1.6 crop digital can [quoted text clipped - 7 lines] >- lenses that don't fill out full frame, only the crop frame. > Cheaper, lighter and --- if build well --- just as good. I understand what you're trying to say, but I'm sorry that's a design issue, and has nothing to do with special glass properties. >- lenses that (e.g. by multicoating) deal "better" with light > entering from the sensor's end.[1] Where "better" is probably > when compared to that manufacturers normal methods or marketing. That one I will "agree" is marketing, nothing more. At first I thought there might be something to it - seemed somewhat logical. But after personally using over a dozen different lenses from both Canon and Nikon that were all designed, coated, and manufactured before digital SLR came into fruition, I believe it's a load of bunk. As I'm sure you remember, some years ago the digital craze really took off, and anything that had the word "digital" slapped on the side of it was a hot seller due to brilliant marketing in the new digital age. It still shows today, but to a lesser degree, when buying something that is not at all digital, like home theater speakers. >> Using a FF lense on a 1.6x cropped camera will use the "sweet spot" of >> the optics in the lense center, usually resulting in better sharpness >> and contrast across the 1.6x cropped frame. > >Compared to that FF lens on a FF sensor. A bad FF lens will not >magically become good on a crop sensor. I didn't say it would. Obviously they will not become great lenses. But they do produce better results at the sensor edge than they do at the film edge. >> Lenses that were made 10+ years ago before DSLR cameras even existed >> will function perfectly on cropped sensor cameras and produce the same >> quality images. > >The 75-300mm is from 1995 ... and most tests do not test >things like "shooting into the sun", flaring, ... That has little to do with digital, and a lot to do with lense coatings and quality of design. Quite simply, a well designed and high performing lense made years ago will work just fine on a digital body. I've read reports of people using Nikon F-mount lenses from the 60-70's on their digital bodies, and they all work just fine. A good lense is a good lense. >[1] digital sensors reflect much stronger than film (which rather > scatters light than reflecting it, too). Which exacerbates the > problem of in-lens reflections back to the sensor/film. That's > the reason TTL flash metering is done with pre-flash(es), > not with light reflected from the film/sensor. I'm sorry, but pre-flash TTL came into existence prior to digitals proliferation, back in 1995 for example for Canons E-TTL. Digital had nothing to do with it. From what I understand, because the AA filter over the sensor is flat, any light it reflects goes straight back instead of scattering light the way film does. So bouncing light to the TTL flash sensor inside the mirror box was no longer practical with digital bodies. That's the reason why we can't use older TTL metering methods. The sensor was moved into the light path prior to exposing the sensor, where it measures the pre-flash and then continues the rest of the capture sequence. >>>>I've wondered about this since I read in an Australian lens review that >>>>putting a full frame lens (like the F2.8L) on a 1.6 crop digital can >>>>actually result in reduced performance over putting a lenses specifically >>>>designed for the smaller sensor. >>> There is no such thing as special "digital" glass, so no, the results >>> will not be reduced at all. >>Ah, but there is: >>- lenses that don't fill out full frame, only the crop frame. >> Cheaper, lighter and --- if build well --- just as good. > I understand what you're trying to say, but I'm sorry that's a design > issue, and has nothing to do with special glass properties. "glass" as in "lens", yes, that's a digital lens, since the APS format is dead. "glass" as in "UD elements" or "flourite elements", no, there's no special "digital elements" stuff. >>- lenses that (e.g. by multicoating) deal "better" with light >> entering from the sensor's end.[1] Where "better" is probably >> when compared to that manufacturers normal methods or marketing. > That one I will "agree" is marketing, nothing more. > At first I thought there might be something to it - seemed somewhat > logical. But after personally using over a dozen different lenses from > both Canon and Nikon that were all designed, coated, and manufactured > before digital SLR came into fruition, I believe it's a load of bunk. Ah, but did you sample good lenses? These would be designed right (instead of cutting corners). Have you also tested them in high contrast situations, e.g. shots where the sun is near or in the frame, possibly without using a lens shade? > As I'm sure you remember, some years ago the digital craze really took > off, and anything that had the word "digital" slapped on the side of it > was a hot seller due to brilliant marketing in the new digital age. > It still shows today, but to a lesser degree, when buying something that > is not at all digital, like home theater speakers. Naah, there are many 'digital' thingies that didn't make it. >>> Using a FF lense on a 1.6x cropped camera will use the "sweet spot" of >>> the optics in the lense center, usually resulting in better sharpness >>> and contrast across the 1.6x cropped frame. >>Compared to that FF lens on a FF sensor. A bad FF lens will not >>magically become good on a crop sensor. > I didn't say it would. I didn't say you did. > Obviously they will not become great lenses. But they do produce better > results at the sensor edge than they do at the film edge. True, unless they were completely broken by design (e.g. being even worse halfway to the border than at the border). However, that only matters if the lens itself is good enough for the pixel peeping possible with todays digital cameras. >>> Lenses that were made 10+ years ago before DSLR cameras even existed >>> will function perfectly on cropped sensor cameras and produce the same >>> quality images. >>The 75-300mm is from 1995 ... and most tests do not test >>things like "shooting into the sun", flaring, ... > That has little to do with digital, and a lot to do with lense coatings > and quality of design. Ay, and there's the rub: does the lens coating handle light reflected from the film/sensor well? > Quite simply, a well designed and high performing lense made years ago > will work just fine on a digital body. *IF* used so that it's strengths come into play, not it's weaknesses. (Every lens has both!) As an example, I wouldn't like trying to focus a MF lens on today's amateur DSLR viewfinders, with just a poor excuse for ground glass and no split screen nor microprism ring --- at least in most circumstances. (Small apertures and hyperfocal distances come to mind.) I also wouldn't like a 28mm for WA on a crop sensor. I also wouldn't like provoking flare and stray light with some lenses (just look at 30 odd elements of some IS zooms of today), even if they are near perfect otherwise. > I've read reports of people using Nikon F-mount lenses from the 60-70's > on their digital bodies, and they all work just fine. And why shouldn't they? > A good lense is a good lense. >>[1] digital sensors reflect much stronger than film (which rather >> scatters light than reflecting it, too). Which exacerbates the >> problem of in-lens reflections back to the sensor/film. That's >> the reason TTL flash metering is done with pre-flash(es), >> not with light reflected from the film/sensor. > I'm sorry, but pre-flash TTL came into existence prior to digitals > proliferation, back in 1995 for example for Canons E-TTL. True --- look up Canon's A-TTL for even earlier use of preflashes. But that doesn't invalidate my statement (ok, I was unclear: "digital sensors reflect much stronger [and don't scatter light]. That's the reason [for] preflashes]." > From what I understand, because the AA filter over the sensor is flat, So is the CCD or CMOS chip. It has to be flat and exactly perpendicular to the optical axis, or you'll have problems with varying sharpness especially wide open. > any light it reflects goes straight back instead of scattering light the > way film does. So bouncing light to the TTL flash sensor inside the > mirror box was no longer practical with digital bodies. Yes. Now, the light is reflected, straight back. If the lens was only coated and designed for light entering from the other side, that might give problems, like stray light being reflected back to the sensor, and not necessarily on the same pixel ... -Wolfgang |
Reply to this Message |
 Sign In
 Join
 My Latest Posts My Monitored Threads My Blog My Photo Gallery My Profile My Homepage Start New Thread Enable EMail Alerts Rate this Thread
|
©2009 Advenet LLC Privacy Policy - Terms of Use
This website includes both content owned or controlled by Advenet as well as content owned or controlled by third parties.