The Pursuit of an Ultra Shallow depth of field by Dirk De Paepe

The pursuit of an ultra shallow depth of field

with the Sony NEX-7 and the Canon FD 85mm 1:1.2 S.S.C. Aspherical.

by Dirk De Paepe

Well, I never owned a Leica and I probably never will. Nevertheless I’ve admired those camera’s for as long as I can remember and I can’t stop dreaming about them. I love their concept. I love their looks. I love their feel. And above all, I adore their image quality. But I’m one of those guys (and I guess there’s a lot like me) that say: no, this is way over the top too expensive. I can’t justify to spend that kind of crazy money for a camera, so I won’t buy it. OK, if it was half the price (that would still be a lot of money), I would go for it. But I like to spend my money on other things too, so I’ll pass. That’s why I have settled for a Sony NEX-7, with some nice Zeiss ZM glass, Instead of an M9 with Leica lenses.

You can’t always get what you want. That’s a fact of life. Until pretty recently, taking pictures just for fun was one of those items for me (besides owning a Leica). I’ve always taken a lot of pictures for my job (product shooting and reportage work within our branch), but that’s different. There’s not too much creative freedom involved in that. Admiring good pictures, and thinking about what I could do, if ever…, was the farthest I got. Pretty recently though, when I got more time in my business, I could really bit by bit realize what I’ve always dreamed about: spending more “quality time” with my camera and reading more about photography. I already renounced the big and heavy DSLR a long time ago, so Steve’s site drew my special attention. Visiting it on a daily basis really broadened my way of thinking about photography, and consequently it changed my way of shooting. Where at first, I disagreed with Steve on a regular basis (this doesn’t mean that I didn’t appreciate his opinion – on the contrary, I always valued it very highly), I found it remarkable that more and more I began to think in the same direction. I always considered reshaping once opinion as one of the most valuable abilities, so this process was a very positive experience for me, since it led me towards new and interesting paths. Reading about those special lenses, like the Noctilux or the SLR Magic 50mm T0.95 and looking at pictures taken with them, made me dream about owning some glass that really could produce this fabulous shallow DOF and the 3D separation that goes along with it. But again, those lenses were out of my league, regarding their prices and the fact that I would only buy one “just for fun”.

One of the nice things about Steve’s site is that it also publishes guest contributions – which I often find very inspiring. Anyway, it was thanks to one of those that I thought of giving my old Canon FD lenses a second life with my NEX-7. And then I discovered the FD 85mm 1:1.2 S.S.C. Aspherical, a lens that I have never before given any attention, because I was shooting in a different way. But now I thought: this will probably offer me an amazingly shallow DOF, close to that of those two great examples, that I mentioned above. Of course, I wouldn’t compare this FD to those famous lenses, but still, whenever I read about this 85mm, it receives the highest appreciation and is by many authors considered as one of the best lenses ever made. I already had the Novoflex NEX/CAN adapter, and knew that the FDs work terrificly well on the NEX. So I went bidding on ebay on a very nice example and all of a sudden got myself a beautiful “ultra shallow DOF” lens for about 1/4 of the price of the above mentioned SLR Magic, or less than 1/10 of the Noctilux! And, as I mentioned, concerning shallow DOF, it probably comes very close to those two great examples, although not being a 50 mm. But still… So I consider this as one of the best buys I did in years! In the meanwhile, I got so enthusiastic about shooting with it, and it works so well together with my NEX-7, that I’d like to share some pictures, made with this camera/lens combination.

Of course, being an 85 mm and the NEX using an APS-C sensor, this makes for a very important crop factor and a compressed perspective, compared to a 50 mm on the M9. When, for instance, I focus on a person that I wanna picture from the waste up, I will have to shoot from a larger distance, thus focussing further away and loosing some of the shallowness in DOF. But on the other hand, the shallowness increases again with the tele factor, so that compensates. Over all I believe the DOF is pretty comparable. It’s mainly the compressed perspective that bothers me sometimes. That’s why I continue dreaming about shooting a 50mm with a full frame camera. But like I said, you can’t always get what you want, still this lens/camera combination can give very nice results. Often I can even benefit from the crop/tele factor, like in portraits or in the shooting of people in general. Being able to stay at a larger distance leaves the person more easily in his comfort zone, which typically results in a more spontaneous facial expression. It’s important to always look on the bright side of life, isn’t it…

My first picture is a good example of this: a non-posed portrait that brings the girl’s face very close, thus creating a great intimacy. Her facial expression stays very natural and makes you wonder about here thoughts. This picture is an out of camera RAW conversion, no cropping nor any other processing was done. It’s absolutely razor-sharp up till 100% enlargement, and introduces the haze already at the cheek. I believe the shallowness is pretty amazing. The background, which has no function at all in this case, is completely blurry beyond recognition, thanks to the wide open shooting. IMO this lens is a dream for portraits. And the focus peaking of the NEX made it possible to focus very fast and accurate, which is an absolute necessity when portraying in that kind of “instant” circumstances. IMO, both lens and camera are perfect for this kind of work.

Despite this, I love those “ultra shallow DOF” lenses the most, when they’re focussed at a medium distance, because then the background, although still blurred, remains recognizable. But at the same time, the object in focus benefits of a very clear separation, not possible to achieve with a less fast lens. With this wonderful 3D effect, you can evoke extra expression. Thanks to the information in the background and the great separation, you can bring a completer story. And again, the combination of the NEX-7 (with its focus peaking and brilliant viewfinder) and this Canon FD 85mm offers total control and very fast action. The “Kissing in the Park” picture is a good example hereof, where one really needs to act fast, to be able to freeze the action in the split second that it has its greatest expression. This picture was taken from a distance of about 10 meters and was cropped to 87,5%, since there was a bit too much grass in the foreground.

I really love this Canon FD 85mm lens in combination with the NEX-7. It’s really razor-sharp when shooting wide open. I find the colors to be rich, well-balanced and very natural and I like its bookeh. Of course it’s a somewhat heavy lens, weighing 737 grams, but it’s a lot of glass and, as far as I can tell, all ultra shallow DOF lenses are pretty heavy. As a matter of fact I always shoot it wide open, because I only use it, when I’m pursuing this ultra shallowness. I also own a Zeiss ZM Tele-Tessar 4/85mm for “normal” work, which comes in much more handy, due to its much lighter weight and compact format. I believe a lens like this Canon is made to shoot wide open, otherwise one wouldn’t wanna carry that weight… By the way, this Canon lens is available on ebay on a regular basis. I read it’s very familiar in terms of construction to the currently available EF 85mm f/1.2 USM that sells for over 2000 Euro in Europe.

I hope you can enjoy some of my pics, and share with me the satisfaction that for about 2000 Euro’s I have bought myself a camera/lens combination, that can bring an atmosphere and an expression to my pictures, for which, until recently, I thought I’d need to spend at least three times more money. Yes, I’m a happy man…


  1. Hey guys! forget about all theese “0.95” and “1.2” toys ) – try Kodak Aero Ektar 12″ (307mm)/2.5 on a 8×10″ camera – it’ll give you equivalent of ~46/0.37 : )

  2. How do you find focussing this lens wide open on a mirrorless camera? Does the NEX-7 give you any help, is there a focus magnifier?

    • The NEX-7 indeed has a focus magnifier. But I find the focus peeking much more convenient and efficient! You need to practice a bit with it, to be able to swiftly find the “peeking center” (that is the exact point of focus) and to know what color and sensitivity you wanna use for the fringes. Practicing your focus speed is very important when using a really fast lens, because, while the DOF is so shallow, you really need to focus spot on. So sometimes you’ll go a bit back and forth to find the exact center – a bit comparable with focusing with a range finder. With some practice, focus peeking really works fast and accurate, IMO faster than with a range finder, because the focus peeking offers the enormous advantage that you can focus in any point in the viewfinder (!), thus eliminating the need to reframe. The pictures of the dog, the portrait and the kissing were taken “on the very moment that it happened”, in a split second. So focussing needed to be done really fast. I’m not an AF photographer, but I really wonder what kind of auto focussing system would be able to take the “kissing in the park” picture and autofocus correctly in a split second. Maybe it’s possible, but I don’t know of one. I’m just curious… If anybody knows it, please tell me.

  3. Great photos … there is of course always more than one way to skin a rat! ….. If using Pentax then a K5 + the inimitable FA77 1.7 will be yours for around Β£1500 …… and well …need I say more ?

    • You are so right, Tom. The FD85 on the NEX-7 is one camera/lens combination that makes me happy. Did I buy the ultimate set? Or is it rather your Pentax suggestion? That’s a question that one can never answer for sure, I guess. And frankly, Tom, I couldn’t care less. What matters to me is: it’s very affordable and it does exactly where I bought it for – and in an outstanding way! It’s a blessing that there are more ways to skin a rat, it only makes more people happy – as long as they can be satisfied with what they got! I really don’t wanna spare the very last Euro, but I don’t wanna pay three or five times more… Big difference!

  4. Try to find a Canon FL 58/1.2. This lens is the closest thing to the 1.0 Noctilux M but only a few people know about it. FL was the predecessor of Canons FD series. The mounts are more or less compatible.

    • I also tried and even bought an FD 55/1.2 aspherical, but it was a lot less sharp then my own FD 50/1.4 (that I own since the 70’s) and not that much shallower, while my 85/1.2 is really in another league, both regarding sharpness as well as DOF. So I was lucky that I could return the 55/1.2 to the seller. It’s probably a good idea to take a closer look before buying. πŸ˜‰

  5. When using a FF lens on a smaller sensor, in order to get the same field of view you need to step back to a distance = (crop factor x distance). By then, the depth of field at the target increases to more than what an equivalent focal length FF lens would deliver at the same speed. Spending money a on fast lens and then not using a FF sensor camera to get the most out of it is not the sensible way to achieve the “ultra” shallow DoF. The ultra fast tele lenses are big and defeats the purpose of having a APS-C sensor small camera body like the Nex-7.

    • You’re absolutely right Hakan. But you know, to me, getting a FF camera has always been my final goal. And I think pretty soon Sony (or Fuji for that matter…) will come with an IC full frame body. And I’m pretty sure Novoflex (and others) will make an M-mount adaptor for that camera. So personally I look upon the M-mount as “my standard”. I also own a Zeiss Ikon film camera, that is M-mount. And I recently made my FDs all M-mount, as I did with my Jupiters. So the NEX/M adaptor stays on my NEX-7 (as it will on my future FF Sony), and I use the two camera’s (three to come after purchasing the FF Sony) with one and the same set of lenses. That’s really convenient! When I will have bought the FF Sony, I’m still gonna keep the NEX-7, exactly because of it’s crop factor, that sometimes really comes in handy. It’s like a tele, without the reduced perspective…

  6. I think the physics of a lens dos not change with the sensor size. An 80 mm lens is 80 mm is a 80 mm it is just that to achive a certain magnification with that lens on a full size camera you have to be in half the distance from the subject as of the same lens on a micro 4/3rd and the dof is more the further you move from the subject so thats why people say dof changes with the sensor size. It dos not in the same distance regardless of your sensor size.

    • Absolutely correct, Benarou. That’s exactly what I meant. The sensor itself doesn’t change the DOF. But when you compose a picture, you need to be further away with a smaller sensor and distance definitely changes DOF. So the sensor size only changes the DOF indirectly, not really in a mathematical way of speaking, but still, in practice, indeed it does. It’s not because of a smaller sensor that I’d want only a part of the girls face in the portrait, isn’t it…

  7. Here’s the best combination for ultra smooth bokeh and shallow DOF:

    Nikon full frame (e.g., D800/E) and the 200mm f/2, especially around 10-15feet.

  8. NIce shoots. With full frame do you not get shallower DOF. If what to go whole hog, 300mm f2.8/200mm f2.00. But nice shoots You have a good eye for isolation of subject.

  9. Dirk,

    First of all many thanks for your inspirational contributions to this site (and of course to Steve for running it )! I’ve enjoyed ‘lurking’ here for many months now, and this is my first post here. I love your enthusiasm for photography in general and for the NEX-7 in particular. Actually, it was your first (I believe) Daily Inspiration post about using the NEX-7 with Zeiss ZM-lenses that motivated me to explore mirrorless cameras in-depth – and I was hooked! As a result, I’ve been the happy owner of an NEX-7 and some ZM-glass for some months now, and I couldn’t be happier!!

    For the past ten years, I have mainly pursued wildlife photography with a focus on birds, using Canon DSLRs and lenses in the 300 to 500mm-range. With a 1.4x tele-converter and when using a camera with an APS-C sensor, this can result in effective focal lengths of over 1000mm. With maximum apertures from f/2.8 to 4.0, these lenses have extremely shallow DOF when used wide open. Despite the difficulty of focussing accurately on a bird’s eye (forget about AF with fidgety subjects like birds – it will focus anywhere else but where you want it to), I really enjoy working with this shallow DOF. Not only does it work wonderfully to isolate the subject from it surroundings; the lenses I’ve used also perform very nicely and offer beautiful bokeh when wide open. So it’s very similar to what you and others are after when pursuing shallow DOF, only with much longer focal length and often with smaller subjects – and with a lot more weight to lug around…

    With my NEX-7, I have thus far mainly done landscape photography, with the Zeiss ZM lenses usually stopped-down to around f/8.0 to achieve both the desired large DOF and maximum optical performance of the lenses. I’ve been blown away by the results, I must admit. These tiny RF-lenses in combination with the hight resolution of the NEX-7’s sensor deliver an image quality that was previously unknown to me even from high-quality Canon L-glass used with a 5DMkII. I am still awed by the fact that I can produce wonderfully detailed 24×36″ prints with a camera and lens combination that weighs less than two pounds. It’s nothing short of a revelation for me…

    I’m afraid this was a little long-winded, but I guess it serves reasonably well for introducing myself to all of you. What I’d really like to express is my gratitude to you and to all of you here (and most of all to Steve, of course) who make this the inspiring community that it is! I’m looking forward to many more enjoyable months or readership – and perhaps even a contribution or two, who knows…



  10. The fd 50mm f1.4 is also great. It’s super sharp and can give you decent shallow depth of field and a 3d look. Here’s one I took of my daughter a while back using a gf1 which gives a fairly good view of he sharpness and depth of field.

    It also makes a great portrait lens on a micro four thirds camera and is a cut above my Olympus 45mm f1.8. For reference, it’s very similar to my modern Canon ef 50mm f1.4 with the exception that it is prone to flare.

  11. You’re dead right about using the Canon 85mm, Dirk: so many people lust after a hugely expensive ‘Noctilux’, but they’d get almost exactly the same results (almost) by just using a longer lens with a slightly less-wide aperture you’ve just shown here! ..Lovely soft background, and sharp faces.

    Canon also made earlier screw-fit lenses for their Leica copies (Canon started life as a company making Leica knock-offs), and their 85mm f1.8 and 100mm f2 in Leica screw thread lenses (easily found on eBay) are terrific your shots here.

      • Michiel, That premise only works if you don’t change the viewing size of the cropped image. If you display the images at the same size (which is what we would normally do) apparent DOF would be different.


      • Sensor size is a valid factor of DOF because it, along with focal length, determines your field of view, which determines your focus distance.

      • Assuming that you use the same equivalent focal lengths, the same aperture, and stand the same distance from the subject (same magnification,) then DOF does change with sensor size.

      • (..Taking a VERY deep breath, and only posting this once, as I really don’t have the time or inclination to get involved in this today – guest’s coming in half an hour, and the cat needs attention to the splint on its leg [repaired Achilles tendon]..)

        Some of what Robert has said, above, quoting the man from Zeiss, could be easily misinterpreted, as the man-from-Zeiss’ English is a little less than perfectly phrased..

        The man from Zeiss says “..If we select the suitable focal length to ensure that we always display the same field with different film formats..” – but this should really be written as: “..If we select the same apparent focal length..”

        That’s to say, we should start by selecting lenses which deliver a similar image on different-sized formats on “full-frame” 36x24mm let’s use a 50mm lens. For APS we’d thus use a 34mm lens (which gives the same view as a 50mm lens when it’s used on the smaller APS sensor), and for (micro)FourThirds we’d use a 25mm lens (which will give the same view on m4/3 as a 50mm lens does on “full frame”.

        So we’re using 50mm, 34mm and 25mm – but they each give the same apparent view when used on their relevant sensor. The sensor itself has nothing to do with it: we’re considering only the size of the sensor, or the piece of film ( you correctly said, Michiel, at point 4, above, when you mentioned “ just take a smaller bite out of the image circle”.)

        So these three lenses, 50mm, 34mm and 25mm, are all set to the same aperture ..let’s say f1.4.

        The 50mm f1.4, on full frame, will have a shallow depth-of-field at this wide aperture – and it gets shallower the closer that you focus, and the closer that you are to something! – so that it gives a nicely blurred background beyond, say, the head of someone you’re shooting at three feet (1 metre) away.

        Now take the same shot with the 34mm f1.4 lens wide open on the APS-sized camera. It will have the same apparent view of the head at 3 feet (1 meter) away, BUT, because this is a 34mm lens (a shorter focal length than 50mm at the same distance and aperture) it will show a greater, deeper depth of field, with more in sharp focus ..because it’s a shorter focal length lens ..all other things being equal – i.e; distance and aperture.

        Now take the same shot with the 25mm f1.4 lens wide open on the m4/3 camera. It will have the same apparent view of the head at 3 feet (1 meter) away as the 50mm lens on full frame and the 34mm lens on APS – BUT, because this is a 25mm lens (an even shorter focal length than 34mm at the same distance and aperture) it will show an even greater depth of field, being sharp quite some distance beyond the head which is three feet (1 metre) away.

        That’s why some people are reluctant to use m4/3, because they say that they can’t get the same shallow depth of field that they can with full frame ..or with the smaller-than-full-frame APS cameras.

        The apparent depth of field depends ONLY on the focal length of a lens, when all other things – such as aperture and distance – are equal.

        To get a similarly shallow depth of field with an m4/3 camera as you’d get with a full frame camera when shooting at the same distance as full frame you’d need to open up the aperture by two stops (from 1.4 to f0.7) to get similarly shallow d–o-f as full frame. That’s because you’re using a 25mm lens (instead of 50mm) which has half the focal length of the 50mm lens. (And with APS, you’d open up the aperture by about one-and-a-bit stops.)

        The man from Zeiss says (in his point 1: ‘Distance’) “..depth of field grows with the square of the focusing distance”. But in our example, if the focusing distance is always the same (i.e; the head you’re shooting is always three feet – 1 metre – away) simply ignore his comment’s irrelevant in this example.

        The man from Zeiss also says (in his point 2: ‘Focal Length’) “..the depth of field with equal focusing distance is inversely proportional to the square of the focal length”. This means that at the same distance away the d-o-f increases as the lens’ focal length decreases you get greater depth with a 25mm (or 34mm) lens than with a 50mm lens at the same distance away ..and he should also have said, but missed saying, “at the same aperture”. ..But I already mentioned “all other things being equal”.

        Finally, the man from Zeiss says (in his point 3: ‘Aperture’) “..the depth of field increases linearly with the f-number..” meaning that as you close down the aperture, the depth of field increases, but he didn’t mention anything about the distance at which you’re shooting, nor the focal length. This business of stopping down is “universal” whatever distance you’re shooting, and with any lens, the d-o-f will double if you decrease the aperture by 2 stops (e.g; from f5.6 to f11). But that, too, is irrelevant here, when comparing different focal lengths for different sensor, or film, sizes.

        Essentially: forget the sensor size! ..except to remember that if you quarter the sensor size (e.g; from full frame 36x24mm to m4/3 17.3x13mm ..that is, if you halve the length of the sensor and also halve the height of the sensor) you must then halve the focal length of the lens you’re using to get a similar image ..i.e; use a 25mm lens on m4/3 for the same image as a 50mm lens on full frame.

        But halving the focal length from 50mm to 25mm will double your depth of field, because you’re using a shorter lens. So because you’ve halved the focal length, you must open up the aperture by two stops to get a similarly shallow d-o-f as the longer focal length (50mm) lens.

        Forget the sensor; just think of the focal length: open the aperture 2 stops if you halve the focal length.

        Some of the geometrical perspective will change if you alter the distance at which you’re shooting, but all other things being equal, if you halve the focal length, open up the aperture 2 stops for the same d-o-f.

        If you use a 50mm at f4 on full frame, then 25mm at f2 will give the same appearance, and d-o-f, on m4/3.

        (Our guest’s arrived, the cat needs food and medicine, so I gotta go now, and won’t be writing any more about this, as there’s already plenty on the web. It’s just a matter of unscrambling it all ..and some of it isn’t too accurate, so tread carefully!) – David.

        • I rest my case, and can not even hope to approach your knowledge of teh subject. thanks for that eloquent explanation of the subject David. Hope and trust that dinner with your guests went well!

          • “I rest my case, […]”

            Which case?

            Please note that David Babsky does not confirm that DoF is independent of sensor size. The only thing he does here is offer you a method for taking sensor size out of the equation. This method explicitly avoids

            a) using the same focal length on different sensor sizes
            b) ‘zooming’ with your feet

            If your ‘case’ does not contain one or both of these two elements either, you can safely rest it indeed. Otherwise I’m afraid the two of you are not on the same wavelength re. all this, and you might still have a gap to bridge.

            The concept of equivalence, which David Babsky uses here to explain himself, is IMO an excellent starting point for approaching DoF and understanding how it fits together with all the other parameters like sensor size, focal length, field (or angle) of view, subject distance, etc. I can warmly recommend this article for further reading.


            Hope to have been of help.

        • I think this relationship between achieved DoF is not linear with the crop factor, but more like the square of it.

          Here is a worked example using Leica’s data.

          A = FF sensor camera with Summicron M 50mm, 2m away from subject.
          B = FF sensor camera with Apo Summicron M 75mm, 3m away from subject.
          C = APS-C sensor camera with Summicron M 50mm, 3m away from subject.

          All three cameras will have the same field of view.
          Aperture is fixed in all cases and it’s f/2.0

          A will have DoF = 187mm
          B will have DoF = 209. (ie. very similar in DoF to that of A)
          but C will have DoF = 427mm

          To get around DoF of 427mm A needs to be set to over f/4.0
          So the increase in DoF is more like square of crop factor.

          So the using a fast tele FF lens on a smaller sensor defeats the purpose of wanting to achieve “ultra” shallow DoF.

          However, combination of using fast lenses (they have high quality glass so diffraction comes into play at smaller apertures) on a cropped sensor is a great way to increase DoF without letting diffraction of glass getting into play.

          I used a Noctilux .95 on a NEX-7 and set it to f/11 to achieve deep DoF, it surely worked well.

          • “All three cameras will have the same field of view.”

            This is incorrect. Field of view is an angle determined by sensor size and focal length. Changing your subject distance doesn’t change your field of view.

          • What the diagram shows is how you think you can zoom with your feet. My point is that you can’t.

            Zooming is what you do when you crop or increase focal length. This changes your field (i.e. angle) of view.

            Getting closer to your subject does not change your field of view. It only changes what you include in it.

            Another difference is that your perspective changes when moving your feet. This might not be very apparent when photographing a flat wall with a door in it (like in your diagram), but it will when you photograph something more 3D, like when you take a portrait of your girlfriend.

    • Absolutely, David. I find the 85mm a very usable lens indeed – especially for shooting people, where I prefer it over a 50mm – because of the somewhat greater distance one can keep. At the same time it’s still very managable, which is certainly no longer the case (IMO) with say a 200mm lens. With “managable” I mean: aiming out of hand, focussing and shooting in a blink and easily carrying it with you around your neck for a few hours. Only minus (again IMO) is the somewhat compressed perspective (but that doesn’t count for portraits with unrecognisable background). That’s why I also would love a very fast 50mm. Maybe I’ll ever go for a VoigtlΓ€nder Nokton 1.1, because it’s still quite “affordable”. But anyway, I will always keep this 85FD, because it IS an incredible peace of glass for sure!

    • Hey guys, there is a motto I impress a million times on my collaborators and that is: “In achieving a goal, the greatest difficulty is to keep it simple.”
      Wow, all this technical stuff is really to much! Really, please don’t expect me to read it all! I work with a few very simple facts. And one can easily read those from the scales of a traditional manual focus lens and looking through a viewfinder.
      1) the lenght of the DOF becomes shorter when the focus distance becomes shorter.
      2) the lenght of the DOF becomes shorter when shooting with a (longer) tele-lense.
      3) to obtain the same framing with a smaller sensor, one must keep a greater distance.
      Maybe, to be really scientifically correct, I should have used a lot more or other words, but with some goodwill, you know exactly what I mean, even though I’m not native English speaking.
      This means that, indirectly (!!) and in practice (!!) DOF with a FF sensor is is indeed shallower (I prefer “shorter”, but who am I when it concerns the English language?) than with an APS-C or other smaller sensor, if you want the same subject (say a portrait) framed in the same way. And really, that’s all I need to know about it. Take the girls portrait, with the same FD85 lens on, say an M9. To get a comparable picture, I would have had to come closer, with a shorter DOF as a result. To me that’s a simple fact. I don’t need complicated arithmetics to see that. Of course, it wouldn’t have been exactly the same picture, with for sure a different blur at the background, but it probably would have been equivalent, only with a shallower DOF, that is: the blur on her face would have occured earlier. Can anybody disagree with that?

      • “Can anybody disagree with that?”

        In principle nobody can, because the post is about what works for you, and you are the only person governing this.

        Moreover, you ‘only’ compare FF vs. APS-C, two formats that in the end differ too little for most people to be really bothered by in practice. It’s not like an 85mm lens turns an APS-C camera into something completely inadequate for taking portraits. One can safely state that in your specific comparison, the different fields of view can, for all intents and purposes, be compensated for by taking a different camera position. Though technically debatable, it’s pretty hard to disagree with in practice.

        Would you on the other hand compare FF against say 1″, things would look considerably different. On a Nikon 1 camera, 28mm is seen as a pretty nice portrait length, but I can assure you that your love wouldn’t like being portraited with a 28mm lens on FF. Don’t worry honey I’ll simply get a bit closer will suddenly not work anymore.

        The gap I see here is that some people will say: hey I don’t have a 1″ camera, whereas others will say hey that’s interesting where does this difference come from? I don’t really mind where you stand on this, but I sincerely hope that you’re not ‘keeping it simple’ because you think the ‘science’ behind all this is too complicated to grasp. It isn’t. It’s only a bitch to get across over the Internet.

        Finally I’d like to tell you how much I like your last shot. Luckily your lens left some of the background intact, because I think it really adds to the romantic feel. The shot is exquisitly timed too. Thank you so much for sharing it.

        • Dear Alexander,
          Steve published my article when I was on vacation, the first days without internet. It was only the 21st in the evening, three days later, that I was surprised by the publication and the number of replies, I wanted to reply asap. So I began reading and writing, my wife went to bed, and I said to follow soon, but stayed for more than another two hours at my screen. πŸ™‚ When I wrote “keep it simple” and “please don’t expect me to read it all”, you have to look at it from the perspective of that situation.
          Of course I’m very interested in the “science” behind it all, but I reckon a forum is not the best place for it, with so many people coming with theories that others contradict. It is very enthralling and entertaining, but up till now, I just haven’t had the time yet to read it all.
          And again I’d like to keep it simple. I have a very clear question: is anybody capable of comparing the DOF of my lens with, say the Noctilux? But then in a meaningfull, practical way. I’l give an example. Focus both lenses at a distance where the frame hight is exactly 1 meter. Determine from that point the lenght of the DOF for both lenses. To me that’s the only kind of comparison that makes sense. (I hope I explained myself clearly.) Can anybody explain how to make this calculation? I guess it doesn’t need to be complicated.
          Further, Alexander, thank you so much for appreciating my last picture. It was exactly for that kind of pictures that I bought this lens, where the background contributes a lot to the story. But at the same time the 3D separation adds a lot of drama to the subject by strongly drawing the attention. It takes a great lens to produce such a great expression. And I believe the FD85/1.2 is part of this select company. You know it was actually one of Steve’s pictures in particular that triggered me into wanting that kind of lens (in his article The SLR Magic Hyperprime 50 LM T0.95 Leica Mount Lens “Rolling Review” it’s the picture of the sitting black gentleman speaking through his cellphone – I was blown away by the 3D and BTW also loved the colors of the M9/SLR combination a lot – increadible picture IMO!, that also benifits a lot from the background and stil IMO has an ultra shallow DOF). So when I went to the park in Hasselt (Belgium), I already had that sort of scene in mind, hoping to find the right figurants. Of course I was lucky to encounter this lovely couple at that beautiful spot. But then again, it’s no coincidence that they’ve exactly chosen this nice spot to “get inspired”…

  12. I think if you’re really interested in shallow DOF and that special 3D look, you should look into the Brenizer method. It’s basically a special panorama technique where you use fast telephoto lenses to capture close objects (like you would with a 35mm for example). If you shoot wide open and stitch the images afterwards, you are basially reversing the crop factor, or increasing the equivalent focal length. This way you can easily create images that are comparable to the Noctilux and even more, but of course it takes some practice and when shooting people they have to be a bit more patient. πŸ™‚

    • I love the thrill of the shooting, Willi. That is: I consider a picture the more perfect, the least I have to process it afterwards. That gives me the most satisfaction. Also, I love it when I was able to really “nail the moment”. All of that is not compatible with the Brenizer method, I guess…

  13. To avoid a possible misunderstanding (ask D Babsky): DoF doesn’t change with a smaller sensor size; you just take a smaller bite out of the image circle.

    As for the images (which I can’t view right now at full res for some reason): I don’t think these are “ultra shallow”; not even really shallow. I find ultra shallow can best be achieved very close up, where the slightest focusing error can make an eyelash pinsharp but the eye you were focusing on visibly unsharp. A 50 (or 85) close up does that at f2.0, 1.4, and 1.2… πŸ™‚

    • For a given focal length, aperture and subject magnification, DOF will change with sensor size.

      A full frame sensor with a 50mm lens at 1.4 will produce exactely half the depth of field compared to a micro 43 camera with a 50mm lens at 1.4.

      The most common misunderstanding is that DOF changes with focal length for a given sensor size.

      Check out “Understanding DOF” at Lumonous Landscape.

      • Robert, you are wrong. DOF is independent of the sensor, it is dependent on focal length and f number.
        The reason you have more effective DOF with a larger sensor is that you have to use a longer lens for the same field of view.

        • Robert is not wrong per se, his comparison only doesn’t assume the same field of view, nor the same subject distance, whereas yours does. One could make the case that he isn’t comparing apples to apples, but apart from that, his calculations are correct.

          • My comparison included a given subject magnification, wich is the same thing as field of view in practical terms (what you see on the sensor). Not? The subject distance will of course be different when comparing f.eks. full frame to m43 with the same focal length.
            Dof is a tricky subject if you do not include all parameters. There is an interesting practical test at luminous landscape where different focal lengths are used at the same aperture (1 camera=constant sensor size). The subject is framed to produce the same field of view (subject magnification) and the shots produce the same degree of background blur (Dof). This was done to show that Dof is not dependent og focal length. This misconception is probably common because wide angle focal lengts in practical use frame motives with a low subject magnification and vice versa.
            Now please tell me i i don`t have this all wrong, because i have to start a lot of studying again.
            Thanks for the support.

          • I went to have a look at this experiment and you are misinterpreting the result. Obviously you can never achieve the same field of view with different focal lengths.

            It is easy to prove you wrong. Grab a 35mm lens set at f/1.8 and photograph a subject with a distant background. Now take a 85mm lens also set at f/1.8 and frame the subject similar to the first photo. The 85mm will have far greater background blur.

          • Meh. Robert is wrong. DoF is defined by 3 parameters: the maximum aperture opening, the distance of the subject and the magnification. The maximum aperture opening is in turn a function of the focal length (i.e. f/1.4).

            If you use the same 50mm lens on both systems, you’d have to take a step backwards on APS-C to fit the subject, but the magnification is also 1.5x larger. Your DoF calculator might give you different numbers, but the amount of out of focus blur will be identical. The full frame might look more impressive, because you are achieving this blur at a wider angle of view.

            It is when you use a 50mm f/1.4 lens on full frame, and a 35mm f/1.4 lens on APS-C, that the 50mm has a small advantage in maximum aperture opening. Obviously, slapping a 200mm lens on your camera rather than a 50mm has a far greater impact than a 50% difference in sensor diameter.

          • I don`t think I`m totaly wrong. My main concern was that sensor size affects DOF. I have done some searching on the web. Check this out:

            5. Different film formats
            with the same object field
            If we select the suitable focal length to ensure
            that we always display the same field with
            different film formats, then things go just the
            other way round: reducing the size of the
            sensor format increases the depth of field,
            and enlarging the sensor format reduces
            the depth of field, as long as we always use
            the same aperture setting. That is because a
            smaller sensor format displays the same object
            field with an accordingly shorter focal length. If
            the same f-number is used, then the entrance
            pupil is reduced by the crop factor and the light
            cones are narrower.

            Now this is comes from an engineer at Zeiss. After reading the article i know i do not understand everything about Dof. But if i got the sensor size part wrong, so does the guy from Zeiss.


          • DoF calculators tell you the depth of the region that is in focus, but tell you nothing about the amount of out of focus blur.

          • Ah I just read your earlier reply to Robert, and understand now where you’re coming from.

            “Grab a 35mm lens set at f/1.8 and photograph a subject with a distant background. Now take a 85mm lens also set at f/1.8 and frame the subject similar to the first photo. The 85mm will have far greater background blur.”

            Sure. And indeed a depth of field calculator doesn’t tell you about it. It simply shows the same DOF value for this particular example and that’s it. However there is so much else it doesn’t tell you about either. For starters, the two shots would have a completely different perspective. This might work for you in practice, but for me it doesn’t when it comes to a theoretical discussion on depth of field.

            I mean the theorist in me would say: hey, we’re discussing depth of field here, so go away with your out of focus blur. The artist in you might answer: to hell with your depth of field, you might not have noticed but out of focus blur is what this Steve Huff article is about.

            All in all I see two different starting points that are not so easy to cover in one discussion. Appreciate and agree with your point though, and sorry for the confusion from my side.

    • Agreed. I’ve taken close-ups of my 2-year-old simply using my NEX E-mount 50mm f1.8, and one eye is razor sharp and the other is visibly soft, because her head was slightly turned. Even a hyperprime at greater distances will give you a larger DoF than something very close at 1.8.

      And the shallowest depth I’ve dealt with as when trying to use Extension Tubes on an old 28mm MF lens. The focal point was about a centimeter from the lens, with a DoF less than a millimeter. In fact, I couldn’t get all of the head of a wooden match stick in focus. I would call that shallow DoF.

    • Hello Michiel,
      I was away for some days, so couldn’t reply earlier.
      Please have a look at
      It shows a 100% crop of the girl’s eye. IMO the DOF is pretty shallow, the focussing was very delicate and was indeed spot on, all done in a blink of an eye, thanks to the focus peeking of the NEX-7 (and quite some practicing from my part, when I just bought the camera, to master the focus peeking).
      You are of course very right, saying that the DOF is narrower (calculated in cm or inches) when the subject is closer, but then you would only picture a part of the girls face, wouldn’t you? Well, that was not my choice. I find the framing to be perfect, and wouldn’t wanna have it any other way. So the focus distance needed to be bigger than what you suggest. The shallowness of the DOF by itself is not my goal, you know. This is not a competition in shallowness to me. I wanted to portray the girl, and use the shallowness only to help creating a beautiful atmosphere.
      Now I look upon shallowness of DOF as a relative notion, that is: the length of the field that’s in focus in relationship to the focus distance. (I hope I express myself well, because I’m not native English speaking.) To me also the last picture has an ultra shallow DOF, because when focussing at 10 meters (33 feet), to get that kind of blurry background, one needs an ultra fast lens. And BTW, I find the focussing in this picture as delicate to perform as with a portrait. I think the in focus zone on the focus ring is as narrow at 10m as it is at 1m.
      I’m not a guy that argues about words, Michiel, but rather about meanings. Maybe I should have written another title above my article, namely “The Pursuit of a really exceptionally shallow depth of field, in relationship to any given focus distance”. πŸ™‚ We have a Flemish saying that goes (I translate) “A good understander only needs half a word.” Keeping that in mind, I reckoned that the shorter version of the title would do fine…
      What’s important to me is: taking the girl’s portrait, I guess the shallowness is really in the same league as with a Noctilux, when portraying the same girl with the same framing. That is: a big step more than say with a 50mm/1.4 lens. And really, I couldn’t care less if one or the other lens is a fraction more or less shallow. What only mattered to me was that, for a fraction of the price of “the big two”, this lens is able to create the right kind of atmosphere: the unique atmosphere of an ultra fast lens. And like I said, I like it the most at medium distances (say 10 meters) and I still call that ultra shallow. But you are so very welcome to call it something else.

    • BTW, Michiel, Steve repared the enlargement feature for the pictures in the meanwhile. Of course they look a lot better when enlarged – especially the colours appear a lot better and you have a much better idea of the sharpness. But it’s also worthwhile to look at the 100% crop on twitter. It really proves the lens’s ability.

    • Sorry but surely the girls expression is ‘Who the hell is pointing a dirty great lens at me!’ The other girl is completely disgusted assuming this is a candid street shot? I have a shot a bit like this when I randomly pointed my camera at a girl at Camden Lock London. But as for shallow depth of field yes it cuts the mustard.

      • The two girls are friends of mine. Those two pictures were taken at our ranch after a horse show. They new and approved that I was taking pictures. So no, they didn’t dislike it at all, on the contrary.

Leave a Reply

Your email address will not be published.