depth-of-field .. f2.8 vs the rest

aperture and depth of field

A question that came up in the forum recently was whether an 70-200mm f4 zoom would give you the same kind of look that an f2.8 zoom would.

For me, the f2.8 aperture is essential, especially with a telephoto zoom. For the same scenario, it gives me a higher shutter speed than the f4 zoom. Or I can use a lower ISO. More importantly, since I often bounce flash in large areas, the f2.8 aperture gives me more chance of successfully bouncing my flash than an f4 aperture would.

But what does the change in depth-of-field look like?

Keeping in mind that the how much the background appears to be blurred, depends on a few things:
– the chosen aperture. The wider the aperture, the shallower the depth-of-field will be, giving a more blurry the background. (This is the main factor that affects how much the background will be blurred.)
– how far your subject is from the background,
– your chosen focal length.
– The bokeh of the lens will also affect whether the background appears more blurred, or less. (Note that bokeh and shallow depth-of-field are not the same thing.)

The change in depth-of-field is incremental. If you had to look at a photograph showing amazingly thin depth-of-field, you wouldn’t be able to tell just from the photograph alone, whether f2 was used or f2.8 … but a side-by-side comparison will reveal the answer. An individual image won’t. However, it will be obvious when an image was shot at f 8 or f11 (on a 200mm lens for example), compared to either f2 or f2.8 … we should be able to recognize that a shallow depth-of-field was not used on the f8 or f11 image.

There are hordes of examples of how depth-of-field changes with aperture. But I thought it might be interesting to see with an overlay of images, just how much the depth-of-field appears to change …

Jessica and I braved the cold today to do a simple sequence of images with the zoom set to 200mm, and used at a variety of apertures. I used a tripod to keep the framing exact, and Jessica tried to keep as still as possible. The background is a clump of trees about 20 meters behind us. The lens used in this example is the Nikon 70-200mm f2.8 VR II (B&H).

DoF example

Ideally, there should be a bunch of these comparative sequences done at different focal lengths and different distances between subject and background.

But just this one example should clearly show how the depth-of-field changes … and that while the change between (for example), f2.8 and f4 is distinct, it might not be enough reason to warrant the purchase of a more expensive lens. But that is up to personal choice and budget.

28 Comments, Add Your Own

  1. 1 says

    Personal choice and budget. Or just what is made.
    I’m a devoted Nikon shooter. I love my Nikon. But Canon’s f/4 series of lenses would be very attractive for me. Nikon appears to only have 3 lines…. budget, enthusiast, and pro. Or “I like taking pictures” and “I pay my bills by taking pictures”. Not that I really object. I get it, and I lust after the 2.8’s. I would happily trade the reach of my 70-300 f/4.5-5.6 for the speed and clarity of the 70-200 f/2.8. But I (personally) would be VERY interested in an f/4 line. I wish they had a semi-pro series. Something to go along with the D7000/D300s bodies.
    To me, putting a $250 55-200 DX VR on my D7000 is underwhelming, but the 70-200 f/2.8 is twice the price of my body.
    If Nikon made a 70-200 f/4 in the $1,000 range (like Canon does) that was sharp from at f/4 all the way through, with reasonably quick AF motors, 77mm filter size and VR, I’d be all over it. It would bridge the gap nicely. Especially if they made a 14-24 f/4, and a 24-70 f/4. I’d have the whole series. Whether that would take business away from the f/2.8, I don’t know. Maybe it would, and that’s why Nikon doesn’t make one. They seem to have a $500 series, and a $1500 series, just nothing in the middle.
    Just please, Dear God, don’t make them in white.

  2. 2 says

    My main reason to choose f2.8 over f4 is autofocus actually. When you deny your AF a whole stop of light, its performance goes down accordingly. My 17-40 f4 L is essentially useless at wedding receptions once the lights are dimmed, the dance floor opens and madness ensues. Anything over f2.8 essentially ceases to focus properly and accurately even with the AF assist on in these conditions.

    As usual, just my two cents.



  3. 4 says

    Thank you Neil. This is one of those experiments I’ve been meaning to do…

    The topic is shallow DOF and bokeh but there’s the flip side of the coin – the lens performs best around f4 to f8 depending on the lens (although I can’t see any difference at this resolution on this great lens). Based on that, WOW bokeh at f2.8 but WOW detail and OK bokeh at f-say-6.3? Depends what you want on the day.

    I’m off to do my own series.

  4. 5Jan says

    Bogdan gives a good point. Many people are not aware, but on Canon DSLRs, apart from the central point, all the others won’t work as cross points with anything slower than f2.8. The Central one works to f5.6. I’m not sure what the non cross ones do.

  5. 6 says

    Your Assistant-with-an-attitude appears to get progressively more irritated as you increase from f2.8 to f22. Is this a side-effect of the aperture or could it be the ambient temperature? :D

  6. 7Roel says

    I’ve been shooting with a D300 for the past year or so. In the beginning coupled with the 18-200 f/3.5-5.6.

    Then I shifted to a tamron 28-75/2.8… and the difference is night and day. In accordance to Bogdan’s post, I got more in-focus shots at +-70mm than with the 18-200.

    Main reason: at f/2.8, I’m taking in 4 times (!) as much light than at f/5.6 at the same focal lenght with the 18-200!!! I’m not talking about the shot aperture, but wide open during focussing.

    Shot my 18-200 again over the weekend, and I was frustrated with the slow focussing – even with its SWM which the Tamron doesn’t have.

    Don’t get me wrong: the 18-200 is a great multi-purpose lens but most of the time it needs a lot of light for proper focussing (say 1/1000 @ f8 @ ISO200 – that much light). Then it gives excellent pictures…

    For anything else, I’ll stick with my Tammy.

    Best regs,


  7. 8 says

    Thank you for the comparison, Neil. The only other thing I think should have been mentioned is the focusing differences between the two lenses. The 2.8 allows more light into the camera body and thus will focus better (faster) especially in dark venues.

  8. 9 says

    Correct me if I’m wrong, but another factor in the “amount” of bokeh in a photo is the distance between the subject and the camera. As the lens focuses closer and closer the disparity between the correct focus for subject and background becomes greater in terms of the proportion.

    Nice comparison. I agree with the above commenters that f/2.8 makes a big difference in focusing speed. As I see it, that’s one of the main benefits second only to low-light performance. To my eye, the difference between f/2.8 and f/4 in the comparison above is negligible from a bokeh perspective.

  9. 13Alex-D says

    Thanks for replying to my question in such an elegant and “usefull to all” way
    To top this off we benefit of a photo of Jessica your “assistant with an attitude” – and thats great ..the french would say ‘C’est la cerise sur le gateau’.
    The slideshow examples from F2.8 to F22 really show how the image changes according to the aperture used ..evidently the backgroud is much more pleasing at f2.8 than f4 even though the difference is slight .

  10. 14Bob says

    I would also be curious to see a comparison from FF to DX sensor. What would the differences be if you used the D3 at 105mm and a D300 at 70mm at 2.8.

  11. 16Aniversari says

    Neil, your idea using gif is excellent! And the comparison from FF to DX sensor is really a good idea. I would be interested also in how sharp is DX compared with FF, at 2.8.

  12. 18 says

    I would like the FX vs. DX comparison to see what the “crop factor” really is like. In other words, take a D300 and DX lens and a D3 and a FX lens and set them to the same focal length. All the examples of crop factor that I’ve seen simply take a single image and draw the FX and DX boundaries on the same image. It would be nice to see either a side by side image comparison or an animate GIF.

  13. 20 says

    I love my collection of 2.8 lenses (16-35, 24-70, 70-200) and have to remind myself, at times, to take off my 24-105 f/4 at weddings once the light has dimmed. Let’s all imagine when Canon finally comes out with a 2.8 version of THAT lens! If they do, it would rarely leave my camera during weddings (provided it was still an L lens). I like the article but I didn’t see any comparison shots; am I missing something? All I see is the first image in B&W, with Neil’s watermark in the lower left corner, and the 2nd image, in color, with “2.8” superimposed over it. I didn’t see any other images (for example, the same image as posted but shot at f/4 or f/5.6, etc).

  14. 21 says

    You have to give the 2nd image time to load. It’s currently a 5.5 Mb animated GIF. Someone emailed me details about a plug-in that will give the same animation but without the overhead. I’ll update it next week since I am a bit crunched for time right now to delve into it.

    But give the graphic some time to load first.

    Neil vN

  15. 22george says

    It is also worth mentioning that the apparent depth of field changes depending on distance to subject due to the magnification effect. F2.8 on a close up of the face will be very shallow (ears will be out of focus for exmaple) but if you are further away the same f2.8 will focus the whole face. It’s like moving your hand close to your face, the magnification results in less of it remaining in focus.
    On a wide angle lens it’s difficult to achieve good depth of field for the same reason, unless you are close to the subject.

  16. 23David van der Merwe says

    Nikon has another current production offering in this range. It is the 80-200mm f2.8. It sells for just over $1,000. As an older design it lacks the latest VR but is still an excellent lens. Does anyone have any experience with both?

  17. 24 says

    Optically it will be superb, although I do think the latest design by Nikon, (the VR II), has the edge on all other versions of the telephoto f2.8 zooms. But I don’t think you’d be disappointed.

    However, I would miss the VR feature. Look at some of the images in the article on using flash during the wedding ceremony in the church, and you’ll see some examples of photos taken at very slow shutter speeds, handheld. For my work, the VR feature is essential.

    But still, I think you’d love the older 80-200mm f2.8 as an entry into wide-aperture telephoto zooms.

    Neil vN

  18. 25Chris says

    Great visualization! Thanks for that. I googled for “difference between 2.8 and 4 aperture at 200mm” ;-) and your website showed up on the first search page.

    I am thinking of selling my 200mm 2.8L II and buy a 70-300mm 4-5.6L. So I was wondering if I would “lose” something regarding IQ with the new lens…especially if the depth of field loss is negligible or not. So I guess it is negligible, or at least I can’t imagine a situation where f/2.8 would make a bigger difference?


  19. 26steven says

    Chris, only if you want or need better low-light focusing and performance, better bokek. If you need and want more reach in good light then go ahead.

  20. 27Trev says


    As Stephen said but don’t forget that the lens you are thinking of is f4 TO f5.6 when zoomed, so you are no longer getting f4 to compare to f2.8, but f5.6 and therefore it certainly would be noticeable.

    However, if you are wanting/needing the extra reach of 100mm that comes with it, no problem. Just be aware there will be a noticeable difference and not something no longer negligible of f2.8 to f4 in your eyes.

    Regarding image quality, I cannot comment, don’t own the 70-300, I should imagine there may be some sort of trade-off, but depends on the photographer, post-production, etc. see if you can get some comparison images.


  21. 28jayb says

    Great article. thank you so much for the experiment. would love to see the same with using standard distance from the main subject (15 feet) and vary the distance for the background objects. But that would be a lot of work. anyways, thanks so much, it was just what i was looking for. You answered my questions.

Leave a Reply

Your email address will not be published. Required fields are marked *