The sensor or the lens? and does it matter?

Canon 550D Nikon 50mm f1.2 MF lens


In a recent poll in a photographic magazine photographers voted by a large majority that the most important element that determined the quality of their images was the lens. Its an often repeated argument, implying that there are significant differences between the optics we use on our cameras. Indeed lens prices seem to reflect this. Standard zooms, 28-75mm f3.5-5.6 on full-frame, 18-55mm f3.5-5.6 on APS-C, 14-45mm f3.5-5.6 on m4/3, if bought with a kit can cost as low as £30. Go to the top of the range standard zooms and prices rocket upwards. Examples such as the Nikon 24-70mm f2.8 costs £1180, the Canon 17-55mm f2.8 costs £700. This is a substantial difference. 


In reality I've always found that the sensor is the most important element in determining the quality of the image. Just as the film was. Also from my experience, the better the sensor, the less the quality of the lens matters. This also seems to work in reverse. When I've used not so good sensors only top quality lenses have produced decent results.


When I was using a Nikon DX3, I tried it with my manual focus Nikon primes and also with a 28-200mm G f3.5-5.6 zoom. The zoom produced excellent results and though it was possible to see a difference using the primes, this difference was much less than you might imagine it to be.


nikon D3X


The sensors in both the Leica M8 and M9 also seem to produce great results no matter what lens I use. Though again I can see a difference if I look hard enough, it is very small. 


Leica M9 Zeiss 50mm f2 Planar


I even tried some non m-mount lenses on my M9 and it produced great results. Even from the Samyang!!


Leica M9 Nikon 50mm f1.2 MF


Leica M9 Samyang 14mm f2.8


On some of the cameras I have used where the sensor is not so good, I've found I need a really good lens to get acceptable results. On the Pentax K-7 I had, the results from the kit zooms were pretty soft.


Pentax K-7


Whereas the primes I used were better.


Pentax k-7 35mm f/2.8 lens


Again with the Canon DSLR's that I've used, I've always used the best lenses that I can and this has usually meant prime lenses, as I've found the zooms, with the exception of the excellent 24-105mm f/4, disappointing. I even took to using adapted Nikon primes on my Canons as I found they gave slightly better results.


canon 5DMkII 85mm f1.8 lens


Yesterday after writing about my Olympus E-P2 I tried it again with some different lenses. One of those I used was the Panasonic 14-42mm kit lens from the Panasonic GH2. When I looked at the results I thought I had done something wrong as the images were so soft. I had got some reasonable results on the GH2 with this lens at f8 & f11 and using a tripod. However the E-P2 results were pretty terrible. I took the pictures again & took some on the GH2 just to see if I had made some error, but the results were the same. 


In some previous posts I have mentioned that I have always been impressed with the Olympus 14-42mm kit zoom. However a substantial part of this opinion is formed by the fact that I have used it on Panasonic cameras! On the E-P2 for example its quite disappointing.


Olympus E-P2 14-42 Kit Lens Zoom


I will admit that much of this is me being hyper-critical and very picky. I am a confirmed "pixel-peeper" and will probably always be so. An example of this is I have recently been considering buying a Pentax K-5 and have been looking at my K-7 images again. I have always been quite disparaging about that camera. However looking at the images again I didn't think they looked that bad. At the time I was using it I had just bought a Leica M8 and was probably comparing it to that.


Another example was over Christmas I saw a print of a portrait I had made for a family member on their wall. I had taken the picture with my Olympus E-P2 and the 14-42mm kit lens. The print looked terrific. Beautifully sharp and detailed.


The differences that we see on our screens are indeed real and some lenses are better than others as indeed are sensors. I've only really had two cameras which I could say never produced a decent picture. One was a Canon 350D and the other was a Fuji S5 Pro. I've often wondered if they were faulty, or three anti-aliasing filters had been fitted by mistake! Both produced incredibly soft images. The rest of the cameras I've used, while certainly having differences, have all been useable. Maybe its the 27" iMac screen I stare at all day, that exaggerates these differences. The printed image, either self printed or in magazines or books certainly levels out the results. I see no problem in trying to get the best results that I can, and while the differences between sensor / lens combinations may be smaller than I sometimes imply, they do matter to me. However that may not be the case for everyone. Many wedding photographers use Fuji S5's and I can see that its combination of low contrast and soft images could be very useful in that context. For the majority of my own wedding photography I used Canon jpgs. These produced less noise and a softer result than shooting raw. I've done lots of portrait work in the past and razor-sharp is definitely not required here. If you are shooting portraits for clients, I doubt there are many of them that will thank you for producing high resolution ultra-defined images of all their blemishes! Seeing your headshot on a huge monitor screen can be very disconcerting for those of us who are not perfect. (i.e. everybody!)


So while I shall continue in my search for the ultimate camera / lens combination, its undertaken with an understanding that this is about the margins. This is about squeezing that extra bit of sharpness, getting 95% instead of 90%, getting that fractional improvement. My best-selling picture was taken on a Pentax K-10 with a 12-24mm zoom. Its of a bathroom! That probably puts a lot of this into perspective. Though we may fret and worry over our gear, ultimately its still the picture that counts. Now if only I could convince myself of that!


Words - David
Images - David and Ann