For years I’ve enjoyed visiting the Griffith Observatory. It’s one of the few places in Los Angeles with absolutely amazing views of the entire city. Beginning several years back, I shot this same scene, first with a Hasselblad H2 and P30 back coupled to a 50-110 zoom lens. The combined weight was about 10 pounds and needless to say, I didn’t carry that camera into the field much, even with a good tripod. The second go around was with a Canon 5D Mark II and a 50mm F1.4 prime lens, not an L series lens but one of Canon’s better primes; not nearly as heavy but still best used with a tripod. The most recent trip I shot this with a Pentax K5 IIs and a 17-70 F4 zoom lens. This new Pentax camera is so light and compact for a DSLR I’m finding that I hand-hold it more than any other camera I’ve owned; aside from a point-and-shoot. The weight is impressive! All three images were shot using a Gitzo Tripod.
When comparing all three images at full resolution, I was amazed. The difference in sensor size and pixels is as follows:
– Hasselblad H2 with P30 back – 30 megapixels in a 6 x 4.5 medium format CCD sensor.
– Canon 5DMK2, 21 megapixels in a full-frame 35mm CMOS sensor.
– Pentax K5 IIs, 16.1 megapixels in an APS-C (1.5 crop) CMOS sensor.
While the Pentax is half the megapixels and virtually half the sensor size as the Hasselblad, the image quality is substantially better. Both of these images were shot with zooms, though a Hasselblad 50-110 zoom can set you back about $4000 compared to Pentax’s 17-70 zoom at around $500 (quite a huge difference). Camera costs are even more skewed with a new Hasselblad being $$,$$$ while the Pentax K5 IIs was around $1300. One is considered Pro grade, the other Prosumer. So why pay more for the H2 over the K5 IIs… noise and file size. That’s sadly about it! Years ago I would have never attempted to compare an APS-C camera to a MF digital camera like Hasselblad but Pentax has seriously hit a home-run with this one.
And what about the image shot on the 5D2 with a good prime lens… actually, the two images when compared at full resolution are very close, while I favor the dynamic range and noise to have faired better with the Pentax. That’s full-frame with a prime being challenged by APS-C with a zoom. Normally that math shouldn’t compute but in this case it does.
The Three images all have noise and each sensor was straining to produce clarity with objects set at near infinity like the buildings in the background. This isn’t unusual however since the time of day and loss of natural light makes any camera strain to bring the image into good focus.
When looking at the detail in the concrete of the Observatory building itself though, the Hasselblad P30 image is blurry. The H series zoom lens has always been said to be soft and it’s possible being an evening shot that it could have been back-focused slightly however I doubt it. I shot the same lens for several years and never had trouble focusing and typically did so manually as it was a heavy and tight turning lens. The Canon F1.4 prime lens was easily set manually and the Pentax 17-70 zoom lens was tested using autofocus first, then checked manually and then tested again using auto-focus.
Some people that are reading this will conclude the test is not fair as two zooms cannot equally be tested against a prime and vice-versa and I would submit that you would have a point… however, as I had what I had at the time each shot was taken, this is the closest comparison I could achieve with each of the three systems, shot at nearly the same time and as close to the same composition and perspective as possible.
While the Canon produces a good image, the Anti-Aliasing filter used by Canon, IMHO is simply too strong. I’ve shot thousands of images and rarely have I had problems with moire. The real problem I had with Canon has been banding! Their settings and/or software design as integrated with the CMOS sensor creates a substantial amount of unwanted noise when shooting at sunset or low-light and in many cases that I noticed, causes banding (faint vertical lines showing up in your final composition). I had heard many other Canon owner’s complain about the issue so I’m not aware if Canon corrected this problem with the 5DMK3, I can only speak of the system I owned.
Yes, I know it has been several years since Canon introduced the 5DMK2 and Pentax has clearly had time to tweak their software to get the most out of the new CMOS sensors and yes the lack of the anti-aliasing filter does give the K5 IIs a slight advantage however… the Hasselblad’s Phase One P30 digital back has no AA filter. They don’t use them in most MF systems.
So what’s the moral here… well, I’m sorry to say… there isn’t one! All three systems are good systems in the right hands with the right composition. If you have money and are willing to buy a new Hasselblad H4 with P60 back for $32,000 then I have no doubt the image quality would be hard to rival by most any DSLR. However, if you are budget conscious and are trying to decide whether to buy an older Canon FF camera vs a new APS-C, you may want to look at the above files at full-size and also consider that DXO labs tested the Pentax K5 II and ranked it’s overall image quality better than the 5DMK2! What about Nikon? Well, I haven’t gotten my hands on the D800E as of yet but I have seen the test results and it looks very promising. Who knows, I may put it to the test in the very near future. For now, when I’m not shooting film on one of my medium format cameras I’ll be keeping the K5 IIs in my bag wherever I go.
This exercise was meant to entertain, complain, enlighten and bewilder. Hopefully it has accomplished all of that ;-).