Pentax vs Canon AF Performance
I’m a little concerned about the focus performance of my Pentax K10D DSLR. Before I get too far into it, let me establish a little background first…
Two years ago I was looking for an upgrade to my then 2 year old Canon Digital Rebel XT (350D). I really didn’t have much more than the kit lens and a cheap-o EF 50mm f/1.8 II, so I didn’t feel too married to Canon. I did my research and decided the then new Pentax K10D represented a better value for the $1,800 I had to spend than the more expensive Canon 30D. So, I gave my Rebel to my sister-in-law and ordered up a K10D body, Sigma 18-50mm f/2.8 EX and Sigma 10-20mm EX from B&H Photo.
The ergonomics, features, IQ, and interface were all so much better than my old Rebel that there wasn’t much time for retrospection. I embraced my new rig and started shooting like mad. Six months and 20K snaps later I had added the excellent DA* 50-135mm f/2.8 to my kit. To my surprise, Santa (in the form of my mom and step dad) gave me a $1,000 Wolf Camera gift certificate for Christmas 2007. Suddenly I found myself in a quandary.
You see, I had become somewhat smitten by the full frame Canon 5D we had at the office for the Marketing team I managed. It had always seemed so out of reach to me and now suddenly it was all-too-close. After a few days of agonizing over what to do, I ended up selling off my K10D kit to a co-worker friend and added the proceeds to my gift certificate. 2008 began with a brand new Canon 5D body (the $3,000 body was all I could afford). Fortunately, my old 50 1.8 was collecting dust in the bottom of a drawer so I was able to actually use the 5D while I saved up for some better glass.
It was hard to draw many comparisons between the 5D and K10D. They really are quite different cameras- with the 5D being much more like a 35mm film camera than a high-tech digital. Plus, the clunky old 50mm prime I started out with on the 5D really wasn’t anything like the Sigma EX zoom lenses I had had on my K10D. What I didn’t know was that I would decide to leave my six figure job in April to return to school and finish my undergraduate degree. This move required scaling back across the board and my beloved 5D (and all the L glass I hoped to own) suddenly seemed like an excessive luxury. I decided to sell it, my EF 50 f/1.8 II, and the EF 28 f/2.8 and EF 85 1.8 USM I had acquired on Craigslist and invest the proceeds into my college fund. For the first time in 7 years I was without a DSLR.
In September of last year my former co-worker felt pity for my situation and gave me back my old K10D with the kit lens off his new K20D! Since then I have been shooting like mad with the kit lens and a 20-year-old SMC-A 50mm f/1.7 manual focus lens I picked up off of eBay. In the past couple of months that former co-worker of mine has also shared his DA* 16-50 f/2.8 and DA 35mm f/2.8 Macro Limited lenses with me to test on my blog. I wasn’t too impressed by the 16-50 (see review) and my experience thus far with the 35mm macro has been mixed (review here).
While I was at PMA a few weeks ago I played around a bit with all the various cameras on display. Both Pentax and Canon had similar demo set-ups for their top-of-the-range telephoto lenses. While using the DA* 300mm f/4 in the bright tradeshow lighting I tried focusing on a subject standing a few meters away and the SDM motor woooshed close then dunk, da-dunk, “beep-beeped” into focus lock. I then swung the lens up and selected a far away subject- woooosh, dunk, dunk, da-dunk, “beep-beep”. Selecting a subject a few feet away from the second the DA* 300 went dunk, dunk, da-dunk, “beep-beep”. This all seemed quite normal to me. From my experience shooting both the DA* 50-135 f/2.8 and DA* 16-50 f/2.8 SDM equipped lenses.
Then, I moseyed over to the Canon booth where they had a very similar display set up with a 1D Mark III and EF 400mm f/2.8 L IS USM lens. Repeating the short-to-long focus test (this time with a much heavier f/2.8 lens) the Canon went woosh, “beep-beep” to the close subject then woosh, “beep-beep” to the far subject. Selecting a nearby subject to the far one I got bump, “beep-beep”. Surprised I tried several more times, panning the camera around and selecting subjects that were just out of focus. Each time I simply got bump, “beep-beep” as the subject popped into near instant focus. At the time I merely chalked this up to the near 10X cost difference between the two set-ups and continued on my way.
Fast forward a few weeks to yesterday and my niece’s 3rd birthday party. I found myself two-fisting my K10D with 35mm f/2.8 Macro Limited and my sister-in-law’s Canon Rebel XT with kit lens (my old camera) while capturing both indoor and outdoor birthday party action. I was immediately aware at how much faster the 4 year old Rebel was locking focus than my K10D was. It was almost exactly like the performance difference I had experienced at PMA.
While the kids were busy taking turns in the kayak (and out of decent photo range for any of my lenses) I took a moment to do a little non-scientific focus testing. I manually set both cameras to infinity focus, selected the center focus point, 35mm focal length (on the Canon) and (in bright sunlight) focused on my hand at around 18″ away, then on the far side of the lake (infinity), then back to a post about 6′ away with the 35mm Macro on the K10D and the 18-55mm kit lens on the Rebel XT. In all cases the Canon nailed the focus point with zero hunting… zip, “beep-beep”, ziiip, “beep-beep”, ziip, “beep-beep”. The Pentax however, struggled a bit. The focus test went weeeee, tick, tick, tick-tick, “beep-beep” for the close subject, then weeeeee, tick, tick-tick, “beep-beep” for the far subject. Then, the camera completely missed the post 6′ away and went weeeeeeeeeeee all the way to it’s closest focus distance than weeeeeeeee, tick, tick, tick-tick, “beep-beep” on the post. (To be fair, this is a macro lens).
Thinking that the Pentax 35mm Macro might have been to blame (which it was), I pulled out my Pentax 18-55 Mark I kit lens and repeated this test again. The results were nearly identical with the exception that the Pentax kit lens did not overshoot the post and merely ticked around a couple of times before locking focus. Also, the lighter less damped focus action on the kit lens did make it seem a tad snappier when focusing between the close subject and the far subject. Incidentally, this time to reach the correct focus distance was basically identical for both the Pentax and the Canon kit lenses. The exception being all the tick, tick, ticking the Pentax lenses did before locking focus. (In case you were wondering I also repeated a similar test inside under poorer lighting with the two kit lenses and got nearly identical results as outside).
Discussing this with my co-worker friend, he relayed the results of a similar test he decided to try with his K20D as follows:
I just did back to back comparisons of the K20d + FA 50/1.4 and my Rebel + EFII 50/1.8. In indirect sunlight, using the center focus point I swapped focus between the front hedge at about 10M away and my left hand outstretched.
The Canon will ziiiiip, stop, confirmation beep.
The Pentax ziiiiips from 10M to my hand quickly just like the Canon but then it quickly zips one (or two!) more times before focus confirmation… Almost like it has momentum from the first movement.
I’ve been fully aware of this (and failure to focus–maddening) but had assumed that after twenty years of screw-driven, TTL autofocus technology this was the best the world could do. Seems it’s the best Pentax can do.
IMHO this is the worst aspect of my K20d.
Keep in mind that his test was done with a state-of-the-art $1,000 K20D + FA 50mm f/1.4 vs a $300 6 year old Canon 300D + 50mm f/1.8 II eBay find.
In hindsight I realize now how many shots I have missed over the years with my Pentax due to its slow autofocus performance. The most frustrating part is that the vast majority of shots I’m missing are while the camera is tick, tick, ticking its way towards focus lock, not because it has mis-focused. Once it gets there it is generally just as accurate as any other camera I have used. I hope Pentax is listening and improves the performance of their next generation of DSLRs. In my opinion, this is much more important than live view, more megapixels, movie recording, higher frame rates, and/or higher ISO performance. Pentax’s DSLRs are some of the best values on the market. Bring the AF speed up to par with the competition would end a lot of arguments in Pentax’s favor.
What do you think?
PS- I excluded Nikon from this comparison because I have very limited experience with their modern DSLRs. However, I did borrow a friend’s D300 + 50mm f/1.8 to take some studio shots of my K10D for my DA* 16-50 f/2.8 review. The camera was set up to use all 50-something focus points in AI mode. Being unfamiliar with Nikon’s focus confirmation implementation I didn’t realize until after I was done shooting that when all those grey boxes popped up in the viewfinder that the camera had locked focus. In hindsight I now realize that it was happening so fast that I was unaware that the camera was focusing. To get my shots BTW, I ended up switching over to manual focus ;-)
UPDATE: First of all I want to make it clear that I am not suggesting that Pentax’s autofocus system is any less accurate than anyone else’s. It is clearly quite effective. Also, I fully understand that Pentax has to pick and choose where they spend their limited development funds and I am happy they put it where they have (weather sealing, in-body shake reduction, user oriented controls, etc). In fact, that is why I chose Pentax over the others in the first place. My main reason for publishing this article was to attempt to draw the same comparisons so many Pentax users have over the past few years in a more objective manner so that those who haven’t experienced the differences themselves could be more fully informed. Too many consumers buy products today based on too narrow a point of view and then blame the product when it doesn’t meet their expectations.