 | Zooms vs. Primes |
| Zooms vs. Primes |
Zooms vs. Primes | ||
| Sure, if you want the best images you use primes. Er..... | ||
| Bob Wallace (BobTrips) Post Count: Joined: | Msg: #1 Added: 2007-06-05 19:50:50 | |
| Perhaps you don't get much improvement with primes.... Here's an interesting article from Popular Photography that sheds some new (to me) light on the topic. "Frankly, we expected that primes would produce visibly contrastier and sharper pictures—if only because they're generally simpler in design, with fewer elements and internal glass surfaces to magnify flare. In fact, as you can see from the charts here, we found scant difference in image quality between the eight tested primes and zooms. Subjective Quality Factor (SQF) tests, which evaluate both resolution and contrast, produced data so close that the differences aren't visible to the eye, especially at common print sizes (i.e., to 11x14). Prime lenses do consistently outscore zooms…but not by much." Is PP blowing smoke? | ||
| Adrian Warren (AdrianW) Post Count: Joined: | Msg: #2 Added: 2007-06-05 23:15:29 | |
| It depends. If you choose the right lenses, and the right parameters, you can prove almost anything ;) You should be familiar with the old line: "there's lies, damn lies, and statistics" I suspect with the same design age, and same manufacturing contraints it should be true that primes are still sharpest. Of my current lenses (all Canon): 100mm f2.8 USM Macro (razor sharp, very sharp wide open) 200mm f2.8 L/II USM (very sharp, but not quite in the same league) 24-105 f4/L IS USM (very sharp, but CA strong at the edges - craw fixes this nicely) 17-40 f4/L USM (significantly less sharp, not bad, but...) Of the old 'uns 50mm Macro (sharp) 200mm Macro (sharp) ^^ they beat the 17-40 in terms of overall sharpness, the following are softer: Sigma 24mm MF f2.8 (sharpish) 35-70 f3.5-4.5 FD (sharpish) Tokina SZX210 70-210 (very soft) Sigma 600mm f/8 mirror (very soft) Sigma 17-35 EX HSM (hopelessly soft) So, in overall subjective sharpness terms P,P,Z,P,P,Z,P,Z,Z,P,Z | ||
| Bob Wallace (BobTrips) Post Count: Joined: | Msg: #3 Added: 2007-06-06 03:35:10 | |
| I'm very familiar with the old line: "there's lies, damn lies, and statistics". One of my undergrad majors was statistics and besides being a published researcher I taught research methodology at the university level. I'm also quite familiar with how people often react when data gores one of their sacred oxes. ;o) I couldn't find where PP had put any of your lenses on their optical bench so I picked a couple of others that they had tested. Both Canon. Both EF-S, which suggests that they were designed/manufacturer in the same time frame. Both tested within a short period. Canon 17-85mm f4-5.6 EF-S USM IS Scores at 50mm Aperture 5x7 8x10 11x14 16x20 20x24 5.6 96.5 95.3 92.9 88.9 84.7 8.0 96.3 95.1 92.5 88.2 83.5 11.0 96.3 95.0 92.4 88.1 83.5 16.0 96.0 94.8 92.0 87.5 82.5 22.0 95.1 93.5 90.0 83.8 76.6 32.0 94.3 92.5 88.4 80.2 70.1 Canon 60mm f2.8 EF-S USM Macro Aperture 5x7 8x10 11x14 16x20 20x24 5.6 96.7 95.6 93.2 88.6 83.2 8.0 96.8 95.8 93.5 89.3 84.2 11.0 96.8 95.8 93.5 89.3 84.3 16.0 96.5 95.4 92.9 88.2 82.6 22.0 95.7 94.4 91.2 85.1 77.6 32.0 94.7 93.2 89.2 80.8 70.2 (I left off the 2.8 and 4.0 data for the 60 as it wasn't given/possible for the 17-85.) I did one more thing. I determined the difference in scores for each lens at each data point. 17-85 minus 60 Aperture 5x7 8x10 11x14 16x20 20x24 5.6 -0.2 -0.3 -0.3 0.3 1.5 8.0 -0.5 -0.7 -1.0 -1.1 -0.7 11.0 -0.5 -0.8 -1.1 -1.2 -0.8 16.0 -0.5 -0.6 -0.9 -0.7 -0.1 22.0 -0.6 -0.9 -1.2 -1.3 -1.0 32.0 -0.4 -0.7 -0.8 -0.6 -0.1 I also calculated an average for each column. Not sure that it's a valid statistic as I don't know if PP is using a ratio scale. But it should be slightly informative. Average -0.5 -0.7 -0.9 -0.8 -0.2 The 60mm prime tests better with most settings but the difference is, in general, less than one percent. (Damn!!! I spent a bunch of time lining up those columns. WYSI*NOT*WYG.) | ||
| Justin Watson (Justin) Post Count: Joined: | Msg: #4 Added: 2007-06-06 14:23:39 | |
| I'm a fan of using the best lens for the job. Now the best lens is selected from the lenses you have of course :-) I've compared a few of my Pentax lenses. My FA*200 isn't noticeably sharper than my DA50-200 zoom at 200 especially around f5.6 to f8. However the images out of my FA*200 are far better rendered and much nicer bokeh. Of course the * lens line is like the Canon L series lenses. So you'd hope its better in some way :-) I also quite like my 100mm macro and 50mm 1.4, however i have the range covered by the zoom.... I find if i'm shooting around the same focal length i'll use a prime. Things like the car rally i went to, a prime was ideal. I was in the same spot and the cars went the same way every time. | ||
| Bob Wallace (BobTrips) Post Count: Joined: | Msg: #5 Added: 2007-06-06 14:40:59 | |
| How about posting samples from the 50-200 @200 and the 200? Same tripod/aperture/etc. sorts of shots. | ||
| Adrian Warren (AdrianW) Post Count: Joined: | Msg: #6 Added: 2007-06-06 15:13:06 | |
| LOL, you would choose the recent prime with the lowest image quality wouldn't you ;) I'm aware that you have a stats background - that was why I made the comment. As for the sacred ox - to be honest I'm not really worried whether or not primes are better. I have both primes and zooms, and I'm sufficiently happy with at least one of the zooms to use it day-to-day. The 17-40 doesn't leave me filled with glee in IQ terms though. We've already ascertained that I'm a bit fussy though. OK, how's this for interesting - lowest rated Canon prime, vs lowest rated Canon zoom. Highest rated C zoom, vs highest rated C prime. Using Photodo the suspects are Canon EF 28-80mm f/3.5-5.6 IV USM: 2.2/5 MTF Canon TS E24mm f/3.5L: 3.3/5 MTF Canon EF 50mm f/1.4 USM: 4.4/5 MTF Canon EF 80-200mm f/2.8L: 4.2/5 MTF Primes win both MTF rounds. That said, there are some notable missing lenses in that line up - I'd expect the 70-200 Ls to be sharper than the 80-200, and I'd expect the 100 USM Macro to be sharper than a 50, but... Also comparing one of the cheapest zooms with a Tilt-shift lens isn't entirely fair either, but it was the lowest rated prime. I would use PopPhoto, but I can't see a way to list reviews by IQ? | ||
| Bob Wallace (BobTrips) Post Count: Joined: | Msg: #7 Added: 2007-06-06 15:56:45 | |
| The point is that primes don't necessarily win by much any longer. Not like back in the days when this bias was so firmly established in the "consciousness" of photography. Take a look at these two lenses... Canon EF 50mm f/1.4 USM: 4.4/5 MTF Canon EF 80-200mm f/2.8L: 4.2/5 MTF The difference between 4.4 and 4.2 on a scale of 5 is only 4%. (Assuming a ratio scale.) And remember that you are most likely going to have to print at 16" x 20" and look at the prints side by side to see that level difference. | ||
| Adrian Warren (AdrianW) Post Count: Joined: | Msg: #8 Added: 2007-06-06 23:05:29 | |
| Really does depend on how linear that scale is I guess. Very hard to say overall, not least because none of the lenses they have MTF data for are lenses I own ;) Also, along those lines, there are a heck of a lot of missing lenses there. Neither of my primes are listed, and one of those is (I feel certain) sharper than almost anything else Canon has to offer. | ||
| Justin Watson (Justin) Post Count: Joined: | Msg: #9 Added: 2007-06-07 13:02:28 | |
| Bob i'll see what i can do. I probably need to take a variety of shots at different apertures, certainly a little more scientific than my visual quick comparisons. Certainly the difference is not as noticeable when the subject is not close and the aperture is stopped down a little. | ||
| Number Users: 3 Number of Posts: 9 | To Reply Join Jeber Photo Clubs | |
