|
mclifford82 posted:Maybe I'm alone, but my first thought when considering a 2.8 vs 4.0 aperture lens is the depth of field difference as well as the 2.8 likely being sharper at f4 than the f4 is wide open. I know that's not always the case, some lenses are sharp as hell wide open, but generally speaking you want to stop down a peg or two to increase sharpness, especially in the corners. As you and others point out, with ISO performance being what it is today, it's not really something I think about when deciding between a 2.8 and 4. I generally think all of the 24-whatever L-series lenses are pretty decent even wide open. Here's Photozone's tests on the 24-70 at f/4 and the 24-105 wide open: At 24mm: 24-70: 3240 / 2512 24-105: 3216 / 2363 At 40mm: 24-70: 3326 / 2776 24-105: 3079 / 2494 At 70mm: 24-70: 3001 / 2566 24-105: 3363 / 2022 At 105mm: 24-105: 3112 / 2515 So at the wide end it's pretty much equal, the 24-70 is a touch sharper in the corners. The 24-105 performs a little worse in the midrange, the 24-70 is a little sharper overall at 40mm (but both are pretty sharp, 3000/2500 is great). At 70mm, the corners are quite a bit softer, but the center is a fair amount sharper than the 24-70. If you put it all the way to its long end it's equal-to-better than the 24-70 was at 70mm. Overall I think they're reasonably equal at f/4 except the 24-105 does a faceplant in corner sharpness at 70mm. They noted some alignment problems in the 24-105, which could be affecting that. He noted that it took him 4 tries to get a good ("good, not 'great'") sample of the 24-70, so it's not a unique problem. That also may be affecting the results a little, to be honest - taking the best of 4 lenses is cherrypicking your results a bit. Paul MaudDib fucked around with this message at 00:56 on Apr 18, 2014 |
#
¿
Apr 17, 2014 23:47
|
|
|
|
| # ¿ May 16, 2024 04:47 |
|
|
Even if I wanted a 24-70 (I prefer primes) I'd still suggest you at least get the old 24-70 mark 1 instead. It's a perfectly fine lens and it's like half the price. The new one is a little better, nowhere near twice as good. For the same money as the new one you could get a used Mark 1 AND the sigma 35/1.4.
|
|
#
¿
Apr 22, 2014 22:37
|
|
|
Drewski posted:I'm betting somewhere in between, and in the end it seems like apples and oranges to me. Each has its intended uses and I personally prefer the versatility of zoom. But the primes provide shallower DOF, more light gathering capability and a lighter setup that the 24-70 physically cannot offer. Sounds like you've generally got the idea. Up until recently zooms have usually been a fair bit softer, especially wide open. Zoom design took a big step forward in the 90s and the "pro" quality zooms (eg 24-70, 24-105) since then have usually been decent performers even wide open (consumer-grade zooms not so much, and even some of the pro zooms do have softer areas in their zoom range). Primes are usually somewhat sharper wide open (primes have gotten better over time too of course, apples-to-apples they're probably something like 10% sharper than a zoom of equivalent vintage), but the big benefit is super-fast apertures like f/1.4. There's no substitute for a fast aperture in low light - wider apertures help the AF system in dark conditions and the shallower depth-of-field will give better focus accuracy even in good conditions. I don't even know if it's really a weight savings - the Sigma 35/1.4 is only a half pound lighter than a 24-70 for instance. If you're carrying two big hefty lenses to replace that one zoom then you'll probably be heavier in the end. But you can at least get light primes (not f/1.4) that are worthwhile, whereas any zoom that's worth its salt is a solid hunk of glass. A 70-200 f4 is "light" for a zoom and that's still a pound and a half, and the heavy versions can be more than double the weight. To continue their recent trend of shaking up the high-end lens market, Sigma is reportedly working on a FF 24-70mm f/2. And if their crop 18-35 f/1.8 and the 35/1.4 and 50/1.4 are anything to judge by, it's gonna be quite a performer. In my opinion Sigma is doing everything right - they're focused on fast lenses that can deliver crazy resolution right from wide open. For a given sensor size, that's the only way to increase megapixel counts while avoiding the diffraction limit, which is starting to become a big issue with crop sensors (eg 24mp APS-C is diffraction-limited at f/5.6). In the last two years they've put out three best-in-class lenses at a fraction of the price of their competition. Canon is very much resting on their laurels in comparison, their lenses are getting one-upped by bit players like Samyang let alone by the stuff Sigma is putting out. Paul MaudDib fucked around with this message at 19:01 on Apr 23, 2014 |
|
#
¿
Apr 23, 2014 18:19
|
|
|
Seamonster posted:I want to know what kind of arcane magic Sigma is using to get such awesome wide open sharpness. Shitloads of aspheric elements and exotic glass, plus more elements gives more chance to correct abberations. Like the new 50/1.4 Art, 13 elements is not an unprecedentedly complex design, that's roughly what any fast wideangle lens has (35/1.4 or 24/1.4 or something). It's just that Canon is still using the same 7-element all-spherical Planar formula that everyone has used since the 60s and that was designed 120 years ago. There's only so much you can do to tweak a basic design like that, and Canon has not even done that much since 1993. I think your idea about the image circle is probably also not too far off. They probably are designing for a slightly larger image circle and then vignetting it down to 35mm, I think that's very likely with some of the primes. Zeiss bragged about the Otus being a "medium format lens" (hence the Distagon name on a normal lens) and I think the logic behind that may be a larger image circle that they sweet-spot on. Also one way to get extra lens speed is a telecompressor - you take a large image circle and squish it down into a smaller one. The same number of photons over less area makes it brighter and since you're taking a big image and squishing it down, it gains sharpness. It's the opposite of how a teleconverter costs speed and makes softness more apparent since you're magnifying the image and showing all the flaws. That's how Olympus gets their superfast zooms - they're stock FF Tamron designs with telecompressors in the back to crush the image down to 4/3 and boost the speed. I haven't seen a lens formula diagram, but I have a hunch that's what Sigma is doing with lenses like the 18-35 f/1.8 and the 24-70 f2. They design the lens for like a 0.7x crop factor for the respective format and then telecompress it to gain speed and sharpness. So half of it is Sigma being willing to think laterally, and the other half is Canon being unwilling to take risks when they can just spin out the 5th incremental update (1 2 3 4 5) of the 24-70L lens and rake in the dough. Actually the 24-105L should probably be in that list too. The EOS-M just perfectly encapsulates Canon's priorities to me - the principal design goal was not to compete with their existing gear in any fashion and risk cannibalizing sales, making it a good tool to take pictures with was not even on the list. Paul MaudDib fucked around with this message at 21:05 on Apr 23, 2014 |
|
#
¿
Apr 23, 2014 20:32
|
|
|
Combat Pretzel posted:I'm starting to get wary about that supposed 24-70mm/2.0. There have only been a few mentions about it long in the past, with nothing new since then. Just a couple of days ago I looked for it and came across some thread on a photography forum where they theorycrafted how much bigger and heavier such a lens would get. While I don't remember the numbers anymore (I think 6 lbs was mentioned), it wasn't pretty. Of course, I suppose that assumes regular lens design. If they're going with the aforementioned telecompressors, then who knows. Well, it won't be a lightweight for sure. The 18-35 f/1.8 is a brick too. Theorycrafting this out myself, my rule of thumb would be to take a medium format 6x4.5 lens and work from there. Relative to 645, full frame is a 1.6x crop factor, so roughly the same amount of speed gain as going from 35mm to APS-C (1.5 stops). For an f/2 lens at full frame, then, we're talking about a 645 lens that is roughly f/3.5. There are a couple 645 zoom lenses but none are quite that fast or cover that focal length. The Pentax-FA 645 45-85 f/4.5 weighs 805g, the Vario-Sonnar T 45-90 f/4.5 weighs 1200g, both of those lenses need to be roughly 2/3 of a stop faster and cover a bit more focal length (they're 30-55 ish). The Pentax-FA 645 55-110mm f/5.6 (500g) is just about the right focal length (35-70) but it needs to gain about 1 1/3 or 1 1/2 stops of speed. So scale that up by how much you think adding an extra stop of light and some zoom range would weigh for a medium format zoom, plus a MF->35mm telecompressor. All of those lenses are pretty old designs and probably could be done better (or lighter) by Sigma, plus the telecompressing will hide minor flaws near wide open (ignoring abberation from the telecompressor, you get 1.6x the resolution), but it'll still be fairly heavy. I think probably 1500g is the bottom end, maybe 2000g on the high end, but that's just a wild-rear end guess. So roughly equal or somewhat heavier than a high-end FF 70-200, and probably not much smaller. 6lbs seems a bit high but not out of the realm of possibility. You're right that they've been pretty quiet on it, though. It has been a while. Paul MaudDib fucked around with this message at 08:06 on Apr 24, 2014 |
|
#
¿
Apr 23, 2014 21:25
|
|
|
Seamonster posted:There are 2 other theories as well: they're actually designing the lenses for even wider apertures than what is being advertised but pre-emptively "stopping down" to avoid the common issues of actually being wide open. Oh and for the record this theory is very testable. Take a long-exposure shot with the lens wide open and look down the barrel. Do you see the iris stopping down? If not, if you disassemble the lens, is there a round waterhouse stop near the iris? A waterhouse stop is like a fixed, circular iris, and they could still be using one to manipulate the maximum allowed aperture of the lens rather than the blades of the iris, although you'd be trading focus accuracy for improved bokeh. I'm personally guessing "no" on that one (especially on primes), but the practice isn't unheard of. Olympus actually does that with their telecompressed zoom lenses. Paul MaudDib fucked around with this message at 23:43 on Apr 23, 2014 |
|
#
¿
Apr 23, 2014 22:56
|
|
|
Thumposaurus posted:I picked up a Canon FD 50mm 1:1.8 at a thrift store for $1. Canon FD stuff is cheap at this point because no one wants it. Adapting it to EF mount doesn't work (for the above reasons), so the only people who want it are people who shoot FD mount on film and people who adapt it to mirrorless. It's a fine lens, but nothing special. There's tons of them out there because they were the standard "kit" lens at the time. If you want to shoot a 50mm go buy a EF 50/1.8 for like $100
|
|
#
¿
Apr 30, 2014 21:26
|
|
|
Seamonster posted:Live view, yo. Live view really kinda sucks for focusing manual lenses, though. It works OK if you're shooting a stationary target off a tripod but it's hard to nail focus on a moving target and handheld there's a lot of shaking at high magnifications. Focus confirmation is a joke, it doesn't work worth a drat at high apertures. It's better than nothing but I didn't find it reliably accurate at all. I really recommend a focus aid of some kind. Either optical (replacement focus screen with a split prism) or digital (focus peaking or an XT-1 style split prism). The lack of user-changeable focus screens is a big reason to prefer a used prosumer body to a used consumer body. More than just the technical specs, there's lots of those little technical niceties that have been stripped out. Paul MaudDib fucked around with this message at 20:43 on May 1, 2014 |
|
#
¿
May 1, 2014 20:40
|
|
|
Tricerapowerbottom posted:I use a T2i and a 50mm 1.8 for most of my shots, with just the center AF point for confirmation. I haven't done the testing to see if I have more OOF photos at 1.8 versus 3.6 or something, but is there anything I can do with those two elements, plus a tripod, to improve focus? I seem to get it right somewhere between 10-20% of the time. Would just going with a smaller aperture improve my chances in situations like parties, etc? Yeah, this: Whirlwind Jones posted:Shoot 2.5+. You'll have a much wider focus plane and won't be missing so often. Focus wide open, then stop down, which will give you a bit more wiggle room. Really though 50mm is just too long for indoors. Canon's greatest weakness is their lack of a 35mm f/1.8 DX equivalent. The Sigma 30/1.4 is decent enough but it's still too expensive to really fill that niche. The 40/2.8 is the most reasonable substitute but it's still a bit long especially with Canon's higher 1.6x crop versus the normal 1.5x. Paul MaudDib fucked around with this message at 22:10 on May 1, 2014 |
|
#
¿
May 1, 2014 22:07
|
|
|
totalnewbie posted:Well, after sending my 50D and Tamron 17-50 to my mother, she's complained about her pictures being fuzzy. I ask her what she means and she tells me that when she takes pictures of flowers from a few inches away, it comes out fuzzy. If she doesn't understand minimum focus distance, FF is probably an unnecessary upgrade for her. That said, I'd look at the Tamron 90mm macro or the 100mm Canon macro lenses.
|
|
#
¿
May 12, 2014 21:43
|
|
|
Tricerapowerbottom posted:Just one of those would really cover landscapes, candids, etc? I know they're well regarded lenses, but I want something to put on my camera for a "I don't know what's going to happen" kind of day. 35mm equivalent is a great "I don't know what's going to happen" FL. Wide enough for most amateur architecture or landscape type stuff, not so wide that portraits are impractical. I find it's possible to cover 95% of my shooting with 2 focal lengths, 3 tops. 28+50 equivalent or 35+105 or something along those lines. In Canon the obvious long lens is the 85mm.
|
|
#
¿
Jun 3, 2014 22:05
|
|
|
|
| # ¿ May 16, 2024 04:47 |
|
|
Tricerapowerbottom posted:So I'm wondering, with this talk of portraits and crop sensors: the sensor size has no effect on the compression effect of a longer focal length, right? When people say that a 'On crop, a 50 would be about the equivalent of the 85." they're only meaning in terms of how wide of a field of view you will get on the sensor itself, right? Sensor size and focal length determine field of view (assuming the lens covers). Yeah, all people are talking about when they're comparing crop/FF is field of view, unless otherwise specified. If people are talking about depth-of-field equivalence, they'll say so. Smaller sensors will have deeper depth of field (for a given focal length), but compression is a function of distance from your subject, not the lens itself. What narrower field of view buys you is the ability to stand further away from your subject and fill the frame, if you take a picture from a certain distance with a 70-200 and a 24mm you will have the same compression effect, the person is just going to be much smaller in the picture with the 24mm. If you cropped it back down to the same FoV the subject would look the same (ignoring distortion, etc). Paul MaudDib fucked around with this message at 22:05 on Jun 17, 2014 |
|
#
¿
Jun 17, 2014 22:02
|
|