Lens Recommendation

[TxHunter]

Looking for a lens that will allow fast shooting of sports photography, have the range for scenery at a distance, but yet have wide angle capabilities for reasonable close up pics.  I have a 70-200 F4/L IS, 55-250 F/4.0-5.6 L and the standard 18-55mm that comes with some ofthe older EOS models.

I find the 55-250 is not wide enough on the low end and soft on the longer zoom.    I would like a lens that can still shoot the sports photography and yet be capable of shooting close ups at wide angle. 

Any recommendations?

[hsbn]

None. There is no FAST super zoom lens. But there are some good super zoom lens, you can check out this article

http://www.adorama.com/alc/0013834/article/7-All-In-One-Lenses-That-Do-It-All

[cicopo]

Actually there are 2 & I have them both. The old 35-350 L and current 28-300 L IS can do it but not at night unless there's really good lighting. Unfortunately the 28-300 is both expensive & HEAVY, so it takes getting used to. You're probably better off considering a 2 body / 2 lens solution.

[TCampbell]

Part of the point of a camera that allows for interchangeable lenses is that you can put the lens on the camera which is optimized for what you need to shoot.

All lenses compromize in some way.  If it were not true, then the camera makers wouldn't need a removeable lens camera -- they could just fit one lens permanently onto the body and be done.  

Super-zooms (lenses that go from wide angle to fairly agressive zooms all in one lens) compromise perhaps more than any other lens on the optical quality and focal ratios.  They do well only in convenience.  If you HAD to pack for a trip and had room to take ONLY one lens then a super-zoom could be that lens.  But if you could bring more than one lens, then you would almost certainly get better performance by doing so.

Super-zooms compromise in optical quality to the point that you may as well just use a point & shoot camera -- you wont see the performance of DSLR capabiltities coming through when you use these lenses.

If the sports photography is daytime outdoor photography then you might look at the Canon EF 100-400mm f/4.5-5.6L.  This really depends upon the sport.  For sports on large fields (e.g. soccer, football, etc.) then it's nice to get to a 300 or 400mm focal length.  For indoor sports (e.g. basketball) then a 70-200mm is probably a good choice and while the EF 70-200mm f/2.8L IS USM II is not inexpensive, it's not "re-mortgage your house" expensive.   The EF 200-400mm f/4L IS USM Extender 1.4x is over $10k.

Sports photography, and in particular indoor games or night-games under artificial lights, can be very demanding on the equipment because you need high shutter speeds and you have to shoot in very poor light.  This involves cranking up the ISO and using lenses that are very good at gathering a lot of light (lowest focal ratio you can possibly afford) while having strong zoom lenses or long prime lenses.  

[cicopo]

"Super-zooms compromise in optical quality to the point that you may as well just use a point & shoot camera -- you wont see the performance of DSLR capabiltities coming through when you use these lenses."

Sorry but I disagree BIG TIME. I get some pretty good shots from superzooms including consumer grade ones. DSLR's have very low shutter lag times, better AF systems & better view finders plus when shooting action they aren't usually held out in front watching the LCD so people are forced to use better technique.  Superzooms may not be perfect when compared to primes but they offer tremendous versatility and that's a big plus to a whole lot of amateur photographers.

[TCampbell]

---

@cicopo wrote:

"Super-zooms compromise in optical quality to the point that you may as well just use a point & shoot camera -- you wont see the performance of DSLR capabiltities coming through when you use these lenses."

Sorry but I disagree BIG TIME. I get some pretty good shots from superzooms including consumer grade ones. DSLR's have very low shutter lag times, better AF systems & better view finders plus when shooting action they aren't usually held out in front watching the LCD so people are forced to use better technique.  Superzooms may not be perfect when compared to primes but they offer tremendous versatility and that's a big plus to a whole lot of amateur photographers.

---

Most zoom lenses tend to offer a zoom range which provides about a 3x magnification factor from the wide end to the narrow end.  A 24-70mm for example... 24 X 3 = 72 -- so that's about a 3x zoom.  A 70-200... 70 X 3 = 210... again about a 3x zoom.  A 3x zoom range is not so ambitious as to create lots of other problems.

When light passes through any single lens element, it bends (which is the whole point) but each wavelength component of the light bends at a different amount.  That means if you only had a single lens element, all the various colors of light would not actually come to focus at the same distance behind the lens.  A lens focused perfectly for "green" light would not also be in focus for the "reds" or "blues".

To counter this, you can use two lens elements arranged as an achromatic doublet -- which helps re-converge the chromatic dispersion problem -- but it's not perfect.  You can use apochromatic triplets to do a better job... and keep going with techniques to make a better image.  

You can optimize this fairly well if you're just trying to create a single focal length.  If you're trying to cover lots of focal lengths, then it becomes a bit more complicated.   You can also control the issues better if the zoom doesn't try to be too ambitious about it's zoom range.  That's why I mentioned the 3x range.  Some slighlty more ambitous zooms may try to go to a 5x zoom range (e.g. a 55-250 would be an example).  

And THEN you get the super-zooms.  These lenses go well beyond a 3x or even 5x range.   A 10x or 11x would not be unheard of and the 28-300 is an example of a lens that exceeds a 10x range.

It's difficult to provide a 10x range and still maintain good optical quality.

Keep in mind that we don't use most of the pixels we shoot.   You'd really need to blow up an image to 100% size such that you can see every pixel.  

The chromatic dispersion is nearly non-existant along the central axis of a lens... and increases the farther away you are from the central axis (corners are worst).  The f-stop you are using also has a lot to do with things.  As the aperture gets smaller, the chromatic abberation problems are reduced and the image appears to get sharper... until you pass the point where diffraction comes into play (usually after you pass f/8 it starts going the other way).

If you go look at the MTF curve charts for the Canon EF 28-300mm f/3.5-5.6L IS USM lens... and then go look at the MTF curve charts for the Canon EF 70-300mm f/4-5.6L IS USM ... the less-ambitious zoom is not just a little better... it's a LOT better.  And you'll find this comparison is pretty consistent.  You'll be hard-pressed to find an exception to it.  (you would need to a comparitively high-end super-zoom (like the Canon "L" series) and then compare it to a very low quality consumer-grade zoom to find an exception.)  The the greater the zoom range (when thought of as magnifciation factors) the more the lens is disadvantaged for being ambitious.

It's a fair generalization to say that super zooms... almost without regard to price tag (becuase the Canon EF 28-300mm f/3.5-5.6L IS USM is not an inexpensive lens) generally cannot compete with their less-ambitous siblings.

This is NOT to say that super-zooms have no place.  They can be tremendously useful.  You have to consider how an image will be used after it's exposed.  If you're publishing them on the web then it's actually very hard to tell the difference between a good lens and a fantastic lens because we don't use most of the pixels we capture. The image is down-sampled to a reasonable web-size and nobody gets to scrutinize the fine-detail.  If the image is going to be printed on very large size and hung in a gallery... you probably want to be fussier about the optical quality of the lens.

[Skirball]

---

@TCampbell wrote:

---

@cicopo wrote:

"Super-zooms compromise in optical quality to the point that you may as well just use a point & shoot camera -- you wont see the performance of DSLR capabiltities coming through when you use these lenses."

Sorry but I disagree BIG TIME. I get some pretty good shots from superzooms including consumer grade ones. DSLR's have very low shutter lag times, better AF systems & better view finders plus when shooting action they aren't usually held out in front watching the LCD so people are forced to use better technique.  Superzooms may not be perfect when compared to primes but they offer tremendous versatility and that's a big plus to a whole lot of amateur photographers.

---

Most zoom lenses tend to offer a zoom range which provides about a 3x magnification factor from the wide end to the narrow end.  A 24-70mm for example... 24 X 3 = 72 -- so that's about a 3x zoom.  A 70-200... 70 X 3 = 210... again about a 3x zoom.  A 3x zoom range is not so ambitious as to create lots of other problems.

When light passes through any single lens element, it bends (which is the whole point) but each wavelength component of the light bends at a different amount.  That means if you only had a single lens element, all the various colors of light would not actually come to focus at the same distance behind the lens.  A lens focused perfectly for "green" light would not also be in focus for the "reds" or "blues".

To counter this, you can use two lens elements arranged as an achromatic doublet -- which helps re-converge the chromatic dispersion problem -- but it's not perfect.  You can use apochromatic triplets to do a better job... and keep going with techniques to make a better image.  

You can optimize this fairly well if you're just trying to create a single focal length.  If you're trying to cover lots of focal lengths, then it becomes a bit more complicated.   You can also control the issues better if the zoom doesn't try to be too ambitious about it's zoom range.  That's why I mentioned the 3x range.  Some slighlty more ambitous zooms may try to go to a 5x zoom range (e.g. a 55-250 would be an example).  

And THEN you get the super-zooms.  These lenses go well beyond a 3x or even 5x range.   A 10x or 11x would not be unheard of and the 28-300 is an example of a lens that exceeds a 10x range.

It's difficult to provide a 10x range and still maintain good optical quality.

---

Truthfully, he might well use a P&S, as well as many others, if it could track moving subjects acceptably.  You seem to imply that IQ is the fundemental difference between dSLR and P&S.  I don't think that that is the case anymore.  The IQ of P&S has increased significantly over the past several years.  The quality can be quite good... for static subjects.  If that was the only difference there would be a lot more "serious photographers" just carrying around P&S these days.  But when I think of the limitation of the handy little portably cameras I think of AF, not IQ.

As to the superzoom: While I mostly agree with your overall point, and don't care for the superzooms myself, your argument against 10X isn't actually based on anything.  Yes the light gets more attenuated the more lenses it passes through, but there isn't anything fundamentally limiting a 10X lens from being optically acceptable, technology just isn't there yet. The same argument could be made against any zoom vs a prime, yet there are now zooms that offer near the same resolving capability.   If you told people 15 years ago that you'd have a zoom that could rival a prime they would have laughed at you.

Your point stands that the IQ of super zooms isn't near as good as many top notch lenses, but that doesn't make them useless.  The value of MTF charts are greatly overstated.  I don't take photos of MTF charts, nor does Cicopo.  It's the quality of the images that it produces that matter.  And if that quality is sufficient for his needs, then the lens serves a purpose for him.

[Skirball]

Truthfully, he might well use a P&S, as well as many others, if it could track moving subjects acceptably.  You seem to imply that IQ is the fundemental difference between dSLR and P&S.  I don't think that that is the case anymore.  The IQ of P&S has increased significantly over the past several years.  The quality can be quite good... for static subjects.  If that was the only difference there would be a lot more "serious photographers" just carrying around P&S these days.  But when I think of the limitation of the handy little portably cameras I think of AF, not IQ.

As to the superzoom: While I mostly agree with your overall point, and don't care for the superzooms myself, your argument against 10X isn't actually based on anything.  Yes the light gets more attenuated the more lenses it passes through, but there isn't anything fundamentally limiting a 10X lens from being optically acceptable, technology just isn't there yet. The same argument could be made against any zoom vs a prime, yet there are now zooms that offer near the same resolving capability.   If you told people 15 years ago that you'd have a zoom that could rival a prime they would have laughed at you.

Your point stands that the IQ of super zooms isn't near as good as many top notch lenses, but that doesn't make them useless.  The value of MTF charts are greatly overstated.  I don't take photos of MTF charts, nor does Cicopo.  It's the quality of the images that it produces that matter.  And if that quality is sufficient for his needs, then the lens serves a purpose for him.

[cicopo]

To me the MTF charts are a great way to pick the better of 2 (or more) similar lenses but that's it. They are done under very controlled conditions and do not take into account shooting hand held nor while panning, which are both huge variables that influence the photos taken through the lens. If you don't shoot from a tripod most of the time just how close are your results between similar lenses used identically.

Also note that most people DON'T print big, and not many even print at all relative to the number of photos they take yearly. Most people need to develop great technique to get good value from high end lenses, and that comes over time but it also comes from experimenting with different focal lengths which superzooms allow. Once someone has improved to the point their equipment is holding them back that's when they can start shopping for the best in class lenses or bodies. I've shot the original Sigma 18-200 & when it was introduced bought the OS version plus I've shot the original Tamron 18-270 VC with some great results, but they don't compare to the Canon L's I mentioned early in this thread. I also have the Nikon 28-300 VR but so far it hasn't done as well as any of the lenses I've mentioned but then again I may not have the micro adjustment as dialed in as I thought I did.  

[TCampbell]

---

@cicopo wrote:

Also note that most people DON'T print big, and not many even print at all relative to the number of photos they take yearly. 

---

This is VERY true.  Most people do not print big.  That's why I mentioned that depending on how you use a lens, not only might the owner not be able to tell the difference between a merely "good" lens and a "fantastic" lens... but even if you showed your images to something with quite an eye for detail -- they may not be able to tell them apart either.

I'm careful to say that it's not that super-zooms aren't useful or have no place ... but I think it's important to understand what you're buying.

The 300mm image that comes out of a 28-300mm lens and the 300mm image that comes out of a 70-300mm lens will not have equivalent optical quality... BUT... depending on how you use that image, nobody may be able to tell the difference.  But it's equally fair to point out that this would not hold true for all uses of the image.  Output size (or heavy cropping) is an important consideration.

I think we are probably all mostly agreeing with each other.  

[TCampbell]

---

@Skirball wrote:

As to the superzoom: While I mostly agree with your overall point, and don't care for the superzooms myself, your argument against 10X isn't actually based on anything.  Yes the light gets more attenuated the more lenses it passes through, but there isn't anything fundamentally limiting a 10X lens from being optically acceptable, technology just isn't there yet. The same argument could be made against any zoom vs a prime, yet there are now zooms that offer near the same resolving capability.   If you told people 15 years ago that you'd have a zoom that could rival a prime they would have laughed at you.

Your point stands that the IQ of super zooms isn't near as good as many top notch lenses, but that doesn't make them useless.  The value of MTF charts are greatly overstated.  I don't take photos of MTF charts, nor does Cicopo.  It's the quality of the images that it produces that matter.  And if that quality is sufficient for his needs, then the lens serves a purpose for him.

---

To say that the technology "isn't there yet" is a bit of a simplification.  It's not that the technology isn't well-understood... to the contrary it's actually understood extremely well.  It's just that the problem is so complex that what you have to do to fix it starts to become ludicrously expensive.

I'll give you an example... 

Canon manufactures some of their larger and better lenses using fluorite crystal.  Fluorite is able to refract light with a low dispersion (the light still splits into a rainbow... but not as much as it would if pushed through more traditional optical grade glass.)   They do this to reduce the CA issues.

BUT... to do this, the fluorite has to be "grown" synthetically in a kiln.  Fluorite occurs in nature, but only in small bits and is loaded with impurities.  To create optical grade fluorite crystal in a large enough size that you can use it to grind a lens out of the crystal, requires that you synthetically grow your own crystals.   But it turns out you cannot do  a rush job... the faster you try to grow the crystal, the more likely that the crystals will have flaws that make them unsuitable for use and you end up with a very high "scrap" rate.    Depending on the size the crystal needs to be to go into the lens, it can take months to grow (and yes, Canon actaully has their own plants growing these things... which is part of why those really long prime and zoom lenses cost ludicrous amounts of money.  You cannot produce these things quickly.)

There are alternatives to fluorite, of course... but this was just one example that it's not fundamentally "easy" to make long lenses to begin with... much less long super-zoom lenses with high optical quality.

Huge telescopes side-step the issue by using mirrors and just avoiding lenses (that was Isaac Newton's brilliant insight... you could focus the light using parabolic mirrors not actually have to pass the light through a glass "lens" element at all and avoid the whole chromatic aberration mess entirely.)

MTF curves are published by Canon at (a) whatever "wide open" for that lens & focal length combo happens to be (and this will vary from lens to lens which makes a direct comparison not so straight forward) and (b) they also publish the f/8 performance of the lens.  Since every lens actually can be used at f/8 (I do not know of any lens that cannot do this) and also since the best performance tends to be one of those middle f-stops (often somewhere between f/5.6 and f/8) this is a good indicator of the what may be the close to the best performance a lens can offer... with one caveat:  if the lens is a zoom lens, they also publish the graph for the wide-angle extreme and the narrow-angle extreme.  But zoom lenses have sweet-spots and that sweet spot may be (and almost always is) somewhere in the middle.  There is generally some focal length that will beat the performance of the published MTF graph (at least by a little).

That all just boils down to the idea that you need to know what you're getting into when you start comparing MTF curves because it's not usually a straight apples-to-apples comparison.  But the curves themselves are useful and as long as you go into it with all the caveats in mind, you can extrapolate some useful information about what you might expect from the lens by looking at those charts.   If they weren't useful, the lens makers wouldn't publish them.  I *think* (and my history may be wrong) that Canon was first to start publishing them -- because frankly they offer bragging rights by saying "our lenses are this good and we're not afraid to let you see this data".  Nikon eventually followed (I think they only publish the wide-open score and not the f/8 score so it's a bit harder to make comparisons using their graphs.)  Now I'm finally starting to see some 3rd party lens makers publish their own graphs.  The big brands (Canon and Nikon) usually beat the perofrmance of the 3rd party lenses -- but everybodies' glass is getting better over time.  

I did mention and I suppose it's worth re-iterating, that whether or not you can notice a difference between the performance of two lenses depends largely on what you plan to do with the exposure once you make it.   If I'm posting a 1024 pixel wide image on a web site you can cover up a lot of optical flaws because it's hard to notice problems when an image has been resampled to such a small size.  But if you plan to make a gallery print... that's something else entirely.

It's also worth mentioning that sometimes we don't care about the quality of an image in the corner.  if you want a soft background and you dont plan to put your subject in the corner of the frame... then soft corners are maybe ok for your purposes.  If you are shooting landscapes and want the entire thing to be tack-sharp from center to corner... then you probaby want to fuss over the optical quality of the glass.

[TCampbell]

As to the superzoom: While I mostly agree with your overall point, and don't care for the superzooms myself, your argument against 10X isn't actually based on anything.  Yes the light gets more attenuated the more lenses it passes through, but there isn't anything fundamentally limiting a 10X lens from being optically acceptable, technology just isn't there yet. The same argument could be made against any zoom vs a prime, yet there are now zooms that offer near the same resolving capability.   If you told people 15 years ago that you'd have a zoom that could rival a prime they would have laughed at you.

Your point stands that the IQ of super zooms isn't near as good as many top notch lenses, but that doesn't make them useless.  The value of MTF charts are greatly overstated.  I don't take photos of MTF charts, nor does Cicopo.  It's the quality of the images that it produces that matter.  And if that quality is sufficient for his needs, then the lens serves a purpose for him.