Sharpness: 300mm vs 500mm & Lepp's tests

In <a href="http://www.leppphoto.com/">George Lepp's</a> lens tests

(the link is to his page, but there's no lens test data there), the decent 100-300mm zooms tend to

rate at about a 7.0 on his scale (1-10, with a 6 being <em>just</em>

good enough for professional use if you do everything right). So

the Canon 100-300 USM gets a 7 at 300mm, f5.6 and the Nikon

75-300 does about the same. OK so far. These lenses are OK, but

not great.

<p>

When he tested the Nikon 500/4 EDIF, the Canon FD 500/4.5L and the

Canon EF 500/4.5L, they all got a score of 7.0 when used wide open.

This suggests that the image quality is the same as that of the

consumer grade x-300 zooms, used wide open at 300mm and f5.6. I find this somewhat suprising (I would have expected and hoped that the

500mm lenses would be better than that).

<p>

Does anyone with field experience of these lenses have a comment?

I've never used any of the 500mm lenses myself. On lenses like the

Canon 300/4L and 300/2.8L, Lepp gives them an 8.0 rating wide

open, i.e. significantly better than the x-300 zooms at 300/5.6.

and this is, of course, correct (i.e. that's what I see too!).

Since nobody else has taken a crack at this, I figured I'd contribute a semi-useful answer.

<p>

I have no first hand experience with any big telephotos(I wish I did) but I know that the lenses you mention are older designs. The FD lens has obviously been around since before the EOS cameras and I'm pretty sure the EF lens is the same optically. The Nikon lens you mention has been around since the mid-eighties unless it's the P version which came out around 1988 I believe. I would be willing to bet that newer designs from Canon or Nikon would do better. I've read that the AFI and AFS lenses from Nikon have been improved quite a bit from previous versions.

<p>

That said, I am also surprised that these lenses didn't do better than even a decent 75-300 zoom. I imagine, however, that the manufacturers are spending a lot more design time on these than the big teles. Canon obviously does since they make about 5 different lenses in this range.

Yes, the EF500/4.5L is the same design as the FD500/4.5L. However

they both use fluorite elements and should be pretty sharp. The

EF300/2.8L came out in 1988 and, I think is the same design as

the earlier FD300/2.8L, and it is <em>very</em> sharp indeed. Lepp did

his tests on the 500mm lenses only a few years ago, so I assume

he tested the recent (current) version of each.

<p>

I don't know how consistant his test numbers are, but for

comparison, the old Sigma 400/5.6 APO got a 6.5 wide open,

and the new Sigma 400/5.6 APO Macro gets a 7+, slightly

better than the 500/4(.5) lenses do. Now the 500s are a full

stop faster and 25% longer, but even so, the numbers suggest

that you are not getting a <em>lot</em> more performance for

your money (a lot more of your money).

<p>

On the other hand, I've never heard of any serious complaints

about the optical quality of the Nikon 500/4 EDIF. In fact

it's one of the most widely used and praised telephoto lenses

around and is used by large numbers of serious amateur and

pro nature photographers worldwide.

<p>

Test data on the big, expensive telephotos seems to be very

thin on the ground. It's a pity because they are a <em>MUCH</em>

bigger investment than the smaler lenses - for which there is

no shortage of test data (for what that's worth!).

I think there might be another reason why the consumer grade x-300 zooms seem to perform as well as the 500mm primes, but since I'm no expert on optics, I'm just guessing here. Anyhow, my reasoning is as follows: It's supposedly pretty easy to make a great 50mm prime. It's almost as easy to make a great 100mm prime. (You can probably see where I'm going with this.) I've heard it said that there was never a really great 300mm lens until the 1980's. So perhaps it's REALLY difficult to make a great 500mm prime -- harder enough that the best 500mm primes perform about as well wide open as the aforementioned consumer grade 300s.<p>

<p>

One of the problems with really long focal lengths is that each color of light comes to focus at a slightly different point. This is a BIG problem for telescope makers, who are essentially designing really long lenses (900mm - 2400mm are common). The best performing refracting telescopes (generally agreed to be Astro-Physics APOs) are very expensive, and I assume were very difficult to design. Then again, as I mentioned above, I'm not an expert on optical design.

Well first, I'm not convinced that 500mm f4(.5) lenses aren't

sharper than consumer 300mm zooms (both wide open). I don't

know if the numbers George Lepp has published are, in fact,

truely reflective of relative optical performance. That's one

reason I asked if anyone had themselves tested such relative

performance (I'd have done it myself if I had one of the

lenses in question!).

<p>

Telescope design is somewhat different. You are usually working

at (photographically) slower speeds. Usually f6 to f8. You

also don't need a flat field for visual use, and you don't

have telephoto design constraints (i.e. physical length less

then the optical focal length). This makes life easier.

<p>

However for telescope design you are looking for 1/10 wave or better

diffraction limited performance in a good scope. You don't

need that in a lens because the film isn't capable of

recording that much detail. For a lens, 1/4 wave is probably

more than sufficent. Even given all this, the fastest 500mm refractor

telescope (the old TeleVue Genesis 500/5 APO Fluorite) cost

about $2000, less than most comparable 500mm lenses.

<p>

BTW the Genesis does make a great lens and blows away the

performance of consumer zooms at 300mm. I know that one

for a fact because I have tested it! However it's far from

being a telephoto, in fact it is significantly <em> longer</em>

than 500mm (close to 1m in physical length). It's actually

a 4" diameter f10 or f12, with a fluorite compressor element

1/2 way down the tube. A Petzval lens design rather than a

traditional refractor telescope.

Since I haven't done side-by-side comparisons, I too doubt that a $6000 500/4.5 couldn't outperform a $350 100-300 zoom. My previous response was on the assumption that Lepp's tests are accurate. Since I can't contradict them with personal experience, I was proposing a possible explanation. If someone wants to buy me a 500/4.5, I'll be happy to do extensive comparisons between it and my consumer-grade gear ;)<p>

<p>

Regarding the telescopes: While you don't need a flat field for <i>visual</i> use of telescopes, it certainly helps to have it for astrophotography. Speaking of which, I think I'll start a new "thread" here about astrophotography...

In fact the Genesis telescope I mentioned <em>is</em> a flat

field instrument (or as flat as fields normally are, even lenses

are not truely falt field). With the Astro-Physics telescopes,

an auxilliary field flattening lens is available, necessary for

attempts at 6x6 work or critical 35mm work. I'm not sure if I'm

remember ing this right, but for a conventional refractor

telescope (i.e. a compound objective at the far end of the tube),

the field curvature is that of a surface whose radius is 2/3 of

the focal length of the objective lens. I believe that <em>strictly</em>

speaking the field curvature is parabolic, not spherical though

(Petzval surface).

I'm not sure that one can conclude that an equal score of

7.0 means equal performance. It could be that they both

show some problems, but maybe different manifestations

or even to differing degrees but of differing importance.

Also, I'm not sure what the difference between a 7 and an

8 represents.

<p>

I do think that the 500/4 Nikkor is quite sharp wide open,

but not as vividly sharp as a 300/2.8 Nikkor. I can't speak

to how this stacks up against a 300mm zoom. I've used

the lens a lot at f/4 with no dissatisfaction in sharpness,

even with a 1.4x converter attached.

<p>

My somewhat parochial guess is that the Canon lens may be

a tad soft (simply judging by a relative lack of widespread

acceptance relative to the Nikkor 500s which are everywhere)

and Lepp applied his algorithm of 'no Nikon lens shall

be rated better than those of my Sponsor.' There have

been some reports that the Nikkor 500s were a bit soft,

but again, I haven't seen it (nor have I done any real

comparisons except versus the f/2.8 which I would expect to

be noticeably better than the 500, and 400/2.8 and 600/4 for

that matter, regardless of brand.) The popularity of

the Nikkor 500 among nature photographers is phenomenal,

I think due to a combination of speed, length, weight,

and overall performance.

Lepp's numbers are "closely related to lp/mm" or words to that

effect. They basically represent sharpness (and are probably

weighted for center sharpness). The difference between a 7 and

an 8 is significant. Lenses that he rates at a 7 (let's say

the consumer zooms, the Sigma 400/5.6 APO and so on) are

significantly less sharp than lenses he rates at an 8 (for

example the Canon 300/4 at f4 or the 300/2.8 at f2.8), based

on my own testing of these lenses.

<p>

I think the reason the Canon 500/4.5L is less widespread than

the Nikon 500/4 is that most Canon users are using AF lenses

these days and the 500/4.5L won't AF with the 1.4x TC (it's

1/3 stop too slow, so the camera won't attempt AF). For Nikon

users, who are still using a lot of MF lenses (not an option

for Canon), the 500/4 is still a very good lens, AF or MF.

And, of course, in the AF version, it should AF with a TC without

problems.

<p>

It could also be that the 500/4.5L is a little soft, though I

have not heard any such complaints (nor have any been posted

here, though this is a small sample). I <em>did</em> hear rumors

that the 400/2.8L wasn't as sharp as it might be, and sure

enough, a "mark II" version was released recently which is

supposed to be significantly better. Of course the old lens

didn't have fluorite elements, which the new one does. I believe

the 500/4.5L does use a fluorite element.

<p>

It's worth noting (but it's only one data point) that the chasseur

d'image test on the 500/4.5L showed it just a slight notch less

sharp then the 300/2.8L and 600/4L (which got equally good ratings).

I don't belive Lepp has ever published numbers on the 600/4L, though

I know he uses one. The 500/4.5L rated about the same as the 300/4L,

which is a *very* sharp lens, so the fact it was slightly below

the rating of the 300/2.8L should be taken as an indication it's

anything but sharp!

Cripes Bob, who knows with George's tests? I've seen all of these and from what I can see there is not a 300mm zoom save perhaps the 50-300 4.5 ED Nikkor (or the 200-400 f4 ED Nikkor) which competes with the 500 f4.5 EF-L Canon and 500 f4 EDIF Nikkor (BTW the word is that the 500 f4 EDIF-AF-I is sharper wide open than the MF version). My best guess is that George tests these lenses at different distances or on different targets => different results. We both know how difficult these sorts of tests are to do w/out a lab bench set-up; I'd guess that the problem is in the experiment, could also be due to vibration, Fritz Poelking ran an interesting series of tests on vibration in long lenses about a year ago -- published in Fotografie Draussen -- just mounting it to a tripod w/MLU in not sufficient beyond 200-300mm!!!

<p>

Grover Larkins

I attended a 4-day workshop with George Lepp at Olympic National Park about 6 years ago. He described his lens/film testing methods and they are quite scientific. He uses the same military resolution chart on all lenses under the same lighting conditions. If you stop a 500 mm lens down to f5.6, you'll already get a significant improvement in sharpness, so if you compare the consumer zooms and the 500 at the same f5.6, then you'd probably see the 500's comparing more favorablly. I haven't used a 500, but I own a Canon EF 300 2.8 and a Sigma 400 5.6. I haven't done any scientific comparisons, but can tell you there is a world of difference (the 300 is MUCH better).

<p>

Sharpness is only one issue when deciding on whether to fork over the big bucks for the big lens. The coatings give the lenses a unique look from one manufacturer to another and for me after having used the Sigma I'd rather stick with Canon's lenses. I just prefer the "look" of the photos from the Canon lenses. Having said that, I'm still thinking of buying a Sigma 28-70 2.8 to use as a "point & shoot" kind of lens because it's much lighter and MUCH cheaper. I wish I could test one first before I buy, but there's not much option to do that here in Taiwan.

<p>

I'd have to say I typically find truth in the addage "you get what you pay for" so I'm very cautious.