Severi Salminen wrote:
> Thank you to all who answered my questions about the combo Eos 30+28-90+75-300.
> I wanted to know that I won't be buying poor or low quality lens - even though
> quality is a very subjective matter.
At the risk of flogging that poor dead horse again, I'd like to offer some remarks
here. First of all, lens performance, while it can be measured and expressed
empirically, is also a relative matter. Canon makes some lenses that are better than
others. Sometimes these better lenses cost significantly more than those that
perform less-well, and sometimes they don't (the 50/1.8 is a good example). This
does not mean that those at the bottom of the scale are poor in an absolute
sense--in the case of zoom lenses, even the worst-performing lens in Canon's current
line-up is probably quite a bit better than most zooms designed and built in the
1970s. Nonetheless, the differences are real, and it doesn't necessarily require
large enlargements or transparencies viewed through an 8x loupe to see these
differences. Make no mistake: depending on the lens and its characteristics,
relatively poor performance can sometimes be seen even on 4x6-inch (10x15 cm)
prints.
As an example, let's take the 28-105/3.5~4.5 Mk.I zoom which I have suggested is a
good buy at current closeout prices. I have been using this lens a lot recently to
check out its "real-world" performance, and I've gotten some nice images with it,
with both negative and slide film. I expect it to perform well in the use for which
I bought it, which is as a lightweight and compact hiking lens. However, it is not
the sharpest lens I have by any means, and it shows significant barrel distortion,
at least at the short end. I noticed this before I looked at any test
results--distant horizons (as when looking out from a mountainside) near the top of
the frame show excessive curvature and a tendency to tilt off-axis; and a quick
visit to . . . Photodo (no anti-Photodo rants please!) showed a tested barrel
distortion of greater than 4% on both axes, which I presume has something to do with
my horizons. This is something that could be easily seen on any minilab print.
Now, the 28-90/4~5.6 is one-half the price of the 28-105 (in the US). With the 28-90
you give up about 10mm of reach, a significant bit of lens speed on the long end,
and full-time manual focusing, in exchange for $100 in your pocket. Photodo has not
yet tested the new lens, so I have no distortion figures for it. However, Chasseur
d'Images conveniently tested both the 28-90 and 28-105 Mk.II (optically identical to
Mk.I) in the same issue (228); they appear to have a different test for distortion
than does Photodo, for CdI's barrel distortion figures at 28mm are 1.3% for the
28-105 and 1.4% for the 28-90--but in both cases they characterize the distortion
(at 28mm) as "annoying" ("gênante"). If we look at the resolution charts, we see
that at 28mm the 28-90 is virtually identical to the 28-105, but as the focal length
is increased, the resolution of the 28-105 improves more than does that of the
28-90: at 50mm the 28-105 is quite a bit better than the 28-90, and it maintains its
edge, although by a smaller margin, at 85mm vs. 90mm on the "kit" lens; at 105mm,
the older zoom also outperforms the kit lens at 90mm, especially wide open (f/4.5
vs. f/5.6.). Distortion figures for both lenses are similar at all tested focal
lengths, and the 28-90 has an edge over the older zoom in terms of vignetting,
especially at maximum aperture.
So, would we see these differences on standard machine prints? Possibly not, but
it's also possible that we would see some or all of these differences--just because
we're looking at machine prints doesn't mean we can simply assume that the optical
characteristics of different lenses will be masked by a general low quality of
reproduction. Is the 28-105 Mk. I worth an extra $100 over the 28-90? Now *there's*
a subjective decision! But, for my purposes it doesn't really matter--the point I
want to make is that it is not, IMO, safe to just assume that lens quality will not
have a discernible impact on our photos if only we stick to machine prints. I would
argue that this is not at all something that can be assumed, even if that assumption
can help us feel sanguine about buying a bottom-of-the-line lens *if we can afford
something better,* either now or in the foreseeable future. And this is the point
where individual budgets come into play, because it serves no purpose to urge
someone to buy a lens he or she can't afford; OTOH, if someone comes along and asks
about the relative qualities of different lenses without indicating their budgetary
constraints, it is IMO completely legitimate to try to steer that person toward
better-performing lenses, if that person shows an interest in making more than mere
snapshots. And from *that* perspective, it seems legitimate to point out that
neither of the lenses I've been discussing can match the performance of the
28-70/2.8L, and that the differences will almost certainly be seen on minilab prints
of the identical scene shot with both (or all three) lenses. Whether the 28-70L
represents a good value at *its* current price is another question! But if your
photos are crappy you may never have the urge to have quality enlargements made, and
so you may never feel the need or desire to have a better performing lens. But that
can be a circular situation--if your lens doesn't perform well you may never want
enlargements, so you'll stick to 4x6 machine prints and never feel a need for a
better lens. But using a better lens might make you dissatisfied with those machine
prints and you might end up with some nice enlargements on your walls instead of a
shoebox full of snapshots. In other words, it may not be safe to assume that if your
4x6 prints don't seem worth enlarging that it's because of lousy printing or poor
technique--it could be your lens.
> Actually I've been quite amazed by how many
> people consider only pro lenses as good quality - anything less is claimed to be
> simply poor.
Well, there's good quality and then there's good quality. Compared to the lenses
(especially zoom lenses) of 25 years ago, Canon's bottom-of-the-line lenses aren't
particularly bad, but that doesn't mean their better lenses aren't noticeably
better. Even on machine prints the 85/1.8 or 200/2.8L (or the inexpensive 50/1.8)
are likely to produce noticeably better results than the kit lenses (or the 28-105).
> People are saying that with "consumer grade" lenses one gets only
> compact camera quality. This I find very hard to believe.
In some cases the lenses on the better point-and-shoots will in fact outperform
low-end SLR lenses; the limiting factors there could be film choice and the ability
to hold a P&S steady.
> Of course it greatly
> depends on what media you are going to use
Does this not imply limiting the use of your photos to a single medium? In other
words, do you really want "4x6 quality" to be your goal? I should think most folks
would not.
> slides projected on wall. In the latter case I do believe that only the best
> lenses are able to provide enough sharpness and contrast.
IMO lens performance may be more important for prints than for projected slides,
because in the latter case you're very often severely limited by the quality of the
projection lens (good ones cost a lot more than a 28-90 zoom) and on whether the
projector is internally aligned properly (I have one that isn't, even after
servicing, and it makes even my best slides look like junk). Viewing slides on a
lightbox with a quality loupe (also more costly than a 28-90 zoom) is a different
matter, as can be scanning. But even so, enlarging negatives with a conventional
enlarger means passing the image through an enlarging lens, which will introduce
some degradation, so the sharper the negative, the sharper the print will be (am I
in contradiction here??).
It is not my intent to persuade Severi or anyone else to buy lenses they can't
afford or to make them dissatisfied with the lenses they can afford. Rather, it's to
attempt to point out what I see as some misconceptions about the extent to which the
optical performance of a lens has a perceptible affect on our photos regardless of
the way the photos are presented, whether it be 4x6 minilab prints or 20x24-inch
LightJet prints from 9000 ppi drum scans--particularly the apparently
often-unexamined assumption that we won't see any difference on well-executed
machine prints. In the end I'm arguing that logic and my personal experience (for
what they are worth) suggest that is not the case. It is indeed unfortunate that
Canon's marketing strategy has been to place almost all their better-performing
lenses in the "L" line (with a few exceptions such as the 85/1.8, 100/2, and 50/1.8
and /1.4), but that doesn't mean those more costly lenses won't produce perceptibly
better results even with negatives destined to be printed on a minilab machine--and
there's nothing to prevent those same negatives from being printed at a higher level
of quality should the photographer choose to do so, in which case the quality of
lens used can be even more important.
Craig
*
****
*******
***********************************************************
* For list instructions, including unsubscribe, see:
* http://www.a1.nl/phomepag/markerink/eos_list.htm
***********************************************************
