EOS FAQ v2.1 (2/6)
robert.m.atkins
check the settings if lighting conditions change.
Note that all bodies after the 630/RT do not support exposure
compensation in manual mode. Manipulation of film speed is
required. (This includes the EOS 1.)
ACS
3.10. A-TTL Questions:
3.10.1. Why doesn't Canon use distance information for flash
exposure calculations?
Distance info is only useful so that the camera can apply the
"guidenumber" formula: f-stop = GN/distance. This is only a rough
approximation. Off-The-Film (OTF) metering measures the true
exposure of the film, which is much more accurate than the guide
number calculations.
In A-TTL, Canon uses a preflash to *actively* judge subject
distance rather than passively. This factors in light loss from
bouncing off different kinds of surfaces (for example, flash
modifiers), and light *gains* from bouncing off different kinds of
surfaces (like ceilings and walls). You simply cannot make the
remotest use the GN calculations for anything but direct flash.
Personally, I avoid direct flash, and most light modifiers are
bound to void the GN calculations, either because of light loss, or
because they require the head to be up (like the Lumiquest pocket
bouncer). So for most shots, the distance information would have to
be thrown away.
ACS
3.10.2. Is it OK to block the A-TTL sensor (with a softbox,
say) with
using plain TTL flash?
It should be fine. TTL metering stands for "Through The Lens"
metering. So, no metering takes place from the flash sensor. The
sensor on the flash is only used for A-TTL metering. If you're
still worried, just shoot some test exposures to make sure [I'm not
sure if TTL exposure compensation works.]
ACS
3.10.3. Is it OK to block the A-TTL sensor (with a softbox,
say) when using A-TTL?
The sensor on the flash is required for A-TTL metering. That's when
the 430 EZ emits a preflash, checks the return signal with its own
sensor, and then computes an aperture that will work. It's best to
think of the A-TTL preflash/sensor combination as a kind of
reflected light flash meter built into the flash itself (and it's
supposedly quite accurate too).
By blocking the sensor, the A-TTL program will definitely get
confused. Similarly, if the A-TTL supporting flash is on a flash
bracket it may give slightly different apertures if the sensor
isn't pointed in the right direction.
ACS
3.10.4. Why does blocking the A-TTL sensor give me *smaller*
apertures?
Some have tried fooling the A-TTL programme by blocking the sensor
and found that it stops down more instead of less! This is because
the IR preflash head is right beside it, and whatever you're using
to block the sensor is probably in front of the IR emitter as well.
So basically, you've put a reflector in front of the
emitter/detector pair.If you carefully cove just the sensor, you
will get the anticipated result; the flash will try to select the
maximum aperture. The easiest way to avoid this is to tilt up the
flash head so that the preflash exits from the main flash head.
ACS
3.10.5. Why can't I use A-TTL with the off-camera TTL accessories?
When you are using the TTL Hot Shoe Adaptor 2 with the Off-Camera
Shoe Adaptor, you lose A-TTL. The A-TTL assumes the flash is in
the camerashoe, or at least near it. It can be wildly inaccurate
when far off camera, so it is disabled when using the off-camera
TTL accessories. There is also the problem of multiple flashes
returning conflicting information on the data bus. [Again, I don't
know how flash exposure compensation is affected by this. Probably
doesn't work unless you are using an EOS A2/5.]
ACS
3.10.6. Can A-TTL cope with filters over the flash?
If you're using a modifier that goes over the main flash head
(filters,Sto-Fen Omnibounce, etc.), note that the preflash in
direct mode comes from the IR sensor below the main flash head.
This can confuse the A-TTL programme since the IR preflash will
not be filtered, giving inaccurate results.
By tilting the head up slightly, or using something to release the
bouncelatch, the 420EZ/430EZ will assume you are bouncing the flash
and will fire the preflash through the main head. This should
yield more accurate results. If the head is being tilted up,
perform tests to confirm evenness of flash coverage. The flash may
require manual zooming to a wider angle to give better coverage.
ACS
3.10.7. Can A-TTL be used with filters over the lens?
The A-TTL sensor, being in the flash and not in the camera may
yield bad apertures if filters are being used over the lens. In
this case, you'll have to find some way to put a similar filter
over the sensor.
ACS
3.10.8. Is the A-TTL program affected by flash exposure
compensation?
In fiddling with my 430EZ, it seems as though it doesn't figure the
exposure compensation into the selected aperture. If you say you
want flash overexposure by 3 stops, it gives the same aperture as
selecting flash underexposure by 3 stops. So, if you really like to
overexpose (for whatever reason), you might not get the expected
result.
This appears to be an oversight on Canon's part. Use manual mode if
lots of flash overexposure is required.
ACS
3.10.9. Why didn't Canon put the A-TTL sensor in the body?
If we assume that A-TTL requires a separate sensor, then there are
a few reasons that Canon may choose to avoid putting the sensor in
the body itself:
1) Cost:
Including an extra sensor would incur an extra cost on every body
manufactured. This could make Canon's bodies marginally less
competitive. This is especially true if the majority of Canon
owners do not wind up using A-TTL.
2) Complexity:
Cramming a 3rd sensor into the body would make the body more
complicated in order to split the optical path yet again in order
to feed the A-TTL sensor.
3) Dim viewfinder:
Bleeding off more light to another sensor would make the viewfinder
dimmer than it already is.
4) History:
A-TTL was developed for the T-90 which did not have a multi-segment
metering sensor. Thus, it was unable to distinguish foreground
exposure from background exposure during the preflash.
Personally, I'm inclined to believe #4, and that Canon has simply
neglected to update its flash system. It is technically quite
feasible to use the existing evaluative metering sensor to perform
preflash exposure computations. [Rumour has it that Canon is
currently throwing around ideas for a new generation of flash
system. Cross your fingers, folks!]
ACS
3.11. Is there a faster way to mount the flash?
The locking pin on the 430EZ is spring loaded (to be compatible
with bodies that didn't support locking pins). If you leave the
locking pin down at the same level as the hot shoe contacts when
the flash is not mounted, you can quickly and securely mount the
flash by just slipping it into the shoe. The pin will rise like the
contacts when being inserted into the shoe, and then pop down into
the little hole when the flash is properly seated. This trick is
especially useful for bodies without a built-in flash.
ACS
3.12. How can I find out more about using flash with my EOS?
Canon publishes the "Speedlite Reference Guide". Stephen Page
posted the following summary:
>From: sdp...@andersen.co.uk (Stephen Page)
Many thanks to someone (sorry, I've lost the posting!) who passed
on a tip about the Canon USA booklet, "Speedlite Reference Guide"
(Chuck Westfall, 1991). This is an excellent little book and it
contains some essential information which Canon have not bothered
to tell the world in their instruction manuals.
The booklet can be obtained from Canon USA, One Canon Plaza, Lake
Success, NY 11042, USA; telephone +1-516-488 6700. They sent mine
immediately and there was no charge.
I've summarized some key points below which may help others. Do get
the book, though; it's got lots more of the same. Also write to
Canon and blast them for a pathetic user manual... I cannot see
how anyone could work out how to use the flash correctly without
this book. My summary is valid for EOS-1/430EZ; other models have
slight differences. Early models of Speedlites work differently.
[.Anyone know if Canon have any other hidden reference guides from
national offices in the US or elsewhere? ]
1. Types of flash exposure.
In "normal" flash exposure, the exposure is calculated for the
foreground subject and the background is underexposed.
In "fill-in" flash exposure, the shutter speed and aperture are set
to expose the BACKGROUND correctly. The flash duration is adjusted
to expose the subject correctly. The Speedlites operate in fill-in
mode most of the time (see below).
2. Flash exposure and the camera modes.
In Program mode, the camera will use fill-in exposure provided that
the light level is EV10 or above. If the ambient light is below
EV10, the shutter speed will be locked at 1/60 sec and the flash
will work in "normal" mode, i.e. the background will be
underexposed.
In Aperture-priority (Av) or Shutter-priority (Tv) modes, the
camera will ALWAYS use fill-in exposure mode. The exposure will be
set to match the background (ambient light level), so if the
ambient light is low then a slow speed and/or large aperture will
be selected.
3. Tinkering with the light levels.
When the ambient light is above EV10, the Speedlite's flash level
is reduced on a linear scale to a maximum of the equivalent of 1.5
EV steps less than the standard exposure level (at EV13 and above).
This is to preserve the natural lighting a bit, by not blasting it
away with flash.
4. Exposure compensation.
The exposure compensation controls on the camera affect BOTH
background and foreground exposure.
To adjust the foreground independently of the background, use the
control on the flash (430EZ) itself.
5. A-TTL vs TTL.
The Speedlites can be switched to use "TTL" rather than the default
"A-TTL" mode. Some flashes, e.g. ring flash ML-3, only use TTL
mode.
TTL mode disables the preflash (near-infrared in normal flash head
position, white in bounce position). This has the advantage that it
doesn't make the subject think that the picture has been taken, and
it doesn't set off any slave units. However, in TTL mode you get no
pre-exposure distance measurement, so the camera can't scream if
you are out of range. You can work it out by using the distance
scale, or by looking at the indicator light AFTER the flash.
I hope this helps some of you to work out the mysterious behaviour
of your equipment!!
Stephen Page, sdp...@andersen.co.uk or uknet!aclondon!sdpage
4. Lenses:
4.1. Lens Specifications?
For a comprehensive chart of Canon lens offerings, please consult
an EF lens guide, such as the one found in the "EF Lenses"
brochure. Per Danielsson has posted data on many EF lenses and his
table is given in Appendix B at the end of this FAQ.
Alvin Shih has also abstracted information on EF lenses and his
table is given in Appendix C at the end of this FAQ.
The following lenses were announced in the USA on Feb 17th 1993.
The information below was supplied by Chuck Westfall of Canon:
Canon U.S.A., Inc. announced 3 new EF lenses at the PMA (Photo
Marketing Assn.) dealer trade show occurring this week in Atlanta.
The new lenses are: EF 20-35mm f/3.5-4.5 USM, EF 50mm f/1.4 USM, &
EF 400mm f/5.6L USM. Here are the basic specifications:
Specification 20-35/3.5-4. 50/1.4 400/5.6L
AF Drive Ring USM Micro USM Ring USM
Construction
(Groups/Elements) 11/12 6/7 6/7
Min Focus Distance 1.1 ft/0.34m 1.5 ft/0.45m 11.5 ft/3.5m
Length 2.7 in/69mm 2.0 in/50.5mm 10.1 in/257mm
Weight 12 oz/340gr 10.2 oz/290gr 43.8oz/1.25kg
Filter Size 77mm 58mm 77mm
List Price $725 TBD TBD
The EF 20-35mm f/3.5-4.5 USM was designed as an easy-to-use
wide-angle zoom lens at a more affordable price. The new lens
features internal zooming and focusing to help maintain its compact
size and light weight, while providing a non-rotating filter thread
for use with a circular polarizer or other special effects filter.
The USM provides quick, silent AF with a full-time manual focus
override.
The EF 50mm f/1.4 USM meets the needs of a wide range of advanced
amateur and professional photographers, due to its large aperture,
light weight and compact design. The optical quality of the new
lens exceeds that of the FD 50mm f/1.4, with even greater contrast
and resolution. The new lens is equipped with Canon's latest Micro
USM focus drive for high-speed, silent AF, with a full-time
mechanical manual focus override for added creative control. Other
features include a metal mount, distance and depth of field scales
and an IR index.
The EF 400mm f/5.6L USM is targeted to professional and advanced
amateur photographers who require a compact, lightweight and
easy-to-handle long telephoto lens for a variety of outdoor
applications. The lens uses a rear-focusing optical system and
Canon USM (Ultrasonic Motor) technology for silent, high-speed
autofocusing. The EF 400mm f/5.6L USM has 7 elements in 6 groups,
including two elements of Canon's Ultra-Low Dispersion (UD) glass.
This is the first EF telephoto to use Canon's newly-developed
"Super UD" glass. The new glass features dispersion characteristics
similar to fluorite, producing outstanding resolution, contrast and
color fidelity.
A full-time mechanical manual focusing mechanism makes it possible
to switch to manual focusing at any time without switching out of
AF mode. Dual rubber manual focusing rings are positioned on both
the large and small sections of the lens barrel to provide improved
operability. A sliding lens hood is built-in and the new lens is
equipped with a focusing range selector switch and a detachable
tripod collar. The EF 400mm f/5.6L USM is compatible with both EF
Extenders for autoexposure and manual focus.
The EF 20-35mm f/3.5-4.5 USM will be available through authorized
Canon USA dealers by the end of March. The EF 50mm f/1.4 USM & EF
400mm f/5.6L USM will be available by the end of June, with pricing
determined at that time.
RMA
4.2. Why does Canon offer so many different lenses, some of
which seem quite similar?
Though Canon offers lenses of similar focal lengths, they are
targeted at different audiences. Canon has 3 broad categories of
lenses: A, B,and L.
The "A-series" consists of low-cost lenses for the "snapshooter"
market.The original A-series consisted of lenses which were AF-only
for use on the 700 and 800 series. They had no manual focus
capability of any sort, and also lacked focusing scales. These
lenses marked with an "A" after the focal length and aperture
designations. For example, "100-200/4.5A".
The new "A-series" does allow for manual focus, and may even sport
USM, but still lack focusing scales, and are not designed to be as
durable as other EF lenses. They no longer have an "A" in their
lens designation. Indeed, the lens mounts of all but the longer
lenses are polycarbonate instead of metal. The optical performance
is reasonable for the money.
At the other end, there is the "L" series. The "L" designates
"Luxury" (or as Mike Coren would say, "Life's savings" :-). These
lenses are targeted towards professionals or advanced amateurs.
L-series lenses typically include special elements of UD or
Fluorite glass for outstanding optical performance. Some lenses of
this series are also designed to withstand the abuse which comes
with professional photography (photojournalism in particular). They
are generally sold complete with case and lens hood.
Any lenses which fall in between are simply in the "B-series".
These lenses are generally reasonable value for the money and
should serve the typical "enthusiast" quite well. They all include
manual focus capability, and distance scales as well.
Outside of the three major categories, there is the class of
"special-purpose" lenses. These include the TS-E lenses (the 24mm
is an "L", but the 45mm and 90mm are not), as well as the macro
lenses.
ACS (RMA)
4.3. Can I use my FD-mount lenses on EOS bodies?
There is a mount converter from EOS to FD, but it is a MACRO
converter only! That is to say, it does not allow the fitted
FD-mount lens to focus to infinity. In addition, it does not work
with cameras that do not have interchangeable focusing screens, and
certain FD lenses might not fit! Ask Canon for the list of
incompatible lenses before investing in an FD-macro setup.
For members of Canon Professional Services ONLY, there is a 1.26x
teleconverter for 200mm+ lenses which also functions as a mount
adaptor. However, this costs 2/3 stop, and only stopped-down,
aperture priority metering is available. (No autofocus, of course.)
ACS
4.4. Can I use EF-mount lenses on FD-mount bodies?
No. First, the EF mount is *bigger* than the FD mount. Second, the
adaptor would have to convert the mechanical aperture setting
action for the old FD lenses into *electrical* information for the
EF lenses. Since you can't stop down an EF lens mechanically, even
if you had stopped-down metering, you'd still be stuck. It would
require quite the contraption to make EF lenses work, and is
probably not worth anyone's effort to research and develop.
ACS
4.5. Why have the focusing motor in the lens?
Having the focusing motor in the lens allows the lens to have a
motor that is "matched" to its focusing characteristics. Heavier
glass can be supplied with more powerful motors. It should also
simplify lens design since there is no need to design a
transmission that takes mechanical energy from a fixed point in the
lens mount to some other point in the lens. With the motor in the
lens, the motor can be placed where ever it will be most effective.
ACS
4.6. Doesn't the motor in the lens make the lenses more expensive?
In some cases, it probably does. If the retail prices of
third-party lenses is any indication, the motor does cost
something (comparing a 3rd-party Nikon AF mount lens to its Canon
EF mount equivalent). However, the lenses from Canon are generally
regarded as faster focusing, so it may be worth the extra money.
ACS
Note that some recent lenses (e.g. Sigma 70-210/2.8 APO) are priced
exactly the same in manual and all AF (Nikon, Minolta, Canon)
mounts, so maybe the 3rd party manufacturers have refined their
design and manufacturing costs to the point where the cost of the
internal lens motor is not significant.
RMA
4.7. Manual focusing with USM lenses?
There are two varieties of USM lenses. One variety, commonly found
on some of the "L" series lenses, converts rotation of the manual
focusing ring into electrical pulses, which in turn are converted
into motor movement. On such lenses, the ring is referred to as an
"electronic focusing ring". The benefit of this system is that some
of the "L" lenses can be set to different focusing ring
sensitivities to allow for speed or precision as the user requires.
By default, the electronic focusing ring is disabled after AF since
leaving it enabled draws significant amounts of power from the
battery. However, on some bodies, the electronic focusing ring can
be left enabled via CF after AF for "fine tuning" of focus by the
user.
Most other USM lenses have a mechanical coupling that works after
AF regardless of the custom function setting. Though this draws no
battery power, the angle of rotation from close-focus to infinity
is usually regarded as too short.
ACS
4.8. Can small animals hear USM lenses?
In a survey taken by Mike Coren, it would appear that some animals
can detect USM noise at close range, but are not annoyed by it.
Indeed, some animals may be even curious. However, there was one
report of a cat who is particularly annoyed by the 28-80/2.8-4L.
This may be an exception.
The lenses in the sample population were:
85mm f/1.2 L - 1
300mm f/4.0 L - 1
28-80mm f/3.5-5.6 - 2
28-80mm f/2.8-4.0 L - 1
28-105mm f/3.5-4.5 - 1
35-105mm f/4-5.6 - 2
70-210mm f/3.5-4.5 - 2
75-300mm f/4.0-5.6 - 1
100-300mm f/4.5-5.6 - 1
4.9. Is there any way to focus a USM lens without making any
ultrasonic noise?
On USM lenses with electronic focusing rings, it appears not. On
USM lenses with mechanical couplings, use manual focus.
ACS
4.10. Canon lenses look too expen$ive. What about third-party?
Because of the built-in motor, 3rd party lenses for EOS are
generally more expensive than their non-EOS counterparts.
Canon has been pretty secretive about their lens interface. There
have been a few postings to the net implying that Sigma lenses
don't AF very well on Canon EOS cameras, or if they do, they only
do so for a limited time before failing.
Generally, Canon lenses are of similar or better quality in
comparison to similarly priced competitors. It is probably better
to assume that you will be staying with Canon lenses.
There are rumours that the Tamron 28-200/3.8-5.6 uses licensed
technology to ensure correct operation on EOS bodies. These rumours
are unconfirmed.
ACS
4.11. Why didn't Canon think of using distance information
(like some other nameless company :-)?
With the release of the EOS 10s in 1990, the following USM lenses
were introduced and return focus distance information:
EF 35-135mm f/4.0-5.6
EF 70-210mm f/3.5-4.5
EF 100-300mm f/4.5-5.6
EF 100mm f/2.8
I imagine all USM lenses since then include focus distance
decoders. However, Canon doesn't claim to use the information for
anything. (Look in the "Flash" section on why distance information
is not too useful.)
ACS
4.12. So what are affordable Canon lenses worth getting?
A list of lens ratings given by the French photo magazine "Chasser
d'Images" and was submitted to the FAQ by Christian Decurnex. It
can be found at the end of this FAQ in Appendix F. This is probably
the most complete survey of Canon EF series lenses. Many thanks to
Christian for sending it in to the FAQ.
Christian has also abstracted the Chasseur d'Images test of the
EF20-35USM lens, and his posting can be found in Appendix H at the
end of this FAQ
###
George Lepp in his "Natural Image" newsletter tests lenses from
time to time. Most of the tests are on Canon and Nikon lenses, but
he also tests 3rd party lenses as well as a few from Minolta,
Pentax etc. He also tests cameras and film and the "Natural Image"
carries articles on all aspects of nature and wildlife photography.
For subscriptions and back issue information contact Lepp and
Associates, P.O. Box 6240, Los Osos, CA 93412 or call (805) 528-
7385. His lens rating scheme is from 0 to 10, with 6.0 representing
the bottom end of his "professional quality" range. 6.0 corresponds
to somewhere around 60 lp/mm. A rating of 10 would usually be for
the best macro lenses, sharp from edge to edge and resolving 130
lp/mm+. None of his tests on any of the Canon lenses have given
numbers less than 6.0. The "regular" lenses usually are in the 7-
7.5 range, while the good "L" series lenses rate in the 8-9 range.
The following numbers are *MY* attempt at averaging *his* test
numbers. He usually tests at several focal lengths (for zooms) and
several apertures. The numbers given here are my averages of his
numbers over various focal lengths (for zooms) and over several
tests (for those lenses tested more than once). Numbers are given
for wide-open (W-O) and at f5.6. Mistakes are mine, not George
Lepp's. For more information see the address and phone numbers
above. Note that you cannot sum up the qualities of a lens in a
single number, so take these numbers as a very general guideline
only. Get the original tests for more information.
Lens W-O f5.6
EF300/2.8L* 8.0 9.0
EF300/4L* 8.0 8.5
EF200/2.8L* 7.5 9.0
EF500/4.5L 7.0 7.5 (est)
EF100-300/5.6L* 8.5/8.0 8.5/8.0
EF100-300/4.5-5.6USM* 7.0/7.0 8.0/7.0
EF70-210/3.5-4.5* 7.0/7.0 8.5/7.5
EF35-350/3.5-5.6L 7.0 7.5
EF80-200/2.8L* 7.5/7.5 9.0/9.0
EF28-70/3.5-4.5II average "overall" rating 8.25
A/B refers to ratings at the short (A) and long (B) ends of the
zoom range.
Many of these lenses (marked "*"), along with Canon, Tamron and
Vivitar teleconverters (see section 4.16) were tested in issue #28
of "The Natural Image" (Winter 1991). Also in the same issue are
tests on similar Nikon lenses, so you can get ammunition for the
Canon vs. Nikon wars! For directly comparable lenses, the Canons
seemed to have a slight edge in sharpness. On repeat tests some
sight variations have appeared. For example in earlier tests, at
300mm and at f5.6, the 100-300L has tested as high as 8.5, while
the 100-300USM has tested as low as 6.5. At the level at which
ordinary mortals (without access to a full optics testing lab) can
test lenses, such small differences are to be expected from test to
test. George has commented though, that in general, the test
results are quite repeatable.
From his test results George has stated that the Canon 300/2.8L and
the Canon 100-300L are the sharpest lenses in their category from
any manufacturer.
Of notable non-Canon lenses, the Sigma 400/5.6 APO rates a 6.5 wide
open (f5.6), and the Sigma 500/7.2 APO rates a 5.5 wide open
(f7.2), and a 6.0 at f8 (1/3 stop down). The 400/5.6 APO represents
good value for money and can yield quite acceptable images. The
Canon 400/5.6L has not yet been released and there are no tests to
date (5/93). I would expect it to significantly outperform the
Sigma lens, but at maybe 3 times the cost.
Again, use these numbers with some thought. They are general
guidelines only and should not be taken as the definitive word on
the quality of these lenses. Sharpness is not the only factor in
choosing a lens (though for me it is of major importance). George
Lepp's rating are weighted strongly by central image resolution
and contrast.
RMA
###
Popular Photography has also tested quite a large number of Canon
EF series lenses (and other 3rd party EOS compatible lenses). The
tests are too detailed to summarize here. Check out back issues
(some libraries may have them). A brief (entirely insufficient)
one-letter summary of their full page tests follows. A = Well above
average, B = Above average, C = Slightly above average, D =
Average. Note that "Average" doesn't really mean average in Popular
Photography, since almost no lenses are below "average". It
probably means "acceptable" in some general sense! Note that there
are many inconsistencies in the Pop Photog tests. The numerical
data does not always agree with the written comments. Look at the
tests in detail (especially the hard data) rather than relying on
their (or my) comments.
EF85/1.8 USM April 1993 A
EF14/2.8L USM January 1993 D
EF35-105/4.5-5.6 January 1992 A(35) to D(wide open at 105)
EF35-135/4-5.6 February 1991 B
EF75-300/4-5.6 August 1992 A(75) to D(300)
EF50/1.8 February 1991 B
EF50/1.8II May 1991 A
EF28-80/3.5-5.6 July 1992 A (B at 80)
EF35-350/3.5-5.6USM June 1993 B to D(at 350)
Interesting Non-Canon Lenses
Sigma 400/5.6 APO May 1992 C
Sigma 21-35/3.5-4.2 February 1992 A
Sigma 70-210/2.8 May 1993 A+
Sigma 28-70/2.8 June 1993 B to D(70mm wide open)
Note however, that the Sigma 400/5.6 actually tested as a 386/6.2.
This does not make it a worse lens, but you should be aware of what
you are likely to actually get if you buy one. Similarly almost all
long lenses and zooms over 200mm are actually significantly shorter
(at the long end for zooms) than they are marked. You can consider
yourself quite lucky if they are within 5% of the marked focal
length, for example the Sigma 70-210/2.8 zoom is actually only
195mm at the long end. This shortfall does not only apply to 3rd
party lenses, but to lenses from almost all the major camera
manufacturers.
A much more detailed summary of all the Popular Photography tests
on lenses which are available in an EOS mount has been submitted to
the FAQ by Leszek Mazur. This summary appears at the end of this
FAQ in Appendix D.
RMA
###
The Sigma 70-210/2.8 (or 70-195/2.8 to be more accurate) does look
like a good optical performer at 1/2 the cost of the Canon 80-
200/2.8. The following item from "Fotomagazin" submitted by Uli
Bodenhausen
From: u...@i13d5.ira.uka.de (Uli Bodenhausen)
Subject: f2.8/telezooms tested in Fotomagzin
Organization: University of Karlsruhe, FRG
The German Fotomagazin (3/93) tested the f2.8 AF telezooms.
The summary is (without permission):
Ratings are from 0(bad) - 10(excellent).
Attention:
These results may not be directly comparable to other ratings of
other zoom lenses in other Fotomagazin tests. The small print in
the test says that the ratings are based on sharpness, contrast,
vignetting, flare, distortions and color. They also say that the
sharpness measurements allow an equal comparison between all lenses
(also in other tests they did). BUT: The contrast ratings are only
comparable between lenses with the same opening (f stop). The
following results should be directly comparable because all have
the same opening. Only one lens of each manufacturer was tested.
Lens rating rating
optical mechanical
Canon AF 80-200 9.2 9.0
Minolta AF APO 80-200 8.8 8.8
Nikkor AF ED 80-200 9.6 9.2
Sigma APO 70-210 9.6 9.2
Tamron AF LD 70-210 8.8 9.4
Tokina AF ATX 80-200 9.0 8.8
###
The following lenses are mentioned on the net repeatedly for being
able to produce good quality images at a reasonable price (maybe
this info should go into the EF lens guide???):
EF 28-70/3.5-4.5 II:
This lens boasts an aspherical element to improve its performance
when shooting wide open. It is regarded as being one of the best
lenses in its focal length and slightly better than the
28-80/3.5-5.6 USM that replaces it. The factors against this lens
are: a) the front element rotates, b) the filter threads retract at
middle focal lengths, c) the filter size is 52mm rather than the
58mm that Canon seems to now favour, d) they are becoming very
scarce. [e) Note that it is now discontinued, so if you want one
and you see one for sale, buy it immediately! (RMA)].
EF 75-300/4-5.6:
Optical performance quite good for the money. Main complaints:
looks a bit flimsy, not the fastest focusing, no distance scales,
and filter threads rotate. It is now available with a USM motor
which provides much faster (and quieter) focusing.
EF 100-300/4.5-5.6 USM:
Noted for its fast focusing, it is supposed to be "quite good but
not outstanding". The main complaint is that it suffers from "zoom
creep".
EF 100-300/5.6L:
This is supposed to be the sharpest lens in its focal length, and
the cheapest of the "L"-series zooms (nearly half the price of the
80- 200/2.8L). Main weaknesses are that it has a push-pull zoom
mechanism, it's slow focusing, the front element rotates, and the
maximum aperture is quite modest. [I have measured better than 80
lp/mm resolution at 300mm wide open (f5.6) on this lens, which is
indeed outstanding performance (RMA)]
ACS
4.13. What does the "II" mean?
Some lenses carry a designation "II" after the focal length and
aperture information. The "II" simply indicates that the lens has
the same focal length and aperture specifications as a previous
lens.
Though the 50/1.8 II is of inferior construction than the original
50/1.8 (plastic mount, no distance scales), the 28-70/3.5-4.5 II
retains the metal mount and distance scales of its predecessor.
However, I do believe that the 28-70 II is optically identical
to its predecessor and that the major change was the replacement of
metal gear components with plastic equivalents. The supposed reason
was to increase the smoothness of the focusing and zooming actions.
[Anybody have an original 28-70/3.5-4.5 who can comment?]
ACS
4.14. What about the Tamron 28-200/3.8-5.6?
Here is a review by Robert French:
From: rfr...@Xenon.Stanford.EDU (Robert S. French)
Subject: Review of Tamron 28-200/3.8-5.6 AF lens for EOS
Organization: Computer Science Department, Stanford University.
Date: Tue, 2 Feb 1993 07:09:00 GMT
As promised, here is my review of the new Tamron AF 28-200/3.8-5.6
for the Canon EOS.
First impressions:
POSITIVES:
- The lens is very short (approx 3 3/8" at minimum zoom)
- The lens has a very positive feel to it - the zoom action dial
(which is most of the lens body) has a reasonable amount of
friction and turns through 45 deg from 28-200mm.
- The lens appears to be very well constructed. It feels rugged
and has a metal mount.
- The minimum aperture at 28mm is f/22, and at 200mm is f/32.
- It comes with a bayonet-mounted lens shade that can be
reverse-mounted for storage.
- The freedom of 28-200 zooming is wonderful.
NEGATIVES:
- The lens weighs a lot more than the claimed 15.2 oz. It feels
like it weighs over 20 oz, but I don't have a scale that measures
over 16. I don't know why the lens weighs so much more than claimed
- the 15.2oz figure is printed on the box and on the info sheet. It
takes a little getting used to after using the extra-light Canon
35-70 and 80-200, but it's not bad.
- There is no focus scale! So much for hyperfocal calculations...
- The minimum focusing distance is ~7 feet at all focal lengths!
- The zoom dial turns in the opposite direction from the 35-70 and
80-200, and I find that very confusing.
- The end of the lens rotates with focusing, so
orientation-specific filters will have problems.
- The focusing motor is fairly slow compared to Canon EF lenses,
and very slow compared to USM lenses.
- At a maximum aperture of 3.8 (which the Canons read as 4.0), the
lens is relatively slow at its wide end, but consistent with most
zooms on the market at the tele end. This doesn't mean it isn't
slow, though.
- 72 mm filters are expensive.
After shooting a roll of film (Kodachrome 200):
The lens appears to perform very well. It appears that the lens
shade DOES interfere with 28mm shooting - there is a slight
darkening around the edges. There is no visible vignetting at 28mm
without the lens shade. The lens also worked well with the Tamron
2x teleconverter. All in all, I like the lens, although it wasn't
the perfect lens I thought it was going to be. The long minimum
focus distance, slow maximum aperture, and relatively slow focusing
motor all make it somewhat annoying. It seems well suited for
outdoor photography, although I'm going to keep some other lenses
around for inside and special outdoor projects. The 28-200 ability
makes much of the problems tolerable, though. I think I'll keep it.
Rob
4.15. My lens cap is too loose. What can I do?
There are a few solutions:
a) Buy a "cap keeper" gizmo that strings your cap to your lens.
(Most people don't like this idea because it reminds them of
"mitten strings"...)
b) Pop the spring mechanism out of the cap, heat with a hair drier
while applying force to mold the ring into the desired shape. An
elastic band or two may be useful.
c) Just screw the cap into the filter threads. My caps stay snugly
attached this way!
d) Buy another (brand) of lens cap! (RMA)
ACS
4.16. Tell me all about teleconverters!
Canon make two EOS compatible teleconverters, a 1.4x and a 2x. They
will only mount on EF telephoto lenses of the "L" series since they
have a protruding front element which prevents mechanical coupling
to (almost) all the other EF series lenses. Even if you could mount
them on other lenses, the results might not be good. The Canon
teleconverters are obviously designed to provide an optimum match
to the optical characteristics of their telephoto lenses. The Canon
teleconverters convey the true aperture of the lens/converter
combination to the camera body, i.e. a 300/2.8 with a 1.4x attached
to it will show a maximum aperture of f4 on an EOS body. Both
converters can be mounted on any EF series lens if you use an EF25
extension tube. This is, of course, only useful for macro work. To
get AF operation with a 1.4x converter you must use a prime lens of
f4 or faster. f4 will become f5.6 with the converter attached.
Similarly to get AF operation with the 2x converter you must start
with a lens that is f2.8 or faster. f2.8 will become f5.6 with the
converter attached. With lenses slower than the values specified
above the camera will not even TRY to autofocus. Thus unfortunately
you cannot get AF operation with the 500/4.5L and a Canon 1.4x
converter.
Tamron, Sigma and Vivitar all make EOS compatible AF teleconverters
(there are also a few "store" brands). These converters will mount
on any EF series lens, or any 3rd party EOS compatible lens. They
do NOT convey the true aperture of the lens/converter combination,
i.e. If you start out with a 300/2.8 and add a 3rd party 1.4x
converter, the EOS body will still allow you to set an aperture of
f2.8, when your true maximum aperture is f4. This will not affect
exposure accuracy however since the TTL metering will take care of
things for you. If you use a hand held meter and transfer readings
to the camera in manual exposure mode you will have to remember to
make the exposure correction yourself. The only problem I can see
is in situations where the lens aperture is used to calculate other
things, such as depth of field or flash range. The Tamron 1.4x
converter claims to provide AF with any lens faster than f4.5 (vs
f4 for the Canon 1.4x). I don't know how accurate this is. It
certainly seems to work fine with an EF300/4L lens. The Tamron 2x
claims to provide AF with any lens of f4 or faster (vs. f2.8 for
the Canon). This is not really true. It does provide some AF when
used with an EF300/4L, but not 100% reliable AF. Under average
daylight conditions with average targets it probably gives
something like a 65% success rate. The reason Canon's specs are
tighter is that they quote numbers where you can expect to rely on
AF operation, not that their converters are not as good (they are
at least as good, and probably better). From some brief
measurements I have made, it looks like both the Canon and Tamron
1.4x converters are indeed very close to 1.4x. The Canon 2x seems
to be very close to 2x, but the Tamron 2x seems a little less
powerful than stated, perhaps closer to 1.8x. This may not be
altogether bad, but it is as well to be aware of it. The difference
should give you a little more aperture, but at the expense of a
little less focal length.
Converter quality. Canon's are the best, but also the most
expensive. George Lepp (in his "Natural Image" newsletter) tested
a number of converters and found the Tamrons to be the best of the
3rd party brands (Tamron, Sigma, Tokina and Vivitar tested with
Nikon, Canon and Sigma lenses) most of the time. They were close to
the Canon converters in center resolution, but were not as good at
the edges of the frame. I have found a similar result comparing
images taken with my Tamron converters with those taken by Jay
Schlegel using the Canon converters. The biggest difference seems
to be between the Tamron 1.4x and Canon 1.4x at the frame edges.
Central resolution (used with a Canon EF300/4L) was good in both
cases at 70+ lp/mm. However the resolution and lateral color at the
edges of the frame were better with the Canon converter. There
seemed to be less difference between the Canon and Tamron 2x
converters in terms of image quality. Contrast with all the 3rd
party converters may also be lower, and they may introduce more
chromatic aberration. They cost less than the Canons, so you have
to decide if you can put up with the lower quality images, though
if you intend to use them with anything other than "L" series
telephotos then the Canon converters are not an option anyway. In
general, converters don't work well with low cost zoom lenses. They
can give acceptable results with some higher quality zooms (e.g.
EF80-200/2.8L) and can give quite good results with high quality
telephotos, especially if you stick to the 1.4x converters. In
general, avoid any 3x converters or "store brand" converters since
the quality of images they produce is not likely to be acceptable.
Converters are a compromise of size, weight, cost and quality. Only
YOU can decide if the quality of the images they provide is
acceptable. Do not expect miracles, and do not use them in critical
applications (like on a one-in-a-lifetime safari) without testing
them on the lenses you will use them with to see if you can live
with the results. A converter which gives great results with one
lens, may not be all that great with another, so test all the
combinations you intend to use.
One final point, despite anything you may read elsewhere,
teleconverters can work no "magic" with respect to depth of field.
Whether you have a 400mm f5.6 lens, or a 200mm f2.8 with a 2x
converter (=400/5.6) you get *exactly* the same depth of field in
each case.
RMA
The following additional information on Canon EF series
teleconverters was supplied by Chuck Westfall of Canon: