Spider in the tank R7 with RF-S 10-18mm hand held |
Canon introduced their mirrorless RF mount cameras and
lenses in 2018 with a range of full frame models.
At the time Canon had a catalogue of crop sensor DSLRs with
EF-S lenses and the separate EF-M system with its own list of bodies and
lenses.
It seemed inevitable that Canon would eventually produce a
range of crop sensor bodies and lenses using the RF mount. But they had to
slowly wind down production of EF-S
DSLRs which were still popular
and EF-M bodles and lenses some of which were also very popular, to make
way for a new mirrorless RF-S system.
In the event the first two RF-S crop sensor cameras, the R10
and R7 were announced in May 2022
together with just two dedicated RF-S lenses, the RF-S 18-45mm basic kit
zoom and the RF-S 18-150mm advanced kit zoom using the same optical design as
the erstwhile EF-M version.
I bought an R7 with the 18-150mm lens when it became
available in Australia and used it for several months on a variety of subjects.
But I was not well pleased with the
unique location of the rear control dial, high ISO files were noisy and I
encountered some autofocus issues. There was prominent rolling shutter
distortion with the E-Shutter and loss of sharpness in some situations due to
shutter shock with the mechanical shutter. I got annoyed by all this, sold the
R7 and concentrated on full frame gear for the next year or so.
In the interim several things changed. I got older and less
inclined to carry anything large or heavy. The R7 got several firmware updates
and Adobe released their Denoise AI which I discovered does a remarkably good
job of removing digital noise from high ISO files while retaining detail. I decided that I had probably over-reacted to
the odd placement of the rear control dial. In addition Canon released two more
RF-S lenses, the 55-210mm utility zoom and the ultrawide 10-18mm zoom.
So I bought another R7 and have been using it for the last
several weeks.
R7 with RF 100-400mm Some photographers avoid this kind of stacked perspective but I like it |
This time around, the experience has been more positive to the extent that I am considering making the R7 my only camera.
But before making this decision I need to discover whether
the negatives about crop sensor gear in general and specifically the R7, can be
managed or for which workarounds can be found.
The first question is: Why not stay with full frame ? Why
consider crop sensor at all ? The
answer to this is a personal one. I just don’t care for the size and mass of
full frame kit, especially the lenses. Yes, I know Canon has a range of compact
full frame lenses of which I have several. Most of these are very good but if I want an even more compact kit that
means going down to APSC.
I should say here that if I was a professional sports or
wedding photographer none of this discussion would be relevant. The most
important thing which full frame cameras with large aperture lenses can do that
crop sensor gear with smaller aperture lenses cannot do is render busy
backgrounds softly out of focus while keeping the subject sharp and clear. Sporting venues both indoor and outdoor often
have very cluttered busy backgrounds which can be visually intrusive unless
rendered very softly by a large aperture lens.
But my use is mainly documentary, street and making a visual
record of places and people. For this I
often want all the elements in the frame to be rendered sharp and clear. And
for this crop sensor gear works very well.
If APSC crop sensor gear can deliver good enough image
quality and performance for my requirements then I will be very tempted to say
thank you very much and go that way.
So I ran a series of tests comparing various full frame kits
with crop sensor kits based on the R7.
R7 with RF-S 55-210mm |
I started with basic one body, one lens combinations.
On the R7 the most versatile single lens is the RF-S
18-150mm. Fortunately this is a very good lens with very good sharpness right
across the focal length range, very good focussing and image stabiliser and
very good close-up capability. In fact this is just about the most versatile
multipurpose lens in the whole RF mount catalogue.
I ran tests comparing the R7 with the 18-150mm versus the R5
with the RF 24-105mm f4-7.1 STM, R5 with RF 24-105mm f4 L and R5 with 24-240mm
superzoom. I used a subjects containing much leafy foliage detail and high
subject brightness range.
I used Raw capture and processed each test file in Adobe
Camera Raw to what I considered a best result.
By the way this does not mean processing all files identically. I find
the R7 files need more Texture, Clarity and Sharpness in Adobe Camera Raw than
those from the R5. The information is in the R7 files, they just requires a
different strategy in Camea Raw to tease out the details so we can see them
clearly in the final output.
When the resulting photos are matched for output pixel
dimensions and compared side by side at 100% on screen I have great difficulty
detecting any significant difference between them.
In the ultrawide range I tested the R7 with the RF-S 10-18mm
f4.5-6.3 against the R5 with the RF 15-30mm f4.5-6.3. I got the same result.
Both combinations are very good and when matched for output size I cannot tell
them apart.
Moving on to the telephoto end of the scale I tested the R7
with the RF 100-400mm f5.6-8 against the R5 with the same lens. Again, when
matched for subject size and output size I cannot tell them apart.
When I compared the R7 with RF100-400mm against the R5 with
the RF100-500mm L lens I found no significant difference between them in the
center of the frame but the big white L lens on the R5 was sharper in the
periphery of the image. We have to spend
a lot of money and lug around a lot more gear to improve on the capability of
the R7 with RF 100-400mm combination.
R7 with RF 100-400mm rolling shutter with E Shutter mode |
Distortion corrected in Photoshop |
As at January 2024 there are two more RF-S lenses.
The 18-45mm f4.5-6.3 (full frame equivalent 29-72mm) is generally sold in a kit with the R10, R50
or R100 bodies. Although it tends to be overlooked by enthusiast photographers
it is ideal when we want to use one of the smaller RF crop sensor bodies as a
compact camera. On my tests the 18-150 is marginally better on the test chart
but with real world subjects the two lenses deliver nearly identical results
with neither offering better subject rendition.
The medium tele companion to the 18-45mm is the RF-S 55-210mm f5-6.3 (full frame
equivalent 88-336mm). My copy of this lens is very nice. Sharp at all focal
lengths, with fast AF and very good IS, it is ideal for occasions when we want
a bit of extra reach without having to carry the much larger RF 100-400mm. Some
reviews have dammed this lens with faint praise. My experience is more positive
so sample variation might be an issue.
Next I ran some stress tests on the R7.
The first was to test the sensor. I selected a very dark place with spotlights. This is an
enormous old navy ship fuel tank now converted into an exhibition space as part
of the Art Gallery of New South Wales. The venue features extreme subject
brightness range with exhibits lit by spotlights while nearby there is so
little light I cannot see my own feet. This place is a big challenge for any
camera. We have to use high ISO settings but we also need high dynamic range
which is difficult as dynamic range declines as ISO setting is increased. I set ISO 3200 and f4.5 at 10mm on the RF-S
10-18mm ultrawide lens which gave me hand held shutter speeds around 0.4
seconds on average.
The results are excellent. The combined body and lens
stabilisers enable consistently sharp pictures at 0.4 seconds. Dynamic range at
ISO 3200 is sufficient to render detail in both highlights and shadows.
The second stress test was the ability of the autofocus system to identify, lock onto and
track and actually render sharply in focus,
birds in flight against a busy background of city structures. Using the RF 100-400mm f5.6-8 lens, the R7
managed this rather well. The camera’s animal (bird) detect and eye detect
system works remarkably well. It can identify a bird against a busy city or
choppy water or foliage background and track it with a high percentage of
in-focus frames. The system is able to lock onto and track the eye of a moving
ibis which is extraordinary, given that the eye, head and neck of the ibis are
all black.
There are many posts on user forums and YouTube videos
reporting inconsistent autofocus with the R7. There are also many published
suggestions by experienced users about settings and strategies to minimise the
problem.
My finding is that the problem is real, it is not just
somebody’s lack of experience with the camera. But I find that although focus
when using [H+] drive mode and Servo AF does wander off target from time to
time it generally corrects itself after a few frames.
I find that shooting in several short bursts rather than one
long sequence minimises the likelihood of the issue appearing. This repeatedly
re-starts the AF subject acquisition sequence. It is good practice anyway to
preserve the buffer and minimise the number of files which must be evaluated
when post processing.
Spiral stairway into the tank. R7 with RF-S 10-18mm |
Next, I sought to unravel the shutter mode puzzle.
The R7 offers three shutter modes,
* Full mechanical (MS) with the focal plane shutter starting
and stopping each exposure.
* Electronic first curtain (EFCS) in which each exposure is
initiated electronically on the sensor then terminated with the focal plane
shutter.
* Fully electronic shutter (ES or E-shutter) in which each
exposure is initiated and terminated electronically on the sensor. This is silent,
although the camera can make little electronic tika-tika sounds to confirm that
the system is working.
The problem with mechanical focal plane shutters (in-lens
leaf shutters as found in bridge cameras are OK) in mirrorless cameras is
a) The undamped
variety are noisy and
b) they can produce shutter shock with some combinations of
lens, focal length and shutter speed. This is a condition in which the
perturbation produced by the shutter blades hitting their stops actually shakes
the camera a bit, in some cases enough to produce visible blurring of image
details.
There are several problems with the electronic shutters in
budget and mid range mirrorless cameras.
The main one relevant to this discussion is their relatively slow sensor
scan speed which is about 30 milliseconds
(1/33 second) in the R7. This leads to severe slanting distortion of
fixed objects when we pan the camera to follow a moving subject and can cause
severe fragmentation of the appearance of the wings of little birds in flight.
This is known as rolling shutter.
Some cameras can use just one shutter mode for everything.
For instance on the R5 I use EFCS for
everything, all the time. I can do this because the focal plane shutter
incorporates a damping mechanism, not present on the R7 or R10, which greatly
reduces the likelihood of shutter shock.
Some cameras such as the R3 can use the E-shutter all the
time because that camera has a stacked sensor which enables a fast E shutter
readout speed of about 5 milliseconds (1/200s).
Some cameras such as the Nikon Z8 and Z9 have an even faster
sensor scan time which allows the designers to dispense with the focal plane
shutter altogether.
And the Sony A9.3 has a sensor which reads out all the
exposure data simultaneously. This is the so-called “global” shutter.
By the way, video always uses the E-Shutter.
But on the R7 (and R10) there is no shutter solution which suits
every situation for still photography.
So based on my own work and reports of other users, my
shutter mode settings for the R7 are:
* I never use MS, there is no need for it and the risk of
shutter shock is higher than with EFCS.
* For all single shot work at any shutter speed, I use EFCS. There is no risk of shutter shock because the shutter blades close at the end of the exposure.
* For most moving subjects using Servo AF I use EFCS and
[H+] drive mode and keep the shutter speed at or above 1/400 second to minimise
the risk of shutter shock. Shutter speeds around 1/50 and 1/100 second do
exhibit shutter shock on my tests affecting a variable percentage of frames.
* I avoid [H] Drive Mode with EFCS because of the
distracting viewfinder blackout with that combination.
* If silent operation is imperative I use E-shutter.
* I sometimes use E-shutter with perched birds with Servo AF
and [H] drive mode. Viewfinder blackout is not an issue with this combination.
Note that [H+] gives nominally 15 fps with EFCS but 30 fps
with E-shutter. In practice the actual frame rate depends on lens, aperture,
light levels and subject features. I often get around 12 fps with EFCS and [H+]
Drive Mode.
Lenses compatible
with the R7
Several photography commentators have criticised Canon for
not providing a greater selection of lenses for their RF mount crop sensor
cameras. I think this criticism is partly justified but over-stated. There is no point in having five versions of
the standard zoom available if just one of them will do the job sufficiently
well in the great majority of circumstances.
Just for interest I trawled through the catalogue of XF
zooms for Fujifilm APSC cameras. I found 12 currently available but many
overlap in focal length range, aperture range and purpose to the extent that
one has to wonder if Fujifans might be better served by fewer models each with
broader appeal.
I have already referred to the four RF-S zooms currently
available. Each of these delivers very good to excellent image quality, is
compact, light and moderately priced.
From the full frame RF catalogue I nominate the RF 100-400mm
f5.6-8 tele zoom and the RF 28mm f2.8 prime as being very suitable for RF mount
crop sensor cameras including the R7. The 100--400 gives us an effective reach
of 640mm at the long end with very good
image quality and performance.
The very sharp 28mm f2.8 provides a full frame equivalent
focal length of 45mm on the R7, ideal for general photography with a very compact kit. We can, if you like, consider this the first
decently fast prime which is well suited to the R7.
The RF16mm f2.8 is soft around the edges on APSC. The RF
24mm f1.8 (full frame equivalent 38mm) and
35mm f1.8 (full frame equivalent 56mm) work just fine on the R7 but might be
considered a bit bulky for the crop sensor genre.
All RF and RF-S mount lenses will function properly on all
RF mount bodies however in the spirit of maintaining compact size, some lenses
are more suitable than others.
I expect that in due course Canon will release the very good
and popular EF-M 22mm f2 in RF mount, or at least offer something similar.
All RF mount cameras can also use EF and EF-S lenses with
the appropriate adapter although not all these allow the camera to run at the
nominated frame rate and autofocus might in some cases not be as fast or
reliable as with RF mount lenses.
Third party EF and EF-S mount lenses can be used but there
are many reports of sub-optimal autofocus performance on RF mount bodies.
The next matter to consider is the ergonomics of the R7 and
specifically the co-located thumb-stick and rear control dial on the upper back
of the camera as seen on no previous Canon or model from any other camera
maker.
My initial reaction
to this was …”what the ***** were you guys thinking… ???”
And after a lot of use and testing my reaction is still….”what
the ****** were you guys thinking….???”
No purpose for this unique control layout has ever been
articulated by any Canon representative as far as I am aware. There has never
been any indication of which problem it was meant to solve or which user action
it was meant to make easier.
In my view it is a folly which I wish had never gotten into
production and which I hope we do not see in any follow-up model.
In general stacked and co-located control modules are an
ergonomic curse because working one of them will sooner or later inevitably
produce accidental activation of the other.
I deal with this by
a) training myself to hold my right palm back, away from the
camera so my right thumb can then flex forward to bear on the thumb stick and
hopefully not turn the dial accidentally. Thie is an un-natural position for my
right hand which impairs the stability of my hold on the camera.
b) allocating [Direct AF point selection] to the Cross-Keys.
This enables me to move the AF area and navigate around various functions with
the cross keys if I wish.
All this is suboptimal but serviceable. I can live with it
for the time being.
Summary
The Canon EOS R7 is basically a very good and versatile
camera with several annoying but non-fatal idiosyncrasies. It also
has considerable appeal for my preferred type of photography. On balance, for
now, I think the upsides outweigh the downsides so I will persevere with the R7
and see where that journey leads.
No comments:
Post a Comment