Each of the photos in this post was made with either the R5+24-240mm kit or the R7+18-150mm kit. Can you pick which photo came from which camera ? |
Superzooms. Lens test afficionados hate them, ordinary photographers love them.
Interchangeable lenses with an 8-10x zoom range and sometimes more, trade off ultimate optical perfection for
convenience and versatility.
In this post I compare two superzoom kits, each using the
Canon RF mount.
Kit 1 is the full frame, high price option which pairs the
EOS R5 with the RF 24-240mm f4-6.3 IS USM lens with a 10x zoom range.
Lower priced full frame options could include the same lens
mounted onto an EOS R6.2 or R8.
Kit 2 is the crop sensor, APSC option which is lighter, more
compact and considerably less expensive. This pairs the EOS R7 with the RF-S
18-150mm f3.5-6.3 IS STM lens with an 8.3x zoom range.
A lower priced option
could replace the R7 with an EOS R10 or R50 camera body.
Any of these options work just fine but for this post I will
compare the R5+24-240mm with the R7+18-150mm, representing the best available
options for full frame and crop sensor sub-systems.
Kit |
Mass w
batt,card filter,hood |
Retail price
new AUD |
Focal length
as FFE |
Aperture |
Filter |
R5 RF24-240 |
1512g |
$6348 |
24-240mm |
f4-6.3 |
72mm |
R7 RF18-150 |
960g |
$2499 |
29-240mm |
F3.5-6.3 |
55mm |
Before getting into the detailed comparison I have to say that either of these kits can produce excellent pictures when used thoughtfully although each has a different spectrum of strengths and weaknesses. In fact on side-by-side testing using the same subject and viewing the output files at the same size, it can be difficult to pick which file came from which source, provided both are post processed to best advantage.
I have used each of these kits in a range of different
conditions over the last two years. I
have done systematic testing on each as well as making many pictures on
location.
So, on with the evaluation.
Size, mass, price. Obviously the R7 kit has a big advantage on
these parameters. So big that the full frame kit will need to be convincingly
superior in other ways to make the extra cost worth while.
Spoiler alert, for many photographic use cases, it’s not.
The full frame kit is wider at the short end of the zoom
than the crop sensor kit but unfortunately that is not quite the bonus we might
hope for because the 24-240mm lens has some significant weaknesses in the
24-30mm focal length range.
Handling and ergonomics. Both kits handle well with no serious problems. If I am holding the gear for an extended period I will prefer the lighter R7 kit but I am not well pleased about the co-located thumb stick and rear control dial on the R7. This arrangement makes it too easy to bump one while adjusting the other. I hope that when Canon is planning a Mk2 version of the R7 they will revert to standard Canon practice for location of the top/rear control dial. Actually I would prefer that Canon simply use the R6.2 body for the R7.2 and stop messing around with ergonomic variations which appear to have no clear purpose.
Autofocus speed and accuracy. Both kits deliver fast, mostly accurate
autofocus on still and moving subjects and both have effective subject
detection capability driven by artificial intelligence. My subjective
impression is that the R5 kit delivers a slightly higher percentage of sharply
in-focus frames when we are using Servo AF and H or H+ drive mode. The R7 is
perhaps not quite as tenacious once the subject has been identified. However
this is hard to evaluate properly as the subjects are running, jumping or
flying so controlled testing is difficult.
Stabiliser. Both
kits utilise both in-body and in-lens image stabilisation and both work very
well. Canon makes some extravagant claims about stabiliser effectiveness. I
have no idea how they figure an 8 stop benefit. In all my tests I find I can
get sharp shots hand held with the stabiliser ON about 3.5 EV steps of shutter
speed slower than with the stabiliser OFF. This applies to most Canon RF mount
body/lens combinations.
The look-at-me
factor. When I am out and about with
people around, the R5 with 24-240mm attracts more attention than the smaller R7
with 18-150mm. For that reason I do not use the R5+24-240 kit for street or
documentary work.
Image quality, camera body.
The R5 has 45 megapixels (8192 x 5464 pixels), the R7 has 32
megapixels (6960 x 4640 pixels). The
linear pixel count of the R5 is 1.17x that of the R7 (8192 divided by
6960). For one camera to be capable of
revealing observably more subject information than another we generally want to see a difference in linear
pixel count of about 1.2x or more. Of course lens quality and in-camera and
post processing activities affect the outcome. But just on pixel count the R5
has only a marginal advantage over the R7.
The full frame kit does have a more convincing advantage of
approximately 1.3 EV step in high ISO/low light noise levels.
My tests show that in low light at around ISO 6400 the R5 has an advantage of about 1.3 EV steps
over the R7 when both are matched for output size and sharpness. So the R7 at
ISO 3200 has about the same level of luminance noise in RAW files as the R5 at
ISO 8000.
Since the advent of noise reduction software using
artificial intelligence the impact on final output of luminance noise appearing
as grain is not as great as it once was. In fact I nominate Adobe Noise
Reduction AI as the most significant factor in boosting the appeal of small
sensor cameras in recent years.
Closely allied to digital noise is the issue of dynamic
range. This determines the ability of the photographic system to render detail
in highlights and shadows when there is high subject brightness range.
Photons to Photos lists the Photographic Dynamic Range at
ISO 100 of the R5 as 11.85 EV and of the R7 at 10.49 EV, a difference of about
1.3 EV which is maintained as ISO settings increase.
The next significant advantage for the full frame kit is it
relatively greater ability to separate subject from background with shallow
depth of field. Neither of these kits is very suitable for this kind of
separation as both lenses have rather small apertures (large f-numbers)
throughout their range. But f5.6 on the crop sensor kit will give approximately
the same amount of background blur as 5.6 x 1.6 = f9 on the full frame kit.
To summarise this section we can show that the full frame R5
has a definite advantage over the R7 in high ISO/low light noise levels,
dynamic range and subject/background separation. Whether
this translates to better pictures or not will depend on the circumstances. My
experience is that in most photographic situations the R7 has enough dynamic
range and low enough noise levels for good output particularly when Adobe
Denoise AI is invoked via the Enhance command in Camera Raw.
Image quality, lens.
The RF-S 18-150mm delivers very good sharpness at all focal
lengths and apertures up to about f10 when some softness due to diffraction at
the aperture diaphragm starts to appear. Sharpness is well distributed across
the frame. With Raw capture and conversion in Adobe Camera Raw, color fringing
and distortion are minimal. Peripheral shading is present at the wide end but
easily corrected. Flare is well controlled.
In summary this is a very well behaved lens capable of
transferring an enormous amount of information from the subject to the sensor.
The RF 24-240mm has a more complex set of optical
characteristics. In the 35-150mm focal length range it behaves very well
indeed, with a very high level of sharpness and contrast and no nasty surprises.
However in the 24-30mm range and especially at 24mm the
corners never get really sharp with details lost and strong blue/purple/green
color fringing becoming obvious at high contrast edges. The loss of peripheral
sharpness and color fringing re-appear at the long end of the zoom although
they are less obvious. In practice I find myself spending more time trying, not
always successfully, to correct color fringing with this lens than any other in
the RF and RF-S catalogue.
So to summarise, the RF-S 18-150mm is more consistently well
behaved across the range of focal lengths and apertures, while the RF 24-240mm
is not so good at the wide end, excellent in the mid range and fades off a bit
at the long end.
The RF-S 18-150mm with a maximum magnification of 0.44x
using AF is much better close-up than the RF 24-240mm which offers only 0.26x
magnification. I frequently photograph wildflowers with the RF-S 18-150mm which
has replaced my macro lens for all but an occasional subject requiring higher
magnification than 0.44x.
Overall both these lenses are good but I prefer the
all-round versatility of the RF-S 18-150mm.
Shutter shock and rolling shutter. Not to labour these issues in this post but the R5 has fewer potential problems with shutter shock and rolling shutter than the R7. However there are strategies for managing both problems with the R7 which I have discussed elsewhere.
Post processing
strategies. Again this is not the place to go into details but I find that
Raw files from the R5 often come off the camera looking almost finished and
needing little help in Adobe Camera Raw.
Files from the R7 have a lot of information in them but it often needs
to be coaxed out with a bit more work in Camera Raw particularly on the
Contrast, Texture, Clarity and Sharpness sliders. I also use the
Enhance-Denoise function much more frequently with R7 files than I do with R5
files.
Summary
We make our choice and pay our money. In Canon-world the
general rule is that if we pay more, we get more. More of everything available
in camera-land. That is true of the two kits in this comparison but sometimes
more is not necessarily better.
These days I find myself using the crop sensor kit most of
the time with very satisfactory results.
However if I had written this comparison prior to the
arrival of Adobe Denoise AI the full frame kit would have been my preferred
option.
No comments:
Post a Comment