When you press the shutter release on a camera, there's usually a lag time
before the shutter actually fires. This time allows the autofocus and autoexposure
mechanisms time to do their work and can amount to a fairly long delay in some
situations. Since this number is rarely reported on (and even less often reported
accurately), and can significantly affect the picture taking experience, I now
routinely measure it, using a custom test system I built for the purpose, accurate
to 0.001 second.
Nikon D70 Timings
Operation
Time (secs)
Non-speed rated card
Time (secs)
Lexar 4GB 40X
Notes
Power On -> First shot
Near Instant
Near Instant
Any startup delay was completely
undetectable. Very fast. (And no apparent dependence on card
size, either.)
Shutdown
Instant
Instant
Shutdown is zero if no images
are in the buffer; otherwise it takes as long as required to save the
files. That could be anything from a second or less to ~9 seconds with
a fast card, to 20 seconds or more with a slow one.
Play to Record, first shot
<0.1
<0.1
No more delay from play
to record than minimum shutter lag in record mode. Very fast.
Record to play (max/min res)
.25
.25
Very fast review, even with
NEF files.
Shutter lag, full autofocus
(no flash)
0.342-0.486
-
Will depend on lens being used, how far the
focus has to traverse from prior shot. Times shown are the minimums we
measured with the 18-70mm lens, with no focus traverse before the shot.
The first time is with the lens set at wide angle, the second with it
set to its telephoto position.
Shutter lag, wide angle, full autofocus
(on-camera TTL flash)
0.382
-
Surprisingly, TTL flash metering adds only 20-40
milliseconds to the shutter lag. (It must issue its metering flash while
the AF system is still doing its work.)
Shutter lag, manual focus
0.155
-
Very fast.
Shutter lag, manual focus
(On-camera TTL flash)
0.168
-
Further evidence that the onboard TTL flash
adds little or no shutter delay.
Shutter lag, manual focus
(Off-camera TTL flash, single SB-800)
0.340
-
I measured shutter lag with the off-camera flash
with the lens set to manual focus, to provide more precise measurement
of the additional lag due to the remote flash communications.
When there's a remote strobe involved, the shutter lag does indeed
increase somewhat. This is due to the time required for the D70 to talk
back and forth with the external flash, to configure it for both the
metering and main flashes. (This time would likely further increase
if more than one remote flash group was in use.)
Shutter lag, manual focus
(Off-camera AA flash, single SB-800)
0.317
-
There's somewhat less delay when running the
remote flash in AA mode, but it's very close, possibly within the margin
of error for these particular measurements.
Shutter lag, manual focus
(Off-camera Manual flash, single SB-800)
0.181
-
When the remote flash is being controlled in
manual mode though, there's almost no increase in shutter lag at all.
(Sports shooters take note - If you can run your remotes with the power
levels set manually, you'll get very good shutter lag performance.)
Shutter lag, prefocus
0.124
-
Very fast.
Cycle time, large/fine files, single shot mode,
auto focus
0.62
(4.7)
0.53
(1.49)
These times for the slow card and the fast Lexar
one look fairly close, but the real story is about what happens after
the buffer fills. The slow card grabs 9 shots, then slowed to 4.7 seconds
between frames, while the 40x Lexar card grabbed 21 frames, then slowed
only to 1.49 seconds per frame. The slow card took 43 seconds to clear
the buffer, the 40x Lexar took only 2 seconds.
Cycle time, NEF files, single shot mode, auto
focus.
-
0.6
(1.65)
I didn't have the patience to test the slow
card with NEF files. The 40x Lexar grabbed two shots at 0.6 second intervals,
then slowed to 1.65 seconds between shots. (Still not bad at all.)
Cycle time, large/fine files, continuous mode
0.35
(4.7)
(2.92 fps)
0.35
(0.43)
(2.92 fps)
The D70 really takes advantage of fast
memory cards. With an old, slow card, the camera runs fast (2.92 frames/second)
for ~9 shots, then slows to 4.7 seconds per shot. With a fast Lexar card
though, it keeps on chugging at about 2.3 frames per second, even after
it runs out of buffer space. Very impressive!
Cycle time, large/fine files, continuous mode
(Noise Reduction enabled)
-
0.68
(0.90)
(1.48 fps)
When I first tested the D70's cycle time, I
was surprised to find it could only manage 1.48 frames/second, far short
of Nikon's claimed 3 fps. The culprit turned out to be the long-exposure
noise reduction option. Even though these shots were captured in bright
lighting with short shutter times, having NR enabled really slowed the
camera's frame rate!
Along with its other excellent features, the D70 is a fast and responsive performer.
Most impressive of course, is its continuous-mode speed and loooong burst lengths
(with appropriately fast memory cards). With many cameras in the past, really
fast memory cards often made a difference only in buffer-clearing speed, not
so much in burst length or shot to shot cycle time. With the D70 though, we
finally have a camera that can really take advantage of fast cards. With
a card rated 60x or so, it can actually shoot at 3 frames/second (fps) in large/normal
mode without stopping, until the memory card is filled. Even with slightly
slower cards in the range of 24-40x, it shows surprisingly long burst capability
and surprisingly fast post-buffer-fill speed. Shooting with a 40x Lexar card,
I found that I could get off 21 large/fine JPEGs before the speed dropped below
~3 fps (2.92 fps by actual measurement), and even then, the decrease in performance
was very modest, with the frame rate dropping only to 2.3 fps.
To make the most of the D70 though, you really do want to have as fast a memory
card as you can afford. With an old, slow non-speed-rated card, I found the
buffer capacity to be "only" 9 frames (still pretty good), and the
post-buffer cycle time to be a sluggish 4.7 seconds/frame. Take my advice, and
buy a good, fast card to use with your D70. It's an accessory that you'll be
living with for a long time, and it can make a huge impact on the camera's responsiveness.
Startup time is another area where the D100 really blows past the competition.
Try as I might, I really couldn't measure any startup delay. Bottom line,
the D70 can get ready to shoot a picture at least as fast as you can.
Shutter response was fast but not blazingly so, actually coming in a bit behind
the previous D100 in this category, and slightly behind the Digital Rebel as
well. Manual-focus and pre-focus lag times were 0.157 and 0.124 seconds, respectively.
This isn't bad (and is a good bit faster than the Digital Rebel), but is a notch
slower than the D100's 0.100 seconds, and well off the mark of Nikon's own higher-end
products. Here's a brief table, comparing the performance of the D70 with that
of the Digital Rebel, the original D100, and Canon's EOS-10D:
d-SLR
Timing Performance
Parameter
Nikon
D70
Canon
Digital Rebel
Nikon
D100
Canon
EOS-10D
Startup
"Instant"
3.09
0.63
2.32
AF Lag
0.34-0.49
0.25-0.28
0.150
0.146
Prefocus Lag
0.124
0.142
0.100
0.104
Cycle Time
0.34
(2.92 fps)
0.40
(2.50 fps)
0.35
(2.88 fps)
0.34
(2.94 fps)
Buffer Depth
9 - 100+
4
7-9
0
Overall, the four come in fairly close to each other, with the D70 and Rebel
being slower than the other two in terms of AF lag. The huge differences with
the D70 are in its effectively zero startup delay, and in its amazing buffer
capacity. The impact of the D70's buffer performance was subtle but dramatic.
(If that's not an oxymoron.) It wasn't so much something that I was particularly
aware of while I was shooting, but rather that I gradually became aware that
I never had to wait for the camera when I was shooting in JPEG mode.