Sony extends their high end to encompass 8(!) megapixels and a sharp 7x Zeiss zoom lens. - And Sony's new RGB+E sensor technology for more accurate color!
Shutter Lag/Cycle Times
When you press the shutter release on a digital camera, there's usually a
delay or "lag time" before the shutter actually fires. This time
allows the autofocus and autoexposure mechanisms to do their work and can
amount to a significant delay in some situations. Since this number is rarely
reported by manufacturers or reviewers, and can significantly affect the
picture-taking experience, I routinely shutter lag and cycle times using
a proprietary electronic test setup with a resolution of 0.001 second.
Sony
DSC-F828 Timings
Operation
Time (secs)
Notes
Power
On -> First shot
1.5
Time from power on to
first shot captured. Very fast. (There's no lens to telescope out,
but even so the startup time is faster than I'd have expected.)
Time measured with Memory Stick Pro.
Shutdown
0-20
No lens to retract, so
"shutdown" time is effectively nil. Depending on the
memory card in use though, it could take 20 seconds or more to
finish writing a large RAW-format file. (Worst case with a Memory
Stick Pro is only 10 seconds though.)
Play
to Record, first shot
0.26
Time to capture image
from quick review mode, lens set to manual focus. Very fast.
(Time would be somewhat longer in autofocus mode, particularly
with the lens set to full telephoto, due to autofocus time.)
Record
to play, quick view mode, camera finished processing previous
shot.
~ 0.4
Initial image display
is very fast in quick review mode, and there doesn't seem to be
any additional time required for the display to sharpen-up. (Playback
zoom also operates very quickly.)
Record
to play, normal playback mode
2.3/0.4
Time from switch to playback
mode to image displayed, large/fine mode images. First time is
for immediate switch after an image is acquired, second is for
switch with camera done processing the last image. Pretty fast
when camera is quiescent, but time for immediate switch is average.
Shutter
lag, full autofocus
0.25
0.44
0.69
First time is with lens
set at wide angle, second time is with lens set to 100mm equivalent
focal length, third time is with lens set to full telephoto. All
times are very fast, wide angle and 100mm times are blazing. (By
far the fastest prosumer camera I've tested.)
Shutter
lag, continuous autofocus
0.42
As is generally the case
with cameras I've tested, continuous autofocus doesn't improve
the shutter lag at all, although it may be helpful for moving subjects.
(This time measured with the zoom set to 100mm.)
Shutter
lag, manual focus
0.20
Blazingly fast.
Shutter
lag, prefocus
0.009
Yes that's 9 milliseconds!
I repeated my measurements 3 times (total of ~40 shots) to be sure.
For this particular parameter, this is the fastest camera of any
sort that I've tested. (I'm fortunate to have a test system
that's fast enough to record this - Most methods of measuring shutter
lag have a resolution of only 0.1 second.)
Cycle
Time, Memory Stick Pro, max/min res
1.27/1.36
These times are for a
Memory Stick Pro card. Almost identical times for large/fine and
small/normal files, the small files taking slightly longer, apparently
as a result the resampling of the image. No detectable buffer,
the camera ran this fast for 27 consecutive shots, until I got
tired of pressing the shutter button. VERY fast!
NOTE in all these cycle time numbers, that the F828 "penalizes"
you for trying to fire the shutter too quickly. If you do so, the
camera will just sit there until you release and re-press the shutter
button.
Cycle
Time, CF card, max/min res
(All CF timings here gathered with Lexar 512MB 24x card)
1.26/1.37
Times with a CF card.
CF cards are slower in other tests, but seemingly no impact in
single-shot mode cycle times. Once again, VERY fast.
Cycle
Time, RAW file format, Memory Stick Pro/CF card
13.49/15.33
First time is for Memory
Stick Pro, second is for CF card. No buffering of files in RAW
mode, you have to wait for each to write to the card before you
can grab the next shot. Here, the Memory Stick Pro showed its higher
transfer rate, although the difference wasn't dramatic.
Cycle
Time, TIFF file format, Memory Stick Pro/CF card
8.82/12.52
First time is for Memory
Stick Pro, second is for CF card. No buffering, you have to wait
for the camera to finish writing to the card before you can snap
the next shot. Oddly, TIFF write times are faster than those for
RAW format. (Usually, the opposite is the case.) Once again, Memory
Stick Pro is the faster media.
Cycle
Time, "S" continuous mode
(In Speed burst mode, the camera viewfinder blanks until
the sequence of shots is completed.)
0.38
(2.65 fps)
Same speed regardless
of card used, and regardless of size image captured. Seven shots
in the buffer, also regardless of image size or card type. After
7 shots, there's a long pause before the next shot (for
large/fine images, it's 11 sec for MS Pro, 13.4 sec for CF, for
small/normal, the numbers are 5.29 and 4.07 respectively). After
that though, the camera will continue snapping additional images
every ~2.7 sec for large/fine, or every ~1.6-1.7 sec for small/normal.
NOTE though, that you have to press the shutter button repeatedly
after the initial burst of 7 shots, to get the shutter to fire.
If you get past the long pause and into the slower shot to shot
time regime, the "buffer" clears very quickly, in a time
equivalent to the cycle time at that point.
Cycle
Time, "F" continuous mode
(In Framing burst mode, the camera viewfinder displays
a "preview" of each image briefly between shots,
slowing the frame rate slightly.)
0.41
(2.43 fps)
Essentially identical
behavior to the "F" continuous mode (7 shots in sequence
regardless of file size or card type, long pause after the 7th
shot, followed by longer shot to shot cycle times), the only difference
being that the shot to shot time stretches a few hundreths of a
second.
Cycle
Time, "M" continuous mode
(In this mode, the camera records 16 tiny 320x240 images
in a single 1280x960 frame.)
0.0334
(29.94 fps)
No surprises here - Only
16 images (played back as a slow-motion sequence on the camera,
but arranged in a 4x4 array in a single image file), but 30 frames/second
capture speed. (Can be set via a menu option to 15 or 7.5 frames/second.)
Camera is ready for the next sequence almost immediately. (About
as fast as you can release and re-press the shutter button.)
Sony makes a great deal of the F828's speed in their marketing
pitch for the camera, and the numbers above show that there's a
good reason for this: The F828 is by far the fastest prosumer-grade
camera I've tested to date, and is the fastest camera at any
price point when it comes to shutter lag in the pre-focused
mode. (That is, the shutter lag after you've half-pressed and held
down the shutter button, prior to the exposure itself.) Other reviewers
have reported this lag time as 0.1 second, but that's probably
just the resolution of their test setup. I was so astonished at
the 0.009 second prefocus lag that I repeated my tests three times,
for a total of about 40 shots, and got the 9 millisecond number
very consistently. Of course, normal full-autofocus lag is what
really matters to most shooters, and the F828 excels there as well.
- At maximum wide angle, it routinely focuses and shoots in only
0.25 seconds. This time stretches to a still-fast 0.69 seconds
at full telephoto, but is a very fast 0.42 seconds with
the lens set to the 100mm equivalent focal length, still a reasonable
telephoto. Shutter lag in manual focus mode is also exceptionally
short, at only 0.20 seconds.
Cycle times are also quite impressive, with a minimum cycle time
of only 1.26-1.27 seconds in manual focus mode with the auto review
display turned off. Add ~0.05 seconds for autofocus at wide angle,
or ~ 0.75 seconds for autofocus at maximum telephoto. Most impressive
is the fact that the F828 appears able to write its image files
to the memory card at this rate, without having to slow down to
wait for the card to catch up. - There's no sign of a buffer, or
if there is one, it holds more than the 27 shots I snapped at maximum
resolution before I got tired of pressing the shutter button. Continuous-mode
operation is also impressive, capturing up to 7 frames (regardless
of image size or JPEG quality setting) at a rate of 2.65 frames/second.
Finally, a "Movie" continuous mode shoots 16 tiny 320x240
images at a selectable frame rate of 7.5, 15, or 30 frames/second,
storing them in a single 1280x960 image file.
The camera slows a fair bit when recording full-resolution RAW
or TIFF files, although its performance there is still very much
in the top echelon, particularly considering that it's saving 8
megapixel images. It really would be nice to see some level of
buffering applied to RAW and TIFF files though, as a lot of prospective
users of the F828 are likely to want to shoot in RAW mode.
Overall, the F828 is easily the fastest "prosumer" non-SLR
camera on the market, by a wide margin. With its long 7x zoom lens,
it'll make an excellent camera for sports shooters. (Although it'd
be nice to see something like Fuji's "Final 5" shooting
mode, for capturing critical moments despite lagging reflexes on
the part of the photographer.
The only complaint I have about the F828's timing performance is
that it's one of the (unfortunately many) cameras that "penalize"
you for pressing the shutter button too quickly after the previous
shot is taken. If you click and hold down the shutter button immediately
after a shot is taken, the camera will save that shot to the memory
card and then just sit there. You have to release the shutter button
and press it again to take the next photo. In practice, you can
develop a pretty good sense of how long to wait before pressing
the shutter button, and so avoid the "penalty," but it's
easy to get a little over-eager in the face of fast-breaking action.
And the camera should really handle this anyway. Why can't the
camera simply notice that the shutter button is still down and
fire off another frame as soon as it's able? As noted, this is
an unfortunately common behavior among even the higher-end digital
cameras I've tested, but one that I'd really like to see
done away with.