Steven wrote:
Greetings all,

I have a question regarding the 'Ultra' / 'High-Speed' SD flash memory
cards. Manufacturers claim "up to a five times boost in performance
over traditional secure digital cards" or even more, however will I
notice a difference in 'Burst' mode frames per second and/or time
between shoots?


I can't speak about SD memory cards, but my experience and experiments
with a few CF cards in two cameras (set out below) give me the
impression that there is little to choose among newly produced memory
cards.

It seems to me most of what differences appear depend on the camera.
Certain dSLRs have integrated with Lexar's WA speedup technology. Most
digital cameras have not. Apart from that, at least one of the more
recent, higher-MP cameras seems to have architecture that accommodates
whatever potential the card may have.

==============================
_I set up my Nikon CP5700 on a tripod, filled the monitor with a view of
a CRT
showing the Windows Time/Date adjustment facility, so the photograph
would include a record, and I could see what was happening in real-and
verifiable, if needed-time.

Camera saw everything at 1/15 second, f2.8, same view for every
exposure.

RAW, full-size exposures after format of a CF card in camera.

I waited until the on-screen digital counter said xx:00 or xx:30,
depending on how long it took me to record data and reformat the card
between shots. As soon as the 00 or 30 digits appeared I pushed the
shutter release button. The effect of this was that the camera monitor
screen went blank at :01 or :31, very consistently.

The camera monitor and the CRT were in a sight line, so I could look at
the time and still see when the camera recovered a view (screen
un-blanked). I recorded that interval for the first few trials, but
ceased when it seemed clear the time was the same independent of card
type or speed: 8 (eight seconds, approx.).

Once the camera monitor acquired a view, I watched the little recording
symbol until it disappeared, and recorded the time I saw on the CRT time
display.

Re-format the CF card, repeat the exposure, three exposures per card,
change cards, repeat for each card. Actually I had to repeat a repeat or
three, as I nodded off and failed to note the time on a few trials.
(more seniorness)

These are the cards I used, and their sources:
Viking 512MB, two years old, no speed marked on card, Amazon.com
SanDisk 512MB #1 plain, one month old, Costco
SanDisk 512MB #2 plain, one month old, Costco
San Disk 512MB Ultra II, two weeks old, Costco
Lexar 512MB 40X #1, eight months old, disremembered online source
Lexar 512MB 40X #2, two months old, BandH
Lexar 1GB 80X, less than a week old, Adorama

These are the times I recorded as elapsed between button-push and
symbol-gone. I chose the mode if times were not identical; otherwise,
they were consistent (identical) among trials.

CP5700 and
Viking 512MB 83 seconds
SanDisk Plain #1 32 seconds
SanDisk Plain #2 39 seconds
SanDisk Ultra II 22 seconds
Lexar 40X #1 25 seconds
Lexar 40X #2 20 seconds
Lexar 80X 22 seconds


Then, with absolutely the same setup, I used the 8MP, ISO 50

CP8700 and
Viking 512MB 23 seconds*
SanDisk Plain #1 16 seconds
SanDisk Plain #2 17 seconds
SanDisk Ultra II 16 seconds
Lexar 40X #1 18 seconds
Lexar 40X #2 16 seconds
Lexar 80X 16 seconds

I take this to mean that identically inscribed cards (two SanDisk
Plains; two Lexar 40Xs) can be as different one to the other as the
differences between comparable but differently branded cards in these
cameras; cards perfomed fairly consistently relative to each other, in
two different but similar cameras; the remarkable differences are
attributable to the camera; the camera can drag a mediocre performer to
the level of much more expensive cards (SanDisk Plains' latencies went
from 32-39 to 16-17 seconds, an improvement of 200-240%, Viking improved
by 360%, and the others were better by 125-140%, camera-to-camera).

* First trial with the Viking card in the CP8700 was a mind-blower: it
lost the view at shutter-release, reacquired it with the writing symbol
on screen, and as near as I could tell, was still writing at five
minutes! Then eight minutes. At ten minutes I started pushing buttons,
but everything was frozen. I had to remove and reinsert the battery tray
to make it come alive again. Reformatted the Viking card (again) and the
rest of the trials went as expected, but quite a bit faster.


This morning I sat in the waiting room while my car was serviced, and
did this:

CP8700, 1GB 80X Lexar CF card, ISO 50, 1/125, f7.2, camera propped up
and as little disturbed as I could manage, framed a glass doorway with
New Car displayed beyond,

I took one dozen RAW photos, releasing the shutter as quickly as the
camera would allow. I counted down between screen blank and view
recovery (about 8 seconds) and looking at EXIF info for actual
button-push times.

The camera would accept a new photo (shutter release) as soon as the
view was reacquired, for the first four photos, even though the writing
symbol was displayed. After that, the hourglass buffer-full symbol would
display for increasing amounts of time, but never exceeded sixteen
seconds after view reacquisition.

After the twelfth shutter release, I let it close itself out, and it
wrote for about 24 seconds.

I take this to mean the CP8700 writes to the 80X Lexar card at 16
seconds per 12,374KB RAW image, and that a user can depend on the camera
to allow about four exposures per minute as long as battery and memory
hold out. Does that sound right?_

===============================
--
Frank ess