Friday, March 30, 2007

K100D SD Card Write Speed Test

Here is a simple SD card write speed test with my K100D, for some SD cards of different brands and models, each of which has different rated read and write speeds, if any specified:-

Card and IdentifierRated SpeedChipset
Write Time for 3 PEFs
(Rounded Up to
the Nearest Second)
Throughput
EagleTec 256MB SD
"SAMSUNG 334
K9K2G08U0M-YCB0
ULG023PX"
Nil
Samsung
17 s
1.79 MB/s
Lexar 1GB SD
"LEXAR MEDIA SD1GB-120M
1405112
A040507B
MADE IN U.S.A."
Nil
Lexar
12 s
2.54 MB/s
Toshiba 1GB SD
"Toshiba SD-M01G
0508TJ8419L
MADE IN JAPAN"
Nil
Toshiba21 s
1.45 MB/s
Kingston 1GB SD
"Toshiba SD-M01G
0514TJ9034L
MADE IN JAPAN"
Nil
Toshiba
20 s
1.53 MB/s
KingMax 1GB SD
"P01GS6GMHLLAG
651054"
18.8 MB/s Read, Maximum
15.4 MB/s Write, Maximum
Unknown
14 s
2.18 MB/s
Corsair 1GB SD 133X
"S51GBA0615KKKT7TWN"
21 MB/s Read, Maximum
15.5 MB/s Write, Maximum
Skymedi
13 s
2.35 MB/s
Toshiba 2GB SD Class 4
"Toshiba SD-M02G
0642TK8283U
MADE IN JAPAN"
4 MB/s Read and Write, Minimum
5 MB/s Write, Maximum
Toshiba
17 s
1.79 MB/s
Toshiba 2GB SD Class 6
"Toshiba SD-F02G
0706TL1531V
MADE IN JAPAN"
6 MB/s Read and Write, Minimum
20 MB/s Write, Maximum
Toshiba12 s
2.54 MB/s

As we can see, a faster card can help the K100D to write faster. However, the increase in performance is not as significant as it should be for such a very fast card used, e.g., the Toshiba Class 6 SD card rated at 20MB/s maximum write speed. Indeed, the K100D still writes rather slowly no matter what card used in it and it is indeed considerably slower than any of the DSLRs on the market. To compare, below are the results of the already slow Nikon D50:-

http://www.robgalbraith.com/bins/multi_page.asp?cid=6007-7905

But yet, the D50 is still relatively faster than the K100D.

Actually, if the same cards are put into a faster DSLR, there will be a huge difference in the write performance:-

http://www.robgalbraith.com/bins/multi_page.asp?cid=6007-8531

Do note that I have carried out my test with a very similar method as Rob Galbraith did and thus the results are quite comparable, I think. I shot 3 consecutive pictures in RAW and the read/write LED lamp's lit-up time is counted, by a stop watch and then the throughput rate for the write speed is calculated where 1MB = 1024^2 Bytes.

It is interesting to note that the claimed Class 4 "guaranteed" write speed by Toshiba according to the SD Association definition for a Class 4 speed card is actually somehow meaningless as the K100D is an example of such a slow device in which a 4MB/s rate can never be reached, so does the Nikon D50, too, although it is somewhere nearer.

In fact, the speed ratings of those SD cards are something that can never be totally trustworthy. In contrast, cards that are not rated may not be really slow, for example, the old Lexar ordinary 1GB SD card (bought in early 2005) has really given me quite a big surprise. It is actually quite as fast as the latest Toshiba 2GB Class 6 card rated at 20MB/s write speed (equivalent to 133X for the pre-"Class" representation), although it is actually not rated in any way.

15 Comments:

AndrewG NY said...

Rather pathetic results. Was this tested with the latest firmware? Is there any chance you'll be able to perform a similar test with K10D?

RiceHigh said...

All tests were carried out with the "latest" firmware version 1.02 and at this moment I have no plan to test the K10D. Nonetheless, I do believe that the K10D is a much faster camera than the K100D for its unlimited large fine jpeg continuous shooting capability in around 3 fps.

Anonymous said...

Wouldn't using a disk bench to write the card via USB while in the camera be a more accurate way to test speed? Lights and a stopwatch.....

markx said...

These results seem to agree with personal experience. Thanks a lot for doing this test, as it saves me the expense and hassle of buying a high-speed card. The improvement is truly minimal. I can imagine JPEG being even slower, given the weak processing power of K100D. Sometimes I find that rather annoying, as the 3 FPS rating becomes rather useless if a huge amount of time is required between each frame burst.

1964 lenses said...

Yes, thanks. Very informative.

Anonymous said...

So if I understand the test well, it tells that there's no need to buy a sandisk extreme III for it (2OMB/S), because it won't use the speed of it and it will be almost as quick if I buy a normal sandisk card (2MB/S) because the pentax wont be able to handle it

Anonymous said...

What is the biggest card the k 100 D can handle

I tried a 4 go and it does not take it.

Thanks

RiceHigh said...

Update to the latest firmware and try with a SDHC card. It should work. Good luck!

Anonymous said...

Hi, which SD for K2000w, thanks.

RiceHigh said...

I bet any current generation of SD card would be indifferent when they are used in the K2000 (K-m). The K-m by itself is a slow camera for continuous shooting, and with a very limited number of continuous shots, too. For more info, see here.

kanio_kanev said...

I did the same test several times with k100d (v1.02) and SanDisk Ultra II 8 Gb (the read/write LED lamp's lit-up time is counted).

Write times for 3 PEFs were around 9.5 seconds.

Anonymous said...

RiceHigh said...
"the K100D still writes rather slowly no matter what card used in it and it is indeed considerably slower than any of the DSLRs on the market."

This is complitely untrue!

k100d writes faster on faster cards for sure.

RiceHigh said...

So what? Even if it is really faster, it is still damn slooow!

Anna said...

SOOOO I just got a K100D and would like to a) what SD card to get and b) where/how to update the firmware.
thanks!

RiceHigh said...

Any SD cards except very large and new ones should be fine for the K100D.

As for firmware update, download and instructions, visit pentaximaging.com and search under the support page.

Post a Comment

Related Posts

 
Creative Commons License
RiceHigh's Pentax Blog by RiceHigh is licensed under a Creative Commons Attribution-NonCommercial-NoDerivs 3.0 Unported License.