[RonHendriks1966]

So I made a quick test on the writespeed.

To my fast messurement (not scientific) the writespeed has a max off about 45 MB/s since there is almost no difference between my 45MB/s and 95MB/s speed card.

[Mikesul]

Boy, if this test is accurate we can save a lot of money. It would also explain why BandH included a 45mbs card with my K3. It is still much faster than the K5iis.

[twitch]

Ron, how did you test? Fill the buffer and time how long it took to empty?

[narual]

You might consider this 64 Gb card...


PNY


It's read speed is supposedly rated at 90Mb/s and I understand its write speed is around 60Mb/s, as reported by others using a USB3 rated reader... Amazon lists them at about $35 each, USA shipping included.


Cheers... M

You sold Amazon out of them. Cheapest one there now is $50.

[RonHendriks1966]

Ron, how did you test? Fill the buffer and time how long it took to empty?

I did a few tests. On both cards.

- Fill buffer with RAW or RAW+jpg and time to clear buffer
- Shoot for 30 seconds on RAW or RAW+jpg and time to clear buffer.

Basicly you get a grip off what is happening with all the data when the buffer is full, withdrawing the time to fill the buffer.

I only got one extra image with the faster card, so that is about to be the margin off error on the messurement.

[twitch]

I did a few tests. On both cards.

- Fill buffer with RAW or RAW+jpg and time to clear buffer
- Shoot for 30 seconds on RAW or RAW+jpg and time to clear buffer.

Basicly you get a grip off what is happening with all the data when the buffer is full, withdrawing the time to fill the buffer.

I only got one extra image with the faster card, so that is about to be the margin off error on the messurement.

thanks, not sure if that is good news or bad, but it's good information

[RonHendriks1966]

thanks, not sure if that is good news or bad, but it's good information

Well the K-5 was well under 19 MB/s so the K-3 is at least twice as fast, so that is a good number. Didn't do the test with all lenscorrections, but that is coming and that is where the PRIME II fall flat on its nose in speed. If the speed is still good then it's not the processor holding things back, but the internal highway for bits and bytes that is limited in traffic. Or the cardreader inside the camera isn't faster. There is Always some bottleneck, but in general the K-3 is faster in processing then the K-5 and that with 50 % more pixels.

[RonHendriks1966]

So with RAW+jpg+lenscorrections the speed goes down. After the full buffer it is around 0,77 fps or so, but still faster then K-5 could do. I don't see the K-3 any faster in writespeed then the 45 MB/s I mentioned earlier.

[mamethot]

Some more info...

One thing I tried yesterday was a comparison of a transfer using a USB 3 cable and a USB 3 card reader (transcend).

The results were (on a 32gb Sandisk Extreme Pro (95MB/s)):

Cable transfer between camera and PC: 18 to 20 MB/s
SDHC card reader: 40 MB/s.

This was for a copy of 3.5gb of files.

[RonHendriks1966]

Some more info...

One thing I tried yesterday was a comparison of a transfer using a USB 3 cable and a USB 3 card reader (transcend).

The results were (on a 32gb Sandisk Extreme Pro (95MB/s)):

Cable transfer between camera and PC: 18 to 20 MB/s
SDHC card reader: 40 MB/s.

This was for a copy of 3.5gb of files.

I forgot to get the cable last weekend.

Still faster then the K-5 that had like 8 MB/s or so with cable and usb 2 port.

[crewl1]

Thanks for the test results Ron.
Sounds like I can save some money and get the Extreme Plus instead of the Pro.

[crewl1]

I have ordered a couple of Toshiba Exceria Type 2 and will report back once I test them.

Toshiba 16GB Exceria Type 2 SD-H16-GR7WA6 SDHC Card r95MB/s w60MB/s Class 10 (CRA-ET2-RT) Retail

Toshiba 16GB Exceria Type 2 SD-H16-GR7WA6 SDHC Card r95MB/s w60MB/s Class 10 (CRA-ET2-RT) Retail

[dosdan]

Toshiba 16GB Exceria Type 2 SD-H16-GR7WA6 SDHC Card r95MB/s w60MB/s Class 10 (CRA-ET2-RT) Retail

The R/W speed of the Toshiba is 95/60 MB/s
The Sandisk Extreme Pro is 95/90 MB/s

The Exceria price is good, but it's the 50% faster writing time of the Extreme Pro that should make a significant difference when the buffer fills. With sports, I currently spend a lot of time waiting for the K-5's buffer to clear. It's very easy to do this when you're taking a lot of bursts in a short period of time e.g. hurdles or a sustained goal attack in soccer. When I get the K-3, with its bigger images and updated SD R/W hardware, I want to give the buffer the best chance.

I'm usually an exclusive raw shooter, but I may even consider shooting JPEG at 14MP, as I don't think my two main zooms (DA* 60-250/F4 & Sigma 70-200/F2.8) will out-resolve this image resolution. (When I get the K-3 I'll perform some tests shots of printed text at a distance to confirm this suspicion.) Buffer depth and speed would then not be an issue.

But if you shoot lots of raw or raw+JPEG bursts with the K-3, then SD write speed will be highly significant.

Dan.

[crewl1]

The R/W speed of the Toshiba is 95/60 MB/s
The Sandisk Extreme Pro is 95/90 MB/s

The Exceria price is good, but it's the 50% faster writing time of the Extreme Pro that should make a significant difference when the buffer fills. With sports, I currently spend a lot of time waiting for the K-5's buffer to clear. It's very easy to do this when you're taking a lot of bursts in a short period of time e.g. hurdles or a sustained goal attack in soccer. When I get the K-3, with its bigger images and updated card R/W hardware, I want to give the buffer the best chance.

I'm usually an exclusive raw shooter, but I may even consider shooting JPEG at 14MP, as I don't think my two main zooms (DA* 60-250/F4 & Sigma 70-200/F2.8) will out-resolve this image resolution. (When I get the K-3 I'll perform some tests shots of printed text at a distance to confirm this suspicion.) Buffer depth and speed would then not be an issue.

But if you shoot lots of raw or raw+JPEG bursts with the K-3, then SD write speed will be highly significant.

Dan.

Sure but according to Ron the camera does not take advantage of the extra write speed of the Extreme Pro, so why pay for it?

[dosdan]

Sure but according to Ron the camera does not take advantage of the extra write speed of the Extreme Pro, so why pay for it?

The statement I can see about buffer performance is:

I only got one extra image with the faster card, so that is about to be the margin off error on the measurement.

But the faster write speed doesn't help much during the burst. It's after the burst that I'm interested in here. That's covered in the PF review:

Pentax K-3 Review - Performance and Burst Mode - PentaxForums.com
Time to clear full buffer
(measured from end of burst)

K-3 RAW 19s 33s

Where 19s was for the SanDisk 95Mb/s 32GB Extreme Pro SDHC Class 10 UHS-I

and 33s was for the SanDisk 45Mb/s 32GB Extreme SDHC Class 10 UHS-1

If using JPEG, the buffer depth looks quite impressive:

Number of frames at max. FPS

K-3 JPEG Fine 52 43

Because the image files are smaller, the write speed becomes more significant for the effective buffer depth, as the buffer can partially empty during the burst.


and the full buffer clearance speed is better:

K-3 JPEG Fine 12s 19s

Dan.