[superpowerpinger]

Pentax K3II Write speeds.

Canon Rebel T6i/T6s.

Nikon D7200.

We see that same cards are giving about the double the speeds on Nikon and Canon compared to Pentax. K3II tops out at 36 MBps while Nikon and Canon give almost double at 73MBPS.

1. Should this be something that pentax look into?
2. Can this be fixed in firmware update or it points to low performance of underlying hardware?

[Fogel70]

I't probably a hardware limitations, most likely Pentax use a different hardware (chip) than Canon and Nikon. The chip Pentax use might only support UHS-I SDR50 (max 50MB/s) and Canon and Nikon use a chip that support UHS-I SDR104 (max 104MB/s).
Pentax has much larger buffer than on Nikon and Canon, so in real life it might not make much difference.

Other camera (FI Sony A7R II) have the same speed limitation as Pentax K-3 II.
Sony A7R II Fastest SD Card Comparison and Recommended Memory Cards for Alpha Mirrorless Digital Camera - Camera Memory Speed Comparison & Performance tests for SD and CF cards

[starbase218]

I't probably a hardware limitations, most likely Pentax use a different hardware (chip) than Canon and Nikon. The chip Pentax use might only support UHS-I SDR50 (max 50MB/s) and Canon and Nikon use a chip that support UHS-I SDR104 (max 104MB/s).
Pentax has much larger buffer than on Nikon and Canon, so in real life it might not make much difference.

Actually, both the D7200 and 7D II can do about 3 second continuous high bursts when shooting RAW. The D7200 does it at 6 fps (5 fps for 14-bit files) and the 7D II at 10fps. The K-3's buffer is probably a little bigger, since it shoots 24MP (same as D7200) images at 8.3 fps (2.3 fps faster than that D7200) and can do this for almost 3 seconds. But it's more or less in the same class.

I think card write speed matters most when you shoot those 3 seconds and have to wait a while before you can shoot another burst. Think BIF and a great moment you may miss.

Anyway, according to these measurements, the K-3 is not too far behind in that area, taking 12 seconds to clear the buffer, as opposed to 10 for the 7D II and 5/6 (ok that's faster, but it's only at 5 or 6 fps) for the D7200:
D7200: Nikon D7200 Review: Digital Photography Review
7D II: Canon EOS 7D Mark II Review: Digital Photography Review
K-3: Pentax K-3 Review: Digital Photography Review

[Alex645]

I'm sure Pentax, knowing this limitation, endowed the K3II with a larger buffer. After all, the K3II is rated to shoot RAW at up to 8.3 fps, the D7200 at 6 fps, and the T6i up to 5 fps, even though they are all 24 MP sensors with approximately the same size files. Of course, once the buffer is full, the fps will drop to the write speed somewhere around 2 fps or less.

The K3II's Prime III Image Processor was first introduced in Oct. 2013 with the K3.
The D7200's EXPEED 4 Image Processor was first introduced in Oct. 2013 with the D5300.
The T6i's Digic 6 Image Processor was first introduced in Sept. 2014 with the 7D Mark II.

To me, it is not a coincidence that both Nikon and Pentax share the exact same size Sony CMOS chip, and new processors were released in the exact same month.

And Pentax has to design for distinction. The Canon T6i is not WR and uses a pentamirror OVF. The Nikon D7200 is the most expensive in your comparison group. Anyone leaning toward Pentax won't be doing so because of superiority in AF, continuous shooting rates, or a cool red accent on the body. Pentax is picking their battles and I like that they are not a Scanikony clone. Pixel shift, astrotracer, a superior SR system all lend itself to distinction from the crowd. And ultimately the fps wars favor SLT and mirrorless cameras and at that point if it matters, you're into a Nikon D4s or Canon 5Ds.

Pentax is picking its battles and I like their choices and when needed, compromises. Should this be something Pentax looks into? I'm sure they have and are fully aware. Does this point to low performance of underlying hardware? You get what you pay for. Arguably, the D7200 cuts the least amount of corners, but there is still room for improvement, but by that time, they're in the FF production costs. The T6i has the newest processor, but the worst buffer. I'd rather get the K3II and spend the money saved on a better lens.

A final way to analyze the situation of Pentax's bigger buffer, slower write speed vs. Canon's smaller buffer, faster write speed: If both cameras give you 50 frames in 10 seconds, would you rather have the K3II giving you 25 in the first 3 seconds vs. the T6i 15 in the first 3 seconds or the K3II's 6 in the last 3 seconds vs. the T6i 15 in the last 3 seconds?

[starbase218]

It occurred to me, that what matters is duty cycle. The K-3 can shoot bursts of 23 frames / 8.9 fps = 2.6 seconds (DPReview says it's 8.9 instead of 8.3). Then though, you have to wait 12 seconds before you can shoot another burst. The burst themselves are of high quality, e.g. 24MP 14-bit RAW files at 8.9 fps. But they don't last too long and the 12 seconds is a long time to wait when you see that bird flying. So duty cycle is 2.6/10.

The 7D II's duty cycle is a bit better: it can shoot for a full 3 seconds and you have to wait 10 seconds for the buffer to fully clear. The burst itself is also of high quality, that is 10 fps (better than K-3), 14-bit (same as K-3), 20MP (worse than K-3, and it has an AA-filter so it won't be as sharp). But duty cycle is 3/10: better than K-3.

The D7200 takes a different approach. When shooting 12-bit RAW (worse than K-3), it can burst at 6.1 fps (worse than K-3), at 24MP AA-filterless (same as K-3). However, its buffer will clear in 6 seconds. And the burst continues for almost 4.6 seconds. So duty cycle is 4.6/6. When shooting 14-bit RAW (same as K-3), it can burst at 5 fps (worse than K-3), again at 24 MP AA-filterless (you guessed it, same as K-3). It can keep this up for 3.4 seconds and you have to wait 6 seconds for the buffer to clear. So duty cycle is 3.4/6.

Overall, if you can plan your bursts, the K-3 or 7D II will probably get you the best results. If you can't do that, the D7200 may be better for you. Or you can set the K-3 to a medium burst rate of 4.6 fps: you can shoot for almost 7.2 seconds. The buffer will still take 12 seconds to clear though. This wil improve the K-3's duty cycle (7.2/12).

edit: I am assuming the same goes for the K-3 II. Not 100% sure though.

[Larrymc]

Pentax K3II Write speeds.

Canon Rebel T6i/T6s.

Nikon D7200.

We see that same cards are giving about the double the speeds on Nikon and Canon compared to Pentax. K3II tops out at 36 MBps while Nikon and Canon give almost double at 73MBPS.

1. Should this be something that pentax look into?
2. Can this be fixed in firmware update or it points to low performance of underlying hardware?

1. Probably not.
2. Again probably not but it gives some of us something to bitch about!

[starbase218]

Personally, the autofocus speed is what bothers me way more than the SD card write speed on Pentax cameras.

[nomadkng]

Personally, the autofocus speed is what bothers me way more than the SD card write speed on Pentax cameras.

well, there's about 2942 threads hear on PF complaining about Pentax AF, so you should feel right at home...lol

let's hope the K1 addresses a lot of these concerns.

[Ron_Man]

In real world experience here: My K-3 is faster doing continuous burst and displaying the preview images in Raw much faster than my friend's D7200 shooting jpegs with Extreme Pro 95 mbps cards. And that right there is what matters more to me

[Larrymc]

In real world experience here: My K-3 is faster doing continuous burst and displaying the preview images in Raw much faster than my friend's D7200 shooting jpegs with Extreme Pro 95 mbps cards. And that right there is what matters more to me

That's good to hear........real world always trumps the theoretical every time.

[Zafar Iqbal]

Writing to a single card is fine for me, but writing the same photo to two cards becomes humiliating slow, even without going crazy with burst shooting.

Besides that....

Personally, the autofocus speed is what bothers me way more than the SD card write speed on Pentax cameras.

.... +1

[starbase218]

well, there's about 2942 threads hear on PF complaining about Pentax AF, so you should feel right at home...lol

let's hope the K1 addresses a lot of these concerns.

If I have to get the K-1 to get better AF speed, I'll be seriously considering moving to Canon or Nikon.

Having said that, I think it's mostly the in-lens motors that are just slow (*cough* SDM *cough*). So Pentax should be able to fix that.

[lm4187]

Pentax K3II Write speeds.

Canon Rebel T6i/T6s.

Nikon D7200.

We see that same cards are giving about the double the speeds on Nikon and Canon compared to Pentax. K3II tops out at 36 MBps while Nikon and Canon give almost double at 73MBPS.

1. Should this be something that pentax look into?
2. Can this be fixed in firmware update or it points to low performance of underlying hardware?

If one scrolls down half-way in the links listed by the OP, to frames written in 30 seconds of continuous shooting, the comparisons are much closer.


Pentax K-3 II: RAW+JPG-> 58 frames, RAW-> 71 frames, JPG-> 127 frames. [The average RAW (DNG) file size was 28.2 MB and the average JPEG file size was 9.6 MB.]
Canon Rebel T6i/T6s: RAW+JPG-> 54 frames, RAW-> 87 frames, JPG-> 142 frames. [The average RAW file size is 26.5 MB, average JPEG file size is 6.2 MB.]
Nikon D7200: RAW+JPG-> 50 frames, RAW-> 77 frames, JPG-> 127 frames. [The detailed subject in this test produces 33.3 MB RAW files and 19.7 MB JPEG files.]

The Pentax has a significantly larger buffer for the initial continuous captures. I, personally don't do many long, continuous-capture sequences, but I prefer the larger buffer to a faster write speed.

The Pentax has the highest frames per second rate, as well.

So, I'm don't think the K-3 II does all that poorly. (OK, I admit to being biased! )

Cheers,
Lou

[souless]

Writing to a single card is fine for me, but writing the same photo to two cards becomes humiliating slow, even without going crazy with burst shooting.

Besides that....



.... +1

Dear zafar,
I notice that you also post a K-3ii review so
May I ask this : do you experience this-one card writing always faster than dual card-behavior on K-3ii as well?
Thank you in advence

[Zafar Iqbal]

Dear zafar,
I notice that you also post a K-3ii review so
May I ask this : do you experience this-one card writing always faster than dual card-behavior on K-3ii as well?
Thank you in advence

Yes, I do - what matters to me is how quickly I can control the camera again. When writing to both cards, it takes much longer before I can do anything.

These are times from the moment the buffer is full, till I can control the camera (change aperture, ISO etc):

Write to single card: 23 secs
Write to both cards: 45 secs

Both cards were Lexar Pro, 95MB/s

The camera takes maybe up to twice as long to clear the entire buffer. This matters too in the long run, but for now and for me, how quickly I can control the camera matters the most and when I'm shooting, not being able to change settings for just few seconds feels like a very long time.