[evilcartman]

Each experiment was done with both K3 and K3ii and in all instances the results were identical so there's no need to separate those two cameras. My methodology was simple, not exhaustive. I just wanted to get an idea of what cards to buy.

For those that don't want to wait the SanDisk SD Extreme Plus 80 MB/s 32GB card is the absolute fastest in my cameras.
This is good news because I won't have to waste money buying premium cards as this is a three generations old card.

Every card I tested was fairly close, which is surprising because their specs were very different.

My methodology was simple. I set the camera on high-speed continuous, single memory card slot, raw DWG, kept lens cap on so every image would be consistently black. I started the timer at the first exposure and I ran it until I filled the buffer plus that first slow shutter click after. I ended he timer when the orange light went off indicating that the buffer was empty.

Almost every card tested took 33 seconds and some change to empty the buffer except for the Extreme Plus 80MB/s which did it in 28 sec.

Cards tested included the latest San Disk Extreme Pro at 280 MB/s! I expected big things from that card and was quite disappointed. All of my San Disk cards from 45 to 95 to 280 all performed identically except for the 80, which gave me a whopping 15% write speed increase.

Also tested a Transcend 60 MB/s card. And my previous champion a Delkin Elite 95 MB/s card that was far and away the fastest thing for my K5s.

I have a Lexar 2000 X coming in the mail I will give an update.

I can't really explain the results nor am I trying to. This was done with cards I already have so they're almost all different sizes from 8 GB to 64GB. I have not had a chance to test same sizes of different speed cards and only a few same speeds of different size cards. I just did the test quickly because I was going to buy a bunch for this back to school tax free weekend. So ended up buying about a dozen of the 64 GB data cards.

Certainly have no problem when it comes to buying SanDisk for reliability and certainly the 80s are not slower than the other cards.

The 280s do read faster in my computer through usb 3 but not that much. Not even sure if my card reader is compatible.

I'm curious. Has is anybody else done real world K3 tests and what were the results?

---

Last edited by evilcartman; 08-04-2015 at 06:13 AM.

[Adam]

Cards tested included the latest San Disk Extreme Pro at 240 MB/s! I expected big things from that card and was quite disappointed. All of my San Disk cards from 45 to 95 to 240 all performed identically except for the 80, which gave me a whopping 15% write speed increase.

Well the 240's are UHS-II cards, which aren't supported (the K-3 lacks the second set of pins). So they revert to UHS-I speeds.

[evilcartman]

Well the 240's are UHS-II cards, which aren't supported (the K-3 lacks the second set of pins). So they revert to UHS-I speeds.

I thought it was actually UHS-3. ( UHS II class 3, great more initials) But I read so much so fast on the SD cards it all started to flow together. I did correct 240 to 280. I was kind of surprised they didn't support UHS-3 with a K3ii. It's new enough.

Cameras never seem to have a chance to keep up with the pace of memory card development. And there's only so much you can do with firmware.

I remember back when the Delkin Elite came out though. Even though the K5 didn't support UHS-1, it wrote to the card like it did. It was easily twice as fast as anything else I had. The other UHS cards took a pretty big penalty and generally performed slower than the SDXC a level down. Never could've find the pixie dust in the Delkin anywhere else. It was a good fine because I was starting to get frustrated at the K5s inability to use any of the newer memory cards. I'm going to guess thats always going to be the case. Though at the current time it seems like both capacity and speed aren't nearly the bottlenecks that they used to be. Then again I don't shoot video.

And not sure what's going on with the 80 MB/s. It would be cool if somebody was able to replicate and verify my results on this.

[Adam]

I thought it was actually UHS-3. ( UHS II class 3, great more initials) But I read so much so fast on the SD cards it all started to flow together. I did correct 240 to 280. I was kind of surprised they didn't support UHS-3 with a K3ii. It's new enough.

Cameras never seem to have a chance to keep up with the pace of memory card development. And there's only so much you can do with firmware.

I remember back when the Delkin Elite came out though. Even though the K5 didn't support UHS-1, it wrote to the card like it did. It was easily twice as fast as anything else I had. The other UHS cards took a pretty big penalty and generally performed slower than the SDXC a level down. Never could've find the pixie dust in the Delkin anywhere else. It was a good fine because I was starting to get frustrated at the K5s inability to use any of the newer memory cards. I'm going to guess thats always going to be the case. Though at the current time it seems like both capacity and speed aren't nearly the bottlenecks that they used to be. Then again I don't shoot video.

And not sure what's going on with the 80 MB/s. It would be cool if somebody was able to replicate and verify my results on this.

UHS class 1 (U1) and UHS class 3 (U3) are different designations than UHS-I and UHS-II. The latter refers to the hardware bus itself, whereas the former is a speed spec. UHS-II cards have a second row of pins that require a UHS-II compatible slot; right now those slots are generally only used in video cameras. See the SD specs here.

As for your findings, they make perfect sense. The maximum write speed most cards is slower than the maximum read speed- i.e. the 80Mb/s Sandisk card has a write speed of 60Mb/s. The other cards probably have even lower write speeds. And since the K-5 maxes out around 30 Mb/s, you don't come to appreciate faster cards in it as much.

Adam
PentaxForums.com Webmaster (Site Usage Guide | Site Help | My Photography)



PentaxForums.com server and development costs are user-supported. You can help cover these costs by donating or purchasing one of our Pentax eBooks. Or, buy your photo gear from our affiliates, Adorama, B&H Photo, KEH, or Topaz Labs, and get FREE Marketplace access - click here to see how! Trusted Pentax retailers:

[Jool]

So what are the read and write speeds of the K-3II? I just purchased a SanDisk 32GB Extreme UHS-I U3 SDHC Memory Card (Class 10).

[wissink]

Im curious as well. What is the best card(s) or card rating available that matches with the K3's performance limits?

[evilcartman]

As I said, I get the fastest buffer clearing from the Extreme Plus 80, not the Extreme Pro 95 or 280. I would expect at least equal performance from the extreme pro cards but they are 15% slower.

Most of my results are consistent in that with any cards over 45 Mb/s write speed have performances that are about equal within 1%. None of this explains the magical %15 boost I get from the Extreme Plus. I'm very much like my old K5s experience with the Delkin. There seems to be no reason for it when looking at the statistics of the cards, but it's just something in real life seems to work best.

In the case of the Delkin it was the fastest regards to offload to the computer (being the first UHS 1 card) so it worked out great. What I find disappointing is that none of the UHS II cards give me that 15% boost that the 80 Plus does. It would be worth the extra expense for how much faster they are at the computer. I just got a UHS II compliant reader and it smokes!

And I guess that's something I wanted to clarify from my original post. I didn't expect the 280 cards to be functional at 280 but I thought I might get a little extra speed from them as I did the Delkin on the K5. Many times faster cards are better optimized and even if you use them in something that's not compliant with the new standard you can often get a little better performance from them on the camera while enjoying much better performance at the computer. This is not the case here.

In short, every single modern card is testing the same on the K-3 and K-3II except for the extreme 80+ which is 15% faster. Because I absolutely must have tethering for a job, I just bought a K5IIs (not happy at the expense) which will arrive this week and I guess I will test on the cards to see if I have similar results.

btw- As an experienced computer professional I will only buy name brand cards. I've seen too many of the off brand ones fail. And do not mistake PNY for a name brand, they are amongst the worst cards of all.

[wissink]

Thanks for the information evilcartman. I need a new card, my ahem... PNY is not useful when my 16 GB SanDisk fills up. When using the PNY 32GB, my K3 constantly puts me on hold showing an hour glass. This includes viewing images, waiting for even one image to unload from the buffer let alone a couple dozen... bracketing is a nightmare... etc.

[Mikesul]

Thanks for the information evilcartman. I need a new card, my ahem... PNY is not useful when my 16 GB SanDisk fills up. When using the PNY 32GB, my K3 constantly puts me on hold showing an hour glass. This includes viewing images, waiting for even one image to unload from the buffer let alone a couple dozen... bracketing is a nightmare... etc.

That is odd. I have been using PNY 64th uhs-1 class I cards for several years and they are just as fast as the equivalent SanDisk.

[evilcartman]

That is odd. I have been using PNY 64th uhs-1 class I cards for several years and they are just as fast as the equivalent SanDisk.

I can refer you to Cnet, Tom's Hardware etc., but rarely does someone have anything good to say about PNY cards. Some flavors of the cards get absolutely terrible Amazon reviews, others not so bad. But if you look at the negative reviews you see very high rates of failure. In my personal experience they've been absolutely abysmal: both unreliable and never even lived up close to their speed rating. In general they are known as a junk brand amongst techies. I would find it hard to take chances with my photographs like that, given what I know. It's not worth the few bucks I would save.

And just when I thought I had this all nailed down, there's a new champion in town. It is the Lexar Professional 633x. With a full 2 second buffer clearing advantage over the Extreme Plus bringing it to about 25% faster than the other cards, it is now my front runner. This makes the Lexar the clear winner as it is 40% cheaper as well as 10% faster than the SanDisk.

It's not like I go around shooting continuous, filling my buffer all the time. But this is one of those things where it really doesn't put you out much to have the absolute best advantage you can have. ( cheaper too!) Then, at the one time when you really need those extra two seconds to shoot, they will be there.

On a side note, I ended up getting this little Fotasy card vault to put my best cards in and protect them. It was very cheap but little thing is built like an absolute tank, with water seals and everything. I have to say I'm kind of impressed. It's about the size of a cigarette case, and it has a slight dome shape and ribbing for extra reinforcement and is very light. It holds 12 cards securely in rubberized slots, but I might put a little piece of foam inside when you close it to discourage the cards from popping out. I'm going to use the left side for new and the right side for used so I will only get six cards a piece in then but that's enough for me.

I can send pictures if anyone's interested. But it's nice to know when you've just been spent hundred bucks on new memory cards that you have a nice place to put them. And I mentioned it because there was an old thread here about somebody looking for something just like this.

[Mikesul]

Well, it is not useful to argue these things extensively but your reference to Amazon reviews is at least misleading. The PNY 64gb which I referred to has 77% 5 stars and 11 percent 4 stars which means that 88 % of the reviewers were pretty happy with the cards. I offered you my actual experience with the cards in my cameras for over two years of use, a total of 6 cards without a single failure. Not scientific but it is at least actual experience rather than speculation. If you have used PNY cards without success you are certainly entitled to avoid them. I first used them on the recommendation of a forum member several years ago. As i said argument is pointless so I will not respond again. Just wanted to call attention to the issue of the Amazon reviews.

[evilcartman]

Well, it is not useful to argue these things extensively but your reference to Amazon reviews is at least misleading. The PNY 64gb which I referred to has 77% 5 stars and 11 percent 4 stars which means that 88 % of the reviewers were pretty happy with the cards. I offered you my actual experience with the cards in my cameras for over two years of use, a total of 6 cards without a single failure. Not scientific but it is at least actual experience rather than speculation. If you have used PNY cards without success you are certainly entitled to avoid them. I first used them on the recommendation of a forum member several years ago. As i said argument is pointless so I will not respond again. Just wanted to call attention to the issue of the Amazon reviews.

Amazon reviews can be very puzzling to me sometimes agreed. When you read the actual reviews there's often a huge disconnect between that and the star rating the same person gave them. You can't look at the just the stars, you actually have to read the reviews and that can be time-consuming. And the way the reviews are grouped, they are often not even of the same product That's why I referred primarily to TH. But PNY still have 7% 1 star ratings, where as name brand cards typically have 2% or 3% at most maximum dissatisfaction. That's more than double the amount of unhappy customers. I'm a fairly picky consumer so I tend to focus on negative reviews because those are the kind of things that are really going to bother me personally. But I read the positives to. And particularly on Amazon, you have to read everything to get the whole picture. You cannot get it at a glance.

That's one of the reasons I started this thread. Just to say "hey this card is the absolute fastest I can fine for this application".

My K5IIs came in today and I ran the same tests. The results were quite surprising. The Lexar 633x Professional, which is the reigning K-3 champion card was the absolute slowest on the K5iis, by almost 70%!

All of the other cards fell in line as they did before, pretty much the fastest being Extreme Plus and then the Extreme Pro which was only slightly faster than the other name brand cards. The old extremes at 45 MB per second along with the 60 Transcends were just as fast, so you can definitely use older cards with no penalty.

The K5iis buffer is a lot smaller than the K-3 though which, meant the timings went by faster and so there was a greater margin for error.

These kind of experiments are the reason that you can't rely just on the cards numbers. It's like DXO Mark. There are always real world factors that make huge differences in performance.

Even my buffer test doesn't mean that that card will be the fastest for you know doing live previews, tethering, or whatever.

But it gives me a little bit more surety when I go to plop down money on cards for a particular camera.

[CarlG]

It seems that you are looking for the optimal card for the camera. Just don't forget that read speed, and thus download time from the card to your PC also is a property with a card. Most PCs are more powerful than cameras. If you are reading up to 64G in one go and there is differences up to 70 per cent in that end, you will take an overall time penalty.

[wissink]

My PNY never had an issue other than speed. I'll be keeping it as a backup. I just bought an SanDisk 80 Plus 32GB. Seems decent. I haven't done any real testing though, too busy having fun shooting.

[evilcartman]

It seems that you are looking for the optimal card for the camera. Just don't forget that read speed, and thus download time from the card to your PC also is a property with a card. Most PCs are more powerful than cameras. If you are reading up to 64G in one go and there is differences up to 70 per cent in that end, you will take an overall time penalty.

That's precisely why i was so disappointed the UHS-II cards are significantly slower in the cameras than the fastest UHS-1 cards. Luckily, all the optimum cards have 80-95 Mbps read speeds.