[AggieDad]

If you are shooting a K-5II or K-5IIs, how long does it take for the hourglass to disappear after you click the shutter? Is your camera/SD card combo fast or slow or average? Do you know? Would you like to know?

To this end I am proposing a little test for K-5II owners to participate in. I am not including the K-5 because I don't know how (or if) its write process varies from the K-5II.

Now I know this wouldn't be a scientific test, given all the potential uncontrollable variables, but if there is good participation then the tyranny of large numbers can offset a considerable amount of potential inaccuracy.

Here is my proposal. Take 10 photos, each of a different subject, and measure the length of time it takes for the hourglass to disappear (i.e. the time it takes to write to the SD card). Remember, most smartphones have a digital timer, so your times should be pretty accurate.

• Be sure you have the latest firmware update in your camera.
• Start with a clean, reformatted card.
• Have the images be different enough that the file sizes with average out to something close to a norm.
• Identify your SD card as to brand, model, capacity, and advertised speed

It will be very interesting to see how our individual cameras stack up when compared to some sort of norm. I hope enough people will be interested and participate to make the results useful.

I'll post my test results tomorrow.

[sergysergy]

Does not that depend a lot on camera processing? (i.e. shadows, corrections, etc)

[Adam]

Speaking of getting the hourglass to disappear in practice- that will depend on a lot of factors, like you mentioned, including the shooting settings selected, buffer usage, file size, and file format. The most notorious setting is distortion correction, which adds over a second of processing time per frame.

If you want to measure the write speed, it's best to shoot a burst of several frames to minimize the impact of timing errors and in-camera processing overhead. If using JPEG mode, lens corrections and highlight corrections should be disabled, though keeping the camera in RAW mode is best. The timer should be started as soon as the SD access light comes on and stopped as soon as it disappears.

Having done my fair share of these tests, I can tell you that the K-5 maxes out at a write speed of roughly 30 Mb/s. Most UHS-I cards these days should easily be able to sustain that speed, thus outperforming the camera. Class 10 cards could be anywhere between 10Mb/s and 30Mb/s, whereas lower class ratings should correspond to their labeling I would strongly recommend using at least a class 10 card for good everyday performance, otherwise the camera will be locked up to write for way longer than necessary.

Here's a little demo featuring the K-5:

You Tube

You Tube

If you use these on a K5IIs, you won't wait long. Samsung 64GB PRO UHS-I SDXC U3 Memory Card (Class 10)

Shooting dance and compared to my old 64GB Sandisk Ultras (rated at 30 MB on their label), it was like getting a new camera. I don't worry about filling buffers and not being able to shoot.
20 bucks per last Black Friday. Disclaimer: Manual mode & DNG.
,

[pathdoc]

If you use these on a K5IIs, you won't wait long. Samsung 64GB PRO UHS-I SDXC U3 Memory Card (Class 10)

Shooting dance and compared to my old 64GB Sandisk Ultras (rated at 30 MB on their label), it was like getting a new camera. I don't worry about filling buffers and not being able to shoot.
20 bucks per last Black Friday. Disclaimer: Manual mode & DNG.
,

This implies that I shouldn't hesitate to buy a card which is nominally rated much higher than the write speed of my camera. If nothing else, I am future-proofing myself against the day the K-1 arrives, and in addition there are benefits when it comes to getting the data off the card and into my laptop (my new laptop arriving soon will have USB3 slots and I have a matching card reader, so even if the inbuilt SD reader isn't on the fast bus I am set to go).

[AggieDad]

Speaking of getting the hourglass to disappear in practice- that will depend on a lot of factors, like you mentioned, including the shooting settings selected, buffer usage, file size, and file format. The most notorious setting is distortion correction, which adds over a second of processing time per frame.

Actually, I think Adam may have answered all my questions and concerns with this one sentence.

I have good, fast SD cards, but I noticed that I was getting a lot of "hourglass time." It seemed this was a new thing, but I couldn't think of why. So after reading this, I checked my camera and, sure enough, at some time I apparently turned the distortion correction on. It was probably a little experiment I did and then neglected to turn it off. I really can't remember.

I will check this out tomorrow.

Thanks to all for your interesting comments.

[djam]

Okay, something must be wrong with my K-5lls. I have the exact same Transcend class 10 card used in the test and yet my write speeds were no where near as fast as the test camera. ...in fact my write times were slower than the test camera with the class 2 card! I tested in RAW format only, and then JPEG only with all corrections, etc turned off. I am running latest firmware 1.07