Testing Shake Reduction on a K100D (camera+washing machine)

Hi everyone

I've browsed this forum for ages, but only now have I had anything to contribute - and it's a question rather than a solution or critique, I'm afraid.

I've been using my K100D for three or four months now, with kit lens, 50/1.4 AF and manual Vivitar 28-70/3.9-4.8 - and I've had great fun with it.

I take lots of hand-held indoor shots, and most of them are soft - in fact, even in bright indoor light (think supermarket), my 'nifty fifty' produces soft images at 1/30s or 1/60s. Outdoors and with onboard or external flash it's pin sharp - even at f/1.4. On the internet I've seen perfectly sharp hand-held images at 1/10 taken with the K100D's shake reduction!

I'm seriously wondering if my shake reduction is functioning. My main question is: is there a good, simple test to determine whether it's working or not? Thanks.

Before you ask: it's always switched on...




I ought to explain the washing machine part. Here it is:

As long as the shutter is open, I hear a funny sound inside the camera - a bit like a tiny washing machine - and I hear it during the noise reduction period afterwards too. Is this noise normal, and does it have anything to do with the shake reduction? This has remained unchanged since it was new.


Thanks loads
Jonathan (slightly confused but ultimately excited Pentax user)

Could you post a few shots for us to look at.

If I look at my own shoots - indoor - I got a similar experience. If I use flash, everything is pin sharp.

I doubt the SR is not working - but lets see some proof

And welcome to the forum!

I think the first thing to test would be whether the results you achieve are actually different with it off.

So pick a subject, focus carefully, switch the SR off, and take ten shots (hand held, obviously) of your subject, all at a shutter speed of, say, 1/4 second. Because of the inevitability of getting a different amount of actual shake each time, you want to pick a shutter speed that would pretty much guarantee blurred photos without SR. Shoot with the best technique you can.

Then take a brief break, so that the practice you've had will wear off, and repeat with SR on. Again, try to ensure your technique is as before. You can't guarantee it to be identical, but taking lots of shots will hopefully compensate for the variability.

Then inspect the shots and see if you can tell the difference.

If the SR is working, I would expect the two sets to be very different.

When SR works, it's brilliant, at least on the K10D. This was taken hand held at 1/3s, and it's not at all a one-off.

Not sure if it's the SR, but it is normal.

Might be obvious, but I thought I'd mention it just in case...

With the SR turned on, you have to half-press the shutter and wait until the little "hand" icon appears in the viewfinder before fully pressing.

If you jab the shutter straight down then SR doesn't have time to work out what's happening!

(Also, according to the manual, there are some modes where SR won't be enabled at all.)

And I'd agree... you need to set up a test scenario that is repeatable and controlled

Matt

I didn't know that

[quote="MattMatic"]Might be obvious, but I thought I'd mention it just in case...

With the SR turned on, you have to half-press the shutter and wait until the little "hand" icon appears in the viewfinder before fully pressing.
....
Matt[/quote]I didn't know that either. Thanks Matt.

Jonathan welcome to the forum and BTW, does it sound like a top or front loader?

Just as well I mentioned it then

Thanks! I'm not sure whether I want it to be faulty or not - if it's functional, great, and if it's not, then there's free improvement to be had with my kit, which is also good.


Done EXACTLY what you said. Used a 'death grip' - strap taut round my elbow, elbows on chest, held breath - and used my 50mm at ISO 200, 1/4s thus somewhere around f/8 (lens sweet spot) on the pink duty-paid seal on the back of a bottle of Drumguish 12-year single malt. (All in the name of good photography, you see.) I then realised that I don't have Photoshop where I am right now, so I can't do the batch resizing and collaging to show you a row of 20 photos just yet - but this is my description:

The first 10 (SR-off) are all motion-blurred in one axis at least. The second ten (SR-on) seem to have a smaller amplitude of shake on average, all about as good as the best two or three of the SR-off 10. But only one of the SR-on ten is acceptably sharp, with no sign of motion blur viewed pixel-on-pixel, but it's still soft for the lens and would be sharper with a shorter exposure no doubt.

So it seems to do *something*, if perhaps not effectively.

Is it worth trying it in more realistic conditions, where it has a reasonable chance of making it sharp (1/15s)?

I'm also going to post the images as soon as I can

Ah - thanks, I didn't know that either!

The big round hole is in the front so I suppose it's a front loader. Are you telling me I've been stuffing my smelly socks in the wrong hole all this time? No wonder they stay smelly, and my images are soft to boot... At least the Persil keeps my sensor clean.

Thanks a lot, everyone

Re-Matts reply;

I tried to snap away before the "SR" had locked on.......and found I couldn't, or maybe I wasn't quick enough.
My K10D doesn't mess about when it comes to SR, it seems to work as soon as the lens is in focus so I dont see how Jonathan's problem could be Matts sugestion.

Jonathan: Are you expecting too much from the SR I wonder? As for it making a noise, the only time I can hear my sensor moving is when It vibrates to clean, your description of been able to hear the sensor while the shutter is open seems a bit odd, I just tried a long exposure and had to put my ear to the back of the camera body to hear anything, in my opinion if your using shutter speeds long enough to detect that sort of niose then your pushing the SR waaaaay beyond its boundries.

If I actually had time to hear it, I wouldn't expect any kind of image hand-held at all no, I'm only expecting sharp images at, for instance, 1/30s at 30mm. I'm gonna try and find some sample images if I can.

But when it is open for long exposures, regardless of SR being on or not (as I just tested), it makes the sound as long as the sensor is exposing - whether shutter/mirror open or closed.

My sensor doesn't have a shake-off-all-the-dust function, unfortunately. I think the K100D Super does.

1/4th is something on the order of 4 stops from where you'd expect to be able to hand hold without SR with a 50mm lens.

I don't know how steady your hand is, and that obviously makes a difference, but 1/4 second at 50mm is asking quite a lot. I think Pentax only claim 3 stops for the system on the K100D, that would be about 1/8th.

On a good day, I used to be able to do about one stop better than 'rule of thumb' in my film days. With the K10D and now a fair bit of practice, I would expect to get mostly good ones at 50mm and 1/4. albeit probably not blown up to A4!

But that does mean the whole 'elbows in, breathe out, relax, squeeeeeze' thing. Which isn't always convenient.

Holding your breath in doesn't help, I believe, since the effort of holding it introduces tension, and a tight strap probably doesn't help either.

In my limited experience, it's the being relaxed that reduces the shake, not the trying to be still, which can make it worse.

I agree Chris, as a rifle shooter I can usually beat the rule of thumb by at least one stop myself, but as I said I know nothing of how steady our new friends hand is.

Sorry about the confusion - I was just using 1/4s as a testing value, as shots are pretty much certain to be shaky without SR at 1/4s. *With* SR, it would be nice if I could rely on, say, 1/30s to produce sharp pictures, and the odd shot at 1/8 or even 1/4 to be acceptable occasionally, but I'm not even sure I get reliably sharp shots at 1/30s, and I'm absolutely amazed that people get usable shots hand-held at 1/4s and 1/3s, judging by the performance of my shake reduction.

Quite frequently I find myself at a party wanting to use available light (my Vivitar 285 is beautiful but makes mortal enemies) so using say, 1/30s @ f/1.4 @ ISO 1600 is not uncommon.

(PS. Then, the problem is that the AF can't cope. Typical!)