Forgot Password
Pentax Camera Forums Home
 

Reply
Show Printable Version Search this Thread
12-04-2008, 09:47 AM   #16
Veteran Member
kristoffon's Avatar

Join Date: Aug 2008
Location: Brazil
Posts: 532
QuoteOriginally posted by Blue Quote
Doesn't that defeat the purpose of 1/500 shutter speed setting?
No, if a single pixel is exposed for no more than 1/500s it won't have time to get blurred.

12-04-2008, 10:00 AM   #17
Moderator
Site Supporter
Blue's Avatar

Join Date: Jun 2008
Location: Florida Hill Country
Photos: Gallery | Albums
Posts: 17,210
QuoteOriginally posted by kristoffon Quote
No, if a single pixel is exposed for no more than 1/500s it won't have time to get blurred.
The part that I don't follow is the part about the slot being exposed for 1/180 regardless of what the speed above that is set at. I'm guessing its the width of the slot that is changing.

Edit: In other words, it seems like a given pixel should be exposed for 1/500 of a sec if that is what the shutter speed is set on.
12-04-2008, 11:08 AM   #18
Inactive Account




Join Date: Nov 2008
Location: Slovakia
Posts: 31
QuoteOriginally posted by Blue Quote
The part that I don't follow is the part about the slot being exposed for 1/180 regardless of what the speed above that is set at. I'm guessing its the width of the slot that is changing.

Edit: In other words, it seems like a given pixel should be exposed for 1/500 of a sec if that is what the shutter speed is set on.
What he probably wanted to say, the speed of shutter curtains is "about 1/180 s". That means, if you select in camera 1/180, then the first curtain will move the entire distance it has, it's the same moment the second curtain is starting to move. So in case you've been shooting vertically (or portrait orientation) a moving object, because of the speed the curtain moves, and the movement of the object that might result in some prolongation/shortening (although surely blurred) of in the resulting picture.

But the shutterspeed (set in camera) is the amount of time the given pixel at sensor is exposed, so it stays effective, but this "1/180" corresponds to the sync speed. I don't know it for sure, but it might be so. But it makes sense. If the flash itself lasts only few 1/10000 of a second, then you have to have fully opened shutter to be avoid get rid of black (underexposed) bar. And the shortest time period that corresponds to fully opened shutter is the sync speed.

QuoteOriginally posted by Ole Quote
It takes about 1/180th of a second to expose the frame no matter what shutterspeed at or above 1/180s is set. At 1/500s, for example, the shutter works like a narrow slit that travels across the sensor in about 1/180s. Each part of the sensor is exposed for only 1/500s thanks to the slit, but it takes 1/180s before the entire frame has been exposed!
Doesn't it take approx. sync speed + shutter speed for any image? The second curtain will start to close after shutter speed, but it takes the same time (sync speed) to travel the entire distance. So practically for fast photos you the sync speed is much of a factor than when doung long exposures.

For this reason any image image stabilisation won't work, if sharp pictures are desired with long lenses using (considerably) higher shutterspeed than the sync speed. So in this case, when 1/1000 is approx. only one fifth of the sync time (1/180). On the other hand, with long exposures (like 1/15 handhold) when the exposure time is 12 times more the SR work (considering also shorter lens). Therefore, focal length is a big factor, considering the fixed 1/180 sync speed.

Last edited by Myn.pheos; 12-04-2008 at 11:52 AM.
12-04-2008, 04:03 PM   #19
Pentaxian
falconeye's Avatar

Join Date: Jan 2008
Location: Munich, Alps, Germany
Photos: Gallery
Posts: 6,863
QuoteOriginally posted by Ole Quote
It takes about 1/180th of a second to expose the frame no matter what shutterspeed at or above 1/180s is set.
This argument ist important when it comes to reason about a possible lag in the start-up of the SR system.

There are 3 possible limitations in the SR system for long focal lengths:
  1. The SR system starts up late and with a short exposure time, the image has already been taken. This partly is the case even with Pentax where the SR system starts with half button press but counter-movements only start at full button press.
  2. There is a latency between detected shake and counter-movement which may be too large with very short exposure times.
  3. The amplitude of required counter-movements is too large to cope with.
As far as #1 is concerned, this should only affect the part of an image which is exposed first (one vs. other half). Obviously, this effect has never been reported so far. So it is safe to ignore it.

As far as #2 is concerned, the shake reduction system does a fourrier decomposition of detected shake and will correct for a detected shake frequency and amplitude using a phase shift which exactly compensates the SR system's electro-mechanical latency. This is the main reason why the SR system needs some time to start up. As it allows to manually enter focal length up to 800mm, I guess that it is designed to work with such large focal lengths, read short exposure times, as well.

As far as #3 is concerned, it is possible to observe in Live View that huge sensor movements are possible (this would probably have to be reduced somewhat with a full frame sensor). Because movements start at full button press only, it should work even at 500mm during the 1/180s taken by the shot.
QuoteOriginally posted by kevbirder Quote
The reason that I think the shutter speed may be important is that there is probably a small time lag between the movement and the SR correction. For a slow wobble as seen in a wide angle or normal lens, if the SR starts to kick in after say 1/500th sec with camera set at 1/30th, then only a very small movement will have occurred prior to the correction. But of course if the shutter speed is set to 1/1000th then the SR wouldnt have even started to correct before some or all of the picture has been taken. With focal plane shutters I suspect that the curtain might be part way across when the SR starts working?
Look at my above argument to understand that those aren't the worries. But of course, the calibration of the entire SR system must be 10x as precise as required for shorter focal lengths.

From published test results, it can be seen that the Pentax SR system excells at 1/15s. To a point that it may perform as well or even better than at 1/30s...

My gut feeling is that the gain of SR drops from 2 EV at 1/15s to maybe only 1EV at shorter exposure times and longer focal lengths.

The idea to make experiments is a great one!

In order to get consistent results, one may use the rule that the SR system does not consistently increase the possible exposure time. It rather increases the percentage of images with acceptable blur at a given exposure time. So, rather than looking at individual images, count the number of images which look sharp enough with SR on and off. The percentage figures should be more consistent then.

BTW, this means that effective use of SR means always making 2-3 shots when SR does matter.

12-04-2008, 05:52 PM   #20
Pentaxian
Lowell Goudge's Avatar

Join Date: Jan 2007
Location: Toronto
Photos: Gallery | Albums
Posts: 15,397
QuoteOriginally posted by Blue Quote
Doesn't that defeat the purpose of 1/500 shutter speed setting?
the moving slit still helps freeze the action. I would have to question a little the math, and I could never get pentax to divulge the settings exactly, but note.

1/180 is the maximum flash sync shutter speed, but that really means that the effective time the full frame is exposed is 1/180 of a second, that includes the leading curtain traverse time, the full frame open period, and the trailing curtain closing time (past any point)

BUT the 1/180 second curtain travel time is too long, because the flash would need to fire in zero time, and it does not.

I seem to recall somewhere that the "open time" at full frame is between 2 and 4 mS, or between 1/250 and 1/500 of a second, therefore the time for the curtain to travel needs to be between about 1/300 and 1/640 to make the whole thing work out to 1/180 of a second.

high shutter speed will not "freeze" totally the shake, and will, if shake is bad enough, distort the image. Years ago people took photos of a speeding race car with a biug vertical shutter (like on a speed graphic) and the wheels turned into ovals as they moved horrizontally while the slit in the shutter moved vertically. Within each narrow section the image was only slightly blurry, but the whole thing wwas distorted. You would get a similar effect even with a high shutter speed on a DSLR and a long lens, due to the relitive speed the image "shakes" or moves. This might be a slight advantage for optical stabilization which keeps the image in one place on the sensor.
12-04-2008, 09:41 PM   #21
Loyal Site Supporter
Canada_Rockies's Avatar

Join Date: Nov 2007
Location: Sparwood, BC, Canada
Photos: Gallery | Albums
Posts: 9,401
It is possible, given various different results reported, that the original poster is not waiting for the SR to be engaged. It will only work when one waits for the little "shaky hand" to show up in the display.

As to the shutter speed discussion, check out some early automotive action shots taken with 4x5 and 8x10 cameras. The wheels on the 1915 racing cars are not perfect circles. The wheel has moved during the time the curtains pass over the wheel. Note that the wheel is still sharp (excepting spokes of course), but misshapen because of the time it took the curtains to pass over the car. If the car is moving right to left, and the curtains go from top to bottom across the film, the bottom of the car will appear further to the left in the frame. The curtains take approximately 1/180 second to pass over the sensor on the k10d. Faster speeds release the trailing curtain before the leading curtain reaches the end of its travel. The slit passes over a given pixel in the amount of time you set on the shutter speed.
12-05-2008, 04:22 AM   #22
Pentaxian
RobA_Oz's Avatar

Join Date: Nov 2008
Location: Tasmania, Australia
Photos: Gallery
Posts: 4,311
Shake isn't just "shake"

For those of an enquiring mind, this is an interesting matter to consider. I believe we may not be considering all aspects of what we simply refer to as camera "shake".

As an occasional long lens user (I have a Pentax 400-600 mirror zoom), I'm aware of the need to support the lens well, to get the best result. Like kevbirder, I've been disappointed with many of my hand-held results, which has dismayed me a little, as I used to pride myself (and still do, to a degree) on my ability to use technique to get a reasonably sharp picture with a slow shutter speed, hand-held.

It shouldn't be a surprise that the blur problem with a long telephoto is greater than with shorter focal length lenses, if you consider the kinematics of the situation.

Usually, we have the back of the camera pressed against our face when we compose and shoot. As our necks aren't perfectly rigid (except for the unlucky few, of course) we effectively mount the camera on a pivot somewhere between the back of the head and the shoulder-line. Therefore, what we call "shake" isn't simply an up-and-down movement, it's also a "nodding" movement, which is mostly rotational, probably combined with side-to-side rotation and a vibration overlaid by muscular tremor. Add the effects of wind and whatever other forces are acting at the time, and you've got a very complex set of motions, all happening at once.

On top of this is the in-camera vibration set up by the mirror movement, the focussing movement and the shutter motion itself. All of the above motions will be in different planes, too - up and down, left and right, back and forth, plus rotations.

Of course, with a long focal length lens, the narrow angle of view means that any of the motions will cause a more significant image movement than with a shorter focal length lens.

I think, therefore, that the "shake" the sensor has to try to cope with is really a very complex set of motions, whose combined effect has to be sensed in just two planes, and compensating movements made in those planes. Any of you who have done any Fourier analysis in their working lives will appreciate just how tricky it can be to analyse real-life motion and turn it into something you can stabilise.

All this comes down to how effective the built-in shake reduction in our cameras is, under these conditions. You can do anything for a price, of course. I've seen very effective aerial reconnaissance motion-stabilisation systems that cost more than a whole Hasselblad (the latest one), but none of us probably wants to spend that much on a DSLR.

I suspect we may be asking a bit much.
12-05-2008, 05:31 PM   #23
Pentaxian
falconeye's Avatar

Join Date: Jan 2008
Location: Munich, Alps, Germany
Photos: Gallery
Posts: 6,863
QuoteOriginally posted by RobA_Oz Quote
Any of you who have done any Fourier analysis in their working lives will appreciate just how tricky it can be to analyse real-life motion
Yes, here!
However, I disagree that this is complex math. And by some fundamental theorem it is possible to do a Fourier analysis for any motion.

Actually, this analysis turns out to be much simpler in the case of long focal lengths. Because compensation based on the measurement of the rotational speeds around two axes only becomes exact in the limit of infinite focal length. And time variations of a given Fourier decomposition of the rotational motion signal is of lesser concern for shorter exposure times.

So, in summary: Shake can be much better compensated for with longer focal lenghts. Provided that the SR system has seen excellent calibration.

Rumors even say that mirror shake, shutter shake and the finger press onto the button have been measured in the lab by Pentax and represent a constant (pre-calibrated) shutter release motion which is assumed to be there by the SR algorithm. Which may be part of the reason why not to switch it on on a tripod...

12-05-2008, 06:10 PM   #24
Site Supporter
jmc7104's Avatar

Join Date: Feb 2008
Location: Rock Island, Illinois 61201
Photos: Gallery
Posts: 437
For my own shooting with the 50-500mm panning to follow Eagels, Hawks ect in flight, I seem to get more keepers by turning SR off also most of my static wild-life photos are shot out the window of a vehicle so use the door as a rest. With SR off you have to shut off engine to eleminate that vibration. High Iso and high shutter speed for panning. I don't have much success with a tripod shooting wild-life, especially birds in flight, However works great at bird feeder. jim
11-29-2009, 01:26 PM   #25
Pentaxian
Class A's Avatar

Join Date: Aug 2008
Location: Wellington, New Zealand
Posts: 9,160
QuoteOriginally posted by kevbirder Quote
I am wondering if this shot was taken before the little hand appeared though and might be blur induced by the SR system.
You can check the EXIF data of the image. The camera records whether shake reduction had stabilised or not.

FWIW, note that Pentax cameras measure rotational movement only. If you set the camera on a tripod and the latter is indirectly moved by other people because of a wobbly ground and the movement is perfectly translational then the camera wouldn't correct for it.

QuoteOriginally posted by Lowell Goudge Quote
This might be a slight advantage for optical stabilization which keeps the image in one place on the sensor.
No, the phenomenon is not affected by different image stabilisation systems as it caused by exposing different parts of the sensor at different times. With motion in the image, even a perfectly stabilised image will have parts or all of it elongated in the direction of the movement.
Reply

Bookmarks
  • Submit Thread to Facebook Facebook
  • Submit Thread to Twitter Twitter
  • Submit Thread to Digg Digg
Tags - Make this thread easier to find by adding keywords to it!
camera, dslr, hand, lens, photography, reduction, shot, shutter, speeds, sr
Thread Tools Search this Thread
Search this Thread:

Advanced Search


Similar Threads
Thread Thread Starter Forum Replies Last Post
SR and fast shutter speeds future_retro Troubleshooting and Beginner Help 10 08-12-2016 01:30 PM
K-7 and shutter speeds in P matthis Pentax DSLR Discussion 10 10-08-2009 02:22 AM
Minimum shutter speed with shake reduction eyou Pentax DSLR Discussion 12 07-30-2009 09:41 AM
Using high shutter speeds swhang Troubleshooting and Beginner Help 14 07-14-2009 12:23 AM
shutter speeds ajaxmgr Troubleshooting and Beginner Help 5 01-03-2009 01:45 AM



All times are GMT -7. The time now is 07:49 AM. | See also: NikonForums.com, CanonForums.com part of our network of photo forums!
  • Red (Default)
  • Green
  • Gray
  • Dark
  • Dark Yellow
  • Dark Blue
  • Old Red
  • Old Green
  • Old Gray
  • Dial-Up Style
Hello! It's great to see you back on the forum! Have you considered joining the community?
register
Creating a FREE ACCOUNT takes under a minute, removes ads, and lets you post! [Dismiss]
Top