IS on and off

[Ron Kruger]

Some weeks ago, we had a discussion going about shooting long focal lengths with IS on and IS off. I was planning to do some experiements, but my camera was in the shop, then the ice storm. Instead of trying to find that old thread, I'll just start a new one.
I just did a number of test with my 200mm f/2.8 of a distant subject comparing handheld with IS on and tripod with IS off (I was also testing to be sure AF was accurate at infinity).
To my surprise, the images with IS on (handheld) were sharper than those from a tripod with IS off.
All shots were with a Pentax K20D (in camera IS) in good, strong light, progam P, a little over 1/600, same f-stop, ISO 100. At normal view, they both looked the same, but when I magnified them 10 clicks on Infranview full screen, those with IS on were noticably better. I actually expected the opposite.
I also tested the same way with ISO set at 800 (1/2000, f/10) to also check for noise difference. The results for IS on and IS off were the same, and I could tell no difference in noise levels.

[Loupey]

It's difficult to make a direct comparison (I know, I've tried ).

What tripod is "equivalent" for testing purposes? How steady are you naturally?

I think each person needs to perform his/her own tests using their own equipment (including the tripod normally used) to see how the IS stacks up on an individual basis. I know that, for me, my 4 IS lenses either handheld or on a monopod is good enough for a majority of my shooting.

I also tested the same way with ISO set at 800 (1/2000, f/10) to also check for noise difference. The results for IS on and IS off were the same, and I could tell no difference in noise levels.

Noise wouldn't change at the same ISO.

[mjs1973]

I'm curious if you used a cable release or the self timer on the tripod shots? For me, I always try to use a cable release if possible when shooting on my tripod. With wildlife, that isn't always possible though. If the critter is moving, I have my ballhead loose, so I can track the subject, and use the tripod as a pivot point. I don't have any IS lenses yet, so trying to find the best technique is key for me.

[Ron Kruger]

Sorry, Loupey, I wasn't very clear. In addition to testing hand-held (IS on) and tripod (IS off) at the same settings, I tested the same way at two different ISO settings to check for noise (which also affects clarity) between the hand-held and tripod shots taken at 100 and 800 ISO--not at the same ISO.
I'm pretty steady off-hand, but I doubt I'm as steady as even a cheap tripod, so I came away from this very confident in my Pentax's in-camera IS.

[Ron Kruger]

No, MJS, I didn't use a cable release. I don't even have one, because I found it impractical for field work. While there is a possibility that I caused some camera shake on my cheap tripod when I pressed the shutter release, I'm pretty smooth and careful about it.
As I mentioned, the difference wasn't noticable until I magnified the images 10 times, but the results were opposite of what I had expected.
Now you've got me wondering if I mixed them up, so I'll do it again for conformation.

[AgingEyes]

Leave it on if you want to. Many people do with no harm to image quality noticed. Mine is on. But Nikon says mine can be on even on tripod

[Ron Kruger]

I did the experiment again--same results.
I did notice some difference, however, in saturation and hue between IS on and IS off. What's up with that?

[AgingEyes]

I did the experiment again--same results.
I did notice some difference, however, in saturation and hue between IS on and IS off. What's up with that?

Never heard of it, never happened to me or people I know. Have you talked to other Pentax users?

[mjs1973]

I did the experiment again--same results.
I did notice some difference, however, in saturation and hue between IS on and IS off. What's up with that?

Could it be a white balance issue? If your camera is set to Auto WB, it could have changed between shots?

[Ron Kruger]

Well, the thing that caused me to investigate this IS on or off thing in the first place, is that I noticed a slightly different look in exposures. It's not a great difference and requires a pretty critical eye. I wasn't sure what it was at first, but I'm fairly certain now it is not noise or clarity, but saturation and hue.
It's a consistent difference I've noticed between various shots taken under the same lighting with the same settings, except IS on and off. I imagine you're right about Auto WB, Mike, but I don't think it is a difference in lighting conditions between shots, but some effect IS on and off is having upon the WB reading.

[Singletracklovr]

Hi Ron,
Could you post an example? Or would the compression to jpeg and resizing skew the results?


Not to change the topic of your thread.
I would love to see some results from your ISO 100/800 experiments. Are you post your results on that experiment?
What exactly do you look for when comparing the two photos? Is it just how black [no noise] the black parts of the photo are?

[mjs1973]

I imagine you're right about Auto WB, Mike, but I don't think it is a difference in lighting conditions between shots, but some effect IS on and off is having upon the WB reading.

With auto WB, the lighting doesn't have to change at all for the camera to change the WB setting. This is reason I don't use Auto WB.

[Ron Kruger]

First to answer some replies, then I'll post the results from my recent tests.
Bob: My tests are for my camera (K20D), and I'm suspect results may vary between brands and even models of the same brand, so I suggest everyone conduct similar tests. I shot all the test images in RAW/JPEG, using exactly the same scene, and saw no appreciable differences between RAW and JPEG images in my camera's computer software imaging program or on Infranview. Besides, I've already deleted the test results, after considerable study and making notes. Maybe later, I'll take the time to shoot some more and process them and post them here, but I have some writing deadlines dogging me now.
The results of the ISO experiements are posted in the Pentax thread, because they are even more specific to the K20D.
MSJ: Because all of my serious shooting is done under the constantly changeing conditions of the outdoors, and because most of my shooting is during spur-of-the-moment situations, I don't have time to shoot test images, study histograms and such such, so I must rely on Auto WB. My subjects move a lot, and none of them want to be famous. Besides, I don't feel I know enough about it to override the auto function.

[mjs1973]

I understand where you're coming from Ron. I don't mess around with WB in the field either. I set mine on daylight, and leave it there. If I need to change it, it's a simple click of a button in my RAW converter. The reason I brought it up was because it could be the cause of the differences you are seeing. If you do another test, you should be able to check the color temperature in your RAW converter to see if they are exactly the same.

[Ron Kruger]

I believe the difference in saturation and hue that I'm seeing between the IS settings has to do with how the auto-exposure is reading when the IS is on and when it is off. I have no idea why, but with the same shots, of the same subject, distance, framing and ISO settings (100), I got 640th sec at F7.1 with IS off and 500th sec at F/5.6 with it off. So I did it again, thinking maybe the light changed between shots (a matter of a few seconds). On the second test I got 500th at F/5.0 with IS off and 400th at F/5.0 with it on.
Now, while this proves the light was changing because of the different exposures between tests at the same IS setting, it does indicated that the readings were varying each time between IS on and IS off.
So I believe the saturation and hue are the same regardless the settings, but the degree of that saturation/hue varies according to the f-stop/shutter speed relationship, which for some reason that eludes me is changing between IS on and IS off. As a general rule, exposures toward the darker side are going to have more saturation (and more dramatic hues), while those on the brighter side have less.These exposure parameters can be controlled by underexposing by -1/3 or over exposing by +1/3 as needed, or by auto-bracketing shots and then picking the one with the perfect exposure/saturation/hue. (Exposure, of course, can also be altered in PP software to get the same thing, but my attitude is to get the best possible shot from the camera and do as little in PP as possible to maintain the integrity of the image, especially with JPEGs.)
As I mentioned above, I shot all the test images in RAW/JPEG, and saw no appreciable differences between RAW and JPEG images in my camera's computer software imaging program or on Infranview.

Generally speaking, I'm getting a slightly darker exposure with IS off than with IS on, and this is why those shots appear to have more saturation and a slightly richer texture, which I actually prefer for images I submit for publication.
The tests I just conducted re-confirm that general image sharpness is better with IS on than with it off and on my tripod. (This time, I might add, I set my camera on mirror-up, 2 sec delay for tripod shots to cut down on possible mirror slap and shake during exposure.) I will admit, however, that my tripod is a WalMart bargain and no were near professional quality.
That’s because I find a tripod impractical for the kind of wildlife photography I do and use a good monopod instead. Most of the time, however, I hand-hold my 200mm F/2.8 for shooting critters.
While my experiments have answered some questions about IS, and have given me some clues to getting the perfect exposure with it either on or off; and while it has renewed my faith in my K20Ds in-camera IS technology, this does raise the puzzling question of why the exposure meter is reading slightly differently between IS on and IS off?

[mjs1973]

Ah, now things make a little more sense for me. A slightly different exposure would have an effect on the colors. I didn't know you were using AE. My guess is that is the reason you are getting slightly different exposure settings. There is no way that you are going to be able to frame the same image exactly by hand holding do to movement. It wouldn't take much movement to introduce a little more highlight, or shadow into the frame when you move to turn the is on or off. Depending on the metering mode, this difference could be quite large.

If you have time, and want to dig deeper into this, I would test in manual mode, so you know that your exposure isn't going to change at all. Eliminate as many variables as you can. Tripod, manual exposure, manual focus, manual WB etc. I like to do my testing in doors, with a subject that isn't going to move, and lighting that I know isn't going to change. Of course this isn't going be the same way you will be shooting in the field, but it may help you get to the bottom of your questions.

[Singletracklovr]

First to answer some replies, then I'll post the results from my recent tests.
Bob: My tests are for my camera (K20D), and I'm suspect results may vary between brands and even models of the same brand, so I suggest everyone conduct similar tests.

I hope I am not hyjacking your thread...I don't mean to be...your test just sound so interesting.
Being new to digital slr's I seem to go out shooting with my camera set to different parameters everytime.
I have come back with all orange photos, photos over exposed/underexposed, out of focus, blotchy everything.
Thank God I did not buy this camera and go on a world tour...
I decided I need to just go out in the backyard with a detailed plan to get these setting figured out.

I shot all the test images in RAW/JPEG, using exactly the same scene, and saw no appreciable differences between RAW and JPEG images in my camera's computer software imaging program or on Infranview.

Interesting, I can usually see a slight difference, mostly in the hues of the shadows if I put the photos up side by side. It might be worth checking your control panel color depth. When my daughter plays a game on my computer it sometimes switches the monitor color depth to 256 colors (8bit) I'm always checking that it's in 32million colors.

Besides, I've already deleted the test results, after considerable study and making notes. Maybe later, I'll take the time to shoot some more and process them and post them here, but I have some writing deadlines dogging me now.

I understand, sorry if I put you on the spot.

The results of the ISO experiements are posted in the Pentax thread, because they are even more specific to the K20D.

Thanks, I can't wait to try this test as well.

[Ron Kruger]

MSJ: Those are excellent suggestions. I realize a controlled experiement in the outdoors is not cont completely controlled, but that is where I live with my camera, so although my results are open to scientific speculation, all I intend to draw from them is general principles that I can apply to actual field work.
Indoor and studio photography is a whole other world, and I think the results I would get with artifical lighting would not be as applicable to outdoor lighting, and that's where I do everything since my wife won't pose nude for me anymore.
I do, however, want to dig deeper into this and will take your suggestion for more controlled exposures and such at a later date.

[Ron Kruger]

FSJ: During my tests, I'm using my 200mm F/2.8 and shooting the front porch of my neighbors house about 1/8 mile away. I'm setting the camera for tripod shots first and putting the exact center of the frame upon his front door. As soon as the shot snaps, I'm grabbing the camera from the tripod, switching IS and taking the IS on shot from the same spot as quickly as possible, but being sure to again center the frame upon his front door.
Given the focal length and the controlled framing, I doubt the exposure would change between shots as it might if I was using a closer subject or a slightly different framing.

[Ron Kruger]

Bob: I certainly not new to photography, but am fairly new to digital, so I'm on a learning curve too.
This is my slow time. I've often said the only good thing about February is it only has 28 days. So it is when I conduct various experiments, test lenses and such to learn things that will help me in the field throughout the rest of the year.
You didn't put me on the spot. I believe, based upon MSJs suggestions, that I'll be conducting further, more controlled tests, and I'll post some shots here next time. I probably should have thought of doing that in the first place, because someone may notice something I'm missing.
I have calabrated my screen and set it to load those paramaters each them the computer is turned on, so I think what I'm getting in my camera is what I'm seeing on screen.
And I sincerely hope more people try these tests so we can compare different makes and models, especially those shooting in-lens IS.
The differences between those technologies posed in posts here a couple of months ago is what prompted my investigation about IS in the first place.

[Ron Kruger]

I am, by the way, shooting everything in aRGB these days. Even though that is a different color space than sRGB, for comparision purposes, I can't imagine how it would make any difference from one shot to the next, all shot in aRGB.

[Ron Kruger]

Another by the way:
I belong to Pentax's Professional Program and have asked one of there Product Specialists to check out these threads and comment, maybe offer some insights--especially about the exposure differences I've noticed with IS on and off.
You know, these guys are pretty clever about designing these products for people who don't know nearly as much about photography as they, and modern cameras are like a computer in your hands. It could be that they realized when someone was shooting with IS off, they would benefit from a 1/3 bump in exposre (slightly higher shutter speed), or when they turned it on, they could get by with a little slower shutter speed and more saturation.
It might be built into the firmware to benefit rubes like us without even knowing it.
I don't know. Just a thought.

[Loupey]

... I got 640th sec at F7.1 with IS off and 500th sec at F/5.6 with it off. So I did it again, thinking maybe the light changed between shots (a matter of a few seconds). On the second test I got 500th at F/5.0 with IS off and 400th at F/5.0 with it on.
...

This defeats the original purpose of your test (comparing IS/handheld verses non-IS/tripod). Whenever you test anything, you always leave everything that you're not testing unchanged so as not to introduce any variables into the test (which you got).

[Canon_Bob]

I see a lot of confusion in this thread. First off, Image Stabilization has absolutely nothing to do with image hue, saturation, or noise. None. It is nothing more than a mechanical counter-movement device that either moves the lens elements (Canon/Nikon), or the sensor (Olympus/Pentax) to compensate for movement in the opposite direction. It's really no more complicated than that. It typically uses a small gyroscopic sensor to tell the camera to adjust the lens element angle or sensor opposite the perceived movement. that's all.

I think the conversation a while back may have come from a statement I made about turning off IS at extreme focal lengths when shooting from a good tripod or rest. The reasoning behind this is that at extreme focal lengths EVERY movement is amplified....even that of the IS mechanism. What was happening to me (typically at focal lengths of 1000mm or more) is that a very slight movement (like pressing the shutter) would cause IS to activate. that would in turn adjust the element in the lens, which would then move the highly magnified single point of focus off target. This would then cause the camera's auto focus to go into "hunting" mode. If you're lucky, 15-20 seconds later you'd be back on target! The best remedy was to simply shut off IS, and thus significantly reduce the lens from going into focus hunt mode.

I think it's a completely irrelevant discussion at focal lengths that are sub 500mm. ANYTHING that is hand held will benefit from IS. Nobody is THAT steady....I don't care how much they brag about how steady handed they are. People wobble (but they don't fall down....sorry....just had a "Weebles" moment there). I don't think I've ever turned it off on my intermediate telephotos.

Now, with all that long winded garbage I just typed.....Canon has just released an incredibel 4 stop IS system with thier new 800mm f/5.6 lens. I stood next to Artie Morris on Saturday morning in dense fog, and watched him nailing shots with a 1.6 CF camera (the 50D) on the 800mm f/5.6....giving him an effective FL of 1280mm...and shooting at 1/30th of a second!!!!!! That system is incredible. If I try that with my 600, all I get are artistic blurs :lol: :lol:

[Ron Kruger]

I doubt there is the same effect in IS lens technology, but I am seeing a difference in IS in-camera technology. They may both be trying to accomplish the same thing, but they are doing it from different angles and with different technologies.
Something is happening to saturation and hue between IS on and IS off, and I believe it is due to slightly different (about 1/3-stop or so) auto-exposure readings between IS off and IS on.
I may be confused, as you suggest, or stupid, as Loupey suggests, but something is going on.
I didn't start taking photos yesterday.