Sponsor Message:
Aviation Photography Forum
My Starred Topics | Profile | New Topic | Forum Index | Help | Search 
When To Use/not Use Image Stabilization  
User currently offlineplanenut767 From United States of America, joined Jun 2009, 66 posts, RR: 0
Posted (3 years 1 day 17 hours ago) and read 6557 times:

Hello I'm still learning the in's and out's of my Canon T1i (EOS 550D) and wanted ask a question. I plan on going to a few airshows this year and plan on taking plenty of pictures. I wanted to know if they are any rules of thumb on using IS especially with regard to high speed aircraft. One of the airshows I will be going to are the Reno Air Races in September and some of the planes can get up to 500 mph (Jets and Unlimited classes) so I wanted to get a few tips before I go out there and maybe get some practice at one of the local airshows. Thanks in advance for your input.

18 replies: All unread, jump to last
 
User currently offlinevikkyvik From United States of America, joined Jul 2003, 9776 posts, RR: 27
Reply 1, posted (3 years 1 day 15 hours ago) and read 6520 times:
Support Airliners.net - become a First Class Member!

Quoting planenut767 (Thread starter):

Depends on your lens(es). My 55-250 only has IS on or off. So when I photograph static airplanes, I have it on, but for moving airplanes, I turn it off (if I remember!).

My 70-300, though, has IS off, mode 1, and mode 2. Mode 1 is the same as the IS on the 55-250. Mode 2 is specifically for moving subjects - it only stabilizes vertically, and not horizontally, to avoid the IS trying to compensate for panning action.

Good explanation of IS by ckw here:

Shooting Technique... What Am I Doing Wrong? (by calibansa333 Mar 11 2011 in Aviation Photography)



"Two and a Half Men" was filmed in front of a live ostrich.
User currently offlineJakTrax From United Kingdom, joined Jun 2005, 4936 posts, RR: 7
Reply 2, posted (3 years 1 day 13 hours ago) and read 6482 times:

Quoting vikkyvik (Reply 1):
Mode 2 is specifically for moving subjects

Nearly, but not quite...

Mode 2 is actually for subjects that are static but begin to move - but as you say, effective for panning, so you're not too far off the mark. Quite why mode 1 is needed on lenses that incorporate mode 2 I'll never know. Over to you, Canon.....

IS is often (mistakenly) thought to stop motion blur, but in reality that's not the case. An object moving within a still viewfinder (e.g. an aircraft not moving from left to right but turning on an axis) will still be subject to movement, and therefore IS is of little help in such circumstances. IS is not a miracle tool and will only work WITH a competent photographer, rather than FOR.

Karl


User currently offlinevikkyvik From United States of America, joined Jul 2003, 9776 posts, RR: 27
Reply 3, posted (3 years 1 day 12 hours ago) and read 6469 times:
Support Airliners.net - become a First Class Member!

Quoting JakTrax (Reply 2):
Nearly, but not quite...

Mode 2 is actually for subjects that are static but begin to move - but as you say, effective for panning, so you're not too far off the mark. Quite why mode 1 is needed on lenses that incorporate mode 2 I'll never know. Over to you, Canon.....

Interesting. I'd be curious to hear where you've seen that, as in the manual for my lens, it says:

Use Mode 2 for continuous shots of moving subjects. Mode 2 provides image stabilization only in the direction of camera movement.

And then later in "Tips on Using the Image Stabilizer":

Mode 2:
If you are taking continuous shots of a moving subject.
If you are tracking an erratically moving subject.


In the manual for the 70-200 F2.8L, it says that Mode 2 compensates for vertical camera shake during horizontal following shots, and horizontal shake during vertical following shots.

(info for both lenses was pulled from the online user manuals on Canon's site)

In my (relatively limited   ) experience, even when shooting a stationary object, Mode 2 only provides IS in one direction.

I'm not trying to say you're wrong or anything, just trying to get some clarification.

Thanks!

EDIT: So now that I'm home, I just did some playing around with my 70-300 in Modes 1 and 2, and it seems you may be right!

I'm still glad Mode 1 is available, though, because the threshold for switching from two-axis to one-axis stabilization in Mode 2 seems to be pretty low - that is, it didn't seem to take much camera movement to make it switch. So if I want to recompose a shot of, say, something that is briefly static, I can do that in Mode 1 and quickly fire off another shot without worrying that it switched to Mode 2 and hasn't switched back to Mode 1.

[Edited 2011-07-27 18:49:29]


"Two and a Half Men" was filmed in front of a live ostrich.
User currently offlinecpd From Australia, joined Jun 2008, 4879 posts, RR: 38
Reply 4, posted (3 years 1 day 9 hours ago) and read 6453 times:

Quoting vikkyvik (Reply 3):
(info for both lenses was pulled from the online user manuals on Canon's site)

That doesn't mean much, people will still argue the point.  

For Nikon, they have two modes on the VR system for most of the higher-end lenses:

1. Active: This is for use when taking photos from an unstable platform, such as a moving vehicle, maybe a car, motorbike (eg, Tour de France style) or helicopter.

2. Normal: This controls unwanted movements (such as vertical movements) when taking panning type photos.

Nikon VR also has tripod detection in some lenses - meaning that you can use the VR while the lens is on the tripod, without drama (contrary to popular opinion). Check the user manual for your Nikon lens to determine if this applies to you or not.


User currently offlineJakTrax From United Kingdom, joined Jun 2005, 4936 posts, RR: 7
Reply 5, posted (3 years 1 day ago) and read 6409 times:

Mode 2 very simply is mode 1 until you start to move the camera - at which time it 'becomes' mode 2. Kinda renders mode 1 redundant really.

Canon's IS also has tripod detection. I wouldn't like to say whose stabilisation is better but Canon generally takes the credit for pioneering it.

Just remember, IS of any description can only limit camera shake by the operator - it can't control subject movement.

Karl


User currently offlineckw From UK - England, joined Aug 2010, 730 posts, RR: 16
Reply 6, posted (3 years 22 hours ago) and read 6377 times:

Quoting JakTrax (Reply 2):
Mode 2 is actually for subjects that are static but begin to move

Not sure where you got that from - it actually sounds like the description of the AI autofocus mode. Whether the subject is moving or not is totally irrelevant to IS. If you are shooting a moving subject but holding the camera still, then mode 1 will work just fine.

Of course if you expect to start panning if a static subject starts moving, then you better have thought ahead and have mode 2 enganged!

The key difference between modes 1 and 2 are the thresholds at which stablisation on one axis shuts down. Personally I generally leave the lens in mode 2 as for me there more risk in forgetting to engage mode 2 when needed than there is of excessive shake when shooting static in mode 2. But that's one for individual trial and error.

Quoting JakTrax (Reply 5):
Canon's IS also has tripod detection. I wouldn't like to say whose stabilisation is better but Canon generally takes the credit for pioneering it.

Not all Canon lenses have tripod detection - perhaps critical for users here, the 100-400 is an example of one that doesn't. IS should be turned off if using a tripod.

Canon was certainly the first, in part due to their changing lens mounts from FD to EF - this allowed a whole bunch of developments that were more challenging for those companies who opted to provide legacy support for older lenses.

(though note there were stabalising systems prior to IS based on powered gyroscopes attached to the camera body)

As to who is best, I think the real contention is between those that use lens based IS and those that use in camera IS (corrections are performed by moving the sensor). In my opinion, simple physics dictates that lens based IS works better with long lenses, simply due to the practical limits on how much the sensor can be moved, but for shorter focal lengths in body IS is probably as good, and of course offers the benefit of stabalising any lens you can fit on the camera.

Cheers,

Colin



Colin K. Work, Pixstel
User currently offlineJakTrax From United Kingdom, joined Jun 2005, 4936 posts, RR: 7
Reply 7, posted (3 years 19 hours ago) and read 6337 times:

Quoting ckw (Reply 6):
If you are shooting a moving subject but holding the camera still, then mode 1 will work just fine

You should try that method with the 24-105L. Mode 1 is intended for still subjects only, and any attempt at panning using IS on this lens will result in blurry shots. The only exception is if you pan very fast - in which case the IS effectively gives up trying to resist the movement.

I don't remember where I learned that mode 2 is basically mode 1 until your subject starts to move, but I can assure you that it was from a Canon source.

Karl


User currently offlineckw From UK - England, joined Aug 2010, 730 posts, RR: 16
Reply 8, posted (3 years 15 hours ago) and read 6311 times:

Quoting JakTrax (Reply 7):
You should try that method with the 24-105L. Mode 1 is intended for still subjects only, and any attempt at panning using IS on this lens will result in blurry shots.

I have   But what I said was - If you are shooting a moving subject but holding the camera STILL (ie let the subject move across the frame. Which is why I think you have got a confused message from Canon - or perhaps read a poor translation of a Japanese text (not an uncommon problem with cameras!), as the point of the AI AF mode is exactly that - if a static object starts moving, the AF switches from single shot to servo mode automatically (or at least its supposed to - in practice its a bit iffy)

Perhaps I misunderstood you when you said

Quoting JakTrax (Reply 2):
Mode 2 is actually for subjects that are static but begin to move



which implies that the IS can detect subject movement which is just not possible - mode 2 can only detect camera movement (which indeed yoy did say in your subsequent post.

But to summarise, and hopefully clarify for those who are now thoroughly confused -

Mode 1 - for use when the CAMERA is static

Mode 2 - can be used when the CAMERA is moved (ie panned horizontally or vertically). It can also be used when the camera is static.

The reason why you may not want to use mode 2 all the time is that it is less tolerant of movement in one plane than mode 2. If the movement beyond a certain threshold is detected on one axis, the lens will assume you are panning and stop stablising that axis.

SUBJECT movement is irrelevant to the IS system.

The complete opposite holds for the AF system where subject movement (or lack of) determines the appropriate mode to use.

BTW - my main source for info on how this stuff works is the EF Lens Work book from Canon - although about 50% is essentially a catalogue of Canon lenses, the rest is a mine of information on how the damn things actually work. I'm sure most people would pick up one or two gems of info which will help them get the most out of their equipment.

Cheers,

Colin



Colin K. Work, Pixstel
User currently offlineJakTrax From United Kingdom, joined Jun 2005, 4936 posts, RR: 7
Reply 9, posted (3 years 15 hours ago) and read 6308 times:

I'm confused now!!!  

Think you've said pretty much what I meant though. Mode 2 kicks in when the camera starts to pan (bad wording when I said, "When the subject starts to move"). So mode 2 is basically just mode 1 until one starts to track a moving object (and of course the IS starts to sense movement of the camera).

The other difference of course is that there's pivotal panning and parallel panning, but I'm confused enough for one evening!

Karl


User currently offlinecpd From Australia, joined Jun 2008, 4879 posts, RR: 38
Reply 10, posted (3 years 14 hours ago) and read 6302 times:

Quoting ckw (Reply 6):
In my opinion, simple physics dictates that lens based IS works better with long lenses, simply due to the practical limits on how much the sensor can be moved

I think so as well, and theoretically, it can be tailored to each lens.

Quoting ckw (Reply 8):
BTW - my main source for info on how this stuff works is the EF Lens Work book from Canon - although about 50% is essentially a catalogue of Canon lenses, the rest is a mine of information on how the damn things actually work. I'm sure most people would pick up one or two gems of info which will help them get the most out of their equipment.

I'm not a full time user of Canon cameras, but I've got that book too, it's really good.


User currently offlinedvincent From United States of America, joined Jan 2007, 1742 posts, RR: 11
Reply 11, posted (3 years 14 hours ago) and read 6298 times:
Support Airliners.net - become a First Class Member!

Quoting ckw (Reply 6):
As to who is best, I think the real contention is between those that use lens based IS and those that use in camera IS (corrections are performed by moving the sensor). In my opinion, simple physics dictates that lens based IS works better with long lenses, simply due to the practical limits on how much the sensor can be moved, but for shorter focal lengths in body IS is probably as good, and of course offers the benefit of stabalising any lens you can fit on the camera.

The amount of movement the sensor has to make to account for shake is minimal even for long lenses. A very small movement of the sensor can account for a very large movement across the final image. This was the same argument made against full frame antishake and it was bogus then and it was bogus now.

http://mhohner.de/sony-minolta/sss_ff.php

Incidentally, in-lens IS helps with its swimmy viewfinder image (which some may like, others dislike). An EVF solves this problem for in-body.



From the Mind of Minolta
User currently offlinevikkyvik From United States of America, joined Jul 2003, 9776 posts, RR: 27
Reply 12, posted (3 years 13 hours ago) and read 6293 times:
Support Airliners.net - become a First Class Member!

Quoting ckw (Reply 8):

The reason why you may not want to use mode 2 all the time is that it is less tolerant of movement in one plane than mode 2. If the movement beyond a certain threshold is detected on one axis, the lens will assume you are panning and stop stablising that axis.

That's exactly what I was trying to say as to why I like having Modes 1 AND 2 on the camera, but you worded it much more clearly.   The threshold for two-axis to one-axis switching seems rather low.

It also confuses my eyes when the image in the viewfinder keeps switching between one-axis and two-axis stabilization. Whether or not it would make a difference to my final image (depending on shutter speed, focal length, and how stable I'm keeping the camera at that instant), it makes me second-guess myself.

Quoting JakTrax (Reply 9):
The other difference of course is that there's pivotal panning and parallel panning, but I'm confused enough for one evening!

Based on Colin's explanation in the thread I linked in Reply 1, the Canon IS uses an angular velocity sensor. If I understand what you're referring to with pivotal vs. parallel panning, then it would seem the IS should only correct for pivotal.



"Two and a Half Men" was filmed in front of a live ostrich.
User currently offlineckw From UK - England, joined Aug 2010, 730 posts, RR: 16
Reply 13, posted (3 years 12 hours ago) and read 6278 times:

Quoting JakTrax (Reply 9):
The other difference of course is that there's pivotal panning and parallel panning, but I'm confused enough for one evening!

I'm pretty sure the current IS system is strictly 2 axis - the camera can cope with either side to side or up and down movements. If you were to rotate the camera or move the camera diagonally while panning it would confuse the system.

This is a bit theoretical, but could have practical applications when, say, panning a rapidly climbing jet - the transition from horizontal to vertical movement may cause problems.

I believe some sensor-shift systems CAN deal with rotational movement.

Quoting dvincent (Reply 11):
The amount of movement the sensor has to make to account for shake is minimal even for long lenses. A very small movement of the sensor can account for a very large movement across the final image. This was the same argument made against full frame antishake and it was bogus then and it was bogus now.

I don't think the article you quote proves anything really - its just an opinion based on the fact that the author doesn't think a 1.5mm movement is much, But this totally ignores the fact that an effective IS is determined by responsiveness, accuracy and speed of the movement. He also ignores the fact that the extra weight (and inertia) of a full frame sensor makes the whole thing more difficult. And of course there are different types of camera shake - it could be rapid small movements, or slower large movements.

This is not to say that sensor based stablisation doesn't work - obviously it does, and may be better suited to offering a wider range of stablisation (3 axis). But as the lens gets longer, sensor systems have to work increasingly hard compared to lens based systems. Lens based IS will always be theoretically more efficient - all else being equal. Try shining a torch at a target, once by holding it at the end, and again by holding it in the middle - which is easier?

Cheers,

Colin



Colin K. Work, Pixstel
User currently offlinedarreno1 From United States of America, joined Jun 2010, 224 posts, RR: 0
Reply 14, posted (3 years 9 hours ago) and read 6258 times:

I've gone back and forth between having VR on and off on my Tam 70-300 and frankly I couldn't tell the difference for the majority of shots and the ones that I could, they were very minute. I usually leave it off unless I'm shooting at low shutter speeds handheld in which case, every little bit helps.


Nikon D7000 / Nikkor 105mm AF f2.8 / Nikkor 35 f1.8G / Nikkor 50 f1.8D / Nikkor 85mm / Nikkor 300mm f4 AF
User currently offlinedvincent From United States of America, joined Jan 2007, 1742 posts, RR: 11
Reply 15, posted (2 years 12 months 4 days ago) and read 6141 times:
Support Airliners.net - become a First Class Member!

Quoting ckw (Reply 13):
This is not to say that sensor based stablisation doesn't work - obviously it does, and may be better suited to offering a wider range of stablisation (3 axis). But as the lens gets longer, sensor systems have to work increasingly hard compared to lens based systems. Lens based IS will always be theoretically more efficient - all else being equal. Try shining a torch at a target, once by holding it at the end, and again by holding it in the middle - which is easier?

And optical based systems don't have to work as hard to correct the same amount of motion? Think about what you're saying here. The devices that perform the movements and sense the movements work on the same principles between both kinds of systems (gyros and piezoelectric motors). Also, you're still holding the camera and body as a system with a long lens. Even then, though, the actual amount of correction done to account for shake (in lens or in body) is surprisingly minimal. The large shakes you're talking about can't be corrected by either kind of system.

Seriously, though - you can go through my whole history of photographs in here to see a lot of interesting low light stuff done with long lenses and in-body IS. WIll I always get 4 stops with the 70-400G? No, and that's mostly because not having the swimmy viewfinder does hurt effectiveness in some cases. But in terms of the ability to correct for shake, there is effectively no difference between the two. And yeah, in theory, the perfect in-lens IS system would be more effective, but in reality the "betterness" is so minimal that the ability to have it on every lens outweighs it.



From the Mind of Minolta
User currently offlinestevemchey From United States of America, joined Feb 2009, 365 posts, RR: 0
Reply 16, posted (2 years 12 months 3 days 13 hours ago) and read 6098 times:

Nikon recently posted an interesting article about why they think in-lens IS is superior to in-camera IS. Of course they are biased, but I still find it very informative. Especially the argument that with in-lens IS you can actually see the effect of the stabilization.

User currently offlineckw From UK - England, joined Aug 2010, 730 posts, RR: 16
Reply 17, posted (2 years 12 months 3 days 12 hours ago) and read 6091 times:

Quoting dvincent (Reply 15):
And optical based systems don't have to work as hard to correct the same amount of motion?

No - if you consider what IS actually does. It does not stop the camera shaking, it corrects the light path. Think in terms of a light ray passing down the lens - any aberration to the path will increase the further it travels. Making the correction sooner rather than later means a less extreme adjustment is required. I found this simple diagram which I think illustrates the point.

http://www.canon.com/camera-museum/tech/report/200406/img/zu2.gif

I would agree that up to say, medium telephoto, the differences are probably negligable, and more to do with the makers implementation of their chosen IS system. (A new Sony sensor based system is probably much better than early generation Canon lens systems), but on the longer lenses an optimum lens system should better a sensor shift system.

Cheers,

Colin



Colin K. Work, Pixstel
User currently offlineXT6Wagon From United States of America, joined Feb 2007, 3392 posts, RR: 4
Reply 18, posted (2 years 11 months 3 weeks 4 days 5 hours ago) and read 5746 times:

It really depends on the generation of IS. The cannon IS with no mode select is known to have issues with errors in its "correction". The early lenses with mode select are far better, but still cough up some extra correction some times. The latest IS is frankly insanely good. Its a technology that has come along way in a short time, and so you need to decide based on the actual lens more than universal rules.

Top Of Page
Forum Index

This topic is archived and can not be replied to any more.

Printer friendly format

Similar topics:More similar topics...
When To Use A UV Filter? posted Tue Feb 21 2006 18:44:12 by Cadet57
Filters... To Use Or Not To Use? posted Fri Nov 2 2007 05:35:36 by Metroliner
What Kind Of Camer To Use To Take A Pics? posted Fri Jul 13 2007 04:38:32 by Boeing743
Which Airline Title To Use? posted Tue Aug 1 2006 10:06:09 by Edoca
Company Wanting To Use My Photo posted Sun Jul 30 2006 19:52:03 by Phxplanes
Northwest Wants To Use One Of My Photos! posted Tue Jan 31 2006 00:31:16 by AlitaliaMD11
What Info To Use For AC With Temp Registration? posted Mon Sep 19 2005 07:45:58 by Seachaz
Reqeust To Use Photo For Calendar posted Tue Sep 13 2005 06:10:08 by Syncmaster
Which Color Profile To Use For Your Screen? posted Mon Aug 15 2005 22:41:32 by FlyingZacko
So Yeah, Wikipedia Wants To Use A Photo Of Mine. posted Wed Aug 3 2005 06:13:02 by Fiveholer