PDA

View Full Version : Blur on part of photos with 1Dm3 + 70-200 2.8



Cary
2010-11-09, 04:08 PM
I've recently started having a problem with my Canon 1Dm3 + 70-200 2.8 (non-IS), where a small part of my photo is blurry, it always seems to be the same area of the photo, and it doesn't happen all the time. Since everyone here shoots airplanes with similar techniques, I'm hoping someone might be able to help. I'll copy the details that I posted on a photography forum:

I've started getting blur (looks like motion blur) on only a small area of some of my photos (in what seems to be the same area each time). Left and right of the blurred area, the subject is sharp, so it's not a DOF issue. I'm also not changing the focal length (it was fixed at 80mm for 3 side shots, and all were blurred). Here's a sample of a 100% crop of the affected area, along with a thumbnail of the entire uncropped photo so you can see the area relative to it. You can see the blur most noticeably on the "UNITED ST" and #2-3 engines.

1/1000, f 9.0

http://i164.photobucket.com/albums/u36/soak2179/crop.jpg

http://i164.photobucket.com/albums/u36/soak2179/crop.jpg (since large images are resized here)

I'm panning right to left with the plane, but it's a smooth and slow pan, and this is the non-IS version of the lens. The strange thing is that some of my other pics from the same day don't exhibit the blurring - the only difference being they were taken at 150-200mm. Could this be a case of lens decentering, or something else? Thanks.

--------

One person said decentering would be soft/blurred only on one side, and not between two sharp areas. Other than that, I think I can rule out:

-Motion blur (shooting at 1/1000, and only one area is blurred)
-Zoom blur (I shot 3 photos in a row, all at 80mm - all have the same problem)
-DOF (f 9.0, plus other areas in the same plane are sharp)
-FOD (I checked the front/rear lens surfaces, along with the 1Dm3's sensor for smudges or other debris)

cancidas
2010-11-09, 04:26 PM
do you have any smudges on the lens or any filters? might be as simple as that.

Cary
2010-11-09, 04:29 PM
do you have any smudges on the lens or any filters? might be as simple as that.

Nope, there's nothing on my lens/filter/sensor. After further examination, it looks like I've had the same issue for years. It just isn't very noticeable unless there are big letters or other high contrast items in that area. However, it's happening in that same area (around the first rule of thirds vertical line) regardless of if I'm panning right to left or left to right, or if it's a static subject.

ANITIX87
2010-11-09, 05:30 PM
Looks like a sensor issue. Also, what ISO was that shot at? Since you're using a 1D, if that's anything less than 400ISO, then you've definitely got a defective sensor. Send it in for repairs, especially if it's still under warranty, though you say you've had the problem "for years".

TIS

Cary
2010-11-09, 06:22 PM
Looks like a sensor issue. Also, what ISO was that shot at? Since you're using a 1D, if that's anything less than 400ISO, then you've definitely got a defective sensor. Send it in for repairs, especially if it's still under warranty, though you say you've had the problem "for years".

TIS

ISO 200, but I don't use in-camera NR, plus I ran Auto Contrast and a 350 amount 0.2 radius Unsharp Mask on it. I've heard of grease dripping onto the sensor (which I'd imagine might cause something like what I'm seeing), but I examined it pretty closely and didn't see any evidence of that. I'm going to shoot without a filter, or with one of my B+W filters next time to see if that makes a difference. If not, I'll visit Canon since they're only 30 minutes away from where I live. Thanks!

gonzalu
2010-11-09, 08:27 PM
Definitely NOT a sensor issue. THANK YOU FROM THE BOTTOM OF MY HEART for posting. I too have the same issue...

Nikon D3 and Nikon D2Xs BOTH exhibit this problem... and at different times without explanation. However, I have tried to narrow it down and I BELIEVE I have some answers.

I take LOTS of photos and not just of planes LOL. But I too became totally frustrated by this when I started to shoot airliners. I am at work now but I can show my own dirty laundry when I get home :-)

First some background and my techniques and reasoning for doing so. Also my gear and how I tend to use it. We together should be able to come to a good conclusion that is beneficial to all.

My gear:

Nikon D3 - Full Frame
Nikon D2Xs - Crop Sensor (1.5x)
Nikkor 70-200mm f/2.8 AF-S VR
Nikkor 200-400mm f/4 AF-S VR

Those are the workhorses for aviation work for me. I mainly have shot airshows in the past and only recently started spotting for real and taking pictures for a purpose of recording and also achieving the highest quality possible.

What I have excluded so far:

IT is NOT my lenses because testing them against controlled test environments returns good images within acceptable or expected boundaries.

--Sensor Size and Pixel Pitch.

This is an interesting aspect. I have read many many articles, end user reports with different observations and opinions in the sensitivity of your [system] to motion blur based on pixel pitch... good one. I am not totally sold on it or against it...

--Magnification ratio

OK, so 1/400sec should be enough to hand hold a 400mm lens. IT DEPENDS on many factors. but I can tell you that rule DOES NOT APPLY when you;re standing at Bayswater with a 40knot cross wind and you;re trying to track a slow plane across your frame. So, the jury is still out on that.

--Heat Distortion.

On side shots, or even head on shots, I have to sometimes remember the ENGINES are spewing tons of hot air... anything in its path will be distorted... so I have come to accept this and take it into consideration.

--Lens Quality and CA and all related things.

I will speak for what I know... Ya'll chime in with your own expertise :-)

- Nikon 200-400mm f/4

A VERY SHARP zoom lens... arguably one of the best zooms in terms of reputation and real world quality. HOWEVER, in the real world, the lens has its bad traits.

= Subject Distance. If the subject is far away it will be DISTORTED. Period. If subject distance is within known limits, watch out, spot on.

=Aperture. SHARP at f/4 all throughout, except for above rule, even at f/16, it will be horrible.

= Focal length: STAY AWAY from 400mm ... unless see above. Otherwise, you will incur the wrath of distortion. It is much worse on the D3 not sure why. At 390mm is is not an issue. Perhaps other lenses are the same in Canon Land, do share. I know in Nikon land, the 70-300mm lens is horrible at 300mm even at f/8 LOL.

=Pixel Pitch - vs. aperture. Yes, this is true... on my D3 I can go to f/16, on my D2Xs, f/8, MAYBE f/11

- 70-200mm f/2.8

Well, my perfect friend. SHARP at all times, no matter what, 70mm, 200mm and anywhere in between... little to fault... EXCEPT only on crop sensor. It is really bad on the D3 at the edges. Understandably so!
:-)

====

Aside from the above, there is technique.

-Panning

You must be PERFECT at it. if you're at an angle to the plane for even a split second, the front of the plane is moving at a different speed than the back. That's hard to deal with... you're talking 1/2000 sec or higher to deal with it. The issue is the rolling shutter curtains... think about it.... if they don't move fast enough, you're recording different parts at different split seconds causing image blur...

- VR element motion induced blur. I feel that if you;re not careful, at times you allow the VR motor to stop and then re-engage in between a sequence of shots. This may or may not contribute to CRITICAL sharpness

- Auto-Focus . I believe this is partly to blame, but not for uneven blur. I do not get CRITICAL sharpness all the time and I feel the AF system to blame. I am still toying with this. For me at an Airshow, my success is with 9 point AF / Dynamic / Continuous.... spot on when trackign an F-15 swooshing by. Why doesn't this work with a REALLY SLOW MOVING Boeing 747-400? Beats me.. .still learning on this front.

-----

I really think it has a lot more to do with the size of the subject and the sensor plane angle relationship to the subject. and the speed across the frame while the shutter curtain travels across the plane of the sensor. Can't prove it, it is just a hunch. Love discussion on this.

Thanks again for this thread. I hope to learn something new.:cool:

gonzalu
2010-11-09, 08:35 PM
Filters --

I NEVER use one unless it is for effect. That means NAKED front elements for me. I will use a polarizer if I want to, or a ND filter if I want to or a colored gel if I want to but not for no reason. The lens DOES HAVE a front element already. I do not believe in "protecting" the front element. In my opinion, one more thing to clean but more importantly, ANOTHER AIR-to-GLASS-to-AIR surface for light to be distorted through... I do not want that... personal choice. I know there is a huge camp that believes in having a front filter ALWAYS.

Sensor Cleaning -- I do anti-static every two days, Rocket Blower on each lens swap and wet cleaning as needed. High Mag loupe reveals nothing odd. More interestingly, why do some pictures come out perfect and others not in the same sequence? Something tells me it is MY FAULT and MY TECHNIQUE or perhaps a faulty camera but I really believe it has to do with combo of physics and technique.

Do you think Canon or Nikon do testing of "our" subject matter? I mean they do a BATTERY of tests using common subjects: Landscapes, Portraiture, Architecture, Sports... sports comes close I guess but never with HUGEOUGOUS subjects, really close to the sensor at varying degrees of angle in respect to the subject and a traveling shutter curtain.

Have you tested this theory? I am toying with it but unable to get a good enough result. Using a shutter speed on your camera like the flash sync where the entire sensor is exposed at once. This requires higher ISO of course or lower apertures which both induce other side effects. It is hard to just isolate this.

gonzalu
2010-11-09, 08:37 PM
Cary sorry, to clarify, my issue is sightly different than yours from the point of view that it is ALWAYS with the very right side of the image if planes are moving right to left in the frame or the very far left of the frame if planes moving in opposite direction!

Yours is in the "middle" somewhere? Wow, touch one. Only the 70-200mm f/2.8?

Cary
2010-11-09, 09:21 PM
Cary sorry, to clarify, my issue is sightly different than yours from the point of view that it is ALWAYS with the very right side of the image if planes are moving right to left in the frame or the very far left of the frame if planes moving in opposite direction!

Yours is in the "middle" somewhere? Wow, touch one. Only the 70-200mm f/2.8?

Hey Manny, yes, it sounds like our issues are a little different. Yours sounds more like what I occasionally get with my 20D/40D/7D + 100-400 IS, which I like to nickname "IS smearing". I think in that case, I'm not panning in a direction that my IS lens wants me to be panning in, or I don't let the IS system stabilize before taking the shot.

After reviewing most of my photos, the problem with my 70-200mm appeared to rear it's ugly head even on static airplanes, like the Icelandair DC-3 I shot from my hotel, so I think I can discount poor panning technique or motion blur as the source of the problem. I do have my worst quality filter on the 70-200mm - a Hoya HMC UV(0) - so when I go out next time, I'll try my B+W and also without a filter. I have pixel peeped a lot at my 7D + 100-400mm shots, but don't see any similar problem (besides the "IS smearing" I mentioned above).

gonzalu
2010-11-10, 03:04 PM
Cary, thank you. First off, your gallery is awesome! inspiring images there.

I am going to show my panties because I need to make sure I am not going crazy. LOL. I thought I knew how to take pictures of planes before but now that I have started taking pictures of behemoth aircraft moving very slowly, sometimes in poor conditions, has totally pwned me! :frown:

What may seem like a lot of keepers, MOST of my images are NOT keepers. I proudly get about 90% keeper rates at an airshow, but sadly, spotting picture outings can sometimes yield 1%. VERY disconcerting. I have become obsessive in seeking out the most quality in photography. I am a tough editor and will discard anything that is not critically sharp. But I am worried when there is not at least ONE shot in FIFTY of a particular pass that at least sharp and across the entire frame.

Sometimes I feel it is the AF system missing, but I have to discount this as I have tried MULTIPLE techniques to no end:

--AF-C : Continuous auto on and off. Neither helps.

--AF mode: While I have not yet researched all combinations (I mean, I am usually out to get a shot, not practice ha!) AF priority vs. Shutter Release pri. Bot hon and off, same results.

--VR motor action: You're right Cary, that must be it. I totally forget that it is actually BETTER for the VR motors/mechanisms to lock on and stay locked rather than being asked to recalculate every second. While panning at an airshow, I am panning extremely fast on the horizontal and it seems to be spot on all the time. But after thinking about what you said, it made me think that I do tend to work too hard to keep the plane centered in the frame while panning and it causes me to make very small adjustments fore and aft. I BET that's causing the VR motors to engage or recalibrate in the WRONG direction (they are going left when I am going right and then back in a very small period of time)

The other situation I noticed yesterday while catching some action on 4L/R at JFK from Bayswater, the 30 knot gusts were really pushing against my setup. If you know my lens, it is quite long and the shade makes it a sail in the wind. Fighting that makes for very small yet sharp moves back and forth. It really must drive the VR motors crazy.

I am going to take your advise and test the theory of turning VR off ... I also need to shut it off when I use my monopod. I always forget to.

Now, I'd also like to discuss with you if you can help me, the whole AF problem I seem to have.

Cary
2010-11-10, 04:15 PM
Thanks, and I know exactly where you're coming from. These days, I immediately review photos from each a/c pass, and if the reg, antennae, or text/titles have any amount of softness, blur, heat haze, or ugly reflections from the sun, I immediately delete them. My 7D seems to have trouble acquiring critical sharpness at longer focal lengths (HoBe, Panera, etc.), but is almost perfect (knocking on wood) when a little closer, like from the mounds. That's why I'll always prefer the mounds over Panera, even if I have to freeze to death, or worry about ticks. And yes, windy days are brutal on long lenses, especially when it's cold on top of that.

As for AF, I like using AI-Servo (AF-C in Nikon-speak) with the center AF point only. I assign it to my back AF button, and leave the shutter to only take the pictures. So for each pass, I'll just hold down my dedicated AF button with my thumb the entire time, and hit the shutter when I want to take pictures. Works well enough for me (at the lower focal lengths). Hopefully, using a technique like that, along with better lighting conditions, less heat haze, and low humidity this time of year will help in getting a higher percentage of critically sharp pics.

gonzalu
2010-11-10, 09:38 PM
Cary, we should definitely swap stories in person LOL. I use exactly that. Yes my settings are:

-- AF-C which is Continuous Autofocus (Servo in Canon) mode. I have tested Dynamic AF mode as well as single sensor modes. The D3 has 51 AF points and can be used singularly or in groups or in total crazy full 3D dynamic mode of all 51 points calculating af. I have NOT tried Manual Focus yet LOL. I get the same or similar results no matter what mode I pick.

On the D2Xs I don;t have many AF points but they do have similar options. I do notice on the D3 when I pick a random subject at a distance, say a traffic sign on the opposite parking lot at HoBe, my AF is going berserk making infinite microscopic adjustments... even if I have a good AF target like a rough pattern. In Nikon land, and I am sure it is similar in Canon, the AF sensors need something to grab on to. A white nose on a Finnair or Lufthansa can drive the AF nuts. But any way, when it locks on, it should lock on... and if it needs to adjust, for god's sake, pick a point and stick to it.

-- I also remove the AF from the shutter. Our pro bodies have a dedicated AF-ON button next to both the main and Vertical shutters. I find it a lot easier to de-couple these two for when for example you need to frame TWO planes in the same shot and the AF sensors will be eating blue sky :wink:

I really think you nailed it in the head with the VR ... I am a BIG proponent of VR and use it a lot and I see it help me. But in this case, I think you;re right andthe more I think of the physics of a Gyro, the more I think that's got to be it. The whole purpose of a Gyro Stabilizer is to dial in a location in 3D space and hold it there... since the REGULAR mode of my VR systems is UP/DOWN sensitive and will compensate only for up or down movement but not left to right movement, it can throw it off when I am not smoothly panning left to right. I see where at Panera I go from low to high to low... I may try ACTIVE mode which compensates for being on a boat like surface which then includes up / down as well as left / right.

Then again, if I am up at 1/1250sec @ f/8 ISO 200, there should be enough shutter to correct for motion blur. More to test. I really love the smoothness of my viewfinder when the VR is engaged LOL. But if it hurts me, well, it needs to go!

As for your 7D, I think the idea of pixel pitch being more sensitive to motion blur is well founded. The pixels are so finely packed that a movement of a photon from a particular spot to another is a lot greater than if the pixels were further apart. I am no expert of course but it does seem to make sense to me. Closer subjects have a lot less fine detail to reprooduce as everything is bigger so it is a wash... you have MORE pixels correct than incorrect so it gets hidden. It is when those little fine lines need to be rendered across a row of pixels and you have a slight movement that you will notice the blur. I look at the little static discharge antennas on the trailing edges of the stabilizer or wing tips etc and look for them to be sharp as well as rivets etc. That's when I know I have critical sharpness.

Oh and to top it off, the D3 has a VERY STRONG low pass filter (anti moire) which equals soft images to begin with :mad:

gonzalu
2010-11-11, 03:17 PM
Cary, update. I was out this morning at Bays and Panera for some testing and trying to catch a Qantas and Kalitta .. Anyway, I may have nailed it for my issues. So far so good. Much testing to still be done but I am getting more keepers.

--VR is not evil! lol. I just have to learn to get around its limitations. I realize that if I try t fill the frame I have to work TOO HARD to keep the lens on track wit hthe plane as it is moving so slowly. If I rack back the zoom a bit I can keep it centered much easier and with less jerkiness to the system which I believe helps the VR work smoother and cause less "smearing" I think it worked for me today. We'll see when I look at the pics full screen.

--I have switched the D2Xs to AF-C, single AF point, not dynamic. I also am trying AF priority over Shutter frames per second. This forces a lock before shutter curtain drops. I think that too will help.

--WAIT AND WAIT until the craft is closer. The AF seems to be badly affected by heat haze. I think it is locking on to something that's just in front of the airplane.

--On the D3 I will keep Dynamic 51 point 3D AF turned on. It seems to lock on fine.

I tried both VR ON and VR OFF today at Bayswater. It was also CONSIDERABLY calmer. African Airways tail number looked very sharp in viewfinder... so fingers crossed.

Thanks for your insight and ear. How's your issue coming along? I have been selfish and have monopolized this thread so I am sorry :-)

Cary
2010-11-11, 04:36 PM
Hey Manny, glad to hear you may have fixed/minimized your issues. I didn't get out to an airport, but I did a little testing, just taking pictures of shoe boxes stacked up without any filter :) No problems, but I really didn't expect any from shooting a static object at close distance. I'll have to get out to PHL/BWI/DCA to test properly (I don't use my 1Dm3 + 70-200 at JFK).

gonzalu
2010-11-11, 06:53 PM
Hmm. I think you may need someone to keep u company out there hahaha. Have not been to Gravely Point in over 5 years and never to the others. Philly is easy for me. Washington a little harder. I'd love to tag along with someone who has experience. Hopefully your issues are as simple as a cheap filter :wink: