Lens VR unit IR leakage affects long exposures - cont'd.

Started Feb 17, 2008 | Discussions
Marianne Oelund Veteran Member • Posts: 7,788
Lens VR unit IR leakage affects long exposures - cont'd.
2

This is the continuation of a thread started by MaxKPhoto, in which he disclosed his findings regarding IR pollution of long-exposure images (30 minutes or more), when using the 18-200 VR lens. See his thread "Design flaw in all Nikkor 18-200 VR's - All models are defective" here http://forums.dpreview.com/forums/read.asp?forum=1039&message=26783756 .

The thread contains a considerable amount of speculation, much of which misses the point, but important contributions were made by Randy Simms here http://forums.dpreview.com/forums/read.asp?forum=1039&message=26796958 , where he showed that the 70-200VR also has a similar problem, and by Chuxter here http://forums.dpreview.com/forums/read.asp?forum=1039&message=26804001 , where he showed that the light pollution could not be from a hot surface.

So far, the only lenses that I am aware of which exhibit this problem, are the 18-200 VR and the 70-200 VR. I have confirmed that the 200 f/2 VR and the 200-400 f/4 VR do not produce any such effects. I will leave it as an exercise to other forum participants to investigate other VR lens models, and I invite them to post their results on this thread.

The Cause

I carried out an investigation using my 70-200 VR. This lens produces a very distinct vertical red stripe on long-exposure images, approximately 7mm to the right of sensor center. I first ruled out reflections from potential sources within the camera body, by imaging the mirror box of my D300 over 30 minutes; the result was a completely black frame showing no IR sources.

Confirming the lens itself as the IR source was done by operating the lens on the camera, rotated away from its normal mounted position slightly so that it was de-powered, and observing that the IR pollution was eliminated.

The final step was to image the lens directly, to pinpoint the source. This was done by mounting the lens on an F100 film body, and setting up a second camera to look at the back of the lens through the open back of the F100 while it was operating. The second camera used was a D3 at 25,600 ISO, micro-Nikkor 105mm lens, at 10-second exposure time. By successive approximation of camera position and focus, I was finally able to obtain this image of the IR sources within the lens:

This is a crop, aligned so that the center of the crop is at the center of the subject lens. We see two IR sources (reflecting off of the diaphragm blades, I believe), one directly left of center, and a second dimmer source directly above center. The sources therefore lie on the horizontal and vertical axes, which of course are the axes along which the VR system moves the elements.

In any servo system such as the VR unit, there must be a means of position feedback to the control electronics. In various applications, we typically find potentiometers, strain gauges, LVDT's or RVDT's, and optical position sensors used for this. Apparently, the latter are employed in the VR unit. These consist of a light source, in this case an IR LED, and a phototransistor or transistor array which uses the light from the LED to read physical position. In the case of the 18-200 and 70-200 VR units, then, there must be a small amount of leakage from the IR LED which is finding its way via reflections, back to the camera's sensor.

Remedy

Unfortunately, this would be beyond simple at-home measures to remedy, as it would require disassembly of the lens, and installation of masking to control the stray IR.

The best solution, at least for now, may simply be awareness of the effect, so that the user can make informed choices regarding lenses for special applications.

SteveL54 Senior Member • Posts: 2,709
Re: Lens VR unit IR leakage affects long exposures - cont'd.

Marianne,

I ran tests on my D70 (more IR sensitive) and 18-200mmVR.

It turns out to get rid of the IR source both lens switches (VR and AF) must be off. If AF is disabled by the body's switch only - leaving the lens in M/A - then the IR source stays on. I suspect the folks reporting this problem with VR off and AF in manual position left the lens switch in the M/A position.

Here's a thread I restarted last night in the lens forum.

http://forums.dpreview.com/forums/read.asp?forum=1030&message=26820887

Marianne Oelund wrote:

In any servo system such as the VR unit, there must be a means of
position feedback to the control electronics. In various
applications, we typically find potentiometers, strain gauges, LVDT's
or RVDT's, and optical position sensors used for this. Apparently,
the latter are employed in the VR unit. These consist of a light
source, in this case an IR LED, and a phototransistor or transistor
array which uses the light from the LED to read physical position.
In the case of the 18-200 and 70-200 VR units, then, there must be a
small amount of leakage from the IR LED which is finding its way via
reflections, back to the camera's sensor.

Your suggestion that IR leakage from an optical position feedback sensor makes sense.

Steve

Thomas Comerford Veteran Member • Posts: 9,745
Re: Lens VR unit IR leakage affects long exposures - cont'd.

Marianne Oelund wrote:

This is the continuation of a thread started by MaxKPhoto, in which
he disclosed his findings regarding IR pollution of long-exposure
images (30 minutes or more), when using the 18-200 VR lens. See his
thread "Design flaw in all Nikkor 18-200 VR's - All models are
defective" here
http://forums.dpreview.com/forums/read.asp?forum=1039&message=26783756 .

The thread contains a considerable amount of speculation, much of
which misses the point, but important contributions were made by
Randy Simms here

http://forums.dpreview.com/forums/read.asp?forum=1039&message=26796958 , where he showed that the 70-200VR also has a similar problem, and by Chuxter here http://forums.dpreview.com/forums/read.asp?forum=1039&message=26804001 , where he showed that the light pollution could not be from a hot surface.

So far, the only lenses that I am aware of which exhibit this
problem, are the 18-200 VR and the 70-200 VR. I have confirmed that
the 200 f/2 VR and the 200-400 f/4 VR do not produce any such
effects. I will leave it as an exercise to other forum participants
to investigate other VR lens models, and I invite them to post their
results on this thread.

I managed to get a post in just before the last thread filledup. The 70-300 VR does it too, but switching the AF switch on the lens from "M/A" to "M" seems to solve the problem:
http://forums.dpreview.com/forums/read.asp?forum=1039&message=26817314

The Cause
I carried out an investigation using my 70-200 VR. This lens
produces a very distinct vertical red stripe on long-exposure images,
approximately 7mm to the right of sensor center. I first ruled out
reflections from potential sources within the camera body, by imaging
the mirror box of my D300 over 30 minutes; the result was a
completely black frame showing no IR sources.
Confirming the lens itself as the IR source was done by operating the
lens on the camera, rotated away from its normal mounted position
slightly so that it was de-powered, and observing that the IR
pollution was eliminated.
The final step was to image the lens directly, to pinpoint the
source. This was done by mounting the lens on an F100 film body, and
setting up a second camera to look at the back of the lens through
the open back of the F100 while it was operating. The second camera
used was a D3 at 25,600 ISO, micro-Nikkor 105mm lens, at 10-second
exposure time. By successive approximation of camera position and
focus, I was finally able to obtain this image of the IR sources
within the lens:

This is a crop, aligned so that the center of the crop is at the
center of the subject lens. We see two IR sources (reflecting off of
the diaphragm blades, I believe), one directly left of center, and a
second dimmer source directly above center. The sources therefore
lie on the horizontal and vertical axes, which of course are the axes
along which the VR system moves the elements.
In any servo system such as the VR unit, there must be a means of
position feedback to the control electronics. In various
applications, we typically find potentiometers, strain gauges, LVDT's
or RVDT's, and optical position sensors used for this. Apparently,
the latter are employed in the VR unit. These consist of a light
source, in this case an IR LED, and a phototransistor or transistor
array which uses the light from the LED to read physical position.
In the case of the 18-200 and 70-200 VR units, then, there must be a
small amount of leakage from the IR LED which is finding its way via
reflections, back to the camera's sensor.

Remedy
Unfortunately, this would be beyond simple at-home measures to
remedy, as it would require disassembly of the lens, and installation
of masking to control the stray IR.
The best solution, at least for now, may simply be awareness of the
effect, so that the user can make informed choices regarding lenses
for special applications.

-- hide signature --
Ravncat Senior Member • Posts: 1,109
Re: Lens VR unit IR leakage affects long exposures - cont'd.

There seems to be a rather large difference between the effect that occurs on the sample above and on the 70-200 when compared to the 18-200 images posted in the prior thread.

The 18-200 image shows a very clear gradient from the right frame to the left, while the above images are much more pinpoints or scratches of light. This would seem to suggest a different cause, or some other design difference that's altering the effects of the same cause.

I have to wonder if it's related to the 18-200's VR/AF servo's and the circuitry between the lense and camera, since the gradient (if you look in photoshop and use the gradient tool to cancel it out) appears to be at a very simmilar angle to the location of the electrical contacts on the lense mount, perhaps a natural continuation of those circuits. If those run hot interior to the camera, providing internal heat on the right side of a greater amount than the other lenses, it could be a cause.

It's not too surprising that the gradient appears to correlate with the heat equation (see calculus / boundary value problems / thermodynamics) Unfortunately, I have not done any kinds of density tests or actual angle checks to verify the above conjecture, so please do remember that it is conjecture.

It's clear that it has something to do with the A/F and VR lenses, unless I missed something on my 18-70 images, I didn't pixel peep those tests, because I was looking for a large obvious gradient effect, that was separate from Ampglow. The 18-70 lacks the VR function but does have a M/a override. I'll have to test again with that third combination. (Camera AF off, M/A on lense, Camera AF on/M on lense, Camera AF off, Lense M/A set to m)

This certainly makes me wonder why the shown effects on the 70-200 and the OP's tests show up as nearly specular. The theory that it's internal IR bouncing off the blades seems to make sense. This however, tends to point to the idea that the 18-200 issue is different from the other lenses, in a fundamental way. Unfortunately without dissasembling the lense, doing proper reflection analysis, it's hard to tell if it's being diffused, but even if that were the case, it seems awfully strange that it's the entire right side of Maxk's photos.

-- hide signature --

A poor photographer blames his tools.

SteveL54 Senior Member • Posts: 2,709
Re: Lens VR unit IR leakage affects long exposures - cont'd.

Thomas Comerford wrote:

I managed to get a post in just before the last thread filledup. The
70-300 VR does it too, but switching the AF switch on the lens from
"M/A" to "M" seems to solve the problem:

Yeah, you caught on to the fix first.

It's an easy workaround for the problem. I'm quite puzzled that Nikon service didn't tell the OP from the first thread about turning the lens AF switch to M/A.

If either switch (VR xor lens AF) is in the active position the IR contamination seems to be the same.

Nikon should put a note in these lens's manuals about configuring the lens for long term exposures.

Steve

Ravncat Senior Member • Posts: 1,109
Well the fix still seems to work on the 18-200 so... nt
-- hide signature --

A poor photographer blames his tools.

SteveL54 Senior Member • Posts: 2,709
Re: Lens VR unit IR leakage affects long exposures - cont'd.

Ravncat wrote:

There seems to be a rather large difference between the effect that
occurs on the sample above and on the 70-200 when compared to the
18-200 images posted in the prior thread.

The 18-200 image shows a very clear gradient from the right frame to
the left, while the above images are much more pinpoints or scratches
of light. This would seem to suggest a different cause, or some other
design difference that's altering the effects of the same cause.

As Marianne described, she actually focused the IR by another camera/lens looking through a film camera with it's back off. The film camera had the lens suspected of generating the IR attached to it. (A very clever setup.) The reason the other IR images are so indistinct is that they were taken directly by the camera with the offending lens and since the IR originates within the lens the IR source(s) won't be focused.

Steve

Ravncat Senior Member • Posts: 1,109
Thanks Steve, originally read through it too quickly. nt
-- hide signature --

A poor photographer blames his tools.

OP Marianne Oelund Veteran Member • Posts: 7,788
Yes, you can turn it off
1

SteveL54 wrote:

It's an easy workaround for the problem. I'm quite puzzled that Nikon
service didn't tell the OP from the first thread about turning the
lens AF switch to M/A.

If either switch (VR xor lens AF) is in the active position the IR
contamination seems to be the same.

Well, this is certainly embarrassing - I shouldn't believe every claim made in posts, without verifying them first. Max insisted that there was no combination of switch settings that would clear the problem. See http://forums.dpreview.com/forums/read.asp?forum=1039&message=26784890 , where he claimed to have been using the M/A switch "M" position.

Clearly, he wasn't, as this will turn off the IR source on both the 18-200 VR and 70-200 VR.

I did check the operation of the 18-200 on the D3, though, and the IR pattern is very interesting. No smooth gradients in the "bigger picture;" there are clear light and shadow areas on the sensor. If anyone's interested, I can post a sample.
Talk about much ado about nothing, though . . .

MOD TOF guy Forum Pro • Posts: 16,841
easy precautions prevent IR leakage to affect pics for long exposures

Marianne :

The best solution, at least for now, may simply be awareness of the
effect, so that the user can make informed choices regarding lenses
for special applications.

Is there such an application? Who needs VR or AFS for a 30 minutes exposure?

However the important point is taken that users should turn off VR and AF for very long exposures, and the latter using the lens switch on specific lenses. Thank you for confirming the problem on the 70-200 f2.8 and the results of your investigation.

Using the AF switch on the camera to manual * should * be enough. So I'm ready to call the fact that the switch on some lenses must be moved as well a design flaw, but a very small one, with an easy workaround. IMO MaxKphoto's angry comments at Nikon were quite over the top... Also, IMO again, the subject of my post is better worded.
--
Thierry

 TOF guy's gear list:TOF guy's gear list
Nikon Z9 Nikon AF-S Nikkor 70-200mm f/2.8G ED VR Nikon AF-S DX Nikkor 12-24mm f/4G ED-IF Nikon AF-S Nikkor 24-70mm f/2.8G ED Nikon AF-S Nikkor 24-120mm F4G ED VR +9 more
OP Marianne Oelund Veteran Member • Posts: 7,788
Re: easy precautions prevent IR leakage to affect pics for long exposures
1

TOF guy wrote:

Is there such an application? Who needs VR or AFS for a 30 minutes
exposure?

Of course, that was written before I realized that MaxKphoto's assertions about his switch settings were incorrect.

However the important point is taken that users should turn off VR
and AF for very long exposures, and the latter using the lens switch
on specific lenses. Thank you for confirming the problem on the
70-200 f2.8 and the results of your investigation.

You're welcome - I enjoy investigating strange phenomenae.

Using the AF switch on the camera to manual * should * be enough. So
I'm ready to call the fact that the switch on some lenses must be
moved as well a design flaw, but a very small one, with an easy
workaround. IMO MaxKphoto's angry comments at Nikon were quite over
the top...

Yes, even if there were a design problem with no immediate workaround, I do not believe Nikon should be held responsible for discovering every possible effect for every possible application of their products. Not many of their customers would be willing to wait for that testing to be carried out before the release of each new product.

Peter iNova Veteran Member • Posts: 3,250
Re: Yes, you can turn it off

I like the factoid that you can take multi-minute images with the D300, and my working habit includes this very important item: When shooting long exposures, lock the focus and turn off the VR. In fact, any time the camera goes onto a tripod--turn off the VR.

Of course, with the camera in the dark, how do you focus? Once, with the lights on, or using a distant detail as a focus target for infinity night shots, or manually with a magnifying right angle finder or similar aid. Then lock that focus. Leaving it live will only invite trouble.

What the value item of this whole thread turned out to be is: To clear the lens of any and all IR contamination via the VR system, switching both camera options off is the big tip.

Thanks for the research.

-iNova
--
http://www.digitalsecrets.net

SteveL54 Senior Member • Posts: 2,709
Re: Yes, you can turn it off

Marianne Oelund wrote:

Well, this is certainly embarrassing - I shouldn't believe every
claim made in posts, without verifying them first. Max insisted that
there was no combination of switch settings that would clear the
problem. See

http://forums.dpreview.com/forums/read.asp?forum=1039&message=26784890 , where he claimed to have been using the M/A switch "M" position.
Clearly, he wasn't, as this will turn off the IR source on both the
18-200 VR and 70-200 VR.

Talk about much ado about nothing, though . . .

LOL... I think you and I were more 'focused' on the nature and source of the image contamination rather than how to configure the camera/lens to simply avoid it.

It's good to know that VR lenses have sources in them that actively radiate IR. It's a new source to add to things such as amp glow and thermal noise as long term exposure image spoilers.

Looking back through that thread it's apparent that Max initially assumed or expected the body's AF switch -- along with the lens VR switch -- should be sufficient to deactivate any lens electronics causing the problem. I don't think Max initially claimed to have explored every switch combination that could have impacted the result. Max just claimed to have disabled AF and VR. He just didn't specify how he disabled AF.

Later Max mentioned that he would check into the lens M/A-M switch setting impact.

http://forums.dpreview.com/forums/read.asp?forum=1039&message=26807156

To be fair to Max, why would it be necessary to turn off the lens focus switch to disable something associated with VR? After all, non-VR AF-S lenses have an M/A-M switch, but one doesn't have to set them into the M position to avoid IR contamination. (I checked this with three of my non-VR AF-S lenses.) It looks like the designers coupled the lens AF with VR in some way -- possibly to save a few pennies.

I did check the operation of the 18-200 on the D3, though, and the IR
pattern is very interesting. No smooth gradients in the "bigger
picture;" there are clear light and shadow areas on the sensor. If
anyone's interested, I can post a sample.

Yes, I'd be interested in seeing such an image. It's useful to be able to identify the signature of such issues.

Thanks again for the excellent investigative work.

Steve

OP Marianne Oelund Veteran Member • Posts: 7,788
VR 18-200 IR leak - a wider view
1

SteveL54 wrote:

Marianne Oelund wrote:

I did check the operation of the 18-200 on the D3, though, and the IR
pattern is very interesting. No smooth gradients in the "bigger
picture;" there are clear light and shadow areas on the sensor. If
anyone's interested, I can post a sample.

Yes, I'd be interested in seeing such an image. It's useful to be
able to identify the signature of such issues.

Here you go - taken with the D3 at ISO 25600 in just 30 seconds, rather than 30 minutes:

HenriHofmeir Junior Member • Posts: 31
Re: Kudos to Marianne

Marianne Oelund wrote:

Well, this is certainly embarrassing

. . .

Talk about much ado about nothing, though . . .

Just wanted to say thanks for the excellent investigative work done on this thread. Great thread! It is very educational to "hear" how one goes about investigating the various issues we discover with our gear. I learn a great deal about my new tools (D300, etc.) by "listening" to those in these threads that attempt to characterize the various anomalies that are discovered.

In particular, Marianne did an excellent job of sifting out the relevant facts, summarizing what was known, then taking the investigation to the next level. Her excellent, patient and thorough descriptions of her tests and methods arm us with the kind of knowledge that is easily applied and generalized. And she did so without inciting any discord.

Marianne: While you may not have been the one to cite a workaround, the workarounds are easily arrived at given the technical details you so aptly laid before us. I was impressed by your testing, methods, and your ability to share your knowledge in a straight forward, coherent and convincing manner. Excellent work that resulted in a clear characterization of the phenomenon. I am glad a workaround was not immediately apparent because it was very useful to learn from the process you used to characterize this phenomenon.

I, for one, am thankful we have contributors like you on this forum. You certainly have nothing to be embarrassed about.

len walshe Junior Member • Posts: 41
Re: easy precautions prevent IR leakage to affect pics for long exposures

Congratulations on a first rate piece of detective work.

Is it your intention, now, to inform Nikon of the results of your excellent diagnosis so they may consider how best to put Nikon user's in the picture regarding this issue.

Clearly, as probably with most lenses, Nikon or otherwise, there is a practical limit to the type of application where good results can be guaranteed. Nevertheless, users should be warned that in certain circumstances unsatisfactory images may be obtained.

 len walshe's gear list:len walshe's gear list
Nikon Df Nikon D500
Yves P. Forum Pro • Posts: 18,674
In other words ...

Turn VR off for long exposures anyways, VR is not going to correct anything slower than say 1/8 of a second anyways ...

Marianne Oelund wrote:

This is the continuation of a thread started by MaxKPhoto, in which
he disclosed his findings regarding IR pollution of long-exposure
images (30 minutes or more), when using the 18-200 VR lens. See his
thread "Design flaw in all Nikkor 18-200 VR's - All models are
defective" here
http://forums.dpreview.com/forums/read.asp?forum=1039&message=26783756 .

The thread contains a considerable amount of speculation, much of
which misses the point, but important contributions were made by
Randy Simms here

http://forums.dpreview.com/forums/read.asp?forum=1039&message=26796958 , where he showed that the 70-200VR also has a similar problem, and by Chuxter here http://forums.dpreview.com/forums/read.asp?forum=1039&message=26804001 , where he showed that the light pollution could not be from a hot surface.

So far, the only lenses that I am aware of which exhibit this
problem, are the 18-200 VR and the 70-200 VR. I have confirmed that
the 200 f/2 VR and the 200-400 f/4 VR do not produce any such
effects. I will leave it as an exercise to other forum participants
to investigate other VR lens models, and I invite them to post their
results on this thread.

The Cause
I carried out an investigation using my 70-200 VR. This lens
produces a very distinct vertical red stripe on long-exposure images,
approximately 7mm to the right of sensor center. I first ruled out
reflections from potential sources within the camera body, by imaging
the mirror box of my D300 over 30 minutes; the result was a
completely black frame showing no IR sources.
Confirming the lens itself as the IR source was done by operating the
lens on the camera, rotated away from its normal mounted position
slightly so that it was de-powered, and observing that the IR
pollution was eliminated.
The final step was to image the lens directly, to pinpoint the
source. This was done by mounting the lens on an F100 film body, and
setting up a second camera to look at the back of the lens through
the open back of the F100 while it was operating. The second camera
used was a D3 at 25,600 ISO, micro-Nikkor 105mm lens, at 10-second
exposure time. By successive approximation of camera position and
focus, I was finally able to obtain this image of the IR sources
within the lens:

This is a crop, aligned so that the center of the crop is at the
center of the subject lens. We see two IR sources (reflecting off of
the diaphragm blades, I believe), one directly left of center, and a
second dimmer source directly above center. The sources therefore
lie on the horizontal and vertical axes, which of course are the axes
along which the VR system moves the elements.
In any servo system such as the VR unit, there must be a means of
position feedback to the control electronics. In various
applications, we typically find potentiometers, strain gauges, LVDT's
or RVDT's, and optical position sensors used for this. Apparently,
the latter are employed in the VR unit. These consist of a light
source, in this case an IR LED, and a phototransistor or transistor
array which uses the light from the LED to read physical position.
In the case of the 18-200 and 70-200 VR units, then, there must be a
small amount of leakage from the IR LED which is finding its way via
reflections, back to the camera's sensor.

Remedy
Unfortunately, this would be beyond simple at-home measures to
remedy, as it would require disassembly of the lens, and installation
of masking to control the stray IR.
The best solution, at least for now, may simply be awareness of the
effect, so that the user can make informed choices regarding lenses
for special applications.

-- hide signature --

Yves P.
Share the Knowledge

NPS Member
PBASE Supporter

Some pictures I like:
http://www.pbase.com/yp8/

 Yves P.'s gear list:Yves P.'s gear list
Apple iPhone 11 Pro
Thomas Comerford Veteran Member • Posts: 9,745
Re: In other words ...

Yves P. wrote:

Turn VR off for long exposures anyways, VR is not going to correct
anything slower than say 1/8 of a second anyways ...

In my case (and this probably applies to the other posters aswell) the VR was already off, but the AF had to be switched to "M" on the lens to stop it from happening.

-- hide signature --
Keyboard shortcuts:
FForum MMy threads