Z7 Mk1 sensor problem?

[Pmrts]

Hello everyone!

Since 2020 i've been shooting whit my Z7 Mk1. I am very careful whit all my equipment. I still use the original bubble protectors when packed.

Since last month i noticed the upper third of the sensor with some bending in the images. It started when i began using silent photography mode. It happened once or twice a week. But it got more frequent. Today i got the bending problem not just in the upper third but thru the entire image. Initially i thought it was a panning problem and blamed the rolling shutter problem. But now i have the problem even in a stationary shoot. It happens in about half the imagens. I cannot understand. If it's sensor error shouldn't it be continuous? Does anyone have this problem? I really don't mistreat the camera: never fell, always well protected, never change lens in wind or sandy situations...

Here are two examples:




in this case the upper third is affected




In this case the entire image is affected . Linear structures are bent in the background



It sounds expensive to repair. Any thoughts?

Thank you for your help!

Pedro Soares

 

[cyuill2007]

Hello. Thank you all for your help.

This started i guess after the firmware update.

I've been shooting with the tamron 150-600mm via the FTZ 1 adaptor. but it never before gave me this problem. I will try with other lenses with and without the silent mode and i'll give feedback.

Once again, thank you all!

I take it you are referring to a camera body upgrade. Does your model of the Tamron 150-600 have compatibility with a dock that can be used for updating the lens firmware?

Also, does anyone know if Nikon bodies can have their firmware rolled back to a previous version?

 

[Pmrts]

Hello everyone!

Just got out this morning to try all proposed solution variants to this problem.

Here are the results:

# Different lens: same problem

# VR off on the lens: less notorious aberration but still noticeable in some cases

# Didn't try turn of IBIS

# Silent photography mode Off: Problem solved !!!

So I guess the read speeds of the sensor are to slow for a shutter speed such as 1/1250s.

This makes this mode absolutely rubishh for nature photography.

I tried to reduce volume on the mechanical shutter but couldn't find how. Does anyone know? Is it even possible? I thought the sounds were artificial but if it is a mechanical shutter then it might not be possible. I am so confused about this...

Thank you all for your incredible help!

 

[ghostfox_1]

# Silent photography mode Off: Problem solved !!!

As I said originally.

So I guess the read speeds of the sensor are to slow for a shutter speed such as 1/1250s.

Slow readout is common on most sensors, barring higher end cameras. The issue is you're moving the camera, or the subject is moving. Shooting a completely static scene the shutter speed is irrelevant.

This makes this mode absolutely rubishh for nature photography.

For moving subjects, generally, I'd say yes.

I tried to reduce volume on the mechanical shutter but couldn't find how. Does anyone know? Is it even possible? I thought the sounds were artificial but if it is a mechanical shutter then it might not be possible. I am so confused about this...

No. You have a mechanical shutter. You can't reduce the volume in camera. You would have to insulate the camera using a sound blimp or similar, but having shot one for a while, wildlife didn't actually care that much about the noise. People shot with much louder sounds for many decades and got shots.

 

[cyuill2007]

I recently came across this YouTube video review of the Fujifilm GFX 100 ii -

. Go to 6:50 and you will see the same issue when electronic/silent shutter is used. This isn’t an issue with only the Z 7.

Congratulations on solving the problem. I wouldn’t worry too much about the sound coming from the mechanical shutter. It’s quieter than the sound from a DSLR.

 

[Horshack]

I tried to reduce volume on the mechanical shutter but couldn't find how. Does anyone know? Is it even possible? I thought the sounds were artificial but if it is a mechanical shutter then it might not be possible. I am so confused about this...

Try the EFCS shutter if not already set - it'll be quieter, at least during the exposure. It's only available for shutter speeds of 1/2000 and slower though, as it produces its own set of artifacts above that speed.

https://onlinemanual.nikonimglib.com/z7_z6/en/09_menu_guide_05_d05.html

 

[Laqup]

Slow readout is common on most sensors, barring higher end cameras. The issue is you're moving the camera, or the subject is moving. Shooting a completely static scene the shutter speed is irrelevant.

He said he isn't moving anything at all, therefore I still suspect some defect. This is not normal behavior.

Quote:

"The shoot was without tripod but i was leaning on a stable wall, truly quiet, no camera shake, no panning."

In a use case like this ES is perfectly useable (in theory) but with such slow sensors EFC is generally the better suggestion if silence is not a "must".

 

[ghostfox_1]

Slow readout is common on most sensors, barring higher end cameras. The issue is you're moving the camera, or the subject is moving. Shooting a completely static scene the shutter speed is irrelevant.

He said he isn't moving anything at all, therefore I still suspect some defect. This is not normal behavior.

He believes he wasn't, yes. That is much different than he actually wasn't moving. This is normal rolling shutter.

Quote:

"The shoot was without tripod but i was leaning on a stable wall, truly quiet, no camera shake, no panning."

In a use case like this ES is perfectly useable (in theory) but with such slow sensors EFC is generally the better suggestion if silence is not a "must".

I'd suggest op try it on an actual tripod to prove this is the case. Hand held movement is extremely likely

 

[Horshack]

Seems to be rolling shutter. Turn silent mode off and shoot with mechanical or EFCS shutter types and it'll solve the problem.

This is pretty common for most cameras.

While the outcome is probably created by sensor readout speed this can't be the classical rolling shutter issue where the user is responsible for the effect by rotating the camera during exposure. This is a static shot and no one will be able to move the camera in such a dramatic way some parts of the image are heavily affected but the image does not exhibit motion blur otherwise at all.

The other explanation found in this thread that it is IBIS/lens VR induced movement seems more plausible but this should be labeled as an IBIS/VR issue then.

It could be memory/overall circuit board related too.

They're long focal range shots, up to 600mm, so simple hand-holding shake can induce these rolling shutter artifacts over the 1/15 it takes to read out the sensor.

 

[zephxiiii]

Hello. Thank you all for your help.

This started i guess after the firmware update.

I've been shooting with the tamron 150-600mm via the FTZ 1 adaptor. but it never before gave me this problem. I will try with other lenses with and without the silent mode and i'll give feedback.

Once again, thank you all!

If the VR of that lens operates anything like my F mount Tarmon 100-400 then I can definitely see this happening on electronic shutter of this sensor.

 

[Laqup]

Slow readout is common on most sensors, barring higher end cameras. The issue is you're moving the camera, or the subject is moving. Shooting a completely static scene the shutter speed is irrelevant.

He said he isn't moving anything at all, therefore I still suspect some defect. This is not normal behavior.

He believes he wasn't, yes. That is much different than he actually wasn't moving. This is normal rolling shutter.

Quote:

"The shoot was without tripod but i was leaning on a stable wall, truly quiet, no camera shake, no panning."

In a use case like this ES is perfectly useable (in theory) but with such slow sensors EFC is generally the better suggestion if silence is not a "must".

I'd suggest op try it on an actual tripod to prove this is the case. Hand held movement is extremely likely

You won't be able to reproduce this effect, despite what you are claiming.

If this were rolling shutter extreme motion would have been in effect. Extreme motion is noticeable. On top only part of the image is affected. Not very likely to exactly time extreme motion to the short exposure times to create an effect like this.

To create something like this:



at 1/1250s shutter speed and with the 66ms readout of the Z7 respectively you would have to spin around siginifcantly..

 

[ghostfox_1]

Slow readout is common on most sensors, barring higher end cameras. The issue is you're moving the camera, or the subject is moving. Shooting a completely static scene the shutter speed is irrelevant.

He said he isn't moving anything at all, therefore I still suspect some defect. This is not normal behavior.

He believes he wasn't, yes. That is much different than he actually wasn't moving. This is normal rolling shutter.

Quote:

"The shoot was without tripod but i was leaning on a stable wall, truly quiet, no camera shake, no panning."

In a use case like this ES is perfectly useable (in theory) but with such slow sensors EFC is generally the better suggestion if silence is not a "must".

I'd suggest op try it on an actual tripod to prove this is the case. Hand held movement is extremely likely

You won't be able to reproduce this effect, despite what you are claiming.

Well, you're right, since I don't have my z7 anymore.

If this were rolling shutter extreme motion would have been in effect.

Not necessarily.

Extreme motion is noticeable. On top only part of the image is affected.

Yes, because he was only moving for part of the exposure (and therefore sensor readout).

Not very likely to exactly time extreme motion to the short exposure times to create an effect like this.

You're welcome to your opinions, but as it was fixed by switching silent shutter off, it seems that was in fact the problem.

In response to your edit, all it would take is slight motion, as at 600mm the fov is extremely small and even a slight movement would move the image that way.

 

[Laqup]

Slow readout is common on most sensors, barring higher end cameras. The issue is you're moving the camera, or the subject is moving. Shooting a completely static scene the shutter speed is irrelevant.

He said he isn't moving anything at all, therefore I still suspect some defect. This is not normal behavior.

He believes he wasn't, yes. That is much different than he actually wasn't moving. This is normal rolling shutter.

Quote:

"The shoot was without tripod but i was leaning on a stable wall, truly quiet, no camera shake, no panning."

In a use case like this ES is perfectly useable (in theory) but with such slow sensors EFC is generally the better suggestion if silence is not a "must".

I'd suggest op try it on an actual tripod to prove this is the case. Hand held movement is extremely likely

You won't be able to reproduce this effect, despite what you are claiming.

Well, you're right, since I don't have my z7 anymore.

If this were rolling shutter extreme motion would have been in effect.

Not necessarily.

Extreme motion is noticeable. On top only part of the image is affected.

Yes, because he was only moving for part of the exposure (and therefore sensor readout).

Not very likely to exactly time extreme motion to the short exposure times to create an effect like this.

You're welcome to your opinions, but as it was fixed by switching silent shutter off, it seems that was in fact the problem.

Ever heard of human physiology? Your average human is not able to be in a rotating motion to this degree and then come to a complete stop all within 66ms. This is simply not possible. Most of us are not robots after all. Even less possible with something "heavy" in hand..

THis is equally funny as the the usual mention of certian board members that some people tend to pitch their camera during the shot and that this causes AF missfocus (even during continuous shooting and w/o fov changes during those shots)..

 

[Horshack]

Slow readout is common on most sensors, barring higher end cameras. The issue is you're moving the camera, or the subject is moving. Shooting a completely static scene the shutter speed is irrelevant.

He said he isn't moving anything at all, therefore I still suspect some defect. This is not normal behavior.

He believes he wasn't, yes. That is much different than he actually wasn't moving. This is normal rolling shutter.

Quote:

"The shoot was without tripod but i was leaning on a stable wall, truly quiet, no camera shake, no panning."

In a use case like this ES is perfectly useable (in theory) but with such slow sensors EFC is generally the better suggestion if silence is not a "must".

I'd suggest op try it on an actual tripod to prove this is the case. Hand held movement is extremely likely

You won't be able to reproduce this effect, despite what you are claiming.

Well, you're right, since I don't have my z7 anymore.

If this were rolling shutter extreme motion would have been in effect.

Not necessarily.

Extreme motion is noticeable. On top only part of the image is affected.

Yes, because he was only moving for part of the exposure (and therefore sensor readout).

Not very likely to exactly time extreme motion to the short exposure times to create an effect like this.

You're welcome to your opinions, but as it was fixed by switching silent shutter off, it seems that was in fact the problem.

Ever heard of human physiology? Your average human is not able to be in a rotating motion to this degree and then come to a complete stop all within 66ms. This is simply not possible. Most of us are not robots after all. Even less possible with something "heavy" in hand..

I guess I'm a robot...

 

[sirhawkeye64]

Hello everyone!

Just got out this morning to try all proposed solution variants to this problem.

Here are the results:

# Different lens: same problem

# VR off on the lens: less notorious aberration but still noticeable in some cases

# Didn't try turn of IBIS

# Silent photography mode Off: Problem solved !!!

So I guess the read speeds of the sensor are to slow for a shutter speed such as 1/1250s.

This makes this mode absolutely rubishh for nature photography.

I tried to reduce volume on the mechanical shutter but couldn't find how. Does anyone know? Is it even possible? I thought the sounds were artificial but if it is a mechanical shutter then it might not be possible. I am so confused about this...

Thank you all for your incredible help!

I don't think you can reduce the sound of the mechanical shutter. It's just the actual sound the shutter makes (unlike the Z8/Z9 which is "electronic" and simulated/a sound playback). But glad you figured out the shutter issue.

If by "shutter" you're thinking IBIS when it engages/disengages (makes a clunk sound), this is also something you really can't control unfortunately and one likely reason why the Zf's shutter doesn't lock down when powered off. If this was your concern, my only suggestion is to leave IBIS off unfortunately is the only workaround I know of to this sound.

 

[ghostfox_1]

Ever heard of human physiology? Your average human is not able to be in a rotating motion to this degree and then come to a complete stop all within 66ms. This is simply not possible. Most of us are not robots after all. Even less possible with something "heavy" in hand..

I guess I'm a robot...

Domo arigato, Mr roboto

 

[Laqup]

Slow readout is common on most sensors, barring higher end cameras. The issue is you're moving the camera, or the subject is moving. Shooting a completely static scene the shutter speed is irrelevant.

He said he isn't moving anything at all, therefore I still suspect some defect. This is not normal behavior.

He believes he wasn't, yes. That is much different than he actually wasn't moving. This is normal rolling shutter.

Quote:

"The shoot was without tripod but i was leaning on a stable wall, truly quiet, no camera shake, no panning."

In a use case like this ES is perfectly useable (in theory) but with such slow sensors EFC is generally the better suggestion if silence is not a "must".

I'd suggest op try it on an actual tripod to prove this is the case. Hand held movement is extremely likely

You won't be able to reproduce this effect, despite what you are claiming.

Well, you're right, since I don't have my z7 anymore.

If this were rolling shutter extreme motion would have been in effect.

Not necessarily.

Extreme motion is noticeable. On top only part of the image is affected.

Yes, because he was only moving for part of the exposure (and therefore sensor readout).

Not very likely to exactly time extreme motion to the short exposure times to create an effect like this.

You're welcome to your opinions, but as it was fixed by switching silent shutter off, it seems that was in fact the problem.

Ever heard of human physiology? Your average human is not able to be in a rotating motion to this degree and then come to a complete stop all within 66ms. This is simply not possible. Most of us are not robots after all. Even less possible with something "heavy" in hand..

I guess I'm a robot...

No clue what that is supposed to be. Can you post an actual non-cropped picture like the one shown in this thread at 150mm (100 is fine as well of course) where the bottom half of the picture is completely sharp and not affected and only the upper half suffers from significant rolling shutter (while being like 100m away -> large structures)?

Your picture looks like you waving the camera around in front of a very close structure while using continuous shutter while praying that you get a picture that has some straight and some bent structures inside. But this is something completely different than the sample shown by the op? Or do you actually claim that you took an intentional single shot during a "stopping motion"?

 

[Horshack]

Slow readout is common on most sensors, barring higher end cameras. The issue is you're moving the camera, or the subject is moving. Shooting a completely static scene the shutter speed is irrelevant.

He said he isn't moving anything at all, therefore I still suspect some defect. This is not normal behavior.

He believes he wasn't, yes. That is much different than he actually wasn't moving. This is normal rolling shutter.

Quote:

"The shoot was without tripod but i was leaning on a stable wall, truly quiet, no camera shake, no panning."

In a use case like this ES is perfectly useable (in theory) but with such slow sensors EFC is generally the better suggestion if silence is not a "must".

I'd suggest op try it on an actual tripod to prove this is the case. Hand held movement is extremely likely

You won't be able to reproduce this effect, despite what you are claiming.

Well, you're right, since I don't have my z7 anymore.

If this were rolling shutter extreme motion would have been in effect.

Not necessarily.

Extreme motion is noticeable. On top only part of the image is affected.

Yes, because he was only moving for part of the exposure (and therefore sensor readout).

Not very likely to exactly time extreme motion to the short exposure times to create an effect like this.

You're welcome to your opinions, but as it was fixed by switching silent shutter off, it seems that was in fact the problem.

Ever heard of human physiology? Your average human is not able to be in a rotating motion to this degree and then come to a complete stop all within 66ms. This is simply not possible. Most of us are not robots after all. Even less possible with something "heavy" in hand..

I guess I'm a robot...

No clue what that is supposed to be. Can you post an actual non-cropped picture like the one shown in this thread at 150mm (100 is fine as well of course) where the bottom half of the picture is completely sharp and not affected and only the upper half suffers from significant rolling shutter (while being like 100m away -> large structures)?

It's an uncropped 100mm hand-held photo of my window shutters while I hand-held the camera unsteadily. The blinds near the bottom 1/3 of the frame are straight relative to the top which is not, same as the OP's second photo.

Your picture looks like you waving the camera around in front of a very close structure while using continuous shutter while praying that you get a picture that has some straight and some bent structures inside. But this is something completely different than the sample shown by the op? Or do you actually claim that you took an intentional single shot during a "stopping motion"?

Are you arguing that any intentional attempt to recreate the OP's distortion is invalid because...it's intentional? Only random acts of photography qualify?

 

[Laqup]

Slow readout is common on most sensors, barring higher end cameras. The issue is you're moving the camera, or the subject is moving. Shooting a completely static scene the shutter speed is irrelevant.

He said he isn't moving anything at all, therefore I still suspect some defect. This is not normal behavior.

He believes he wasn't, yes. That is much different than he actually wasn't moving. This is normal rolling shutter.

Quote:

"The shoot was without tripod but i was leaning on a stable wall, truly quiet, no camera shake, no panning."

In a use case like this ES is perfectly useable (in theory) but with such slow sensors EFC is generally the better suggestion if silence is not a "must".

I'd suggest op try it on an actual tripod to prove this is the case. Hand held movement is extremely likely

You won't be able to reproduce this effect, despite what you are claiming.

Well, you're right, since I don't have my z7 anymore.

If this were rolling shutter extreme motion would have been in effect.

Not necessarily.

Extreme motion is noticeable. On top only part of the image is affected.

Yes, because he was only moving for part of the exposure (and therefore sensor readout).

Not very likely to exactly time extreme motion to the short exposure times to create an effect like this.

You're welcome to your opinions, but as it was fixed by switching silent shutter off, it seems that was in fact the problem.

Ever heard of human physiology? Your average human is not able to be in a rotating motion to this degree and then come to a complete stop all within 66ms. This is simply not possible. Most of us are not robots after all. Even less possible with something "heavy" in hand..

I guess I'm a robot...

No clue what that is supposed to be. Can you post an actual non-cropped picture like the one shown in this thread at 150mm (100 is fine as well of course) where the bottom half of the picture is completely sharp and not affected and only the upper half suffers from significant rolling shutter (while being like 100m away -> large structures)?

It's an uncropped 100mm hand-held photo of my window shutters while I hand-held the camera unsteadily. The blinds near the bottom 1/3 of the frame are straight relative to the top which is not, same as the OP's second photo.

Your picture looks like you waving the camera around in front of a very close structure while using continuous shutter while praying that you get a picture that has some straight and some bent structures inside. But this is something completely different than the sample shown by the op? Or do you actually claim that you took an intentional single shot during a "stopping motion"?

Are you arguing that any intentional attempt to recreate the OP's distortion is invalid because...it's intentional? Only random acts of photography qualify?

I don't get it. I thought that you are the guy that normally likes to test things properly as your sensor readout tests show. Thanks for that btw!

Do you really believe that your example is a good approximation of the effect shown in the OPs examples? What about subject distance and "required motion" to bent structures of a certain scale in a certain distance? What about other parts of the same image being actually sharp and unaffected? I don't actually see that reflected well in your example, which is a mushy, noisy, 1MP mess without any real detail and of course a very close structure.. I'll try this myself with a proper setup. Maybe it is rolling shutter after all.

 

[ghostfox_1]

I don't get it. I thought that you are the guy that normally likes to test things properly as your sensor readout tests show. Thanks for that btw!

Do you really believe that your example is a good approximation of the effect shown in the OPs examples? What about subject distance and "required motion" to bent structures of a certain scale in a certain distance? What about other parts of the same image being actually sharp and unaffected? I don't actually see that reflected well in your example, which is a mushy, noisy, 1MP mess without any real detail and of course a very close structure..

So let's assume that horshacks test was bad.

That doesn't explain how you can believe you're correct about it not being rolling shutter when switching from silent mode solved the problem.

What's your theory?

 

[Laqup]

I don't get it. I thought that you are the guy that normally likes to test things properly as your sensor readout tests show. Thanks for that btw!

Do you really believe that your example is a good approximation of the effect shown in the OPs examples? What about subject distance and "required motion" to bent structures of a certain scale in a certain distance? What about other parts of the same image being actually sharp and unaffected? I don't actually see that reflected well in your example, which is a mushy, noisy, 1MP mess without any real detail and of course a very close structure..

So let's assume that horshacks test was bad.

That doesn't explain how you can believe you're correct about it not being rolling shutter when switching from silent mode solved the problem.

What's your theory?

I'd like to see a test without IBIS + VR.

I'm shooting some cameras exclusively with ES (faster sensor than the Z6/Z7/Zf though) and for a "static" shot this effect looks really crazy. So if those are actually static shots and this effect can be repeated any time (while really concentrating on camera motion) I suspect either an actual defect or something induced by IBIS/VR.

Even if you get lucky and somehow manage to take a similar shot like the ones shown by the OP, I am pretty sure that you need many tries to do so and that you cannot easily replicate this behaviour. OP seems to be able to do so, therefore something else seems to be amiss (or he was not specific enough, e.g. about percentage of occurence).