A7RV Camera problem? Left side image blur

ttusteveo

Forum Enthusiast
Messages
415
Reaction score
279
I just bought a new lens (Batis 85mm f/1.8) and in testing it, noticed that at full magnification the left side of the image seems softer than the right.

I then mounted my 20-70 f/4, and took the same photograph of my neighbor's house. I shot both lenses wide open, but I'm shooting from my yard across the street - so I'd expect sufficient DOF. I see the same effect on the 20-70 which isn't that fast.

If I hold the camera upside down and reshoot, the blur follows - the right side then seems soft.

See attached images, one taken with each lens. The brick sharpness starts to fall apart on both images on the far left side. I've not noticed an issue in normal practice, but I haven't had the camera all that long.

Do I have an issue with the camera or lenses?

View attachment fbce2ec0aefe4c539bc3fb9cc692dbe8.jpg
Far left side sharpness falloff , Batis 85mm

View attachment a3f7686d113245e2910db772a010947f.jpg
Far left side sharpness falloff, Sony 20-70
 
Last edited:
What’s likely is that the sensor is improperly shimmed or the lens mount has been knocked out of alignment. Has the camera ever had to be serviced for any reason?
 
Never serviced. The camera is only a few months old and has probably less than 2000 shutter actuations.
 
Hard for me to tell from the images on my phone, but if the plane of focus trades front to back, left to right, then opposite when shot upside down (and rotated in post) with all lenses, then it would mean the sensor and mount are out of alignment. There’s no service centre that will fix this, unfortunately. Sony only calibrates this from the factory (sensor shims) and does not provide any tooling or procedures for their licensed service centres to correct it.

On the other hand, it is only two lenses, and could very well be a coincidentally similar lens tilt/decentering issue with both lenses
 
That is an odd one for sure... you would think if it was a "tilt" issue, it would be more of a gradual shift in sharpness and not quite as abrupt as it seems in your photos. Did you try it with IBIS off and on? The RV has multi axis IBIS, I'd be curious if it the same when stationary vs active.

Maybe its something simple like some cleaning solution residue on one edge of the sensor from the last wet clean someone did on it? Could the rear elements of both lenses have a partial palm print on the edge? We are creatures of habit, and how we hold one lens when changing lenses and caps is probably how we hold all of them, heh.

I guess you could check the screws on the mounts involved while you are at it and make sure they are reasonably snug (don't over tighten)

Please report back your findings and might be something that could help others someday.
 
I just bought a new lens (Batis 85mm f/1.8) and in testing it, noticed that at full magnification the left side of the image seems softer than the right.

I then mounted my 20-70 f/4, and took the same photograph of my neighbor's house. I shot both lenses wide open, but I'm shooting from my yard across the street - so I'd expect sufficient DOF. I see the same effect on the 20-70 which isn't that fast.

If I hold the camera upside down and reshoot, the blur follows - the right side then seems soft.

See attached images, one taken with each lens. The brick sharpness starts to fall apart on both images on the far left side. I've not noticed an issue in normal practice, but I haven't had the camera all that long.

Do I have an issue with the camera or lenses?
I wonder if that's due to depth of field. The left corner of the house is further away than the right corner. Choose a smaller aperture or a flatter shooting surface to test and always focus on the same point.
 
  • Like
Reactions: Lan
Never serviced. The camera is only a few months old and has probably less than 2000 shutter actuations.
I'd do a few more tests on different walls first, but then contact Sony with the images as evidence.

It might be that somehow your sensor or lens seat (or both) are misaligned. Perhaps the IBIS suspension is damaged?

BTW.. I've never seen this before. Did you buy the camera new or used? Have you (or someone else) swabbed the sensor with excessive force perhaps?
 
Last edited:
Never serviced. The camera is only a few months old and has probably less than 2000 shutter actuations.
I'd do a few more tests on different walls first, but then contact Sony with the images as evidence.

It might be that somehow your sensor or lens seat (or both) are misaligned. Perhaps the IBIS suspension is damaged?

BTW.. I've never seen this before. Did you buy the camera new or used? Have you (or someone else) swabbed the sensor with excessive force perhaps?
Don't use walls. Use a far horizon line. Go some high grounds in a calm, dry, not too hot day. Aim for something that is like 100 or 300 meters away, go manual focus, shoot with tripod and rotate with trypod.
 
Hard for me to tell from the images on my phone, but if the plane of focus trades front to back, left to right, then opposite when shot upside down (and rotated in post) with all lenses, then it would mean the sensor and mount are out of alignment. There’s no service centre that will fix this, unfortunately. Sony only calibrates this from the factory (sensor shims) and does not provide any tooling or procedures for their licensed service centres to correct it.

On the other hand, it is only two lenses, and could very well be a coincidentally similar lens tilt/decentering issue with both lenses
And given that the camera is new it is in warranty. Your testing seems adequate to me. Get Sony involved. Good luck!
 
I think I sorted this out... Operator error on multiple fronts.

I shot these (and other) test pictures wide open to exploit any flaws with the lenses. I expected a shallow DOF but figured it would be deep enough for the subject.

When I raise this camera to my face and look through the viewfinder, I finally realized that I'm not holding it perfectly perpendicular to the subject. I'm maybe a few degrees off. It's a new camera for me and handles completely differently from my Fuji. I need to roll my right wrist/hand outward a bit to square the camera better on the subject. Realistically I'd never take this shot wide open, so this misalignment probably wouldn't matter in practice.

Second, my DOF expectations are skewed by both the change from APS-C to FF and by the massive resolution increase from the X-T2 to the A7RV. The A7RV is cruel in that regard - you can zoom in so much further and it will reveal things you wouldn't see at lower resolutions.

I was asked if I'm happy with the 20-70. I'm thrilled with the versatility of it. The lens correction profiles - especially DXO - deal with the expected distortions well. My jaw dropped when I saw the first pictures from this body and lens, coming from the X-T2.

I'm very excited about the Batis. I've wanted this lens for a long time (years ago when I had Sony APS-C before), and I got what I thought was an amazing Amazon Prime day deal on it, about $600 new. Used ones seem to go for around that.
 
Last edited:
I think I sorted this out... Operator error on multiple fronts.

I shot these (and other) test pictures wide open to exploit any flaws with the lenses. I expected a shallow DOF but figured it would be deep enough for the subject.

When I raise this camera to my face and look through the viewfinder, I finally realized that I'm not holding it perfectly perpendicular to the subject. I'm maybe a few degrees off. It's a new camera for me and handles completely differently from my Fuji. I need to roll my right wrist/hand outward a bit to square the camera better on the subject. Realistically I'd never take this shot wide open, so this misalignment probably wouldn't matter in practice.

Second, my DOF expectations are skewed by both the change from APS-C to FF and by the massive resolution increase from the X-T2 to the A7RV. The A7RV is cruel in that regard - you can zoom in so much further and it will reveal things you wouldn't see at lower resolutions.

I was asked if I'm happy with the 20-70. I'm thrilled with the versatility of it. The lens correction profiles - especially DXO - deal with the expected distortions well. My jaw dropped when I saw the first pictures from this body and lens, coming from the X-T2.

I'm very excited about the Batis. I've wanted this lens for a long time (years ago when I had Sony APS-C before), and I got what I thought was an amazing Amazon Prime day deal on it, about $600 new. Used ones seem to go for around that.
If I understand correctly you noticed that you pointed the camera not completely perpendicular to the subject. Did you use the artificial leveling of the camera? Anyway, you repeated the test and now the image is left-right equally sharp?

Btw, thanks for your comment on the combi A7Rv with the 20-70 f/4.0 G lens. It is my preferred combi for the near future. I am already using DxO PL(9)/FP7/VP4 since 2023 (and before that time PureRAW) with my RX10iv.
 
I think I sorted this out... Operator error on multiple fronts.

I shot these (and other) test pictures wide open to exploit any flaws with the lenses. I expected a shallow DOF but figured it would be deep enough for the subject.

When I raise this camera to my face and look through the viewfinder, I finally realized that I'm not holding it perfectly perpendicular to the subject. I'm maybe a few degrees off.
There have been whole threads of people "testing" lenses incorrectly like this.

Don't feel bad.. it's extremely difficult to be precisely geometrically normal to an (almost) flat plane like a wall and fast glass wide open has an extremely shallow DOF (and it may not have a perfectly flat focal plane anyway).

Furthermore when you are close to the subject any tiny misalignment error on your part is further attenuated.

That's why I will never even go there.. my lenses are plenty good for me.

All lenses are softer near the edges and especially in the corners.

(And who shoots walls wide open anyway .. LOL :)
 
Last edited:
I think I sorted this out... Operator error on multiple fronts.

I shot these (and other) test pictures wide open to exploit any flaws with the lenses. I expected a shallow DOF but figured it would be deep enough for the subject.

When I raise this camera to my face and look through the viewfinder, I finally realized that I'm not holding it perfectly perpendicular to the subject. I'm maybe a few degrees off. It's a new camera for me and handles completely differently from my Fuji. I need to roll my right wrist/hand outward a bit to square the camera better on the subject. Realistically I'd never take this shot wide open, so this misalignment probably wouldn't matter in practice.

Second, my DOF expectations are skewed by both the change from APS-C to FF and by the massive resolution increase from the X-T2 to the A7RV. The A7RV is cruel in that regard - you can zoom in so much further and it will reveal things you wouldn't see at lower resolutions.

I was asked if I'm happy with the 20-70. I'm thrilled with the versatility of it. The lens correction profiles - especially DXO - deal with the expected distortions well. My jaw dropped when I saw the first pictures from this body and lens, coming from the X-T2.

I'm very excited about the Batis. I've wanted this lens for a long time (years ago when I had Sony APS-C before), and I got what I thought was an amazing Amazon Prime day deal on it, about $600 new. Used ones seem to go for around that.
If I understand correctly you noticed that you pointed the camera not completely perpendicular to the subject. Did you use the artificial leveling of the camera? Anyway, you repeated the test and now the image is left-right equally sharp?

Btw, thanks for your comment on the combi A7Rv with the 20-70 f/4.0 G lens. It is my preferred combi for the near future. I am already using DxO PL(9)/FP7/VP4 since 2023 (and before that time PureRAW) with my RX10iv.
The level meter will show you rotation, and pitch up/down, but not whether you're off axis left/right. (Unless there is a setting for this I haven't discovered yet.). You can have everything lit green but still not be perpendicular.

Taking more care, I'd say I got a better result (attached)... noting that you have to zoom far beyond 100% to notice differences between the hard left and right sides. Stopped down a bit, and you don't see any difference each side.

My guess is that if I had a camera body issue (lens/mount alignment problem), I could not achieve the sharp results in the beige brick photo, wide open, at close distance. The DOF is very shallow at such close distance.

View attachment ef6108a0248442df9ea14fc9dd3ddea8.jpg
Batis 85 at f/1.8, attempt No. 2. Edges are softer than center, as expected, but more even to each other.

View attachment ed405b40a6b84382b40a48277b9e6290.jpg
Batis 85, stopped down - much better result

View attachment 1a860ba7ffea4c49b0a3e76de513a4c1.jpg
Batis 85 at f/1.8, shot much closer - so the DOF would be very shallow.
 
Last edited:
I think I sorted this out... Operator error on multiple fronts.

I shot these (and other) test pictures wide open to exploit any flaws with the lenses. I expected a shallow DOF but figured it would be deep enough for the subject.

When I raise this camera to my face and look through the viewfinder, I finally realized that I'm not holding it perfectly perpendicular to the subject. I'm maybe a few degrees off.
There have been whole threads of people "testing" lenses incorrectly like this.

Don't feel bad.. it's extremely difficult to be precisely geometrically normal to an (almost) flat plane like a wall and fast glass wide open has an extremely shallow DOF (and it may not have a perfectly flat focal plane anyway).

Furthermore when you are close to the subject any tiny misalignment error on your part is further attenuated.

That's why I will never even go there.. my lenses are plenty good for me.

All lenses are softer near the edges and especially in the corners.

(And who shoots walls wide open anyway .. LOL :)
I agree :) . I was just wanting to know whether I had a good lens before I lost the opportunity to exchange it. In my other reply, I include another captivating brick picture shot just a few feet away. To me those results look very good, so I think really everything is okay with the body and lens. We're looking at sand grains in mortar now, and I'll be using the new lens for portraits which will be far sharper than what's required anyway. (Still, I'm amazed by what I can see with 60mp.)
 
I think I sorted this out... Operator error on multiple fronts.

I shot these (and other) test pictures wide open to exploit any flaws with the lenses. I expected a shallow DOF but figured it would be deep enough for the subject.

When I raise this camera to my face and look through the viewfinder, I finally realized that I'm not holding it perfectly perpendicular to the subject. I'm maybe a few degrees off.
There have been whole threads of people "testing" lenses incorrectly like this.

Don't feel bad.. it's extremely difficult to be precisely geometrically normal to an (almost) flat plane like a wall and fast glass wide open has an extremely shallow DOF (and it may not have a perfectly flat focal plane anyway).

Furthermore when you are close to the subject any tiny misalignment error on your part is further attenuated.

That's why I will never even go there.. my lenses are plenty good for me.

All lenses are softer near the edges and especially in the corners.

(And who shoots walls wide open anyway .. LOL :)
I agree :) . I was just wanting to know whether I had a good lens before I lost the opportunity to exchange it. In my other reply, I include another captivating brick picture shot just a few feet away. To me those results look very good, so I think really everything is okay with the body and lens. We're looking at sand grains in mortar now, and I'll be using the new lens for portraits which will be far sharper than what's required anyway. (Still, I'm amazed by what I can see with 60mp.)
Exactly my experience. That sensor is simply amazing.
 
I think I sorted this out... Operator error on multiple fronts.

I shot these (and other) test pictures wide open to exploit any flaws with the lenses. I expected a shallow DOF but figured it would be deep enough for the subject.

When I raise this camera to my face and look through the viewfinder, I finally realized that I'm not holding it perfectly perpendicular to the subject. I'm maybe a few degrees off.
There have been whole threads of people "testing" lenses incorrectly like this.

Don't feel bad.. it's extremely difficult to be precisely geometrically normal to an (almost) flat plane like a wall and fast glass wide open has an extremely shallow DOF (and it may not have a perfectly flat focal plane anyway).

Furthermore when you are close to the subject any tiny misalignment error on your part is further attenuated.

That's why I will never even go there.. my lenses are plenty good for me.

All lenses are softer near the edges and especially in the corners.

(And who shoots walls wide open anyway .. LOL :)
I agree :) . I was just wanting to know whether I had a good lens before I lost the opportunity to exchange it. In my other reply, I include another captivating brick picture shot just a few feet away. To me those results look very good, so I think really everything is okay with the body and lens. We're looking at sand grains in mortar now, and I'll be using the new lens for portraits which will be far sharper than what's required anyway. (Still, I'm amazed by what I can see with 60mp.)
BTW, I meant to say accentuated and not attenuated .. but I think you got the point ;)
 
When I raise this camera to my face and look through the viewfinder, I finally realized that I'm not holding it perfectly perpendicular to the subject. I'm maybe a few degrees off.
This is why I suggested you to aim for something very far away, to remove this difference
 
When I raise this camera to my face and look through the viewfinder, I finally realized that I'm not holding it perfectly perpendicular to the subject. I'm maybe a few degrees off.
This is why I suggested you to aim for something very far away, to remove this difference
Understood. It's a test I'll perform when I'm able.
 

Keyboard shortcuts

Back
Top