Autofocus issue after firmware update ( D800 )

Hi,

Just thought I'd update this with the fact that the same AF(S) error occurs when using my Quantum Trio via cable to D800 hotshoe. If I use the red focus assist beam from the Quantum Trio, back focus is clearly evident. If I turn off the Trio focus assist lamp, the AF is perfect, subject is tack sharp. (This is using the 24-70mm at 70mm at F2.8 with a subject 6 feet away) At this distance (using the focus assist lamp) focus is achieved at about 1 inch behind the subject.

Regards,

Richard

Zorpie said:

Hi,

Just thought I'd update this with the fact that the same AF(S) error occurs when using my Quantum Trio via cable to D800 hotshoe. If I use the red focus assist beam from the Quantum Trio, back focus is clearly evident. If I turn off the Trio focus assist lamp, the AF is perfect, subject is tack sharp. (This is using the 24-70mm at 70mm at F2.8 with a subject 6 feet away) At this distance (using the focus assist lamp) focus is achieved at about 1 inch behind the subject.

Regards,

Richard

Click to expand...

Thanks for the update.

Would you like to try another thing? Try to cover the AF assist LED completely so that no red light is reaching the focus target. Does it still back focus? It has been said that the wavelength of red light is causing the problem, but I doubt that since if I cover the AF assist LEDs of the SB-900 the back focus is still there.

olyflyer said:

Zorpie said:

Hi,

Just thought I'd update this with the fact that the same AF(S) error occurs when using my Quantum Trio via cable to D800 hotshoe. If I use the red focus assist beam from the Quantum Trio, back focus is clearly evident. If I turn off the Trio focus assist lamp, the AF is perfect, subject is tack sharp. (This is using the 24-70mm at 70mm at F2.8 with a subject 6 feet away) At this distance (using the focus assist lamp) focus is achieved at about 1 inch behind the subject.

Regards,

Richard

Click to expand...

Thanks for the update.

Would you like to try another thing? Try to cover the AF assist LED completely so that no red light is reaching the focus target. Does it still back focus? It has been said that the wavelength of red light is causing the problem, but I doubt that since if I cover the AF assist LEDs of the SB-900 the back focus is still there.

Click to expand...

The D800E and the D4 should have the same problem if the problem is connected to the AF module since they have the same AF module, the D600/610 should also have it if it has anything to do with some common firmware module connected to the control of on/off function of the AF assist LED. All those cameras have the same DOF since they have the same sensor size and the back focus is so severe that it is not related to high megapixels. If those cameras are perfectly OK then there is a clear bug in the D800.

Hi Olyflyer,

Thanks for the info - I left the focus assist lamp on the Quantum Trio ON and blocked the light with my finger so it wasn't hitting the target. Still most definitely back focussed. Turned the focus assist off - subject perfectly sharp. So it isn't anything to do with the frequency shift of the light itself causing the issue.

I don't have another Nikon body to try - used to have the D3 and the D700 but have 'scaled down' operations here and only now have the D800.

Regards,

Richard

Zorpie said:

Hi Olyflyer,

Thanks for the info - I left the focus assist lamp on the Quantum Trio ON and blocked the light with my finger so it wasn't hitting the target. Still most definitely back focussed. Turned the focus assist off - subject perfectly sharp. So it isn't anything to do with the frequency shift of the light itself causing the issue.

I don't have another Nikon body to try - used to have the D3 and the D700 but have 'scaled down' operations here and only now have the D800.

Regards,

Richard

Click to expand...

Hi, and thank you for doing more tests.

I have contacted Nikon, directed them to this thread and a blog, and they have also initially said that it is a known issue and is caused by the colour of the light, but when I asked them to READ the blog and explained that the problem is there even when the light is covered they asked me to do some more tests, so I did that last night and sent them the raw images.

As a matter of fact, during my tests last night I discovered that the issue is even bigger. You can't seem to be able to get rid of the back focus, if you stop down the focus moves further back. I took some shots in A and P mode as well, just in case it only existed in M, which I used before, and seen that in A there is no difference, but in P there is a HUGE difference. The camera chose f/5.6 and 1/50s (I had the 50/1.4G on the camera) and with this small aperture the focus error was expected to be hidden by the DOF. Guess how surprised I was when I saw that the focus moved so much to the back that the actual focus area was worse than at f/1.4 with the original back focus error. It is really horrible.

I think Nikon must take this very seriously because this error renders the D800 useless as the number one event shooter camera. It is impossible to use the camera in situations where the light is low and you need to use flash and the AF assist.

Okay now i was able to reproduce the issue using a 50mm 1.8G and a 24-70 2.8..
The red beam clearly causes back focusing in 3 out of 4 shots.. When i turn it off and still have the flash on but without the AF assist, the issue doesnt happen.

I tried the same on my D600 using the same lenses and the 85mm ( it was the OP lens used on the D800 ) and everything was fine, no focusing issues whatsoever with read beam from the SB900..

I have yet to try a SB910 or SB800 and see.. but this is clearly an issue with the D800.

imAvalanche said:

Okay now i was able to reproduce the issue using a 50mm 1.8G and a 24-70 2.8..
The red beam clearly causes back focusing in 3 out of 4 shots.. When i turn it off and still have the flash on but without the AF assist, the issue doesnt happen.

Click to expand...

I am now quite sure that the issue is not lens related, it is related to the fact that the camera turns on the AF assist LEDs of the flash. Even if you cover the LEDs with a thick cardboard the AF will be off (back focus) even though the light level is enough for focusing without AF assist, meaning that covering the AF assist LED should not prevent the camera from focusing accurately. So the problem is not connected to the light colour, the pattern or anything else than the fact that it is turned on. Without the AF assist there is no problem at all.

imAvalanche said:

I tried the same on my D600 using the same lenses and the 85mm ( it was the OP lens used on the D800 ) and everything was fine, no focusing issues whatsoever with read beam from the SB900..

Click to expand...

That should exclude the possibility of the AF assist light colour temperature or pattern causing the back focus even more. If the D600 works well with the SB-900 AF assist then it is fairly obvious that the problem is elsewhere.

imAvalanche said:

I have yet to try a SB910 or SB800 and see.. but this is clearly an issue with the D800.

Click to expand...

We can still not say so. Remember that if it is connected to the AF module then both the D800 and the D800E, as well as the D4 is effected, since all three are using the same AF module. If it is connected to firmware then we don't know yet. There might be shared parts between the cameras, if this part is shared or not, I have no idea.

olyflyer said:

olyflyer said:

I mean, is it really possible to downgrade the firmware?

Click to expand...

Click to expand...

A few members have tried and report that it works.

olyflyer said:

olyflyer said:

Isn't the old release identified and update aborted automatically with a message that you already have a new firmware and that there is no update done?

Click to expand...

Click to expand...

You'd think that it would check the version but apparently it does not.

Maybe this is to preempt the following scenario :Nikon comes out with a new version but it has a bug that has not been noticed by Nikon. Not checking the version allows users to go back to a tried and true firmware while waiting for a fix for the new program.

TOF guy said:

olyflyer said:

TOF guy said:

I mean, is it really possible to downgrade the firmware?

Click to expand...

Click to expand...

A few members have tried and report that it works.

TOF guy said:

TOF guy said:

Isn't the old release identified and update aborted automatically with a message that you already have a new firmware and that there is no update done?

Click to expand...

Click to expand...

You'd think that it would check the version but apparently it does not.

Maybe this is to preempt the following scenario :Nikon comes out with a new version but it has a bug that has not been noticed by Nikon. Not checking the version allows users to go back to a tried and true firmware while waiting for a fix for the new program.

Click to expand...

OK, all that is fine, but where is a link to old firmware and how can it be done?