Full spectrum mod - gone bad?

[JCDowdy]

Mark,

 

Wow this has certainly turned into an expensive research project, so sorry.

 

I keep coming back to the fact that DaveO has a D750 which apparently does not suffer this problem.

 

Perhaps his conversion service knows something about this?

[Cadmium]

Has anyone sent this link to LifePixel yet?

[Andrea B.]

Mark, your first step is to have Life Pixel remedy this situation --

• by determining the cause of the light leak and "plugging" it up,
  (like, should the clear replacement filter have some kind of tape around the edges?)
  
• OR by restoring the camera to its original condition by replacing the internal filter set.
  

Life Pixel should be covering shipping fees at this point via a refund or pre-paid shipping labels.

They are nice folks there, so I would get on the phone to them and work this out.

 

As for further use of this D750? This whole thing is really a puzzlement for sure.

We can try to think this thing through a bit further and gather all observations we can.

 

Questions:

• Were the tests you showed above made in the dark?
  I am assuming they were in the dark with lens cap on, but it would be useful to know.
  

• What happens when you make the same test outdoors with lens cap on?
  Does the light leak worsen or is it the same?
  

• Does the amount or placement of the light leak change with changes in exposure length?
  Run an in-the-dark series at a fixed ISO for increments of 5 seconds.
  

• Does the amount or placement of the light leak change with changes in ISO?
  Run an in-the-dark series at a fixed exposure time for 1-stop ISO increments.
  

• Open every port door and make a long exposure indoors in the dark and then outdoors in the light.
  

 

I'm looking for the screwdrivers to apply to my D750 to see what the heck is in there. :)

[nfoto]

At this point, Lifepixel do have to step in some way or other to help solve the poor customer's issue. We can forgive them easily that the first mod went bad if this was a camera unknown to them, but not the second time around.

 

Either Lifepixel discovers why the camera misbehaves and make the modification asked for by the customer, or they should restore the camera to factory specification - free of charge. Additionally reimbursement of the extra shipping cost(s) is not too much to ask for.

[Guest]

I would love to get this light flare problem fixed, and if its something I can do then I will, otherwise I think I've decided to finally draw the line on this one and move on. And by move on I mean, either sell the camera as 'damaged goods', or use it for tinkering and learning/practice for doing camera surgeries (i.e., taking the thing apart, putting it back together, repeat, etc). Either option is overly expensive, and it will be some time before I can afford a new camera again - which means I'll be going back to shooting UV with my D70 (which has been good to me thus far).

 

@ Andrea: in response to your questions...

• All test shots were in a completely darkened room/closet/bag/blanket/combination of all the aforementioned (in short, there was no stray light). And yes, the body cap was on; except for test shots I tried with different lenses on (which showed some variability of their own, but otherwise still does not address the main problem).
  

• I didn't try making any of the tests outdoors, but in a roughly equivalent situation I made tests in a very well lit room, pointing the camera at various angles to/away from lights (again, this showed some variability in the results, but did not change the light flare, and still does not impact main problem).
  

• The placement and pattern of the light flare does not change with changes in exposure length and/or time, only the intensity. That is, greater exposure setting of any kind (time or ISO) result in brighter light flare. Page one of this post has these results (ISO setting simply intensifies or reduces this effect).
  

Note that I am not calling this a light 'leak' because this light is present, internally (evidenced when the camera has been wrapped in black plastic, and a blanket, put into a closet, door closed, all lights off, at night - and its still there, unchanged in the least). Thus, I have taken to calling it a light 'flare', pointing more to how it looks in the produced images.

 

Now, there is one interesting thing I found which does impact the light flare: Exposure Delay. In this model camera there is a setting to wait some amount of time after flipping up the mirror before the exposure is made (to compensate for blur from mirror bounce in long exposures). I took a couple shots to show the difference here (the image inlays are pushed, simply to help show the difference between shots).

 

This delay appears to remedy the problem. There is still just a hint of the light flare, but it is so minimal it is practically gone.

 

Now, while this does present a way of compensating for the problem I certainly wouldn't call it a "fix". For now, until I can get a new camera, it may just have to do. I like this camera a lot (aside from the flare), because of its functionality, performance, and UV sensitivity. I'm going to see if there is a way I can do some kind of automatic subtraction to remove/negate the light flare in-camera, or PP, using a control shot with the flare. Otherwise, I'll be posting contaminated photos for a while here on the forums.

 

I'll do my best to edit out the flare whenever I can.

[Andrea B.]

Good work Mark in finding this observation about the exposure delay. I now tentatively think that this is telling us we might have a IR shutter monitor problem. At this point there doesn't seem to be anything else it could be! Thanks for confirming your test shots were all made "in the dark".

 

If this is an IR shutter monitor problem, then there may be a way to patch it up in there? Somehow the removal of the internal filter set and dust cleaner frame has let shutter monitor light leak around the replacement filter?

 

Why is this seen in your camera and not in others? It possibly could have something to do with the way Life Pixel performs conversion? Or it could be that in your particular D750 there is some mis-placement of the IR shutter monitor?

• Here is something to try which might help you get some use from the D750 until you can get some help with the light contamination : Set the D750 on DX mode. This will cut the number of pixels in in the finished shot, but it will also force the shot to the center and thus avoid the lower right area of contamination. There is also a 1.2X mode which is a bit larger than DX.
  
  • Photo Shooting Menu > Image area > Choose image area > DX > OK button
    

• Here is another thing to try: Mup, mirror up, on the upper left dial. When in Mup mode, the first shutter press raises the mirror and the second shutter press opens the shutter to make the shot. The purpose is to let vibrations die down and prevent motion blur. I have this permanently set on my UV cams as a personal choice because UV shots always seem to be in the range where mirror slap blur is a problem. But I am wondering if this might also "turn off" the IR shutter monitor? We need to investigate that by Googling around, so will do and get back to everyone.
  

Hey, don't give up !!!!! I went through this kind of thing in June of 2010. Seemed to take forever to figure it out. But eventually it served the greater good to learn about shutter monitor flare. I hope it all resolves more easily here for your D750 and that you can do something about it. We'll do what we can to help.

 

I'm going now to write someone who might be able to give us some insights.

[Andrea B.]

Got the email written to ask some questions about your D750. Will eventually report back. :)

 

When you get a moment, please test for the leak when using Mup mode? Thanks!!!

[Andrea B.]

Hi Mark -

 

I have been talking to Dan Llewellyn of MaxMax (LDP LLC) about your D750 conversion. He does not think that the contamination is from an IR shutter monitor. In Dan's experience (and mine), any IR shutter monitor contamination appears as a large, faded, "foggy" area in the photograph.

 

Given that you are sure there is no light leak from port doors, from around the top LCD or from camera cracks,

Dan mentioned two other possible sources of light leak that he has seen:

• LED light from viewfinder status display. Might be able to disconnect this?
  
• Light from status LEDs such as the write-to-memory-card LED. Tape it over inside camera. (This might be fairly easy.)
  

Given that setting Exposure Delay prevented the light leak, I do want you to please also test whether setting Mirror Up (Mup) on the top left dial prevents the light leak?

 

And then one more little test, does the light leak occur when using Live View (both with and without Mup)?

 

The goal here is to solve the problem so you can use the D750. I'm still thinking this might still be a factory problem and that ultimately we should press Nikon for a replacement. But first try those two tests.

 

One more check: Is the AF Illuminator turned off? Just in case. Custom Setting Menu > a Autofocus > a9 Built-in AF-assist illuminator > OFF > OK.

[Guest]

I just took the test shots you asked about. I'd like to note, in response to Dan's suggestion that it may not be from an IR shutter monitor... while there is an obvious large flare area, there is also a background foggy area (much easier to see in the originals). This could be a result of IR shutter contamination, or simply scatter from the flare area - I don't know that it is possible to discern which at this time.

 

First to note - turning off the AF illuminator did not solve the problem, but it did in fact have some impact on the image, albeit very very slight (much easier to see the difference when compared in layers in an image editor). That there is a difference at all says something.

 

Next to note is that in the Shutter Delay and Mup modes the exposures above are 3x's longer, further showing how effective those modes are - and further evidence to support just what the problem is. Though I still don't know yet. I am curious about the possibility of the culprit being something like a status LED or something like that. Do you know of any online instruction I might find to aid me in getting inside the camera to tape it off? Maybe there is some way to simply shut it off?

[Andrea B.]

AHA !!! The problem all but disappears when using Mup or Exposure Delay. Very interesting.

So, thank you very much for performing those tests, Mark.

 

Shooting 15 seconds at ISO 12800 is a fairly serious stress test.

If your editor has an Auto-Levels tool, try that and see if it opens up the photo a little more to see the foggy area?

 

[bTW, just for WIW, artifacts may be seen in a stressed foto which may never appear in ordinary conversion and editing. We hafta use common sense. Not a reminder to you, just for someone reading who may not know this. :) ]

 

---

 

The Clues, Part I

• Prior to conversion, the Nikon-repaired D750 shows no light contamination. Post #23, middle photo.
  
• After conversion, the D750 has light contamination. Post #23, bottom photo.
  
• Conclusion:
  Conversion creates a contamination problem in this particular D750.
  We also know that other converted D750s do not exhibit this contamination.
  

The Clues, Part II

• Using 1-second shutter delay, contamination is almost supressed.
  
• In Mup mode, contamination is supressed even more.
  
• Conclusion: Danged if I know !?!
  The mirror is always raised prior to opening the shutter. So when attempting to suppress this light leak, why would it matter if the mirror was raised immediately before shutter-open or a couple of seconds before shutter-open? Mup mode might turn off the IR shutter monitor, but if so that seems irrelevant because the contamination is still happening.
  

 

I don't think it is the LED write light causing a problem because that would not be lit up until after the shutter closes.

 

I'm thinking that there is some gap around the replacement filter. Things are not fitting entirely properly there post conversion. Then some internal light leaks through.

 

---

 

Mark, please make sure you have the following items set to turn off as many lights (or certain active circuitry) as possible which might be active when the shutter is raised. These settings are just temporary. If there is no improvement, then of course restore your desired settings. :)

• Photo Shooting Menu > High ISO NR > OFF
  
• Photo Shooting Menu > Long exposure NR > OFF
  
• Custom Setting Menu > a Autofocus > a3 Focus tracking with lock-on > OFF
  
• Custom Setting Menu > a Autofocus > a5 Focus point illumination > OFF
  
• Custom Setting Menu > a Autofocus > a9 Built-in AF-assist illuminator > OFF
  
• Custom Setting Menu > d Shooting/display > d7 Viewfinder grid display > OFF
  
• Custom Setting Menu > d Shooting/display > d10 LCD illumination > OFF
  
• Setup Menu > Wi-Fi > Network connection > Disable
  

Please run just one more test with those settings and report any differences. Thank you!

---

 

I'm looking for the D750 tear down link. But I feel obliged to warn you that you can damage your camera if you open it. These days I won't perform a DIY conversion anymore after frying two D7000s in a row in spite of previous successful Kitchen Table Conversions. The typical e-bracelet and e-mat are absolutely required along with special screwdrivers. And there are flash capacitors in these cams which can fry you.

[Andrea B.]

 

Link from Life Pixel for DIY on D750:

https://www.lifepixel.com/tutorials/infrared-diy-tutorials/life-pixel-nikon-d750-diy-digital-infrared-conversion-tutorial

 

This will get you inside the D750 down to the sensor level. Just please be careful!!

[colinbm]

Screws are one thing.....ZIF cables are another !

[Cadmium]

I am going to quote Andrea, and ask this question again:

"And then one more little test, does the light leak occur when using Live View (both with and without Mup)?"

Just curious.

Thanks.

[Guest]

Hi everyone,

 

Sorry for the delayed reply. I just got back from a business trip to Germany, and I'm still getting reacclimated to this time zone (happens every time!).

 

I just did the requested tests - under the usual conditions (body cap on, covered camera, fully darkened room: ISO12800, 15s). In the resulting images presented here I overlaid a copy of each image with an enhanced version, again to help show results. The enhanced overlay is simply the image with +2.00 exposure setting during RAW import. The number in the overlay is the mean pixel value for the enhanced exposure version of the image - just to put some numbers to the relative results.

 

All of the following images use the adjusted camera settings recommended by Andrea.

 

As you can see, the adjusted camera settings did not have an effect. Using a self-timer had no effect either. Again, what did practically remedy the problem was using either an exposure delay, or Mup mode. The results from each are effectively the same (note though that even in these images some trace of the flare is still there - which I only mention for the sake of helping to diagnose the root cause). Live View with Mup is equivalent to simply using Mup mode. Interestingly, Live View itself (no Mup) did have an impact on the flare problem. If I may speculate at this point, perhaps it is due to some delay in the exposure between the time Live View shuts off and opens the mirror before taking the photo and the time the shutter is actuated? I'm guessing this based solely on the sequence of clicking and switching sounds the camera makes when taking a shot from Live View mode.

 

I don't know if any of these results helps point to the cause. Also, while I do have the tools and access to an ISO 5 workstation for dust-free tinkering, I don't think I want this camera taken apart yet another time. There are only so many times such a piece of equipment can go into deep surgery before something goes awry.

 

Lastly, @Andrea, in reference to your comment that stressed photo artifacts may not appear in ordinary photo conversion and editing... I agree - but, just to be clear, this problem does show up in my 'ordinary' UV images. It depends strongly on the signal ratio of the source (incoming 'light') vs. the internal contamination (whatever it may be). In other words, when I have a good strong light source/image then, for example, my exposure settings can be low (i.e., quick shutter, low ISO, etc) - which doesn't allow for much of the contamination 'signal' into the image (or, rather, it swamps the contamination signal). Conversely, a weak image source (requiring long exposure, high ISO) allows for build up of the contamination signal instead. In practice this is exactly what I've seen. And the typical conditions required for the UV images I'm taking land somewhere right on the edge of this signal ratio balance. The result is that I typically get some blended, hazy, discolored area which is the contamination flare. So it becomes a tug of war for me - decent exposure settings with contamination flare, or sub par exposure settings to reduce the flare. Either way the results slide up or down together, so there really isn't much of any real option anyway. At least not YET! (I haven't quite given up yet...).

[Andrea B.]

Mark, you rock! Thank you for these further tests -- while experiencing jet laggies.

 

It was important to know that the artifacts show up in your 'ordinary' UV images. (Interestingly, I had heard differently from another source. But we will discuss that later.)

 

Your test results are sooooooo interesting. This should give us all the "evidence" we need to diagnose the problem but for the fact that we do not really know how the D750 is operating internally. If this light contamination is from the internal shutter monitor, then it is the outlier case because other converted D750s do not seem to show it. So the question becomes what is different about your D750?? Does the new filter fit badly? Is there some manufacturing quirk so that some part somehow does not fit properly? I don't know if we can answer this without dismantling two D750s side-by-side -- yours and some other one.

 

But, let's consider this: I think you are legally within your rights to claim LEMON on this D750.

And then two things happen:

(1) LifePixel puts the camera back in working order. I think all you have to do is ask them nicely. It is obvious to everyone that this is a problem camera. Not your fault and not their fault that this problem exists. [However LifePixel perhaps should have caught this contamination problem from the get-go, in my opinion? All it takes is a simple 15-30 second darkfield test. LifePixel if you are reading this, please do consider this simple darkfield test. Thank you!!!!!]

and

(2) You return the camera to your vendor and ask for your money back because the camera does not work for your intended purpose and it cannot be fixed. Your vendor should give you a refund because yours was in the series of factory-defective D750s. You tell the vendor that Nikon repair simply could not fix it and show the vendor (a copy of) the repair statement.

 

So what do you think about trying to get your money back? Then you could buy a nice used D610 and have that converted and finally get on with things. Yes, it has been a looooong time and you missed a season. But there are so many beautiful seasons to come and you'll get to enjoy your UV photography.

 

We could always gather up some torches & pitchforks and go storm the gates of Nikon Melville to demand a leak free D750. They do exist -- somewhere.

(I don't know where I've put my pitchforks. They're probably in some dark corner of the basement. But give me the word and I'll go looking for them.)

 

---

 

On a side note.....I always use Mup when shooting UV for the usual reasons when on tripod -- wait for internal vibrations to die down.. So now I wonder if perhaps I have failed to also see such a light contamination in my current conversions? I don't think I've missed anything like what you are seeing, but I should double-check.

 

On a side side note...... :) ....My D600 conversion recently did develop a light leak around the upper LCD. In spite of our best efforts cameras do get little bumps and jostles over time. This was an easily curable light leak though. Taping around edges of the LCD did the trick.

[Guest]

I probably could have claimed LEMON on this camera, but in fact I did (knowingly) void my warranty by having the conversion service done. So I don't honestly have good ground to stand on. Also, just to be clear, my camera was not listed in the series of factory-defective D750s (according to the serial number), but they fixed it anyway; which was very nice of them (though I really have no idea what they did, if anything, as I got little to no information from them about the repair).

 

And, to be clear, LifePixel did offer to put the camera back to stock condition. I've just had it up to here [gestures apropos] with this camera. I'll shoot with it, and do my best to work with its 'quirk'. Maybe I can figure out some kind of auto-subtraction/cancellation of the flare in post, or something to that effect.

 

I agree that a quick test by LifePixel should have caught the problem, and I was told that they did test the camera after conversion and didn't see the problem. Of course, I wasn't there to see the test, and they don't have the images from the test, so I still have no idea what could have happened.

 

Some day a super-camera of some kind will hit the market - one that is sensitive from 200-2000nm! - and I will have been saving my pennies all the while. Or maybe I'll take up making pitchforks as a hobby instead.

[Andrea B.]

I hear you! I'm just bummed for you that this happened. :( :( :( I'm going to continue to ask a couple of questions elsewhere about this because in a way it affects us all. Any one of us could have had this happen. (Recall it did happen to me in '10.)

 

I still think there might be a solution in some kind of foam baffling or taping effort. After you have been shooting awhile, maybe you will want to pursue that.

Meanwhile I hope you get sufficient good out of the camera for learning the shooting and editing of UV photographs that you will not feel too cheated by the way it turned out.

 

Finally I thank you for sharing the problem with us and performing all those tests. It is important for any UV/IR shooter to learn that things can and do go wrong with conversion. I will let you know if I find out anything else which might be relevant.

[JCDowdy]

This topic on NikonGear.net might be somehow relevant, D750 with MuP-sickness?

[mathieucarbou]

Hi,

 

I got a Full Spectrum Nikon D750 (converted by Life Pixel) since December 2019 (1 year ago). At that time, this camera was known to be a good candidate for conversion (based on the information from Life Pixel and Kolari Vision websites). This week, going back to their website, I saw some big disclaimers in RED saying that the D750 is not suitable for long exposures and astrophotography (???). What a surprise!!!

 

So after some email chats with the guys at Life Pixel, Kolari Vision and MaxMax (which all have the same warning), plus some Internet Search, I've found that the images from the D750 are somehow impacted by the shutter monitoring IR LED. Since nobody was able to give me a workaround, I did a lot of testing and came up with some workarounds to either reduce or cancel this effect.

 

Here is the article: https://www.mathieu....itoring-IR-LED.

 

Now, we are in 2020... And what a surprise to see this discussion thread, dated from 2016, which describes exactly the same symptoms and pretty much the same workarounds I've found!

 

Life Pixel remembered Mark's problem: they told me it was related to a Nikon recall, but after reading the discussion here, it really seems to me that the problem Mark had was due to the monitoring LED right ?

 

I concur with what I've see in this forum: I can only speak about Like Pixel, but I feel that hey just convert the cameras. They do not bother to test before proposing the conversion if the models will really correctly work, otherwise the D750, Z6, Z7 would have never been propose, or, at least, with a disclaimer.

 

FYI, here is an extract from my article (what I have concluded after some testing):

 

To reduce (or even cancel) the impact of the internal shutter monitoring LED light of the D750, here are my suggestions in order of priority that you can combine, whatever the mode you use:

1. Use Mirror Up

2. Use Exposure Delay Mode 3 seconds. Switching from 1 second delay to a 3 seconds delay has definitively reduced the purple glow. So I would suggest you test with different combinations to see. It seems that the IR LED light has lower impact when the delay increases.

3. Use Live View(in last resort if you cannot use the other ones)

 

For Bulb mode (2, 3, 4 mins)

1. Stay at ISO between 100 and 1600

2. For longer than 3-4 min exposures, stay at ISO between 100 and 800

3. Use with Mirror Up with Exposure Delay Mode 3 sec

 

For High ISO shots (<= 30 sec)

1. Use Mirror Up

2. Use Exposure Delay Mode 3 sec

3. If not possible, use Live View

(from: https://www.mathieu.photography/Articles/Nikon-D750-and-Internal-Shutter-Monitoring-IR-LED)

[Andy Perrin]

In UV the problem is bad enough that they should just be avoided.

[mathieucarbou]

Hi Andy!

Just for my knowledge, what would make the problem worse in UV from what I was able to see the caps on ?

I have tested with a *full spectrum* D750. UV is just a part of the larger band the camera sees.

If you can provided me with some insights so that I can update this page.

Thanks!

[Andy Perrin]

Just exposure time, usually -- many of us expose for ridiculously long times in certain scenarios. Pinholes come to mind and also singlet quartz lenses which need to be used at high F-num to get a decent picture. Also to reach the shorter end of UVA or UVB, long exposures are needed, sometimes on the order of minutes. We push our cameras to the max in this forum!

[mathieucarbou]

But long exposures are fine with the D750 and IR Led at low ISO.

This is only at high ISO that the issue occurs. And what use case could prevent one to instead of doing a long exposure at high ISO, use exposure stacking ? This would also overcome most of the sensor read noise.

This seems to me the use case is similar to astrophotography: where you would ideally shot on a tracked mount several frames 1-4min each at an ISO that is high enough (800-3200). But an alternative is to take more exposures at lower iso and / or shorter time to avoid the IR led impact and read noise.

[Andy Perrin]

You could do that...or you could just convert a camera that isn't messed up to begin with. Definitely going with the latter, personally. Also, I don't think exposure stacking would get rid of a leak. You would just end up with a less noisy leak which would reappear when you pushed the photo in the editor.

[mathieucarbou]

Also, I don't think exposure stacking would get rid of a leak.

 

You probably didn't read and/or understand my article and what I was saying ?

 

It won't if you stack exposures taken with bulb mode at lower shutter speed but high ISO, but it will be unaffected when stacking long exposures at lower ISO and using Mirror Up / Delay Mode with bulb mode, or by stacking exposures taken without bulb mode at shutter speed 30sec by using Mirror Up / Delay Mode. These combinations (as I pointed out) are not affected by the IR Led light.

 

As long the exposures all contains enough data (more signal than noise - histogram spikes not on the complete left side), stacking will improve the signal and reduce the noise and won't improve the inexistent or nearly inexistant effect from this internal light.

 

The D750 is a camera that is ISO invariant and has its top DR at ISO 100-200. So better use it at low ISO with long exposures or stacking, and eventually increase the EV after in post, instead of stacking or doing long exposures at high ISO which is just adding more and more noise.