Nikon D800 astro hack

Wouldn't it be cool if we could make our camera do ________ ?
Patching issues and other general help requests here please.

Nikon D800 astro hack

Postby whwang » Fri Jan 24, 2014 5:06 am

Hi,

I am an astrophotographer using Nikon D800. The nonlinearity in the NEF file has been
bothering me for a long while and I cannot achieve the ideal result that the great
sensor is capable of.

I think it will be nice if we can disable the black level adjustment in long exposures,
and well as disable the HPS filter.

Cheers,
Wei-Hao
whwang
 
Posts: 13
Joined: Tue Jan 21, 2014 7:17 am
Been thanked: 2 times

Re: Nikon D800 astro hack

Postby Simeon » Fri Jan 24, 2014 8:40 pm

Hi, the HPS should be locatatable in the D800 firmware, but we have yet to understand how the black level is configured. So that is on going investigation.
Simeon
Core Developer
 
Posts: 2613
Joined: Wed Nov 30, 2011 6:12 am
Location: Christchurch, New Zealand
Been thanked: 614 times

Re: Nikon D800 astro hack

Postby astronomer » Fri Jan 24, 2014 10:35 pm

Simeon wrote:Hi, the HPS should be locatatable in the D800 firmware, but we have yet to understand how the black level is configured. So that is on going investigation.


Simeon,

I've talked to whwang regarding the issue with HPS. For the time being outputting overscan region is well acceptable and suitable for temperature and linearity estimation. If I remember correctly, it took you sometimes to figure out turning off the HPS without saving the dark pixels. So if you're going to make a patch for D800, no need to hassle on the overscan issue.

Best,
astronomer
User avatar
astronomer
Developer
 
Posts: 722
Joined: Fri Apr 06, 2012 10:06 am
Location: US
Been thanked: 292 times

Re: Nikon D800 astro hack

Postby coderat » Wed Feb 26, 2014 3:49 am

I am wodered if someone tried D800E for astro ? Any results ?

Regards,
coderat
coderat
Core Developer
 
Posts: 2283
Joined: Fri Apr 26, 2013 10:21 am
Been thanked: 450 times

Re: Nikon D800 astro hack

Postby D80 » Wed Feb 26, 2014 12:16 pm

Yes a friend of mine are living on his astrophotagraphy and I am pretty sure that he have a D800E.

http://www.astrofotografen.se/

He have some really great picture here to.

http://astronet.se/showimages.php?id=37
D80
 
Posts: 52
Joined: Mon Apr 09, 2012 4:30 am
Been thanked: 9 times

Re: Nikon D800 astro hack

Postby whwang » Tue Mar 04, 2014 12:49 am

coderat wrote:I am wodered if someone tried D800E for astro ? Any results ?

Regards,
coderat


Hi,

These are my D800 astrophotos:
http://www.astrobin.com/79059/
http://www.astrobin.com/79256/
http://www.astrobin.com/77363/
http://www.astrobin.com/78770/
http://www.astrobin.com/79055/
http://www.astrobin.com/77963/

Because of the nonlinearity problem I mentioned above, flat-field correction for D800 is really a headache.
Because of this, for all the pictures above, especially the first two, I can't confidently say everything in the
pictures is something real from the sky. Some of the faint "clouds" you see in the image may be simply
flat-field artifacts. This prevents me from getting even better pictures. So I really hope we can solve this
via some firmware hack.

EDIT: HPS filter is already disabled for D7000.

Cheers,
Wei-Hao
whwang
 
Posts: 13
Joined: Tue Jan 21, 2014 7:17 am
Been thanked: 2 times

Re: Nikon D800 astro hack

Postby coderat » Tue Mar 04, 2014 6:24 am

whwang wrote:These are my D800 astrophotos:

Great ! I was interested in D800E, because it has no AA-filter , so may be higher resolution.
Because of the nonlinearity problem I mentioned above, flat-field correction for D800 is really a headache.
Because of this, for all the pictures above, especially the first two, I can't confidently say everything in the
pictures is something real from the sky. Some of the faint "clouds" you see in the image may be simply
flat-field artifacts. This prevents me from getting even better pictures.

May I ask you for newbie (as I am) to post short explanation of problem, may be with some pictures as astronomer did for HPS (HPS we know already). So that I can keep my eyes open. Please add sample test procedures description, so even not-astronomer can test it.

@Simeon
Can you may be get HPS off for D800, similar to D7000/D300, please ?

@whwang
Would you provide us feedback on this patch ?

Thanks,
coderat
coderat
Core Developer
 
Posts: 2283
Joined: Fri Apr 26, 2013 10:21 am
Been thanked: 450 times

Re: Nikon D800 astro hack

Postby whwang » Tue Mar 04, 2014 8:15 am

Hi,

Are you considering using D800E for astrophoto only? If a DSLR's primary (>90%) use is astrophoto,
we usually recommend to remove its IR-blocking filter to (greatly) enhance its sensitivity on red
nebulas. My D800 had undergone such a surgery. And in the same modification, I also had its
AA filter removed. So its resolution is essentially the same as D800E. So, if you don't mind having
your camera go under the knife, you may just buy a D800 and enjoy D800E's resolution. On the
other hand, if you also want to use the camera for daylight photo, such modification is usually
not a good idea. In such a case, D800E is the way to go if you want the resolution.

coderat wrote:@whwang
Would you provide us feedback on this patch ?
coderat


Do you mean the patch about the HPS?
A while ago, I made some tests on my D800:
http://www3.asiaa.sinica.edu.tw/~whwang ... 800_mode3/
It looks like the HPS filter on D800 no longer removes anything like real stars. I even could not
confirm whether there is such a filter at all.

The linearity issue is that, imagine we shoot a perfectly uniform object. Although the target is uniform,
the image we get will not have uniform brightness. Because of the vignetting caused by the optics and
by the mirror box of the camera, the image will be brightest at its center, and darker at the four corners
and along the the four boundaries. Now, we shoot the uniform object twice, with different exposure
times and perhaps even different ISOs, but using identical optics. If we decode the raw files linearly,
and then divide the two images with each other, the result should be an image with perfectly uniform brightness.
This is because the two images suffer from exactly the same vignetting.

Such a procedure is called flat fielding. Astrophotographers first shoot their target celestial objects,
then shoot an object with uniform brightness, and divide the target image with the image of the uniform
object (the latter is call "flat"). The result should be an image with no vignetting at all.

On Canon's EOS DSLRs and cooled CCD cameras dedicated for astronomy, the above procedure works
well. We can easily get perfectly flat (no vignetting) images. Unfortunately, this does not work on
D800. Here are three examples:
http://www3.asiaa.sinica.edu.tw/~whwang ... arison.jpg
The three images show results after the flat-field procedure. They all should have image backgrounds
of perfectly uniform brightness. However, depending on how the flats are taken, the background show
different degrees of residual vignetting at the corners. The top one of the three even have bright corners
instead of dark corners (over-correction of the vignetting).

This result implies that the "response" of the camera in the raw file is not linear. Increase the incoming
light by a certain factor does not increase the pixel value by exactly the same factor. This makes it
extremely difficult to remove image artifacts including dark current, and vignetting effect. The above
is a real on-sky example. I also made lab exposures using D800, and confirmed that the response in
D800's NEF files does not go linearly with the strength of incoming light.

I am not 100% sure what causes the nonlinear behavior. However, after some discussion with astronomer,
we both agree that the black-level clipping may have something to do with this. This is why we would
like to have a patch that disables the black clipping. In addition, even if the black clipping has nothing
to do with the non-linearity, it definitely affects the accuracy of the subtraction of dark current and hot
pixels for long exposures. So disabling it is a good thing to do anyway.

Cheers,
Wei-Hao
whwang
 
Posts: 13
Joined: Tue Jan 21, 2014 7:17 am
Been thanked: 2 times

Re: Nikon D800 astro hack

Postby Simeon » Tue Mar 04, 2014 12:35 pm

Understood, it's on the things to keep an out for, the hard part is, with no A/B testing where it's on/off then it's hard trace where the "magic happens"
Simeon
Core Developer
 
Posts: 2613
Joined: Wed Nov 30, 2011 6:12 am
Location: Christchurch, New Zealand
Been thanked: 614 times

Re: Nikon D800 astro hack

Postby whwang » Tue Mar 04, 2014 11:57 pm

Do you think there is anything I can help?
I am not a programmer, so there is not much I can do on the programming part.
I can offer to test something on my D800 if the risk is low.
(It's not my own D800. It belongs to our lab. So I need to be careful if I am going to do something on it.)
whwang
 
Posts: 13
Joined: Tue Jan 21, 2014 7:17 am
Been thanked: 2 times

Next

Return to Feature Requests and Help

Who is online

Users browsing this forum: Google [Bot] and 1 guest

cron