D750 "unexpected end of file" for NEF that loads fine into Capture NX-D

[Horshack]

I have a NEF that produces an "unexpected end of file" error in both Lightroom 5.7 and PS CC (both with Camera Raw 8.7). The same NEF loads and converts fine in Nikon's Capture NX-D V1.0.2. The NEF also converts fine into a JPEG using the D750's in-camera NEF processing function. I can provide the NEF to Adobe for analysis.

[engerim]

so it seems one of your cards is broken (I get the error in both slots). Can you post the raw files somewhere?

[Lattefarsan]

Here are the two NEF. Hope you are able to fetch them http://tbf.me/a/BWVKPg

[engerim]

Thanks, some analysis:

DSC1855 in your SD Card Slot 2 is OK

DSC1855 in your SD Card Slot 1 has errors starting row 3962 (according to dcraw) (not at row 4031 as my images).

(Again talking 0 based for rows). So this is a different error and also fails in Nikon tools.

I checked and View NX can open the files but I assume View NX is only using the preview JPG of the NEF file.

In Capture NX-D you can see a clear "green line" at the bottom of the image.

So in your case it should be an SD Card writing issue. You could swap the cards and see if you get errors in the same slot to be certain.

(Note that in my case the error is in both slots).

Your images in NX-D: http://stuff.inxsoft.net/tmp/201501-test.png

[Lattefarsan]

Thanks. I thought my error would be another when the error occurred in only one slot. I've swopped the cards and get back if a new eof occur

[Horshack]

Lattefarsan, thanks for the NEFs. I analyzed the files and see 57 instances of mostly single-bit corruption (ie, a single bit is either stuck on or off). This implies a likely problem with that SD card. I don't see a pattern to the address and/or bit that is wrong. Here are offsets and byte values - correct NEF on the left, corrupt NEF on the right:

01B66400: B8 98

01B66419: 61 21

01B66439: 4D 4F

01B66458: 48 58

01B6647D: 75 74

01B66486: 7C 3C

01B6648A: 3E 3F

01B6648B: 4F 4E

01B664A2: B6 B4

01B664AF: DC D4

01B664D3: F3 F7

01B664E9: 23 33

01B664F1: B8 B9

01B66507: BC B8

01B66532: 39 3D

01B66538: 53 51

01B66541: 5A 4A

01B66544: 4F 4B

01B6655A: 0D 2D

01B6655D: 2A AA

01B6655F: 23 20

01B66580: 04 14

01B66590: B3 A3

01B66591: 8B 9B

01B66598: 8F 8B

01B665A3: 2D 2C

01B665B3: AB AA

01B665DB: 9B BB

01B665EA: DE FE

01B66631: 36 32

01B66634: 39 29

01B6664D: 45 05

01B66652: B0 B8

01B66699: 25 27

01B6669B: 5C 7C

01B666A5: 1C 1E

01B666A9: 38 B8

01B666D2: 4A 4B

01B666D6: 10 12

01B666D8: 22 A2

01B666DE: 12 13

01B666E8: 6E 6F

01B66702: F6 FE

01B6671C: 21 01

01B66725: E6 C6

01B66731: F1 D1

01B66737: F3 F2

01B6673F: 7D 7C

01B66740: A5 84

01B66773: 73 53

01B6678B: FE BE

01B66791: 37 B7

01B667A6: DA 5A

01B667AE: 1A 3A

01B667B2: 19 99

01B667C9: 59 49

01B667F8: B2 92

[climbbike@gmail.com]

I'm having the same problem. Hopefully Nikon or Adobe fix this quickly. I'm pretty annoyed with Nikon. First the d600 fiasco (which I had) and now this and the flaring issue .  .

[climbbike@gmail.com]

I'm getting the problem with 12 bit RAW. I'd say approx 1 out of 100 files have the problem

[Horshack]

climbbike1@comcast.net, can you make one of your affected NEFs available to download so that we can see if it's the issue discussed in this thread vs more typical SD corruption? You can use https://www.transferbigfiles.com/ to send the file to yourself and then post the resulting link here. Thanks!

[climbbike@gmail.com]

I'll upload my file this evening.

[Horshack]

engerim, I analyzed DSC_4244. It's the same corruption - 32 bytes of replicated data near the end of the file. For all the NEFs I analyzed (yours and mine), the replicated data is on the last row in the image (4031), which again is beyond the output image dimensions of ACR/LR/Capture-NX of 6016x4016. I don't see a correlation yet between the pixel column of the original data vs the pixel column of the replicated data. For example, two of your EOF NEFs compared:

DSC_3992.NEF

Original data block @ filepos 0x1c200070, corresponds to pixel 5270x4031

Replica data block @ filepos 0x1c2000b0, corresponds to pixel 5326x4031

Delta x-position between original (correct) block and corrupt (replica) is 56 pixels

DSC_4244.NEF:

Original data block @ filepos 0x1fa8070, corresponds to pixel 5492x4031

Replica data block @ filepos 0x1fa80b0, corresponds to pixel 5538x4031

Delta x-position between original (correct) block and corrupt (replica) is 46 pixels

[engerim]

ok, you need to be certain to rule out an SD Card issue. I'm using the same SD Cards in both slots (brandnew Transcend 32 GB SDHX cards) and I get the same problematic files in both slots at the same time (so I rule out an SD card issue in my case).

[Horshack]

engerim, do you have another model of SD card you can try to reproduce the EOF with? I'm wondering if certain models are exhibiting unique sensitivity to the D750's media write timing. I plan to analyze the NEFs in your zip to see if I can suss out the corruption, although it's harder to find with the images from both cards being identical.

[engerim]

Well the images are identical except one bit (the slot indicator in the meta data). I might test it with a mixed combination of cards during next week.

Generally its easy to reproduce. around 2 of 100 images. I just use interval shooting to see it sooner.

[Horshack]

Thanks. I had the exact same repro rate with my Sony SD - 2 out of 100 images, on three separate 100-shot sequences. Although I'm stating my Sony card is "bad", that is not conclusive at this point. It might a timing margin issue inherent to the D750's media interface and the intersection of that with particular cards causing the issue. Time will tell as more test theirs. I have 5 other copies of the Sony card I can try later today - just need to find them all

[adriana ohlmeyer]

I am curious to know if you format the card in camera before shooting? If not, could you try and see if you can reproduce this issue, please?

[Lattefarsan]

I've formatted both cards in the camera

[adriana ohlmeyer]

Thank you!

[climbbike@gmail.com]

I uploaded one of my problem files: Transfer Big Files Free - Email or Send Large Files up to 20 GB

Transfer Big Files Free - Email or Send Large Files up to 20 GB

[engerim]

Your sample file has the same issue (in row 4032) as my files.

Row: 4032 Col: 5921

dcraw has it as "Corrupt data near 0x1a881fd"

So the error also appears in 12 bit raw, lossless.

I updated the ticket at Nikon that you also have this error.

You may want to open a ticket there yourself and refer them to my case ID "141229-001985":

Title: "D750 bad huff data in last row of RAW file (14 bit) (row 4032)"

I doubt Adobe will do anything to circumvent the error as its clearly a camera bug.

I may open a ticket with Adobe Support...

[ssprengel]

As far as opening a ticket with Adobe, you may not need to.  Adriana Ohlmeyer, from a few replies above, is an Adobe staff member who usually jumps on the forums after a new Camera Raw release to look for issues being reported.  I assume she's either a developer or a QA person.  There must have been enough replies to have caught her attention after the multiple-holiday week just ended.

[Horshack]

engerim, I analyzed climbbike1@comcast.net's NEF and didn't find any replicated data corruption that exists in our NEFs - I wrote a quick app to do the search programmatically vs trying to eyeball. The fact that NEF decoder craps out on the last row of his file might be the result of regular SD corruption instead of the D750 issue we're chasing - corruption anywhere within the huffman stream can lead to a bit underrun that doesn't get detected until near the end of the NEF. That said, there's always the possibility that his file does exhibit the D750 bug we're chasing and that the corruption signature may not always be the replicated data that we see in ours...but for now I would say his issue is unproven to be the one we're chasing until we get more evidence.

climbbike1@comcast.net, any chance you were using two SD cards in backup mode and have the NEF from the other slot? I noticed you posted two unique file links in your post but they appear to be for the same file since the EXIF is identical (slot indicator same for both). If  not, can you populate the second SD slot with another card and run them both in backup mode and then post both NEFs the next time you get an NEF with the end-of-file error?

[climbbike@gmail.com]

I was running with just a single card. I'll try to run with 2, 1 in backup mode. I suppose it could be a coincidence that this is happening now that i just switched to the D750 but I've never had a problem before with this card. It is getting to be a year or two old and I use it a decent amount . ..

[engerim]

try "Copy" mode with two cards please so you have two RAW files (one on each card)

[Lattefarsan]

Hi! Any response from Nikon?

[engerim]

just that they check and I should wait for their feedback