Any D750 users seeing EOF errors in ACR/Lightroom?

Gerick14 wrote:

Hey, I have had a similar issue, it occurs randomly and I'm thinking it is caused by a corrupted sector on the SD card as I am unable to pinpoint it at fixed intervals. I have been running two SD cards on mine, a previously used 16GB Sandisk and a Lexar 32GB. The Sandisk gets me random errors while the Lexar doesn't have any errors. I suggest swapping out with other memory cards to see if the issue persists.

It's not an SD card issue for reasons I described in the OP. Also, the other user posting to the Adobe thread was able to reproduce it using both SD slots+cards in backup mode - the same NEFs on both cards are identical and produce the end-of-file error in ACR/LR but can be processed successfully in Capture NX-D.

Unlike the other poster on the Adobe ACR forum, I was able to isolate my EOF NEF issue to a specific card, whereas his EOF occurs on both cards (in backup mode). Here is an analysis of my EOF:

I was able to reproduce the EOF on my D750. This time I populated both SD slots - the first slot had my original Sony 32GB 94 MB/s card that produced the first EOF NEF from two weeks ago, the second SD slot had a Sandisk 32GB 45 MB/s. Unlike engerim's setup, my two NEFs didn't match when I reproduced the EOF - the Sony card produced another EOF NEF but the Sandisk's equivalent/"backup" NEF in the other slot could be read by ACR/LR without error. To isolate the problem to the card vs the specific D750 slot, I then swapped the cards between the two slots and reproduced the EOF NEF again - the EOF issue followed the Sony card to the other slot. To isolate the problem to the specific copy of this Sony card vs this model of card, I then replaced the Sony card with another of the same make and the new card did not produce an EOF NEF. So it looks like my original Sony SD card is bad.

I then analyzed the EOF by comparing the EOF NEF from the Sony card to to the equivalent good NEF from the Sandisk and found that the Sony card has data corruption near the end of the file. Because the corruption is at the end of the file it appears Capture NX-D (and the D750's in-camera NEF->JPEG conversion) is not encountering during its decoding, perhaps because it's converting a smaller image dimension from the NEF vs ACR/LR.

Here is an analysis of EOF/corruption. Based on the data-replication I see it appears one of the parallel NAND channels within the card is incorrectly sourcing data from another NAND channel.

I have analyzed a corrupt D750 NEF from another user on the Adobe forum that produces the same EOF error in ACR/LR and found that it has almost an identical corruption profile as the one I had, even though he's using a totally different make of SD card and his happened on both SD cards in the camera at the same time (backup mode). This implies a common bug and likely not a problem exclusively with the cards themselves. You can read more about my findings here.

If anyone else experiences an "end of file error" loading their D750 NEF into ACR/LR please post to this thread. The problem so far appears unique to 14-bit lossy compressed NEFs.

I just shot a wedding with my new D750 along side my old D600... Of the 1120 images taken in 14-bit RAW on the D750, I am unable to process 10 of them in Lightroom/Photoshop because of the 'end of file' error...

On reading another thread online I tried opening them in Capture NX-D (which is a pain to use) and they did work... I was able to recover the images, but it is far from ideal... I am not sure if the error is because of the way Adobe is reading the new NEFs or whether it's an error because of the camera...

I shot the entire wedding on Backup mode... The same NEF files on both camera slots show the same error... hence, it is unlikely this is due to the SD cards...

Hopefully a fix comes in soon...

Sunil Thakkar wrote:

I just shot a wedding with my new D750 along side my old D600... Of the 1120 images taken in 14-bit RAW on the D750, I am unable to process 10 of them in Lightroom/Photoshop because of the 'end of file' error...

On reading another thread online I tried opening them in Capture NX-D (which is a pain to use) and they did work... I was able to recover the images, but it is far from ideal... I am not sure if the error is because of the way Adobe is reading the new NEFs or whether it's an error because of the camera...

I shot the entire wedding on Backup mode... The same NEF files on both camera slots show the same error... hence, it is unlikely this is due to the SD cards...

Hopefully a fix comes in soon...

Thanks Sunil. I posted a response to the Adobe ACR thread. If you could provide one of the NEFs that would be helpful to confirm it's the same issue, which based on your description sounds like it is.

Again, for anyone else who is experiencing this end-of-file error in ACR/LR, please post here or to the Adobe ACR thread.

I had the same issue on a brand new D750 I just returned. All four of my SD cards showed the issue. They were completely fine on my D800 and Fuji X T1. We'll see how the new one performs, it arrived on Monday.

My problem also showed up in ViewNX so it wasn't just limited to Lightroom.

This is a serious issue. I'm wondering if this could be the real reason for the silent recall and not the flare issue. It seems more prevalent than it should be.

By the way my problem is exactly the same as described in this YouTube video.

http://youtu.be/6vMegk_IXCI

jrubio2 wrote:

I had the same issue on a brand new D750 I just returned. All four of my SD cards showed the issue. They were completely fine on my D800 and Fuji X T1. We'll see how the new one performs, it arrived on Monday.

My problem also showed up in ViewNX so it wasn't just limited to Lightroom.

This is a serious issue. I'm wondering if this could be the real reason for the silent recall and not the flare issue. It seems more prevalent than it should be.

Thanks. Do you have a NEF you can share with me? I'd like to make sure it has the same problem signature as the others. You can use https://www.transferbigfiles.com/ and post the resulting link here (or private-message me the link on dpreview). Thanks!

OK, done.

I added two files.

https://www.transferbigfiles.com/ccda8dbe-0a62-4738-9462-2f5ea64292b0/oHNcMuJX8l5aZGLnIkWgfQ2

The first one doesn't give me the EOF error, but is heavily corrupted.  The second give EOF error.

Very curious to find-out if we have the same issue.

Here is what the first one looks like when exported at JPG from ACR.  The second one won't even load into ACR.

Thanks!

Horshack wrote:

I created a thread two weeks ago on Adobe's ACR forum (here) about an "End of File" error I get on a particular compressed NEF I took with my D750. Normally an EOF occurs as a result of SD corruption, where the file size of the NEF is smaller than the size of the image data indicated in the EXIF. However in my case the file size matched the EXIF-indicated compressed data size thus all of the image data was there. Moreover, this same NEF can be read and processed fine by Nikon's Capture NX-D and also by the D750's in-camera NEF->JPEG conversion feature. This implies that Nikon maybe changed some aspect of the encoding that third-party processors are not accounting for or that those processors are attempting to decode too-large an image area (slight dimension mismatch) or that the D750 has a bug that causes it to corrupt the final columns of the final row(s) of the NEF data for certain scenes and it just happens that Nikon's decoders ignore that final row.

So far one other user on the Adobe thread has produced a D750 NEF with the same symptom as mine. If anyone ever sees an EOF error on a compressed D750 NEF can you please post to this thread, hopefully with a link to your NEF so that I can analyze it and update the running Adobe thread? You can use https://www.transferbigfiles.com/ to upload the file and post the resulting link here (or PM it to me).

Thanks!

That's a very interesting thread. Judging from the samples I would say this is a firmware bug of some kind, and Nikon will certainly need to address it. I suspect that the reason Nikon's software doesn't complain is just because it fails "silently". I don't think it is due to a change in the compression scheme.

I'm getting close to releasing the next version of Pixel Fixer which will support the D750, and it will be possible to use it to "repair" files where the corruption is very close to the end of the file.  There will still be missing image data of course, but you won't see it in most raw converters since they don't use all the image data.

In the meantime, you'll probably find it useful to experiment with settings like bit-depth, compression type, Active D-Lighting, Long Exposure NR, Lens Correction and (I think) High ISO NR. They all affect how the buffer is used internally (you'll notice the "rXX" number in the viewfinder will change) and I suspect that some combinations will either make the problem worse (because the corruption will appear earlier in the raw data) or prevent the problem from happening at all.

jrubio2 wrote:

OK, done.

I added two files.

https://www.transferbigfiles.com/ccda8dbe-0a62-4738-9462-2f5ea64292b0/oHNcMuJX8l5aZGLnIkWgfQ2

The first one doesn't give me the EOF error, but is heavily corrupted. The second give EOF error.

Very curious to find-out if we have the same issue.

Here is what the first one looks like when exported at JPG from ACR. The second one won't even load into ACR.

Thanks!

Thanks for the NEFs! The corruption on your NEFs is different than the issue I'm chasing. On your DSC_0055.NEF, the corrupt data starts at row 18 and column 5018 of the image, near file offset 0x3fcc74. For DSC_0014.NEF, the corruption begins at row 133 and column 5946 of the image, near file offset 0x3b7a2c. This places the corruption near the beginning of the image, whereas the issue I'm following is corrupt data at the very end of the image, beyond the visible portion that is rendered by raw processors. Btw the embedded JPEGs in your raws are intact, so that means at least the frame buffer for the image on the body you returned was correct (ie, the corruption happened downstream in the I/O path and the card).

Thomas Comerford wrote:

Horshack wrote:

I created a thread two weeks ago on Adobe's ACR forum (here) about an "End of File" error I get on a particular compressed NEF I took with my D750. Normally an EOF occurs as a result of SD corruption, where the file size of the NEF is smaller than the size of the image data indicated in the EXIF. However in my case the file size matched the EXIF-indicated compressed data size thus all of the image data was there. Moreover, this same NEF can be read and processed fine by Nikon's Capture NX-D and also by the D750's in-camera NEF->JPEG conversion feature. This implies that Nikon maybe changed some aspect of the encoding that third-party processors are not accounting for or that those processors are attempting to decode too-large an image area (slight dimension mismatch) or that the D750 has a bug that causes it to corrupt the final columns of the final row(s) of the NEF data for certain scenes and it just happens that Nikon's decoders ignore that final row.

So far one other user on the Adobe thread has produced a D750 NEF with the same symptom as mine. If anyone ever sees an EOF error on a compressed D750 NEF can you please post to this thread, hopefully with a link to your NEF so that I can analyze it and update the running Adobe thread? You can use https://www.transferbigfiles.com/ to upload the file and post the resulting link here (or PM it to me).

Thanks!

That's a very interesting thread. Judging from the samples I would say this is a firmware bug of some kind, and Nikon will certainly need to address it. I suspect that the reason Nikon's software doesn't complain is just because it fails "silently". I don't think it is due to a change in the compression scheme.

I'm getting close to releasing the next version of Pixel Fixer which will support the D750, and it will be possible to use it to "repair" files where the corruption is very close to the end of the file. There will still be missing image data of course, but you won't see it in most raw converters since they don't use all the image data.

In the meantime, you'll probably find it useful to experiment with settings like bit-depth, compression type, Active D-Lighting, Long Exposure NR, Lens Correction and (I think) High ISO NR. They all affect how the buffer is used internally (you'll notice the "rXX" number in the viewfinder will change) and I suspect that some combinations will either make the problem worse (because the corruption will appear earlier in the raw data) or prevent the problem from happening at all.

Definitely appears to be a firmware bug. The reason Nikon's software doesn't complain (nor the D750's in-body NEF conversion) is because they don't appear to decode the huffman stream of the NEF beyond the rendered output dimensions of 6016x4016, whereas the replica corruption occurs on the last row of raw data encoded in the file (row 4031).

Ok. Glad it appears to be something different. I'll see how my new body does once it shows up--odds are overwhelmingly in the direction that it will be fine. For sure though I am not the only person with my issue. See YouTube link above and:

http://www.dpreview.com/forums/thread/3731881

Many thanks btw. Excellent.

Hi all,

Would like to report that the replacement camera doesn't exhibit any file corruption and send 100% fine.

Thanks.

danmitch wrote:

Hi - Just got a new D750 and experiencing the same problem as described in this post, were any conclusions reached on this?

Many thanks - Dan

Hi Dan,

If you can send me one of the affected raws I can report back about whether your issue is the same as the one I described in the OP. You can use https://www.transferbigfiles.com/ to generate a link you can post here or via PM.