HELP!!! WEIRD LINES, SPOTS AND ARTIFACTS ON PHOTOS!!!!!

[Judobreaker]

What flash are you using with the camera? It looks kind of like a Vivitar 283, but I cannot be sure from the image.

Some older flashes dump the entire charge through the camera. This can damage some cameras, others protect against the full-current discharge.

Shoot some pictures without the flash. Sometimes- induced current finds it's way into digital electronics and has an affect on processing. Looking at the JPEG images, it looks like the image is corrupt- as in bits in the pixels either being stuck, or being dropped.

Seeing as the flash is visible in the image I'm going make the wild assumption that the flash was fired remotely.
I would like to see a remote flash frying up a camera because of full-current discharge.

This looks like a mighty fine in-camera processing problem to me.
Seeing as this doesn't happen with RAW I'd suspect the in-camera JPG processing.
This pattern repeats itself every 8 pixels, that's one heck of a nice computer number for ya (2 to the power of 3, also happens to be the length of one bit).
Have you tried adjusting the JPG quality to see what happens? It might be that adjusting the quality causes the processing to skip a process or do it differently which could make the problem disappear.
If that doesn't help you could also try to upgrade the firmware... If there's a software problem somewhere it might be fixed that way.

I'll ask my dad if he has an idea tomorrow, his job is hard-core image processing (like automatically analyzing security-camera footage kind of processing). He might just have some more ideas.
However, if a firmware update won't work I'm thinking this might be a problem you can't fix which would mean sending it back to Nikon for repairs or replacement.

Remember that the Vivitar 283 has an off-flash sensor that connects to the camera's Hot Shoe and can be used remotely, AND dump the whole capacitor through the camera. The sensor unplugs from the flash, plugs into a special shoe that goes on the camera, and the cord plugs into the socket of the flash. The capacitor dumps through the cord, into the camera. Never put your finger over the PC socket of the camera if it has both a hot shoe and PC socket.

Send me your camera, and I will arrange for a Demo.

Ok, I'm not familiar with that kind of flash so I guess it is possible to get that current into the camera. However, if that would be the case we're talking about a hardware issue and then you'd expect RAW files to be affected just as much as JPG files.

 

[BrianV]

You never know what is going to get affected, the data/processing path for the JPEG is different from the RAW. The problem is in the JPEG path, which includes demosaicing and interpolation of the color planes. After that, the image is processed as 8x8 tiles. SO- back to the OP to do some more testing. Try with/without flash, and at different JPEG image sizes. If it is a permanent problem, it will occur all the time. If a problem from inducted current, will go away when used without flash. Also- try a different memory card, some are known to induce noise and cause banding in some cameras. If it is a data path problem, changing the size of the stored JPEGS should affect the placement of the lines in the image. I spent years working with data acquisition systems, was easier to diagnose when this stuff took a rack of equipment to implement. Put the Oscope/ logic analyzer/ and PRBS generator, find the bad card. Eliminate noise sources, you could track the problem.

I found the remote cord for the Vivitar. I used the Vivitar in the Lab to induce current on digital circuits to test isolation. And used it for the Elvis concert in '75.

 

[Judobreaker]

Ok, I had a talk with my dad about it and he suspects this would be a hardware problem, probably not software.

 

[480sparky41 40]

Another possibility: Corrupt SD card.

 

[BrianV]

Ok, I had a talk with my dad about it and he suspects this would be a hardware problem, probably not software.

Your Dad and I would have a hex geek out over this one.

Still waiting to hear back from the OP. At worst, just shoot Raw if the camera is out of warranty. If under Warranty, send it in.

 

[ericdao]

Thanks for all the suggestions and advice. The flash doesnt seem to be the problem, this artifacting happens all the time for jpegs. I tried all the different jpeg formats and all have the same results. I actually started testing every single feature of my camera and i noticed i cant record video. When i press the button the green light lights up and nothing happens and sometimes when it does record it records in this strange unviewable format. I recall my brother dropping my camera a while ago, at the time everything seemed alright. How much do you guys thing it will cost to repair and what seems to be the problem?

 

[ericdao]

Also should i even send it to get repaired?? ive been shooting a variety of things for months and never even suspected an issue. I also just bought a sigma 70-200 2.8 os so im also a little broke haha. I do think i should be safe and get it serviced since im going to college for photography and also shadowing a wedding photographer... Any thoughts??

 

[BrianV]

There is always the chance that dropping the camera knocked a connector out of place, and the electronic boards need to be reseated. Chances are something got damaged on a main circuit board, and it needs to be replaced.

Many manufacturers have a flat-price repair, usually about 1/2 the price of a refurbed camera. Taking a digital camera apart and getting it back together- labor intensive, especially getting everyting realigned. "Probably a few Hundred bucks".

The camera works in raw mode, for still photography. So it still provides the main function.

I would not use this camera for paying jobs, even as a backup for a wedding photographer until after it is checked by a professional repairshop and having it repaired to full functionality. depending on the repair cost, you might end up buying a new camera and using this one in a pinch. A professional needs to trust their gear.

 

[BrianV]

I went as far as outputting the sample image provided as a demosaiced raw image and ran a hex dump of it. That way you get a decent represenation of the red-green-blue image planes. Specifically, it looked like bit 6 of the first pixel of every 8-pixel group was corrupt. The values were "dragged low" when the intensity got above 63 on a scale of 255. That is when "Bit 6" would need to go high. You have to deal with interpolation of pixels, compression, etc- but that is what it looked like on a quick look and WAG. Too funny: using software written in FORTRAN running on DOS to look at an image of this size.

So- it looks like a hardware problem with the jpeg processing path, which would also explain the corrupt video.