p.1 #1 · Nikon D800/e users - a little help please ?
I've been a Sony NEX7 user for the last 22 months. Because of my old eyes I pretty much require an EVF solution like the NEX7 or now A7r.
I went to buy/look an 800/e several times but couldn't really get past the usability of the live view situation. I shoot manually almost 100% of the time so critical focus is important to my process.
Like many, I was excited when the A7r was introduced and it delivers on resolution and MTFs as long as it's in B&W.
However, when I went out to shoot my first shots in the field and in color, some issues immediately came to light.
More than anything (even shutter blur) the Sony 'lossless compression' is causing artifacts and bad tonal transitions. It's not too visible in the 1st iteration but when you go to post (or zero post) and then JPEG it's glaring. I haven't been able to create a color JPEG without the issue.
It's all in LR 5.3 but I've tried most variances of the options without success (other than export as 200mb TIFF) Import sRGB RAW from camera into copy, copy as DNG, no process, no output sharpening, export into sRGB, AdobeRGB, ProPhotoRGB.
If you have a good viewing setup (large screen, high resolution) have you seen such problems with tonal gradients in the 800(e) ?
p.1 #2 · Nikon D800/e users - a little help please ?
MaxBerlin wrote:
If you have a good viewing setup (large screen, high resolution) have you seen such problems with tonal gradients in the 800(e) ?
No such issues with the D800/D800E, as long as you shoot in RAW with no compression applied.
While many fanboys may claim otherwise, I can see the subtle differences in tonality and color rendition between non-compressed and "lossless" compression (an oxymoron...) in NEF files; the difference is even more striking between 14-bit and 12-bit capture.
YMMV depending on your workflow and your viewing setup, but if you're spending all that money for the best quality camera and lenses, it seems rather foolish to throw away so much of the data you're capturing before you even process the files.
p.1 #3 · Nikon D800/e users - a little help please ?
I really don't look at things as close as you. but I do have a d800, and if you want a full raw / tiff of something I could take a test photo for you to look at
p.1 #4 · Nikon D800/e users - a little help please ?
thanks for that too, I think I might try shooting non compressed raw. I did a lot of searching and found mixed comments on the subject, but I guess its always best to just try things for yourself and see if you can tell a difference.
memory is cheap and so is processing power anyways right....
molson wrote:
No such issues with the D800/D800E, as long as you shoot in RAW with no compression applied.
While many fanboys may claim otherwise, I can see the subtle differences in tonality and color rendition between non-compressed and "lossless" compression (an oxymoron...) in NEF files; the difference is even more striking between 14-bit and 12-bit capture.
YMMV depending on your workflow and your viewing setup, but if you're spending all that money for the best quality camera and lenses, it seems rather foolish to throw away so much of the data you're capturing before you even process the files.
p.1 #5 · Nikon D800/e users - a little help please ?
This issue is showing up in the transition of greens and blues in a stream reflection. I am going out to retake the photos later today with the NEX7 and A7r.
I don't know if a sky gradient will show the issue.
Extremely complicated scenes don't seem to show the issue either.
I appreciate your offer to help but as a Sony user I have always marveled at some of the Nikon and Canon images. I thought the difference was the lenses but now with the Zeiss 135mm APO and the 55mm Otus I realize that the Sony compression is at issue.
p.1 #6 · Nikon D800/e users - a little help please ?
molson wrote:
I can see the subtle differences in tonality and color rendition between non-compressed and "lossless" compression (an oxymoron...) in NEF files
p.1 #7 · Nikon D800/e users - a little help please ?
Will compressing a file to a .zip make the files contain less information when you unzip it?
I don't see the slightest difference between non-compressed and lossless compressed when viewing in a editing program (essential, because it has to be decompressed). I viewed them on a 10-bit screen with a 10-bit videocard to make sure there would be no difference and I also took some prints (full 14-bit). Don't see any difference.
And if you would see a difference, using a other program would solve it (I do see a difference between Lightroom, NX2 and Gimp (used just to check)).
PS: What monitor are you using? A 8bit monitor WILL have problems with banding, it's not because of the image.
When using a 10bit (+) monitor, make sure you have a videocard that supports 10bit (+). If your videocard is not workstation-grade and under 1000$ it's probably not outputting 10bit.
If you want to be sure, print the full 14-bit file that has banding on your screen. 99% sure that you won't see any banding.
p.1 #8 · Nikon D800/e users - a little help please ?
playerofwar wrote:
Will compressing a file to a .zip make the files contain less information when you unzip it?
I don't see the slightest difference between non-compressed and lossless compressed when viewing in a editing program (essential, because it has to be decompressed). I viewed them on a 10-bit screen with a 10-bit videocard to make sure there would be no difference and I also took some prints (full 14-bit). Don't see any difference.
And if you would see a difference, using a other program would solve it (I do see a difference between Lightroom, NX2 and Gimp (used just to check)).
PS: What monitor are you using? A 8bit monitor WILL have problems with banding, it's not because of the image.
When using a 10bit (+) monitor, make sure you have a videocard that supports 10bit (+). If your videocard is not workstation-grade and under 1000$ it's probably not outputting 10bit.
If you want to be sure, print the full 14-bit file that has banding on your screen. 99% sure that you won't see any banding....Show more →
This may be the issue but the Sony RAW is only 8 bits to start with.
Is that the root of the problem ?
Can you view this RAW in your system and see if you see an issue? http://tbf.me/a/BjiCrW
I am ok with getting a better monitor workstation.
p.1 #10 · Nikon D800/e users - a little help please ?
molson wrote:
While many fanboys may claim otherwise, I can see the subtle differences in tonality and color rendition between non-compressed and "lossless" compression (an oxymoron...) in NEF files
This is incorrect. Lossless compression encodes the data and the RAW converter decodes the data. The resulting files are exactly the same as the uncompressed RAW file. Not kind of the same, or similar, or in the same ballpark. Exactly the same. Understanding how something actually works doesn't make anyone a fanboy.
p.1 #14 · Nikon D800/e users - a little help please ?
playerofwar wrote:
Will compressing a file to a .zip make the files contain less information when you unzip it?
I don't see the slightest difference between non-compressed and lossless compressed when viewing in a editing program (essential, because it has to be decompressed). I viewed them on a 10-bit screen with a 10-bit videocard to make sure there would be no difference and I also took some prints (full 14-bit). Don't see any difference.
And if you would see a difference, using a other program would solve it (I do see a difference between Lightroom, NX2 and Gimp (used just to check)).
PS: What monitor are you using? A 8bit monitor WILL have problems with banding, it's not because of the image.
When using a 10bit (+) monitor, make sure you have a videocard that supports 10bit (+). If your videocard is not workstation-grade and under 1000$ it's probably not outputting 10bit.
If you want to be sure, print the full 14-bit file that has banding on your screen. 99% sure that you won't see any banding....Show more →
There's two kinds of lossless compression. The first is true Lossless, a la LZW (what Zip files use). It's slower, processor intensive and light on the compression.
The second is Visually lossless. That's what is used in modern cameras and it requires a certain standard that does not allow for block compression or similar (what generates JPEG artifacts). Nikon's visually lossless compression is not very aggressive, so there's only very minor data loss and you really have to push the files to run into it. Sony certainly uses a more aggressive compression algorithm which verges on a lossy algorithm in certain cases. This is not aided by Sony's processing chain leaving some noise in large areas of smooth tonality (the NEX-7 is notorious for this in its skies).
p.1 #17 · Nikon D800/e users - a little help please ?
MaxBerlin wrote:
This may be the issue but the Sony RAW is only 8 bits to start with.
Is that the root of the problem ?
Can you view this RAW in your system and see if you see an issue? http://tbf.me/a/BjiCrW
I am ok with getting a better monitor workstation.
That can't be true... Sony raws are also 12 or 14 bit I think. 8 bit in a modern camera is just impossible.
In my D800 I can select the bitrate (12 or 14 bit) and compression of the raw's (uncomp, loscomp, comp,...).
Will check your file tonight (on phone right now).
You don't really need a 10-bit monitor if you know you can get some banding, it costs a lot of money compared to 8bit and offers an advantage to you that your clients won't see (unless you are printing everything). New standard for 4K and 8K tv's are 10 and 12-bit so 12 bit low-end gpu's are around the corner.
I know several people with a 14-bit Eizo hooked up to an Imac, and they think they are getting 14 bit. They are just using 8-bit of that wonderfull monitor. Personally I concidered it for a long time, but eventually went with a workstation gpu for the sake of 10-bit. My prints have improved (because I was able to see more of what was actually going to be printed) and it's a nicer experience overal ( never any banding again).
p.1 #18 · Nikon D800/e users - a little help please ?
JoshI wrote:
Uncompressed NEF's write faster to your card and open faster in your RAW converter ...
Uhmmm ... jpeg files write faster to cards usually. They are smaller (less MB to transfer to the card), however the data have to be subjected to demosaicing and then compression to obtain a jpeg.
Have you a reference showing the write time of uncompressed and lossless compressed NEFs of the same image? Thanx.
p.1 #19 · Nikon D800/e users - a little help please ?
Just checked with a 150mb/s card:
Uncompressed takes about twice as long to write to card (green light goes off).
1.5s vs 2.5s or so (didn't use stopwatch), but it was close to these timings.
Jpegs are even faster (under 0.5s), but that's obvious.
p.1 #20 · Nikon D800/e users - a little help please ?
Ripolini wrote:
Uhmmm ... jpeg files write faster to cards usually. They are smaller (less MB to transfer to the card), however the data have to be subjected to demosaicing and then compression to obtain a jpeg.
Have you a reference showing the write time of uncompressed and lossless compressed NEFs of the same image? Thanx.
That's a good question and you caught me out a little with my wording. All files write to the card at the same speed, either dictated by the write speed of the camera processor or the card (whichever is slowest). Per MB, an uncompressed file will move through the processor faster because the processor does not have to apply the compression algorithm. Overall, the lossless compressed file will write faster because the files are smaller. That's the reason why my D800 manual states that the RAW buffer with 74.4MB uncompressed NEFs is only one file smaller than with 41.3MB lossless compressed NEFs. Logic would dictate that if the buffer can hold 1190MB (74.4 x 16) it should hold 28 lossless compressed NEFs.
And one more correction note on what I said above. Lossless compression is only truly lossless when you open the file in ViewNX or CaptureNX. Theoretically they should be the same in other converters, but the other software manufacturers don't have direct access to Nikon's compression algorithm, so they may not be 100% accurate. At least that's my understanding. I get identical file sizes when I open in Capture or ACR, so I'm going to venture a guess that ACR is accurate, but I have no way of proving that.
This is incorrect. Lossless compression encodes the data and the RAW converter decodes the data. The resulting files are exactly the same as the uncompressed RAW file. Not kind of the same, or similar, or in the same ballpark. Exactly the same. Understanding how something actually works doesn't make anyone a fanboy. 
