NEF to TIFF sizing

I cannot find an explanation of the image size resulting from converting a NEF image (D7000, 14 bit color depth) to a TIFF image. I am told only the TIFF image is 48 bit per pixel. What is that structure?

Same question for 12-but color depth.

Nikon phone support cannot answer this question. The Wikipedia article on TIFF gives the world history of the TIFF format, without providing the information for answering this question.

Thank you.

The TIFF is 16 bits per channel per pixel. That is, each pixel comprises three 16-bit values, one each for the red, green and blue channels, for a total of 48 bits. This is true whether the conversion was performed on 12-bit or 14-bit sensor data from the NEF file.

The NEF file contains recorded values from the sensels (the individual light-measuring elements of the camera's image sensor). These can be stored as either 12- or 14-bit values in the NEF. It's not technically accurate to refer to that as "color" depth since raw values haven't been converted to color values yet. That's the function of the raw converter program.

Thanks, Jon!

PS. Underlying this, I was told that I should use TIFF for lossless conversion from lossless compressed NEF (D7000), 14 bit. But i was also told that I could use high quality JPEG for 12 bit. Is this a subjective intepretation of what adequate quality preservation (upon conversion) would be? I assume that JPEG is never completely lossless. Correct?

> Is this a subjective intepretation of what adequate
> quality preservation (upon conversion) would be?

Very subjective.

The want of lossless or lossy storage should have nothing to do with the input format in the first place.

In theory, some apps support a lossless JPEG format.

---------+---------+---------+---------+
Joseph K
Seattle, WA, USA

D700, D200, D70S, 24-70mm f/2.8, VR 70-200mm f/2.8 II,
50mm f/1.4 D, 70-300mm f/4.5-5.6 VR, 18-70mm f/3.5-4.5 DX

What will you be doing with the converted files? I ask that because if you will be opening them in an editing tool, and if that editing tool has its own raw converter (such as Adobe Camera Raw in Lightroom, Photoshop or Photoshop Elements), you may be better off using your editing tool's raw conversion and entirely skipping the intermediate file format.

That said, yes, JPEG generally involves some loss, although at maximum quality settings I personally can't perceive any degradation.

I am interested in planning for where I might need more than one editing tool, such as CS6 and DxO. I am assuming that I would need a common format to pass the image between them. such as TIFF.

Otherwise, opening with ViewNX2 and, similarly, outputting TIFF.

Re: "in theory some apps support a lossless JPEG format". Is the max size & quality JPEG straight from the D7000 card lossless, compared to lossless NEF, or not? I thought all JPEG was more or less lossy.


Comments?

>I am interested in planning for where I might need more than
>one editing tool, such as CS6 and DxO. I am assuming that I
>would need a common format to pass the image between them.
>such as TIFF.

For that case, yes, using ViewNX 2 to convert to 16-bit TIFF with LZW compression is probably a good way to go.

Along with issues of compression, note that JPEG is an 8-bit format. Whereas the 16-bit TIFF will give you 16 bits per channel for each pixel, JPEG gives you only 8 bits per channel. So even if the compression is lossless, you are still losing some data. This becomes especially problematic if you are processing the image through multiple editing steps. I would not use JPEG as an intermediate format at all.

I belatedly noticed your reference to CS6. What I do (using CS5) for my workflow is perform the raw conversion in CS5 using ACR. The resulting combination of the NEF and the XMP file that ACR creates provides a lossless raw-conversion image that can be recreated at any time by simply opening the NEF in CS5. At that point, I can edit the image directly in CS5 or save it in a lossless format such as PSD or TIFF if I need to work on it with another tool.

Very helpful! Thanks you!

PS.

Re: "Cpmbination of that NEF & XMP that ACR creates" ... "recreate by reopening the NEF in CS5." Is that the original NEF? What is the role of that ACR file when then re-opening the (original?) NEF?

I am brand new to CS (obviously).

Re: "using ViewNX2 to convert to 16-buit TIDD with LZW compression"

Is the LZW compression an option (default or otherwsie?) in the ViewNX2 conversion from NEF to 16-bit TIFF?, or is it a fixed standard?

When you open a NEF in CS, it launches Adobe Camera Raw to perform the raw conversion. In ACR, you can make changes to various settings. Once you accept the ACR settings -- changed by you or not -- and continue on, by default ACR creates a small "sidecar" file in XMP format stored in the same folder as the NEF. The next time you open the NEF in CS, ACR reads that XMP file and applies the settings.

The upshot is that you can reopen the NEF in CS as many times as you want, applying the original ACR settings (or changing them if you like) to get the image the way you originally adjusted it in ACR. Now, if you edit the image in Photoshop CS itself, you have to save it as a copy in another format in order to preserve your changes. Typically, you would save it in Photoshop's native format (.PSD) so you can later reopen it with all of the edits intact.

But you'll still have the "negative" -- the NEF file -- and the raw conversion settings -- the XMP file -- needed to reopen the original image without the Photoshop edits.

Note that in this workflow, you never make any change to the original NEF, only to the conversion settings in the XMP file.

Regarding ViewNX 2 and LZW compression: Yes, you can choose to save the 16-bit TIFF with or without LZW compression. LZW compression is totally lossless, so the only cost of using it is a bit of CPU time needed to perform the compression when saving and then decompression when opening the file. Given the speed of modern CPUs, this seems a small price to pay for the saving in disk space (20-30% or so, I think).

Very helpful & clear. Thanks, Jon!

Hi,

here is a sample image of my daughter taken with my D800. I had noticed that the full resolution JPEG is slightly softer than what I could see in Lightroom when zoomed to 1:1.

The original:

A screenshot of 100% zoom in Internet Explorer:



A JPEG generated from a cropped image:



Please note the difference between the second and third image, especially in the eye lashes above the eye.

Even when exporting with the best quality, a large JPEG image (shown in the middle) seems to come out slightly softer in details, i.e., it loses some detail.

The image was not sharpened from the default value of '25' in Lightroom.

Best regards,
Kari


Attachment#1 (jpg file)
Attachment#2 (jpg file)
Attachment#3 (jpg file)

Clearly visible but, to me, still subtle difference!

I’m not entirely clear how your posted images were generated.

It has been my experience that exporting a 100% JPEG from a NEF or TIFF does not significantly affect fine detail even after 10 consecutive imports and re-exports. Where the image does degrade on successive saves to 100% JPEG is in posterization of solid colors.

This shows up on my website under Button 10 at the bottom of the page. Click on Item 1 and look at slides 7,8,9 in full screen mode.

It’s a moot point. I certainly would not recommend use of the JPEG format for anything other than exporting images.

Barry
http://art2printimages.com

Hi Barry,

The first 100% zoomed image (image # 2) was copied bit-by-bit from a portion of the full image with minimum compression by capturing the window content of a web browser. The full JPEG image loaded to the browser was generated from a NEF (raw) image. So, all the 36 megapixels from D800 were in use in the browser image.

The second JPEG image, on the other hand, is created from Lightroom after the original image was FIRST CROPPED to a similar area (769 x 437 pixels).

The difference between the images indicates that even with minimum compression setting (at least in Lightroom terms), a JPEG file loses some details - unless the details are lost by the browser. If this is not a browser problem, I guess the larger the image, the clearer the loss is when zoomed to 100%.

Yes, there is an additional screen-to-JPEG conversion from the screenshot; Nikonians does not support PNG files. Nevertheless, the difference was quite visible in the browser. It's easy to test: create a JPEG file from your sharpest image with minimum compression, load it to your web browser and compare it to the original image with 1:1 zoom factor.

Best regards,
Kari