bwana Loc: Bergen, Alberta, Canada

Yes, why tempt fate...

If you must start from JPG files, at least save them in a lossless format and retain them in this format throughout the editing process. You might consider saving them as TIF or PNG, per:

"Portable Network Graphics (PNG) is a raster graphics file format that supports lossless data compression. PNG is the most used lossless image compression format on the Internet."

bwa

Apaflo Loc: Anchorage, Alaska

There is no point in not using Lossless Compressed for NEF files. (And for that matter, using either LZW or Adobe Deflate (aka BZIP) compression with TIFF files.) The term "lossless" can be taken literally, and the data you have before saving will be exactly the data recovered when the compressed file is loaded.

The file sizes, absent compression, are pretty straight forward. If a camera has 12 MP, for example, the data for those pixels will be either 12 bit or 14 bit in a RAW file. 12 bit RAW data is saved as 2 locations per every 3 bytes. So a 12 bit file will be 12 * 3 / 2 bytes for the sensor data. There is also a full sized JPEG image embedded in the file, so depending on the quality that is configured the NEF file would be maybe 22 MB. A 14 bit file is coded as 2 bytes per location, so the data set would take up 12 * 2 bytes. If the JPEG is also 4 MB, that mean a 28 MB NEF file.

But a TIFF format, while it doesn't have an embedded JPEG, will either have 8 bit or 16 bit values in each of three (Red, Green and Blue) channels. So it is 3 * 12 for 36 MB, or 3 * 12 * 2 for 72 MB.

And compression will change those values greatly, but usually to about the same effect. If an uncompressed 14-bit NEF file from a D800 is maybe 74MB, a Lossless compressed NEF file will be about 41MB. The uncompressed 8-bit TIFF file might be 108MB, and if compressed may be about 74MB. An uncompressed 16-bit TIFF may be about 218MB, and compressed might be 194MB.

That's a lot of numbers, and while perhaps interesting they really don't mean much in terms of any effect on photography. Shoot RAW in Lossless Compress mode, convert to compressed TIFF files for editing and saving as intermediate. Convert to JPEG for display on the web or transferring to a print lab over the Internet. Use either TIFF or JPEG for local printing.

RWR Loc: La Mesa, CA

I should have stated that the 34mb NEF is from my particular camera. Other sensor sizes and mp will vary.
No, I do not know Lee. Enjoy your trip, I have several friends from Thailand. Good people.

R.G. Loc: Scotland

Absolutely - and there's no shortage of options.

I notice that lots of ISPs have a 20MB per email upload limit. If PNG or TIF files were too big, DNG might be a workable option. And lossy compression may be a possibility with some of these options.

Peterff Loc: O'er The Hills and Far Away, in Themyscira.

Lang may yer lum reek!

Iechyd da!

RWR Loc: La Mesa, CA

You're reading my mind! As soon as I made the comment I changed the setting to lossless compressed. No idea why I hadn't done that long ago (only had the camera for 10 months!). Should have mentioned, too, that I save as 16 bit TIFFs, & convert to JPEG for email or uploading.

jenny Loc: in hiding:)

* * *
In other words, your camera's factory default settings need to be processed no matter whether your shot was in RAW or JPEG... You might consider either way as merely processing. If you have to do additional work to one or two specific areas, that will require some additional, i.e., post processing.
Some of you seem to have missed the point re. processing the JPEG however. Forget the overworked criticism about JPEG being "lossy" for awhile. Make a copy, work on that for your processing and you are obviously not "losing" anything if saving the original. It will be smaller than a RAW file but just as lossless if not worked on.
Trust JPEG, it is not just a term for the state of a photo, it is the name of a panel of experts who made it possible for you to get fantastic results these days with digital exposure.
You risk more noise, especially in high contrast shots, working a RAW file than you do with simple adjustments of JPEG in something as basic as Windows Photo Gallery or Picasa.
Keep in mind that most of us don't make huge poster prints or work for someone who wants to change out things
that require post processing, and unless you have to do that sort of work you may not really need Raw anyway.Be willing to experiment.

bwana Loc: Bergen, Alberta, Canada

Right off the bat you've lost dynamic range using an 8 bit JPG file... Even if you want to assume JPG artifacts don't have an impact on your pictures.

Just saying...

Peterff Loc: O'er The Hills and Far Away, in Themyscira.

"Canons to the right of them, Nikons to the left of them, ...

... theirs not to reason why...."

With apologies to Alfred, Lord Tennyson.

Such is the result of poor or misguided communication....

Sometimes you just have to stand back and shake your head.

sletter Loc: Daytona Beach Area

The JPEG standard was created long ago, in technology time. 8 bit images were state of the art. Storage space was limited and very expensive. Processors were 8, 16 and some 32 bit machines. State of the art has changed. 64 bit processors are now the standard. Multi terabyte disks abound.

The standard was formed in the mid '80's. It has recently been updatedto overcome many of the disadvantages of the original standard. The JPEG committee itself recognizes that the current standard in use has become obsolete. see http://blogs.msdn.com/b/billcrow/archive/2009/07/29/jpeg-xr-is-now-an-international-standard.aspx

Ranjan Loc: Currently Cyber-Nation!

If I understand correctly, all sensors produce RAW information (data from pixels); some cameras make that available to the user (as RAW files and camera-specific nomenclature etc) while others only make the post-processed 'raw' (as jpeg) available as a file.

So, even the earliest digital camera potentially had the RAW information generated by the sensor, though not saved by the camera.

Did I get that correctly?

Apaflo Loc: Anchorage, Alaska

:thumbup: :thumbup:

bwana Loc: Bergen, Alberta, Canada

A win, win decision! Smaller files and still lossless :)

bwa

Apaflo Loc: Anchorage, Alaska

Do not ever forget the significance of JPEG being an inherently lossy format. That is of vast significance and cannot be overstated.

But along the same lines there is another aspect you have not caught yet. The raw sensor data in a RAW file is not an image. It does not specify one single image, but rather it is the material from which one can correctly generate nearly an infinite number of slightly different images. The data does not specifiy one single specific image. Using very correct demosiacing, a large number of very very different images can be produced from one RAW file.

Compare that to a JPEG file, that has data defining just exactly one image. It might show up different on different screens or printers, but that is because the display mechanism is incorrect. And all those other images that were available with the RAW data from which the JPEG was created... are not available in the data set that is saved with a JPEG. Worse, the data that is saved cannot be exactly recovered either. That is the lossy compression. You get something "close enough". Close enough that you can't see the difference, but it is there and assuming it is otherwise will cause the differences to become visible (by editing, for example).


But when that image was first saved as a JPEG it did not retain the huge amount of data that was available in the RAW file. You've already lost 10 times what you are using. And each time it is saved again, you lose more, again.


Not true. When a JPEG file is save it is damaged. If you have a JPEG, it was already damaged. You can make 10 copies, and they are all equally damaged.

If you resave any of those copies it will be further damaged.


That is backwards. The RAW file, with either 12 or 14 bit depth, has more dynamic range. That is Signal to Noise Ratio, which means if you set the white level to any given whiteness, the noise will be buried deeper with a greater bit depth.

Now, if you shoot at less than about ISO 800 or 1600 with modern digital cameras, that additional dynamic range is used, and is greater than what a JPEG encoded image can have. If you shoot with a higher ISO it is true that the RAW file will not have any more dynamic range that what the JPEG can encode.


But it also happens that some do want poster sized images. They also want to crop down to half the original image. But more likely they are a typical parent that comes here to UHH and asks about how to shoot a basketball game, or how to shoot a Christmas program. Both are "event" photography in it's worst disguise, and every advantage helps. Shooting RAW is one of the least expensive advantages that actually helps. Expensive lenses and expensive cameras help too, but...

TheDman Loc: USA

No, that has nothing to do with what I said.



It's hard to write that much stuff and have 100% of it be wrong, but you've managed to pull it off.