Questions on Nikon Capture

Okey Dokey, finally got around to shooting some RAW photos.

Now what?

I have downloaded them to the computer and opened them up using Nikon Capture. Am i supposed to be doing something with them in Capture?

All i did so far was import them into PS7 and a little curve and sharpening there before saving as tiff files.

Am i missing some major steps here?

I have no idea what i am using the Capture for other than using it to open NEF files.

Ahh the technology exceeds me...

There's an awful lot of stuff you can do in the RAW conversion but I'm not familiar with Nikon Capture. I've used ACR and wrote some suggestions on how to run it. James converted those to a tutorial that you can find here. Generally speaking, most things that can be done in the raw converter can also be done in PS but I believe it's best to do them in the converter. First, you might possibly minimize quantization errors. Second, controls in the converter (esp. white balance) are easier than PS counterparts. Finally, everything is done in linear encoding in the converter so it's more accurate. Once you've gone into a color space you have to accept a gamma curve and the errors that may come from editing with it. Still, ACR controls are nowhere near as complete as the PS ones. I strongly recommend you edit entirely in 16 bit so, unfortunately, I think you should consider upgrading to PS CS (which has far more complete 16 bit support). Good luck!

Thanks Craig, I'll read that article. Just got CS but haven't downloaded it yet. As for the controls, i just figured out how to use them! I thought there was something missing...

As i said, technology exceeds me and often eludes too. eg quantization error...Quanta who?

PS CS has an excellent raw converter. You want to adjust your images as much as possible in your raw converter before you go to PS. You will lose less information from your images that way.

I'm sorry. Did I say that?

Quantization is the process of converting continuous things (like tones in an image) into discreet quantities that can be represented by numbers. Computers create digital images by quantizing in two ways. First, by creating discreet pixels as everyone knows. Second, by assigning some number of discreet bits for each pixel. The result is that every time you edit your image small amounts of data are lost. These losses can be minimized by making sure you work in high resolution. That means lots of pixels and 16 bits per pixel. 8 bits is not enough!

The reason I advocate adjusting in the raw converter is that it's the place where no information has been lost yet so it's most accurate. The image hasn't been mapped onto a dreaded gamma curve yet either. The goal in any case is to produce the best 8 bit image you can for your desired output by carrying through the extra information and discarding it at the very end.

A note about bits per pixel. A camera produces between 9 and 12 usable bits per pixel with some as high as 14 bits. Ours typically offer 12 bits in the raw file although the D100 only has 10. 16 bit standard images only actually offer 14-15 bits but that's still enough to hold everything that the camera produces. An 8 bit image holds only 7 bits of camera data due to the gamma coding so when you work in 8 bit you immediately discard 5 of your 12 bits of hard earned data and chew further into that while editing. Don't do it! Sure, you only need 7 bits for output but make sure it's the right 7 by keeping the extra ones around as long as possible.

There's a common belief that you can adjust as much as 2-3 stops of exposure error in a raw converter. Well, this is where it comes from. You aren't really able to do that so much as you can choose which bits from the camera go to making your final 8 bit image. If you are going to throw away bits (which you are) make sure you're the one picking them; not the camera and not the computer. The camera is far better and more forgiving of our inability to operate it than we realize. They are amazingly powerful devices that offer plenty of usable resolution but we need to make sure we don't waste it through foolish processing techniques that are all too common.

Hope that makes sense.

Good morning Craig - man, you are up early! I'm on the night shift, so I have an excuse.

Good post - thanks for explaining in such detail.

Cheers

James

Craig -

although the D100 only has 10

I've only seen the D100 spec'ed as 12 bits. What happened to the other 2 bits?

Chris

They're faked! The D100 only offers 10 bits out of their ADC but produces a 12 bit raw file. The D100 compressed raw file is only 9 bits so not only is it slow but it's lossy as well. I believe the D70 is only 10 bits but the compressed version does not have the same flaw. At any rate, 10 bits is not really a problem for a camera that only has about 8 stops of dynamic range so this isn't the issue that it may seem to be.

Hi Craig,

Thanks for the follow up explanations. Well written and easily understandable.

Thanks

Mike

Mike -

You may notice a slight difference in color accuracy between Photoshop's raw converter versus Nikon Capture. I prefer Nikon's program (better red saturation) to read the NEF files before bringing into PS CS. There was a previous thread on this and someone posted some very impressive differences between the two converters.

M.