DSLRUsers.com

I have a play today with couple pics from the D70 "Go into Nikon Capture 4.1 or higher and change the size / resolution to 8 mp equivalent (3504 x 2320), It works great!" So I tried this trick on both three camera D70/ D100 & D2h, and the resolution is real perfect on the photos.
Have anyone try this before? it does work great, image quality does not go away, especially for portraits!!!

This is how I did:
Click view and open tool pallets 1,2,&quick tools. Open a raw image in Nikon capture and in the "Tool Pallet 1" open the tab that says size / res. and manually enter the pixel resolution there. I think this only works with raw files. and then process as normal, sharpening ( iusually start off with sharpening set to NONE in the advanced RAW pallet) Hope this will help?

Plus, if you have a very noisy image to clean up, upsize the NEF, clean it up, make your adjustments, shrink, and voila, a lot of the noise goes

Birddog
That's very interesting. I will try it out and see how it goes. Thanks.

Nice! I never did notice that feature before (as I don't predominately shoot NEF, I hardly use Capture). I thought it was just to upsize D1x files from squished to interpolated 10+MP.

But ok, so you got 8MP at your disposal. Why stop there? Why not wind it up with 12MP, or 16, or even 20?

Onyx wrote:Nice! I never did notice that feature before (as I don't predominately shoot NEF, I hardly use Capture). I thought it was just to upsize D1x files from squished to interpolated 10+MP.

But ok, so you got 8MP at your disposal. Why stop there? Why not wind it up with 12MP, or 16, or even 20?

Onyx,
Yes, I tried with higher pixel as well as in the previously post but don't go too far outer the limit of technologies same as overlocking the CPU or graphic card, photo will be burn.
The reason I post this thread is someone asked me a question: to use D70 and print photo as the billboard, they worried about the limitation of 6MP on D70 & D100 and 4MP on D2h, that why I tried to get something done by "cheating" with NC and PS, finally it works for them and they can see the pixel at the state they want.
Just something interesting discovers in today technologies and I would like to share

Please remember: Only RAW files can do this/ not JPEG

Wohoo!
This sounds great!

Thank you Birddog, maybe no need for the D2x now?

What are the pixel values for each megapixel value? Just playing around I doubled the height and width of my image and that seemed to work really well.

PlatinumWeaver wrote:What are the pixel values for each megapixel value? Just playing around I doubled the height and width of my image and that seemed to work really well.

It's 3504 x 2320 for 8MP

Ta - I just used 6016x4000..

PlatinumWeaver wrote:Ta - I just used 6016x4000..

Your choice of creatively guy!

Platinum - I hope you realise, 6016x4000 is 24MP!! You even beat the D2x in filesize!

When I sat down and thought about it that's what I realised. to the naked eye the image appears identical to the original. But this is unprinted and with my naked eye, which is perhaps not the best trained in the world.

One odd thing, if you save the NEF with the updated size , the filesize is the same as the picture 1/4 the size. I guess that's something specific to .nef and other RAW files I don't really understand.

Greetiings from Indiana

You gotta love the internet for making the exchange of ideas so easy.

I've read the previous posts in this thread and need some help in understanding what's happening. We all know that the D70 weighs in with about 6 megapixels. So no image can be larger than the pixels available. So it must be that the change in resolution is the result of an interpolated upsize. That is "extra" pixels in the photo are added to the image ehich should result in a softer image.
Here's where I'm confused. If additional pixels are added, why would the file size remain the same? And in conjunction with that, why isns't there an apparent difference in the look of the photo?
Does anyone have any idea what Capture uses in terms of upsizing interpolation? I know that PSElements uses bicubic while Qimage Pro uses either vector or pyramidal.

Greatly interested in any answers!

Best - Phil

maybe its to do with bitrate??!!!