Why many pixels really help

March 4th, 2012

The new Nikon D800 has a 36MP sensor. Bummer

Well it is relative. When you compare it to a D7000 - the pixel-density is the same: Crop-factor 1.5 yields 2.25 the surface and 16*2.25 =36).

So no worry about the lenses, when we disregard the border at larger apertures. Frankly, when do you need sharp corner when shooting wide open?

One inconvenience is the larger file size (up to 70MB), but considering that I bought my computer with a 500GB drive 2 years ago, and now a 2TB turns along with it - do we really have to care?

The D800 also features crop-modes that utilize only a smaller section of the frame, this will keep the file-size down when the resolution is really not necessary (beware: a DX crop transforms the D800 into a heavy D7000, it is not downsampling the whole sensor!).

Yes, it is that easy, but there are many out there who think (or don’t think at all) that so many pixel must come for a price. Indeed, they come, but the only price is the list-price of the D800(e).

Besides the storage issue, two arguments float around:

  1. The pixel density is too high, diffraction will ruin the image, it is better to shoot with a lower pixel-count
  2. Larger pixels perform better at high ISOs, better a lower pixel-count like in the D4

You know my drill, first the short answer and then the lengthy derivation:

  1. Diffraction will only compromise details a lesser sensor won’t even resolve
  2. You can downsample the file to reduce the noise almost perfectly equivalent to a sensor with larger photo-sites

Yes, I use pixels and photo-sites synonymously in this article

Read the rest of this entry »