EXR ..Debunking the Mystery .. Pt 2

The EXR sensor… HR & P modes.

If you had a peek at the Fuji website you will know that the sensor in EXR cameras does some cool stuff to help make your photos better.

HR (High Resolution) , DR (Wide Dynamic Range) & SN (High Sensitivity & Low Noise). I will use the short form of the modes for this discussion.

HR mode is fairly self explanatory, as it is simply the full 16Mps being used to take the image. The intent here is to supply as much resolution to the image that the sensor is capable of. The following images are all shot with DR100% + ISO 100% + Flash set to +1EV. White balance is set to tungsten as we are under lights, focus point is the wording on the top of my s5700. I shot one image in P mode and one in EXR HR mode @  16 Mp for maximum resolution.

16Mp P Mode

For these exercises I will, where appropriate use P mode for comparison. By extension these results should apply to the other PASM modes as well. As always I would encourage users of the HS20 & HS30 to experiment for themselves.

I will take a moment here to discuss “P” mode. A lot of users use P mode or A mode, partly because they are used to doing so from previous cameras they have used. Others like these modes because the thinking is that the EXR performance will be gained by switching to an 8mp image size, and thus reap the gains of better dynamic range performance. Personally I’m at a loss as to why people think this.

To any logical way of thinking it would seem counter productive to build a camera with special EXR functionality, advertise this technology as the prime mode for shooting to gain the advantages of the new EXR technology, and yet apply the same technology to the PASM modes. I cant see any reason as to why you would do this.

It would seem to me at least that the camera when set in PASM modes should function like any other 16 Mp camera that doesn’t have built in EXR. Essentially two cameras in one. As DR 100 to 400% is available in PASM and EXR modes, surely then there is something else at play here.

It would seem so.  The two crops below show a portion of an image cropped at 100%, one shot in P mode and one shot in HR mode. This is why some camera owners “pixel peep” and why its useful to do so. Its at the 100% level where you find out just how well your camera is performing. Exif data is maintained in the shots but here how the camera was set. Note identical settings for both P mode & HR mode.

  • ISO 100
  • Image size 4:3 Large
  • Image quality – Fine
  • Dynamic Range DR100%
  • Film Sim = Velvia (V)
  • Color – High
  • Tone = Std
  • Sharpness – Hard
  • Noise reduction = Low
  • Image stabilisation = shooting only
  • White balance = Auto
  • Metering = Average
  • Ev = 0.33+

Click the two images above to see a full sized image. Hover the mouse over the image to see which mode it is.

At first glance it would seem that there isn’t a lot of difference if any. On further examination there are some noticeable differences, which are by no means easy to see. The most prominent clue comes when examining the lettering in each image. There is more refinement in the edge of the lettering in the HR image. Its not a huge difference, but does show that the image is showing a small benefit from the EXR processor. Both images would be near identical after post processing. On balance I think its safe to say that with these settings both modes would provide good images.

Another thing that comes from this test is clear evidence that in good light at 100 ISO there is very little if any noise evident in either mode, so the sensor is reasonably quiet at ISO100, and negates numerous claims I’ve heard to the effect that using the full 16Mp resolution of the HS20 was only ever going to give poor  and noisy images. This is clearly a fallacy and one I have long known to be so.

And just in case anyone thinks that this is only true of base ISO here some shots  showing both before and after at ISO 800 @ 16Mp. Click the thumbnails for a larger image.

So what do we make of this cursory test?

While not exhaustive it does serve to show that there is a small gain in the HR images. As I shot these with sharpness set to hard, I wasnt expecting a great deal of difference. Had they been shot with sharpness set to low there would be room for a larger amount of differential between the two modes.

Un processed pmode
Processed P mode
Unprocessed HR mode
Processed HR mode.