Fuji x10 6MP RAW DR400 EXR mode vs old 6MP DSLR

Started 4 months ago | Discussions thread
Trevor G
Veteran MemberPosts: 6,460Gear list
Like?
Re: Real DR vs pseudo DR
In reply to photoreddi, 4 months ago

photoreddi wrote:

The X20 has a newer, more modern, much improved sensor. Comparing the X10 to the X20 is not only foolish, it directly contradicts your claim about DR performance being primarily based on the sensor size. When the X10 was introduced there was no X20 to compare it to. But

There is an X20 now, and it is good at revealing dynamic range as the X10.

the F200EXR and other EXR cameras existed, and they were widely recognized as have the ability to capture much wider DR images than other cameras with the same size sensors. Real DR that it, not the shifted tone curve DR techniques that Fuji, Nikon, Canon and most other cameras had been using for years.

Except that all DR "captured" in whatever way can only be applied to the ooc JPEGs through tone curve techniques.  That's why the JPEG histogram doesn't match the equivalent's RAW highlights histogram.

There is no such thing as "real DR" with EXR - it is no different to ISO amplification in the need to use tone curves to squeeze the available image data into the possible JPEG space.

Here's an example - the "REAL" highlights with histogram are revealed in the crop from the RAW image on the left.

The matching ooc DR400 JPEG simply cannot handle all the information, some of which is clipped up against the right hand edge.  In addition, the low lights (note the R G and B peaks in each histogram) are squeezed up into the upper 1/3 of the image space, instead of being almost all the way to the bottom.

Note the reduced contrast in the ooc JPEG compared to the RAW as 2/3 of the image data is squeezed into 1/3 of the space.  Blue sky becomes cyan as B channel is compressed towards R channel.  This is the result of EXR Hardware processing in-camera.

If we reduce exposure of the JPEG slight;ly we see that the B channel is actually clipped as well:

EXR Hardware exposures are ultimately processed in the same way as EXR software exposures - via tone curves.  This reduces the available dynamic range information at the upper end.

Even if I use a DR400 ooc JPEG that didn't obviously clip you will still see that the highlights have been compressed. That can only happen through tone curve manipulation via the in-camera processing.

Ultimately EXR Hardware is no better than EXR Software in squeezinging more dynamic range
 information into the available image space.  The only advantage could come with not having to raise ISO - here we can wait for DMillier to show us what happens on the X10.

We know that the X10 has no "real" dynamic range advantage over the X20.  Different techniques, similar results.

How exactly does a cloud actually present an accurate, measurable light intensity that a step wedge couldn't duplicate?

I am going to suggest that studio test results are either not readily visible in real life images or, probably more likely, that EXR Software has taken over from EXR Hardware. It's easier to implement and more reliable, as the X20 shows.

I don't need to measure the light intensity at the clouds - the camera inteprets what it sees and, as I adjust EV compensation, either clips or reveals the details to give a starting point. It's relatively easy to detect what is clipping in an image and what's not.

That's all we need - the clipping point for peak white.  Friom there we lift shadows until we get no further detail provided, or noise intrudes.  That's all there is to real-life Dynamic range limit points.

More importantly, how do you accurately measure the darker subjects at the opposite end of the scale? Do you measure with real, reproducable numbers or just tell yourself what you want to hear?

Waste time trying to convince you? You're like someone else here, where "actual results" are either ignored or denied if they show what you don't want to see.

Once again, I don't think there is any need for you to be offensive.  You have yet to show one piece of evidence that the X10 has any more dynamic rnage than the X20.  It's just hearsay.  No evidence from you whatsoever.

On the other hand, I can show that the X20 shows as much highlight information as the X10 at the same exposure setting, without compromising shadow details.

-- hide signature --

Cheers
Trevor G
Silkypix tutorials at: http://photo.computerwyse.com

Reply   Reply with quote   Complain
Post (hide subjects)Posted by
(unknown member)
(unknown member)
(unknown member)
Keyboard shortcuts:
FForum PPrevious NNext WNext unread UUpvote SSubscribe RReply QQuote BBookmark post MMy threads
Color scheme? Blue / Yellow