K-x Shadow Compensation - loss of contrast

SRT201

Veteran Member
Messages
2,713
Reaction score
316
Location
US
I have been using the Shadow compensation technology on my Olympus cameras for a while and decided to test it against the Pentax K-x implementation. The following five shots illustrate what I have seen. The first shot was taken with no compensation and then adjusted using Apical's shadow compensation in Olympus Viewer 2 to create the IMGP1105_at.jpg. The other three shots where done using the in-camera shadow compensation levels.

In general I would have to say that I prefer Apical's local tone curve technology to the method used in the K-x. The K-x method provides a bit of shadow recovery at the low and medium settings but to significant degree flattens the image and reduces contrast on Hi. They both do a decent job of recovering detail on the tree trunk for instance but look at the shrub in the lower left. The Apical version enhances contrast in the shrub while the K-x slowly reduces it as the shadow compensation level increases. Also, even on high the K-x method does not recover the tree trunk as well as the Apical version.

Judge for yourself.

Pentax and Samsung licensed Apical's technology right about the same time the K-x came out so subsequent bodies may already have the Apical technology built in. I would like to think that.

Here is the link to Apical's web site and their dynamic range technology.

http://www.apical-imaging.com/DRC

Click the image to see the full difference. Rather amazing.

I guess the point of this is to offer a bit of advice. From what I have seen the K-x shadow compensation provides a bit of advantage at low or medium but I would avoid using high because of the loss of contrast and overall "punch".

Original





Apical local tone curve process applied





K-x shadow compensation LO





K-x shadow compensation MED





K-x shadow compensation HI





Scott

--
Happiness is a want... Contentment is a choice.
 
If you have any interest in the shadow adjustment technology then it would be worth your while to download the originals above and compare them in detail.

--
Happiness is a want... Contentment is a choice.
 
I had no idea this would be such a controversial topic. :-)

Actually, if you have any interest, the View NX 2 program can be downloaded from Nikon for free and has some nice shadow adjustment editing built in. Works for Pentax files no problem. May be based on the Apical technology as Nikon licenses it.

http://nikonimglib.com/nvnx/#

--
Happiness is a want... Contentment is a choice.
 
The one stop underexpose works great for protecting your highlights but as for Shadow recovery, the average post operation is pretty much a global. Check out the Apical technology. It's nothing you could do manually and the results are quite impressive. It can be applied in post as well with the right software.

--
Happiness is a want... Contentment is a choice.
 
I had no idea this would be such a controversial topic. :-)
I don't really understand what your point is in this thread; perhaps other people don't understand either so they don't know how to respond.

You say you've been comparing K-x shots with Olympus shots (camera(s) unnamed) but show images only from K-x. You talk of an editing system that you have applied in Olympus software on the K-x file but don't say if you have to apply in to Olympus files or if it is an in-camera option for Olympus.

If you are saying that computer software can produce more sophisticated conversions than in-camera software (which is what you have posted) then that is not controversial nor is it news. The link you give uses as its example a landscape shot that looks very unconvincing - OK, it's very small, but if it's intended to convince anyone of the merits of the software it fails. The trees look too bright and the cyan-tinted area above them just looks wrong. There are plenty of other editing programs that can do a better job than the example shows.

Coming back to the images you posted, I don't actually think any of them are very good: what they really show is that shooting JPG in high-contrast light is unlikely to give satisfactory results. The Apical version looks better overall than the compensated versions but the shrub you talk about looks worse in the Apical version: it looks as if excessive fill flash was used on it.
--
---

Gerry


First camera 1953, first Pentax 1983, first DSLR 2006
http://www.pbase.com/gerrywinterbourne
 
How do you know that Nikon ViewNX 2™ and Pentax has licenced the Apical Imaging Dynamic Range?

How does the results from ViewNX 2™ compare to Apical Imaging Dynamic Range and K-x build in shadow compensation?
 
Is it not just about adjusting the tone curve, in the low exposure ('left side") region, or is there more, e.g. that automatically shadow areas are defined on which tone-curves are applied separately ?

Anyhow, if you flatten the contrast in the darker areas and make them lighter, it will always be at the expense of a more 'dull'-looking image; it's the contrast including dark areas that makes a picture look bright.

I think the "shadow compensation" is not more than applying a different tone curve and the advantage is that it takes a proven one automatically and it saves you the time to find it out yourselve.

If different softwares do this different I think it's a matter of taste and balance. I suppose it's not likely that one software does this "better" than the other under all circumstances.
 
How do you know that Nikon ViewNX 2™ and Pentax has licenced the Apical Imaging Dynamic Range?

How does the results from ViewNX 2™ compare to Apical Imaging Dynamic Range and K-x build in shadow compensation?
I don't know that Nikon ViewNX 2 is using the Apical technology. Nikon does license Apical's technology but as to whether that is exactly how they implement Active D-lighting is something probably few know.

A quick web search will reveal when Samsung licensed the Apical technology.

As to how they compare... I have not done such a comparison. But Nikon's Active D-Lighting (which is available for free) if you simply download ViewNX 2 is a region based curve adjustment much like Apical's

I do not know what Shadow adjustment tech is built into Pentax DCU4 either.

--
Happiness is a want... Contentment is a choice.
 
I had no idea this would be such a controversial topic. :-)
I don't really understand what your point is in this thread; perhaps other people don't understand either so they don't know how to respond.

You say you've been comparing K-x shots with Olympus shots (camera(s) unnamed) but show images only from K-x. You talk of an editing system that you have applied in Olympus software on the K-x file but don't say if you have to apply in to Olympus files or if it is an in-camera option for Olympus.
Okay... sorry for the lack of detail. The Olympus bodies I use are the 420 and 520 both of which have the Apical technology built in. The software is Olympus Viewer 2 which also has the Apical technology built in for use on any file. Unfortunately the Olympus Viewer 2 software (which is quite a bit faster than DCU4) is only available if you have a valid Olympus DSLR serial number.
If you are saying that computer software can produce more sophisticated conversions than in-camera software (which is what you have posted) then that is not controversial nor is it news.
Well of course that's not news. My point was to compare a couple different technologies for users of Pentax cameras who may not have tried different Shadow compensation technologies.
Coming back to the images you posted, I don't actually think any of them are very good: what they really show is that shooting JPG in high-contrast light is unlikely to give satisfactory results. The Apical version looks better overall than the compensated versions but the shrub you talk about looks worse in the Apical version: it looks as if excessive fill flash was used on it.
--
Gerry
I'm deeply offended Gerry, those images positively stink! :-) They are there simply because it was an area in my yard with blue sky and plenty of shadow thus providing a wide exposure situation camera's routinely have trouble rendering in a natural way.

The whole point of my post was to point out that the Pentax K-x shadow adjustment technology seems to render progressively flatter (lower contrast) images as you push the level of shadow adjustment. I believe this to be the case because the K-x shadow adjustment is a global curve adjustment that brings up the low end of the histogram globally. I don't know this for sure but that's certainly what it looks like.

On the other hand... there are technologies available that do a more complex local analysis of the image applying local curve adjustments. The human eye dynamically adjusts the thresholds of retinal cells according to the local illumination levels thus allowing us a rather large dynamic range. Apical and others are trying to better reproduce what we see and in my opinion are producing more convincing shadow adjustment technologies than global curve adjustments.

As Samsung licensed the Apical technology it's possible that the K-5 may have more sophisticated shadow adjustment built in.

The K-x technology isn't terrible but another part of my post was to point out the loss of contrast that becomes apparent when the in camera Shadow adjustment is pushed to the HI setting and to point out that there are alternatives like Nikon's ViewNX 2 which is free.

I should probably do the same three shadow level applications using DCU4 and see if the Pentax software appears to be doing the same thing as the K-x body.

Here is an interview with the Apical people at our very own DPReview.

http://www.dpreview.com/news/0903/09031801apical.asp

Scott

--
Happiness is a want... Contentment is a choice.
 
I should also note that I wouldn't recommend using any shadow adjustment in camera unless you're shooting RAW+JPG.

It would probably be fine however if you're just shooting family snapshots in which case you will probably get more natural results without all the fuss of post processing every image from little Jimmy's 4th birthday.

--
Happiness is a want... Contentment is a choice.
 
Note that the Apical interview took place in 2009. I'll bet the shadow adjustment tech in many DSLR's is now a variant of Apical's technology.

--
Happiness is a want... Contentment is a choice.
 
If they were all simply applying different global curve adjustments I would agree with you that it's all just a matter of taste. However, as more advanced technologies like Apical's are evaluating the image and adjusting tone curves locally in order to better model how the human eye adjusts to wide dynamic range I would have to say that IMO there are better (local/region-based) approaches and worse (simple global curve adjustments).

As even Apical notes, we don't always want natural images. But there are certainly times we do and that's when a technology like that can really shine. I hope we begin to see this kind of tech built into the JPEG engines of most future DSLR's so that the curve adjustments can all be done from the RAW data.

--
Happiness is a want... Contentment is a choice.
 
Hopefully Pentax still has access to the Apical technology for their own use. If not they should license it.

--
Happiness is a want... Contentment is a choice.
 
As even Apical notes, we don't always want natural images. But there are certainly times we do and that's when a technology like that can really shine. I hope we begin to see this kind of tech built into the JPEG engines of most future DSLR's so that the curve adjustments can all be done from the RAW data.
It's a while since I looked at the Apical site, but the effect of their shadow enhancement seemed similar to what may be achieved by shooting in RAW and using local 'contrast tools enhancemant' such as 'clarity' in LR, and 'adjust shadows and highlights' in Photoshop.

The downside is that they alter the tonal hierarchy within the image, i.e. a pixel which was lighter than its neighbour may, after editing, end up being darker. You can't undo this in post processing, which IMO is a good reason for applying it manually as above (and preferably shooting in RAW too).
--
Mike
http://flickr.com/rc-soar
 
Agreed. Unless I'm just shooting snapshots I would not want any in-camera image enhancement turned on.

--
Happiness is a want... Contentment is a choice.
 
I don't think any of us would want to apply an enhancement technology like this when we're shooting for art. However there are many times I am just taking family snapshots at one event or another and getting more natural shots out of the camera is certainly preferable to spending hours editing or even batch editing shots with a tool like LR.

I think Apical understands that's the real utility of a technology like this.

--
Happiness is a want... Contentment is a choice.
 
I had no idea this would be such a controversial topic. :-)
I don't really understand what your point is in this thread; perhaps other people don't understand either so they don't know how to respond.

You say you've been comparing K-x shots with Olympus shots (camera(s) unnamed) but show images only from K-x. You talk of an editing system that you have applied in Olympus software on the K-x file but don't say if you have to apply in to Olympus files or if it is an in-camera option for Olympus.
The whole point of my post was to point out that the Pentax K-x shadow adjustment technology seems to render progressively flatter (lower contrast) images as you push the level of shadow adjustment. I believe this to be the case because the K-x shadow adjustment is a global curve adjustment that brings up the low end of the histogram globally. I don't know this for sure but that's certainly what it looks like.
You are quite right; it does. Anything that puts a wider DR into a given output range inevitably reduces overall contrast. However, this is what most users are happy with. As I said in the long section edited out here, I personally think the Apical version you posted looks worse than the in-camera versions.
On the other hand... there are technologies available that do a more complex local analysis of the image applying local curve adjustments. The human eye dynamically adjusts the thresholds of retinal cells according to the local illumination levels thus allowing us a rather large dynamic range. Apical and others are trying to better reproduce what we see and in my opinion are producing more convincing shadow adjustment technologies than global curve adjustments.
Mike has explained why some of these technologies can be dangerous (by reversing local brightness differences). The Apical site shows a landscape picture as an example of what it does which looks as if it suffers from exactly the problem Mike describes. To that extent it doesn't mimic the eye so if that's what Apical is trying to do it hasn't yet succeeded.

Your opinion of Apical is that it's better than a simple global curve; my opinion is that it's worse. As it's always posible to improve the simple version later but impossible to remove the Apical problems, Pentax is wise to follow the simple course.

--
---

Gerry


First camera 1953, first Pentax 1983, first DSLR 2006
http://www.pbase.com/gerrywinterbourne
 

Keyboard shortcuts

Back
Top