k20d Expanded Dynamic Range opinions

[photog]

pentaxpoke, you asked:"I'm still at a loss to see the 200%. Wouldn't a 200% increase mean a 2-stop increase in DR?"
Surely 200% means double or 2x the original? 1 stop is also 2x (or 1/2) the original.
Allowing for the usual poetic license that all manuals tend to use (or is it wishful thinking?) 1/2 to 1 stop is not too far off the truth.
I must say that I am more concerned over the finding on a number of sites that there is slightly less dynamic range in normal mode.
Until this has been resolved my old K-20D is safe!
Ron

[DAZ]

I don’t like the use of percentages like 200%. It is an un-defined number with no real meaning. 200% of what? My guess is this. If the normal range is 100% than 100% more is 1 stop more (because stops are logarithmic 1 stop is twice the dynamic range) so 100 plus 100 is 200%. Saying 200% more is not the way Dilbert would say it but it is the way his pointy haired boss would.

Is 1 stop important? Would you pay $1000-$2000 more for it? Put another way, if we say the camera has a 9 stop dynamic range adding 1 stop makes it 10 stops. Doesn’t seem like much. Now if it were 14-15 stops total that would be something. But one stop is what some people are screaming for, begging for, and are just waiting to pay $2000 for. If you go from APS-C to FF (all other things being equal) 1 stop more dynamic range is about what you get. Put that way 1 stop is not bad for something you get for basically free.

OK, is it good for anything? For me yes, it is just one more tool in the camera box that I can make use of. I shoot RAW but if someone is shooting JPG then it may not be much use to them. Not that you can’t use 200% D-range in JPG but you can get about 1 stop more range just using RAW instead of JPG. If they get what they want with JPG they may not have a need for more.

Some thing that it accomplishes is subtle like in this photo Flickr Photo Download: DAZ_0362. When I PP it was basically simple with no extra work.

With this photo Flickr Photo Download: DAZ_0346 I had to work at it much more and I am not totally satisfied with it. Same lens at about the same time.

Just using the 200% D-range made it easer to keep the sky colors smooth and bring up the shadows a little to make them interesting. I like the second photo better but that is because of the composition with the trees then the sky.

Now with this photo Flickr Photo Download: GreenBridgeDAZ_0393 I did not use 200% D-range. Using 200% D-range would not have added anything and may have made things a little worse. Maybe only subtlety worse but a lot of photography is about subtlety.

So like any other photography tool it can be of some use in the right situation and a detriment in the wrong one.

DAZ

[Sergei]

This has been an interesting thread and exercise. good points made by everyone.
I found that under the pressures of a wedding shoot, is not the best time to experiment with tests
Managed these two shots at 800 ISO, both at f8 and 1/60.
A lot different conditions than the controlled ones I did at home.
Results are there, but appear even more subtle, judge for yourselves.

Both shots resized only no USM or PP, full frame.
First OFF and second ON.


Thanks for all the comments to previous post.

[WMBP]

Sergei's pictures seem to me like good test shots, but I confess I can't see much of a difference. I just spent the weekend taking portraits outside in afternoon light. Very problematic exposures. I used the extended dynamic range on the K20D, but to be honest, I can't say that its effects are significant enough to be noticed. It certainly does NOT make it impossible for me to blow the highlights!

I was shooting raw (DNG), if it matters. I'm still uncertain whether the feature is designed only to help your in-camera jpeg shots, or whether it's supposed to affect the raw capture, as well.

My very tentative sense at the moment is that this is a bit of a gimmick. It may be doing something, but it's very far from obvious to me.

I will continue to be interested in the feature, though, and I'd like to be wrong. Perhaps I should be doing something else that would help this feature work properly.

Will

[PentaxPoke]

I agree with you about the gimick. There are now more examples of this feature here on Pentax Forums than I have seen cumulative through searching the web. The effect is almost unobserveable. It does affect the actual exposure because the RAW file is changed.

A 100% increase in DR would be a 1 stop improvement. 200% would be a 2 stop improvement. 400% would be 3 stops, etc. Based on the histograms the EDR does not even come close to a 1 stop improvement in DR, so I am still at a loss as to what Pentax means with the 200% number.

Thank Sergei for taking the time to do this during a critical shoot.

Anyone else with more examples? Did anyone look at my examples? Are we missing something?

[DAZ]

Sergei did not have time to play; he had a job to do. I hope he had some fun anyway.

I am going to speculate here. Sergei metered and exposed for that great looking mural on the wall. He took the shot and turned on the D-Range and used the same settings for the next shot. This is a great example that if you use the same settings you will get the same photo. Now if Sergei had had time he could have tried this. Change one of the following: F5.6, 1/30 or ISO1600. That would have been my preferred order of changes but it would have been totally up to Sergei. This may have let him keep the mural and bring up the rest of the dark church. D-Range will not make a curve change so the photo is going to look the same if you don’t clip when the D-range is off.

I have been told that I make a great Monday morning quarterback. I don’t under stand that, as I never like watching football.

In a way this is different from the way we learn to take photos. What ever you do DON’T clip! With D-Range when you start to clip then turn D-Range on. If we do this then we may see some change but if we don’t there is no change to see.

Some have complained that D-Range adds noise to the dark parts. Sergei has shown that the noise is small and hard to see.

Thanks for the great example of this Sergei.

DAZ

[ivanp]

Comparing two pictures sent by Sergei, then on the picture with DR-ON I can see slightly more details in the dark parts, and at the same time better quality of the two brightest "candles" on the left wall (they seem overburnt with DR-Off). Which is exactly the expected result - overall exposition is slightly higher to enhance dark parts, and too bright parts are exposed using dynamically lowered ISO.

[Okami]

Hey guys, Do you think the camera does something similar to the Recovery slider in Lightroom?

[PentaxPoke]

Originally Posted by Okami

Hey guys, Do you think the camera does something similar to the Recovery slider in Lightroom?

Interesting theory. If it does, it is a very slight recovery.

[WMBP]

Originally Posted by PentaxPoke

Interesting theory. If it does, it is a very slight recovery.

Yeah, I agree with the Poke. The recovery slider in Lightroom allows me to recover a significant amount of detail from ostensibly blown highlights. I say "ostensibly," because they're obviously not really blown or recovery would be impossible, and when they are really blown, moving the slider to 100% still doesn't recover everything.

But the differences I see between EDR on and EDR off photos, here and in my own tests, are so minor as to make the feature hardly worth talking about. It might be recovering something, but it ain't recovering much.

Really, I keep hoping that we get some other info from Pentax that proves us all wrong, you know, like you're supposed to use this feature only if your aperture is smaller than f/8 or something weird like that.

Will

[PentaxPoke]

Exactly WMPB. I feel the same way. I have been just hoping that someone would show me I am doing something wrong, but I think we have so much evidence now that you are correct in that the effect is too minor to be even considered a feature. This camera is so wonderful, but I really thing the EDR must be some marketing gimick to compete with the enhanced dynamic range on the Sony.

[Sergei]

Good morning from Down Under,
just started teaching a new evenng course last night, so still a bit zonked, but a quick read of this interesting thread has brough to light (no pun intended) some interesting issues. Daz makes some very good points, and you're right, I had precious little time to think through the issues you raise, will have to give it a look on the weekend.
One point I notice, to the left of the frame, where there is a cross and two lamps and some colord light spill, in the 2nd shot I can see more detail in the bulbs and the actual red highlights are more intense.
Exactly the same exposure value used on both shots.
So it appears again that any benefit, favours highlight recovery. I agree with another comment that very faint detail is better defined in the shadows on parts of some of the murals.
Thanks for all the replys, and yes;

I did have fun, it was a big Russian wedding, 12 bridesmaids etc, I know the grooms dad for yonks, Vodka flowed, need I say more, except that Sunday nights and wedding partys don't always mix
Might post a selection when I get a chance.

[PentaxPoke]

I read a theory on how this EDR might work and I wonder if any of you might comment on it. The theory is that when you turn EDR on, it internally drops the ISO of the camera by 1 stop, thus underexposing the image. It then "pushes" the exposure back up to a point where you don't get clipping.

That makes the most sense of anything I have read. It is also cinsistent with Pentax's warning that EDR may increase shadow noise. That makes sense since you are underexposing the image, and "pushing" ISO after the fact increases noise. It also makes sense why when you turn on EDR, min ISO increases to 200; since if you set ISO 200, the camera would really be taking the picture at ISO 100.

So if this is true (and it is the most consistent theory I have seen yet) the EDR feature really doesn't increase DR. What it does is prevent clipping.

[WMBP]

Originally Posted by PentaxPoke

I read a theory on how this EDR might work and I wonder if any of you might comment on it. The theory is that when you turn EDR on, it internally drops the ISO of the camera by 1 stop, thus underexposing the image. It then "pushes" the exposure back up to a point where you don't get clipping.
... So if this is true (and it is the most consistent theory I have seen yet) the EDR feature really doesn't increase DR. What it does is prevent clipping.

Poke,

I have heard a similar (perhaps identical theory) somewhere else, probably here in this forum. Can't remember where. It's an interesting theory. I'm unable to evaluate it and I do wish Pentax Inc would throw us a clue about this feature.

But in the meantime, I have two responses, no, three.

First, whatever the EDR feature does, it definitely does NOT prevent clipping. I've been using it a lot lately, in photos taken on bright days, and I still end up with clipped highlights. Perhaps it reduces clipping, but it sure does not prevent or eliminate it.

Second, if all it does it push the captured data up one stop, heck, I can do that myself on the computer, can't I?

Third, if the answer to my second question is yes (I can do it just as well on the computer), then it doesn't seem like much of a feature.

Will

[PentaxPoke]

Thanks Will. Check this out. I just found it:
Pentax K20D review Cameralabs features: lenses

Scroll down and they discuss the EDR. Seems to support what we are saying.

"Expanded Dynamic Range" must have been a term coined by someone in Marketing, not Engineering.