PDA

View Full Version : Dynamic range



Ken.
08-22-2008, 03:35 AM
Every magazine review makes a point about the E series dynamic range, or lack thereof. I'm not seeing this as an issue. I'm just curious: is anyone seeing blown highlights and weak shadow detail as a problem?

raven15
08-23-2008, 02:32 AM
Yes, I frequently wish for more dynamic range on the white end. That is my largest complaint from the IQ department. I have many times taken pictures of white waterfowl, and had a choice between seriously underexposing the background (thus losing nice reflections and leaves), or completely blowing out the birds, and losing all the feather detail. It is particularly irritating if they are far enough away as to be part of the scenery instead of the subject. I have a picture with two swans, and you can't tell which is which, it's just a white shape with two heads, the picture is otherwise beautifully exposed. I also wish my skies would tend more toward the blue than the white side at times. And quite a few other instances where I wish for more.

I shoot exposure bracketing enough to know the camera doesn't have far to go. Just 1 stop extra toward the white would take care of 50% of my problems, particularly with blue skies.

Of course, if they doubled the dynamic range, that would be cool too :) I would be quite happy with, say, 3 stops on the dark end and 4 on the bright end.

Beowulff
08-23-2008, 08:13 AM
Interesting comment Ken...

Most of the reviews of the E-520 that I've read have said something along the lines of "It boasts a greater dynamic range than the E-510 and that's something we can confirm. Many of our test shots showed how the E-520 can really hang onto those highlights while still bringing out plenty of shadow detail."

I guess highlight clipping is one of those difficult things to quantify, and everyone's gonna have a differing degree of what's "acceptable" or not? I agree that DR seems to be one of the major IQ issues that camera manufacturers seem to be ignoring; I'm assuming it's because of the R&D costs of better sensors (which ultimately limit DR).

Cheers :)

Ken.
08-24-2008, 07:56 AM
I don't find it to be much of an issue but it may well be my shooting style and training and my software. I tend to like inky shadows. There's a fine line where a shot can be too perfect, too technically correct and dull as hell. I've done thousands of those shots. Very boring, very "me too". I know my setup well enough to make it work. Then again, the dynamic range means not a heck of a lot for printing. Anything too light or dark will tend to be a problem anyway.

raven15
08-24-2008, 02:58 PM
One thing I wonder about is altitude. I generally live and shoot at a fairly high altitude, pretty much never below 4,500 feet, and just as likely to be over 8,000 feet. I think this makes for harsher lighting, with less atmosphere to dull the sun.

Training... yes, but I tend to make a whole lot of shots of contrasty subjects varying the exposure on each, and have found that frequently I still can't get it right.


True, I don't usually print. It is true clipped highlights on small objects such as glass, leaves, etc. wouldn't be noticeable. However, I think white skies and entire birds are easier to spot.

Razr
08-25-2008, 09:35 PM
"Dynamic range" is much ado about nothing.
Moreover, DR is a "digital" term, bearing little weight in the overall practice of the craft.

Photography demands competent operators and proper equipment and the knowledge to properly expose a scene or venue.

Improper use of the viewfinder by neophytes (how many new shooters ever check the four corners before they shoot?), overly aggressive use of ISO, not knowng the limitations of their gear leads to DR blowouts.

So few of today's shooters use filters** for example, thinking they can do every thing to "fix"*** their work in post-processing.
**The properly used filter for example, will absolutely control most DR blowouts. You don't use filters at your tedious, post-processing peril.
***The very idea we cannot shoot digital without doing at least some post-processing should alert many to the need to use pre-production tools like filters, camera supports, lens shades-etc.

raven15
08-26-2008, 09:44 PM
So, is there any way to correctly expose this swan without making the grass look like twilight, or bringing up noise? If there is, I''d like to know.

In my estimation the lawn is currently underexposed around 1 stop. Post processing might work, but this bird is completely blown, I'd be underexposing the lawn around 3-4 stops to get it right. At that point noise would be visible on the lawn if I brought up the exposure. Granted ISO was 400, so there'd definitely be noise, but in my experience even ISO 100 would suffer serious noise in dark areas brought up three stops. Assume the lawn is ok being underexposed 1 stop, like it currently is, so really only 2-3 stops are necessary to get an acceptable swan

Settings:
E-410 with 14-54mm lens
ISO: 400
Aperture: f/3.5
Shutter: 1/1250 s (I know at that point I could have easily used ISO 100, but I didn't so that's how it is)
JPEG shot in natural mode with contrast and sharpness both -2, saturation +1

If this is a case where you answer "don't take pictures of swans on a sunny morning" then I say, I am quite familiar with my gear and how to use it, but I want 1 extra stop on the white end.

But, if you know how to fix this as-is, I'd be very happy to know.

raven15
08-26-2008, 10:01 PM
I am actually not very concerned with this issue, don't let me seem to be blowing it out of proportion. However, since he asked if anyone had a problem with dynamic range in their shooting, I have to say yes; here is my example. It is not a problem in about 90% of my shots, and is only critical in under 5%. It may be my high altitude is a contributor, I just learned than film photographers shooting at altitude sometimes "pull" (or push?) their shots in the dark room when working at higher altitudes. Of course, my tendency is to push my gear in every direction, so if you ask about any given thing chances are I have experienced it. In fact, just by opening the box I usually violate the recommended operating environment on all of my electronic gear, and I only press its limits further from there.

It is true extra dynamic range is not required to get good pictures, but in my experience it is required to get good shots of certain things. Which can be invaluable if it is not practical to replicate the situation. Just like people who say you don't need more than a 50mm lens to get good photographs. Those people are definitely not wildlife photographers, and clearly don't shoot a variety of other subjects either. It all depends on what you want in your frame.

Razr
08-27-2008, 01:36 AM
So, is there any way to correctly expose this swan without making the grass look like twilight, or bringing up noise? If there is, I''d like to know.

The easiest way was to spot meter the Swan and let the rest of the shot go where it would: at least you'd have the Swan.
The lack of contrast in the venue was also working against you:
medium going on low contrast, which didn't help.


In my estimation the lawn is currently underexposed around 1 stop.

Stop "guesstimating" and trust your gear. Spot metering would have made your need to "guesstimate" moot.


If this is a case where you answer "don't take pictures of swans on a sunny morning" then I say, I am quite familiar with my gear and how to use it, but I want 1 extra stop on the white end.

You would be and are dead wrong. You were shooting in a low contrast venue, the bird the only well-lit subject.
Again, spot metering the Swan would have more correctly exposed the shot.


But, if you know how to fix this as-is, I'd be very happy to know.

The Swan is overexposed and thus all details were and are unrecoverable.
As already noted, there are times when spot-metering is demanded and that was one.

Eric A
08-27-2008, 06:40 AM
Anybody who whines about dynamic range never had to make a living shooting color transparency film.

raven15
08-27-2008, 07:53 AM
Actually, it's not a guess. I metered the lawn then used -1 stop EV to attempt to get the swan. As I said, I wanted both visible. I have tried metering the swan as well, but I wanted to see what was behind it instead of inky black, so here is the result.

My other reason for saying this is that I sometimes shoot a +1/-1 exposure bracket. I have never used the +1, only the 0 or -1. True, that is metering to the whole scene, but that's what first brought it to my attention.

Ken.
08-27-2008, 07:21 PM
Anybody who whines about dynamic range never had to make a living shooting color transparency film.

This took me by surprise. How true that statement is. I kept a supply of film from just about every vendor because no one film covered every base. Ultimately I used Fuji's films since it was damn hard to blow the white areas out.

On the digital side I found too many cameras were emulating chrome film when shooting JPEG and it's now why I shoot RAW almost exclusively.

Ken.
08-27-2008, 07:30 PM
Actually, it's not a guess. I metered the lawn then used -1 stop EV to attempt to get the swan. As I said, I wanted both visible. I have tried metering the swan as well, but I wanted to see what was behind it instead of inky black, so here is the result.

My other reason for saying this is that I sometimes shoot a +1/-1 exposure bracket. I have never used the +1, only the 0 or -1. True, that is metering to the whole scene, but that's what first brought it to my attention.

It's an interesting shot and a scene I come across often down by the river with gulls. I meter for the gulls, let the background go where it goes and post process retouch with Lightroom. Adobe's RAW conversion is a bit different than Olympus's.

When I have a free day I may try some HDR trickery for my shots. I'm not fond of heavy digital darkroom work.

Razr
08-28-2008, 02:25 PM
Actually, it's not a guess. I metered the lawn then used -1 stop EV to attempt to get the swan. As I said, I wanted both visible. I have tried metering the swan as well, but I wanted to see what was behind it instead of inky black, so here is the result.

My other reason for saying this is that I sometimes shoot a +1/-1 exposure bracket. I have never used the +1, only the 0 or -1. True, that is metering to the whole scene, but that's what first brought it to my attention.

Spot metering the Swan would have brought it in regardless of other camera settings.
-30-

raven15
10-23-2008, 07:42 PM
Sooo....

I was just showed a new way to increase dynamic range. All you need is a raw file with dynamic range issues, photoshop, and a little time on your hands. Now, when you first open the RAW file in photoshop you will see a little slider bar that says "recovery." It refers to highlights, and maximizing that bar can gain you around 1/2 to 2/3 stops of highlight recovery.

But the real way is to open the RAW file twice. One you treat like you normally do. The other, you under expose on the raw slider bar (when you open it) by 4 stops, and reduce the brightness and contrast by as much as required. Paste the normally exposed one over the under exposed one on different layers. Then, select the eraser tool and set it to about 10% effectiveness. Start carefully erasing the areas with highlight problems, if the eraser has a broad enough fade radius you won't even be able to see the boundary between the two.

I believe you can gain about 3 stops into the white using this method vs. jpeg ! :eek:
In fact, because I have a hard time quantifying this (just learned an hour ago), I'd say anywhere from 2-6 stops would be just as likely. Don't try it on the blacks though, things get noisy.

Granted, it'd be nice to not go through the trouble, but I do have a new grad ND filter for use on landscapes.

JohnKK
10-24-2008, 04:21 PM
[QUOTE=Beowulff;323820]Interesting comment Ken...

Most of the reviews of the E-520 that I've read have said something along the lines of "It boasts a greater dynamic range than the E-510 and that's something we can confirm. Many of our test shots showed how the E-520 can really hang onto those highlights while still bringing out plenty of shadow detail."

I guess highlight clipping is one of those difficult things to quantify, and everyone's gonna have a differing degree of what's "acceptable" or not? I agree that DR seems to be one of the major IQ issues that camera manufacturers seem to be ignoring; I'm assuming it's because of the R&D costs of better sensors (which ultimately limit DR).





Guys.....

If you are talking about highlight DR....the E520 JPEGS hold onto about (depends on what tests you look at) +2.8 ev (but certainly under 3.0)above the mid grey, which does trounse the E510 JPEGS.....however, RAW files of E520 are also capped at +2.8 ev whereas default RAW conversion with ACR or any other good convertor will yeild about +3.0 ev from the E510.

Hence the issues with E510 & highlight DR vs E520 are that of tone curves for JPEG conversion.

I love my E510 but anyone whom claims the highlight DR problem is a metering issue is just being "difficult"......the white swan example is perfect....ran into it a pile of times when something bright is not the main feature.

For white skys in landscapes though there is a simple solution....p/u a B+W Graduated Neutral Density Filter (No. 502 - 2 stops)....I don't leave home withouit it.

Regards;

John K

Ken.
10-25-2008, 10:07 PM
I've lost track of this topic. I find the dynamic range "thing" to be a bit of a non-issue since I shoot primarily raw and my taste in exposure runs to dark images. What I have found is that ISO 100 is a bit off by about 1/3 to 1/2 stop. It's more 125 - 150. But then again it could just be my shooting style. I still compose color as though I'm composing for monochrome. Old habits, and all that. I have been dabbling with LREnfuse and Lightroom.

I'm beginning to reevaluate the whole JPEG thing. By now, JPEG's certainly should be a heck of a lot better than they are now. This standard has been around forever. I think the manufacturers are going a bit off the rails with their conversion algorithms.

JohnKK
10-26-2008, 11:04 AM
dpreview and others have tested this thoroughly and ISO 100 on the E510/410 is actually ISO 125.....sharp observation!

John K

Ken.
10-26-2008, 06:20 PM
You'd think a firmware update would correct this by now.

JohnKK
11-02-2008, 03:53 PM
you'd think they would have issued firmware update for ISO correction as well as the steep JPEG tone......but go figure......IMHO OLY really is negligent to it's SLR users in the firmware department compared to say Pentax or Sony.

.....its quite disappointing.....I'm not sure if they just thought E510 users would abondone them by the hords for the really minor tweaks of the E520.

They really would have been better off by issuing some firmware updates, publicizing the heck out of them and releasing a significantly improved model at Photokinia.

OLY obviously doesn't pay much attention to user forums or pro-reviews for that matter.

Cheers;

John K

Ken.
11-02-2008, 08:02 PM
As megapixels have climbed, actually pixel density, the quality has dropped. This is happening with all manufacturers. We've reached a point of diminishing returns. JPEG was never designed to be a high quality format and that's the reasons I look for cameras that can shoot RAW.

Overall, I like the E510. The E520 body is not different enough to matter to me.

I've contacted Olympus several times since getting the E510. It's definitely not like the OM days...

JohnKK
11-03-2008, 02:15 PM
I'm not trying to be smart but what are the "OM days"....perhaps I'm showing my ignorance, age or both.

Thanks JK

jeisner
11-05-2008, 04:08 PM
I expected an issue but haven't really had one.. though I did move from the Pentax K10d to the e3 and the e3 has MUCH more DR than the K10d.. Actually if you believe Dpreview's tests the DR is almost the same (like 0.2 of a stop different) to the K20d which isn't half as put down for it's DR.. Having used both now I would agree they are about equal...

kgosden
11-05-2008, 08:37 PM
The OM was the classic Olympus film SLR.

JohnKK
11-06-2008, 03:24 PM
my first camera was not nearly so impressive......flat rectangular silver & black Kodak 110 with a cube flash.....I still look for on ebay to this day.

Ken.
11-09-2008, 08:23 AM
We were talking about this yesterday. My first camera was a Brownie! It was big time when I hit 35mm rangefinders. SLR's were pretty crude when I started out. I had a Topcon Super D, a brick of a thing, but big time then. When the OM1 came out it was big news. In perspective, the OM would be like the E420 today. Small, fast, quiet and rugged. I still look for Rollei 35's, my favorite camera next to Leica's.