Getting best Dynamic Range with OMD.

Started Jan 17, 2013 | Discussions thread
Anders W Forum Pro • Posts: 21,468
Re: But not exact ...

Detail Man wrote:

Anders W wrote:

Detail Man wrote:

Anders W wrote:

Detail Man wrote:

dcassat wrote:

Good question DM, unfortunately the answer is "sometimes."

Obviously the processing is not entirely linear or I would be able to duplicate my successful results but in low light or low contrast situations hitting the mark usually results in slight under or slight overexposure.

I'm rather careful in my processes and have tried time and again to get absolute results but I feel there is some interpolation occurring that I cannot control.

I had read some statements that the Live Histogram and the Blinkies threshold appeared to (both) change with contrast-ratio of the previewed scene. But that did not make much sense to me.

That the Live Histogram would show the (JPG preview referenced) "luminance" regardless of whether the Blinkies threshold was being somehow altered (due to contrast-ratio of the previewed scene) would make more sense. While all Live Histograms appear to auto-scale the Y-axis scaling, it would surprise me if the X-axis scaling would be altered by the contrast-ratio of the previewed scene.

One might think that the E-M5 Live Histogram (with a UniWB configuration) could potentially be used to indicate maximum (linear) recordable RAW-channel levels. GW appears to have indicated that it seems to him that the Y-axis (auto) scaling is not sufficient in order to cause the display to light-up "dots" to warn of the existence of upper highlights.

Don't know about UniWB. But with auto WB everything works as it should with the LV blinkies as well as the LV histogram as long as you are not in one of the peculiar scenarios described here where neither works as it should.

Thanks for the link to that post where you described certain scenes where the amount of contrast affects the Blinkies thresholds. I do not at all understand why the 100% full-scale reading of a Live Histogram would change as a result of scene-contrast. Yes, the Y-axis amplitude is commonly scaled on all Live Histograms in order to make them more visible. And yes, different previewed scene contrast-ratios are going to lead to differing histogram displays.

Exactly why the live-view histogram (and live-view blinkies) behave the way they do in certain special situations is still not perfectly clear to me (except that it is related to the fact that the camera tries to keep the EVF/LCD brightness within certain brightness limits, i.e., scene visibility has priority).

At any rate, thanks to a prior discussion with you here, I bothered to check how the LV histogram actually works on the E-M5 (I usually don't use it) and found that it behaved in perfect correspondance with the LV blinkies, including the special cases where the LV blinkies fail.

I have read one poster on this forum assert that the E-M5 has a shooting mode setting whereby the previewing can be caused to reflect user-selected F-Number and Shutter Speed. Is that the case ?

There's a preview function whereby you can stop down to the selected aperture, yes. Shutter speed doesn't come into play here as far as I can see. I use it (assigned to the video button) but for DoF preview only.

If such is the case, can it then (also) true that the preview is (still) gain-ranged based upon some sort of measurement of estimated scene contrast-ratio ? Such would seem to be a contradiction in terms.

Yes, preview appears to be gain-ranged as well. The impact on the LV histogram (I just tried) is such that preview is of no use for the histogram reading (i.e., the histogram changes when you preview at an f-stop other than wide open). Ordinary metering seems to work in preview though (no change in the reading).

However, when it comes to what maximum "luminance" (which appears to be largely Green channel, notably less Red channel, and almost no Blue channel image-data derived from the gamma-corrected RGB channels of a JPG-referenced preview mapped into the selected sRGB or Adone RGB color-space) results in a 100% reading on the Live Histogram display, I do not understand how that has (or should have) anything to do with (relative) scene-contrast (unless the X-axis scaling is being dynamically manipulated by the camera) ?

It is true that one may end up missing some lit-up "dots" due to the amount of Y-axis scaling. That is a vertical display resolution issue in itself which may influence interpretation of Live Histogram display.

But, if the X-axis of the Live Histogram is actually being dynamically manipulated as a result of some sort of measurement of overall scene-contrast, then just what kind of Live Histogram is that ?

Perhaps one as seemingly (if not more) unreliable and close to useless as those which we have previously discussed in cameras that lack the "Constant Preview" functionality (of the GH2 and GH3) ?

That's a shame, indeed (if so). My GH2 Live Histogram can (under some circumstances, due to the auto-scaling of the Y-axis) indicate as few as 10 photo-sites - which is a very senstive indicator.

 Anders W's gear list:Anders W's gear list
Panasonic Lumix DMC-G1 Olympus OM-D E-M5 Olympus E-M1 Panasonic Lumix G Vario 14-45mm F3.5-5.6 ASPH OIS Panasonic Lumix G Vario 7-14mm F4 ASPH +28 more
Post (hide subjects) Posted by
(unknown member)
Keyboard shortcuts:
FForum PPrevious NNext WNext unread UUpvote SSubscribe RReply QQuote BBookmark MMy threads
Color scheme? Blue / Yellow