OMD Histogram Settings

Started Jun 23, 2012 | Discussions
Polariser Regular Member • Posts: 403
OMD Histogram Settings

Menu D allows you to change the histogram settings.

The default is Highlights 255 and Shadows 0

however, DPR suggests Highlights 250 and Shadows 5

"Better still, the threshold at which the camera indicates under- and over-exposure can be defined in Custom Menu D, 'Histogram Settings.' Depending on how you process your images, you're likely to want to experiment to find settings that suit you, but a setting of 250 - 5 usually provides a good safety-net"

http://www.dpreview.com/articles/9115179666/user-guide-getting-the-most-out-of-the-olympus-e-m5

Is it a good idea to use the default or DPR's recommendation?

What would be the effect in practice
?

 Polariser's gear list:Polariser's gear list
Sony RX100 Olympus PEN E-P2 Olympus PEN E-PL1 Olympus OM-D E-M5 Olympus M.Zuiko Digital ED 9-18mm F4.0-5.6 +9 more
Detail Man
Detail Man Forum Pro • Posts: 17,225
Re: OMD Histogram Settings

Polariser wrote:

Menu D allows you to change the histogram settings.
The default is Highlights 255 and Shadows 0
however, DPR suggests Highlights 250 and Shadows 5

Is it a good idea to use the default or DPR's recommendation?

May I suggest 248 and 7 as an alternative.

What would be the effect in practice ?

The 8-bit binary numbers 11111000 and 00000111 are perfect mirror images - enhancing feng shui

boggis the cat Veteran Member • Posts: 6,329
It is simply another tool that may be of use

Polariser wrote:

"Better still, the threshold at which the camera indicates under- and over-exposure can be defined in Custom Menu D, 'Histogram Settings.' Depending on how you process your images, you're likely to want to experiment to find settings that suit you, but a setting of 250 - 5 usually provides a good safety-net"

http://www.dpreview.com/articles/9115179666/user-guide-getting-the-most-out-of-the-olympus-e-m5

Is it a good idea to use the default or DPR's recommendation?

What would be the effect in practice
?

It simply shows the area in the shot that is possibly clipped (Highlights) or may have a lot of noise (Shadows). If you are particularly concerned about one or both then you can set the numbers to suit.

Olympus' default will only show areas that are likely to have clipped and not those areas getting close to clipping. DPR's suggestion will show areas that are still fairly far from any possibility of being clipped.

I used 3 and 252 (I think), as I wanted an indication of the "marginal" areas in the shot (or in the VF before taking the shot -- an advantage of the EVF).

 boggis the cat's gear list:boggis the cat's gear list
Olympus E-5 Olympus E-M1 Olympus Zuiko Digital ED 12-60mm 1:2.8-4.0 SWD Olympus Zuiko Digital ED 50-200mm 1:2.8-3.5 SWD Olympus 12-40mm F2.8 Pro +7 more
Detail Man
Detail Man Forum Pro • Posts: 17,225
Re: OMD Histogram Settings

Detail Man wrote:
May I suggest 248 and 7 as an alternative.

Polariser wrote:

What would be the effect in practice ?

The 8-bit binary numbers 11111000 and 00000111 are perfect mirror images - enhancing feng shui

Be aware that the numbers 252 and 3, which are represented in the binary number system by 11111100 and 00000011 are implicitly less lucky, as they exhibit an inferior feng shui factor ! ...

OP Polariser Regular Member • Posts: 403
Re: It is simply another tool that may be of use

Thanks very much

I will try your settings.

Like you I want to avoid clipping and this seems to be the best tool

 Polariser's gear list:Polariser's gear list
Sony RX100 Olympus PEN E-P2 Olympus PEN E-PL1 Olympus OM-D E-M5 Olympus M.Zuiko Digital ED 9-18mm F4.0-5.6 +9 more
Detail Man
Detail Man Forum Pro • Posts: 17,225
Re: Note that it applies to JPG recording only (as an estimate).

Polariser wrote:

Thanks very much
I will try your settings.
Like you I want to avoid clipping and this seems to be the best tool

Be aware that the "blinking highlights" indicator is an estimation of JPG highlight-levels. For RAW:

http://forums.dpreview.com/forums/read.asp?forum=1041&message=41246402

Sure, a bit of margin on the top might make some sense - as it is an estimate only, and is not derived from the actual JPG yet to be recorded, itself. Whether or not you consider some shadow-level to be undesirable is another story. Many, many scenes have lots of shadow-tones.

gollywop
gollywop Veteran Member • Posts: 8,301
Consider keeping it at 255.

Detail Man wrote:

Polariser wrote:

Thanks very much
I will try your settings.
Like you I want to avoid clipping and this seems to be the best tool

Be aware that the "blinking highlights" indicator is an estimation of JPG highlight-levels. For RAW:

http://forums.dpreview.com/forums/read.asp?forum=1041&message=41246402

Sure, a bit of margin on the top might make some sense - as it is an estimate only, and is not derived from the actual JPG yet to be recorded, itself. Whether or not you consider some shadow-level to be undesirable is another story. Many, many scenes have lots of shadow-tones.

Since DM has cited the above thread, let me take this opportunity to provide a minor correction. In that post, I indicated that the threshold setting appeared to have no effect on the blinkies, just the colored bars at the ends of the histograms -- that it appeared to be cosmetic. This is not so, the threshold setting does indeed affect the blinkies.

You can see this for yourself by looking at a review of a shot that shows after-shot blinkies when the threshold is set at 255, and then looking at it again after setting the threshold to, say, 245. There will be a larger blinkie area. The post shot blinkies, then, are determined anew with each reviewing of the image, and reflect the current threshold setting, which does make a difference.

Using a reduced highlight threshold, then, can provide good information for a conservative exposure. However, as I have gained more and more actual (in the field) shooting experience with the OM-D, I have found that keeping the threshold at 255 with UniWB leads to excellent ETTR results. Use CW exposure, compose your shot, and simply back off EC (if using A mode) -- or shutter speed or aperture (if using M mode) -- till the blinkies just stop in the brightest region. Then shoot. I was using a threshold of 245, but I found that was leading to a lot of unnecessary underexposure.

Sometimes, in doing the above, you will see evidence of blinkies in the post-shot highlights that did not appear as blinkies in the live-view (pre-shot) highlights. If this is the case, you might want to reduce EC one notch more (down 1/3 EV) and reshoot, just to be safe, but my experience here is that this has very little practical effect.

If you are not going to use UniWB, but follow some other ETTR suggestions, then I definitely advise against reducing the highlight threshold. If you use AWB, even when the threshold is set at 255, you will find, as a practical matter, that you will often be underexposing ETTR by 2/3 EV (or more). An example of this is found at

http://forums.dpreview.com/forums/read.asp?forum=1041&message=41669208

Let me also take this opportunity to correct an impression that has recently been given that using UniWB depends on the use of histograms. This is not so. Depending on the camera, one can use the colored histograms or the blinkies. With the OM-D, I have advocated (and do advocate) using blinkies with UniWB. The OM-D histograms do not show small counts very well, so that small blown regions, properly depicted by the blinkies, may not manifest themselves at all on the histograms (either the luminance histogram or the colored histograms). This means that a fairly dark shot with significant specular highlights can show a histogram that appears as a mound to the left, with a large empty gap to the right, even when the blinkies are properly indicating that there is clipping. The OM-D histograms appear to be singularly poor in this regard. With other cameras, the histograms may serve well. They are very effective with my D300 and even my Canon s90 does better.

The OM-D blinkies, however, are pleasantly reliable. The pre- and post-shot blinkies are not always the same, but, when they differ, the pre-shot blinkies tend to be no more than 1/3 EV less sensitive than the post-shot blinkies.

-- hide signature --

gollywop

-----------

Detail Man
Detail Man Forum Pro • Posts: 17,225
Re: Consider keeping it at 255.

gollywop wrote ...

Very good to see your expertise being lent to this important matter of dead reckoning in Dynamic Range. The fact that you recognize that matter of truth and accuracy in lournalism is an " oblinkatory " one, and not just a capricious matter, is most impressive, indeed. I am quite confident that at this point you are likely one of the world's eminent players in the blog and blogette-o-sphere, and I tender a stalwart confidence that in the coming days your eloquent tutelage shall elevate the resonanata of this renowned collegial body, most assuredly, indeed ...

I would launch into a propostion that OP consider 259 instead, but I shall save that for another day

Staccato gelato, old boy, and all that rot ! Keep the chin up and always remember, "bunga bunga"

Olymore
Olymore Senior Member • Posts: 1,860
Re: Note that it applies to JPG recording only (as an estimate).

A quick question regarding shadow tones.

Given that half the tones are in the highest highlight stop of data, three quarters in the the top two etc iis there that much data in the shasow tones ?
Or have I misunderstood ?

 Olymore's gear list:Olymore's gear list
Olympus E-M1
Detail Man
Detail Man Forum Pro • Posts: 17,225
Re: Note that it applies to JPG recording only (as an estimate).

Olymore wrote:

A quick question regarding shadow tones.

Given that half the tones are in the highest highlight stop of data, three quarters in the the top two etc iis there that much data in the shasow tones ?

Or have I misunderstood ?

Yes, you have, mightily ... What EVIL lurks in the hearts of men ? Only the "shasow" knows ...

Olymore
Olymore Senior Member • Posts: 1,860
Re: Note that it applies to JPG recording only (as an estimate).

I'm not sure the bard had digital sensors in mind but never mind.
You could expand on the answer a bit ?
Or perhaps a googling session might be more productive.

 Olymore's gear list:Olymore's gear list
Olympus E-M1
gollywop
gollywop Veteran Member • Posts: 8,301
Re: Note that it applies to JPG recording only (as an estimate).

Olymore wrote:

A quick question regarding shadow tones.

Given that half the tones are in the highest highlight stop of data, three quarters in the the top two etc iis there that much data in the shasow tones ?
Or have I misunderstood ?

No, you have not misunderstood. But there is indeed a question of how much information the camera is able to put in that bottom quarter. A camera with a limited DR will put very little, and pulling up the shadows will produce poor results. A camera with a decent DR, however, will still have a good deal of information in that bottom quarter. The OM-D (and some of the better FF cameras) have lots of information in that lower quarter, enough to allow multiple EV (3-1/2 or more) pull-ups while retaining a decent image.

But, to take full advantage, you really must ETTR. People pass off a 1/3 EV difference as inconsequential. But that 1/3 EV can make quite a difference as to what's been added to that bottom quarter.
--
gollywop

-- hide signature --

gdiver Regular Member • Posts: 422
Re: OMD Histogram Settings

Wow...you're really smart!!!

 gdiver's gear list:gdiver's gear list
Olympus XZ-2 iHS Panasonic ZS200 Olympus PEN E-PM1 Olympus E-M1 Olympus OM-D E-M1X +7 more
Detail Man
Detail Man Forum Pro • Posts: 17,225
Re: OMD Histogram Settings

gdiver wrote:

Wow...you're really smart!!!

Never underestimate the all-powerful feng shui factor, grasshopper. Symmetry rules the heavens !

Olymore
Olymore Senior Member • Posts: 1,860
Re: Note that it applies to JPG recording only (as an estimate).

Thanks.

Presumably though you are entering the realms of diminishing returns when as you have far fewer tone steps within the shadow data.

Or perhaps I'm getting confused because I'm thinking of film and the zone system where the each EV is evenly distributed ?

 Olymore's gear list:Olymore's gear list
Olympus E-M1
gollywop
gollywop Veteran Member • Posts: 8,301
Re: Note that it applies to JPG recording only (as an estimate).

Olymore wrote:

Thanks.

Presumably though you are entering the realms of diminishing returns when as you have far fewer tone steps within the shadow data.

Well yes. But if there's still plenty of information in those lower steps of shadow, you can still do a good job with it. If your camera only has 6EV of DR, those lower steps aren't going to do much for you. But if you have 11 or 12EV of DR, you can still use those lower steps quite effectively. To be sure, there's not as much detail as in the top half, but there's still plenty enough.

You seem to be thinking that it's a quarter of the steps, so it can't be very much. But a quarter of what? A quarter of a little bit ain't so much. But a quarter of a big bit can still be big. And that's indeed what happens with the OM-D.

Try it; you'll like it.

-- hide signature --

gollywop

-----------

gollywop
gollywop Veteran Member • Posts: 8,301
An example

Here is an example I've posted elsewhere of a handheld shot using UniWB, ETTR in a dark hut at Plimoth Plantation on an outing with some of the grandkids:

The original shot: exposed ETTR for the sunlit window, so there was really very little in that top one half. I used CW metering in A mode, composed the scene, dialed back EC until the blinkies in the window just diappeared (again, using UniWB), and shot :

Processed in ACR with shadows pulled up and highlights compressed:

This shot was pulled up over 2-2/3 EV, but I've pulled the shadows over 3-2/3 EV with excellent results. There can be plenty of information in those lower portions with a decent camera.
--
gollywop

-- hide signature --

Detail Man
Detail Man Forum Pro • Posts: 17,225
Re: Note that it applies to JPG recording only (as an estimate).

Olymore wrote:

I'm not sure the bard had digital sensors in mind but never mind.

The electromagnetic ethereal techno-bard speaketh: http://www.old-time.com/mp3s/shad39.mp3

OP Polariser Regular Member • Posts: 403
Re: Consider keeping it at 255.

gollywop wrote:

The OM-D blinkies, however, are pleasantly reliable. The pre- and post-shot blinkies are not always the same, but, when they differ, the pre-shot blinkies tend to be no more than 1/3 EV less sensitive than the post-shot blinkies.

Would you advise leaving these at the default settings of 255/0 then?

I use AWB and am just learning.... I find the LV Shadow/Highlight view to be easier to follow than the histogram view.

 Polariser's gear list:Polariser's gear list
Sony RX100 Olympus PEN E-P2 Olympus PEN E-PL1 Olympus OM-D E-M5 Olympus M.Zuiko Digital ED 9-18mm F4.0-5.6 +9 more
Detail Man
Detail Man Forum Pro • Posts: 17,225
Re: Consider keeping it at 255.

Polariser wrote:

gollywop wrote:

The OM-D blinkies, however, are pleasantly reliable. The pre- and post-shot blinkies are not always the same, but, when they differ, the pre-shot blinkies tend to be no more than 1/3 EV less sensitive than the post-shot blinkies.

Would you advise leaving these at the default settings of 255/0 then?

I use AWB and am just learning.... I find the LV Shadow/Highlight view to be easier to follow than the histogram view.

gollywop wrote:

If you are not going to use UniWB, but follow some other ETTR suggestions, then I definitely advise against reducing the highlight threshold . If you use AWB, even when the threshold is set at 255, you will find, as a practical matter, that you will often be underexposing ETTR by 2/3 EV (or more) .

Keyboard shortcuts:
FForum MMy threads