Hi Dave, some of us have already walked through the paths you are following time ago. What you call ERADR is not DR but highlight headroom between the camera JPEG and the RAW file, and will strongly depend on camera processing. This includes adjusted WB, output color profile, contrast/saturation settings,... Also the kind of light will have an influence becase the spectral components will impact in the RAW channel distribution itself.
More logical than measuring "JPEG vs RAW" headroom, is to measure the "Metering vs RAW" headroom, i.e. know how many stops you can add to the exposure meter 0.0 point before having actual RAW clipping. Still the scene lighting will have a strong impact here.
E.g. RAW file metered over a gray uniform surface under tungsten light and exposed accordingly:
RAW histograms:
Canon 350D:
Canon 5D:
Even with that reddish light the G channel captures photons the most, having around 3,5 stops of RAW headroom vs camera metering (apparently a bit more in the Canon 350D, but remember we are exposing in 1/3EV steps so there is room for a slight deviation).
Regards
Ye Gads...Take off for the weekend and return to find everyone's been busy refuting much of what I may have long held as carefully considered reliable opinion, if not actually irrefutable fact!
Guillermo states:
" ....some of us have already walked through the paths you are following time ago. What you call ERADR is not DR but highlight headroom between the camera JPEG and the RAW file, and will strongly depend on camera processing. This includes adjusted WB, output color profile, contrast/saturation settings,... Also the kind of light will have an influence because the spectral components will impact in the RAW channel distribution itself."
I do find it strange that, having trudged these same paths for so long, we have arrived at such differing conclusions. Being firmly of the opinion that if something swims, flies, and quacks like a duck I am highly likely to conclude it to be not a zebra, but in fact , a duck.
Hence, what some peremptorily proclaim to be merely highlight "headroom" tucked discretely just past the JPEG frame, I (and others of the EBTR persuasion) have concluded to be a fully
functional patch of extra raw-accessible dynamic range that does, in fact, characterize the complete raw dynamic range, the full employment of which is the aim of those hoping to capture files of image data of the highest possible quality, without sacrificing either the tonal or chromatic spectra that are the legacy of their bit-depth. Certainly, somewhere out near the ragged edge there is a zone of functional "highlight overhead" incidental to one or two still striving and not-quite-blown color channels. But...out to that point, dynamic range seems still to be dynamic range. Easily performed exposure tests readily confirm its presence and amount.
The actual amount of the unacknowledged and formally unexplained extra DR clearly varies from sensor to sensor. The most reasonable explanation, may likely be the pervasive phenomenon that besets the semiconductor production industry, "process variance". Of the twenty or twenty-one cameras I have helped test none have had less than 1/3 stop of ERADR, the great majority have at least one full stop of ERADR, and the greatest ERADR I've personally tested was two and 2/3 stops! Rumors of a camera with three full stops of ERADR are bruited about. Now, given that even 1/3 stop of ERADR can demonstrably reduce the amount of noise captured, it seems unconscionable that even greater amounts of dynamic range should be allowed to languish incognito... incommunicado...and worst, un-used.
We users of ERADR via EBTR have long seen the term "headroom" being used as a means of subtly diminishing the significance of DR that exceeds that allotted to the JPEG format image data files, and making it easier to sweep that irregular excess" "under the carpet" so it need not be dealt with.
And so, in accordance with the advice of Michael Reichmann, Bruce Fraser and Jeff Schewe to recognize the exposure of raw data as a distinctly different, unique process from that used to expose film and the 8-bit depth JPEG image file. Those required...and still demand to expose a detail of a given value in order to specifically render its same value in the print.
On the other hand, raw image data is a revolutionary, different imaging medium requiring exposing each value with the maximum number of photons proportionate to the brightness of each in their ultimate, brightest possible image, and to the extent of coming as close as possible to blowing highlight detail...without actually doing so. And this is accomplished ONLY by utilizing the absolute maximum available dynamic range. Leave out the brightest possible stop of exposure sacrificed 59% or more of the tonal spectra available to,the image! Leaving out the send-brightest stop of exposure sacrifices 25% more of the image’s tonal allotment.. So if your camera has two full stops of un-used ERADR, you have relegated your final image to a mere 25% of its brightness and color tones!
The brightest and darkest image values are exposed proportionately, regardless of whether exposed darkly (to the Left, wher the maximum amounts of noise are captures) or exposed brightly and as far to the right as possible.
The correctly exposed raw image capture, that can only be accomplished by full use of each sensor’s individually determined ERADR, is an imaging resource that far exceeds the potential
of the oft’ mentioned photographic negative, or even of the unprocessed latent image. The creative, artistic potential of the properly exposed raw image file is most like an infinite collection of the greatest number of un-processed latent image imagineable! Now THAT’S one helluva creative resource. The phototographer’s previsualized image is, of course, guaranteed. As are an almost infinite variety of other creative imaginings!
And much of Guillermo's statement has merit but for one exception:
It fails to mention that among the listed confounding factors should also be included the significant effects of "process variance" that plagues production of a wide variety of semiconductor devices.
It would seem that the observed ...and productively utilized ...phenomenon of additional
stops of exposure "beyond the right" are supposed to be deemed imaginary (?) because some results from RawDigger/FastRawViewer imply that such additional stops don't exist? Indeed...problematic! ...particularly when RawDigger results are used to reach the opposite conclusions by others.
http://diglloyd.com/blog/2012/20121213_1-D800E-ETTR.htmlGuillermo has cogently suggested that:
"More logical than measuring "JPEG vs RAW" headroom, is to measure the "Metering vs RAW" headroom, i.e. know how many stops you can add to the exposure meter 0.0 point before having actual RAW clipping. Still the scene lighting will have a strong impact here."
Indeed, some adherents of EBTR do as Guillermo suggests, and are fully as successful as those of us who use the "ETTR" exposure as our touchstone for adding the ERADR stops at gaining full use of their camera's available extra raw DR...be it 2/3 stop or two and 2/3 stops (or more).
Well, if, as I now hear, all that ERADER that's been perceived to exist and been used as DR for the past decade or more is simply "headroom" of ragged ends of saturated color channels just playing their hearts out in the role of extra dynamic range. Well, if that's actually the case, all that ERADR oughta get an Oscar...or a golden Globe at the very least. Certainly, out at the tattered edge of its effort it becomes obvious even to the untrained eye by resultant muddied hues that it gets to the point where merely two...or maybe one color channel is indulging in a masquerade, but, depending on the camera, there may be anywhere from 1/3 stop to two and 2/3 stops .(or more) available before that "tattered edge" is obviously reached.
The point that those of us long-time-users of EBTR ( Expose Beyond the Right) have relied upon is that from-camera-to-camera it is, on a practical basis, impossible to predict without individual testing how much of that seemingly perfectly valid dynamic range of exposure actually exists beyond and to the right of the clipping signal(s) at the right end of the manufacturer's selected JPEG histogram frame...that ostensibly represents the sensor's dynamic range.
Some of us who use EBTR got started on that course with the late Michael Reichmann's essay "Expose Right" here in LuLa in 2003 and were, in less than a year, strongly and pointedly encouraged by the late Bruce Fraser's "Raw Capture, Linear Gamma, and Exposure" Adobe White Paper, in 2004
https://www.adobe.com/digitalimag/pdfs/linear_gamma.pdfAnd, with time, additional fuel has been added to the supportive fires by a well-accomplished and most inquisitive photographer, one of my friends, Bob Malarz (who is included under the non-specific "we" when I refer, in a non-specific manner, to those who practice EBTR) who posted an excellent example of inadvertent utilization of ERADR:
http://www.uglyhedgehog.com/t-133857-1.htmlAnother thoughtful and knowledgeable photographer of reasonable note, Jeff Schewe, has published another classic example of accidental, inadvertent EBTR in
"Un-debunking ETTR"
http://schewephoto.com/ETTR/In fact, Jeff's example of accidental EBTR has undoubtedly sparked the interest of more than a few to pursue testing their own their own cameras for "whatever it is" beyond the right of the JPEG clipping spot that does a superlative job of standing in for extra raw-accessible dynamic range when, to our dismay, we hear that extra raw-accessible dynamic range being proclaimed not to exist.
And further, the extensive writings on ETTR ( actually EBTR) at
http://diglloyd.com/blog/2012/20121213_1-D800E-ETTR.htmlare particularly illuminating, and seem to use results from RawDigger to strong supportive effect.
Indeed, the prophetic words of Winston Churchill come
to mind when we think of the number of times we have
stretched a bit the clipping hazards beyond the right of the JPEG histogram frame and discovered, to our immense relief, that supposedly "overexposed" and therefore "clipped" highlight details were recoverable with "overhead DR". "Wow! Dodged a bullet that time! Gotta be more careful....next time...." (..right....?)
Churchill said:
"Man occasionally stumbles over the truth,
but usually manages to pick himself up,
walk over or around it,
and carry on."
Further writings by Jeff Schewe in "the Digital Negative" and "The Digital Print" have simply reinforced the trend to at least investigate the implied promise of ( something incredibly similar to...) extra dynamic range beyond that available for the 8-bit, "pre-cooked" JPEG image file.
And it continues to impress some of us that out to an impressive distance beyond the JPEG clipping point that so-called "headroom" does one helluva job making like good-ol'-fashioned dynamic range!
To the extent that it results in reduced capture of noise in the shadows, improves shadow detail definition, and even reduces noise in the face of ridiculously high ISOs, whatever it actually is that we call ERADR succeeds, to an exceptional degree, in providing the functional advantages that would seem, otherwise, to accrue only to maximal use of extra raw-accessible dynamic range.
"Curiouser and curiouser...."
"A rose by any other name...."
A few examples below clearly illustrate why some of us have a hard time giving up use of the ERADR of each of our cameras simply because we are told it's not official dynamic range. It would seem, we are told, that it can't do what we who have been using the technique of EBTR for ten to twelve years have concluded that it actually does do....doesn't it....? ....or does it ?
IMAGES:
Again, I'm not proselytizing for EBTR, simply hoping to increase awareness of its functional and productive capabilities so that those who may be interested in accomplish the aim of maximizing photon capture to produce the brightest raw capture image possible without clipping highlights, but coming as close to that as possible, can actually succeed in that aim.
I should add that beyond the tests on my own cameras and those of a few local friends and evaluations of some ERADR exposure series received burned on DVDs from other photographers, the majority of the camera's I've helped test have been owned by photographers encountered where photographers gather: camera clubs, workshops, classes,and National Park overlooks and campgrounds. The camera owners have made the requisite ERADR exposure series at "native" ISO, loaded the images into their laptops where we have then evaluated them. I've simply added the findings to a list in my wallet. That's how I encountered two camera models represented by more than a single camera. So no, I don't have image files to display and distribute to those insisting on them as a pre-requisite assuring their continuing interest in the phenomenon of EBTR, but it should be no problem for anyone interested to do some testing themselves. I suspect that most who have adopted the technique of EBTR have, as have I, ceased to maintain files proving its utility...we just use it reliably, like gravity, and... water's tendency to flow downhill...and the Sun rising in the East...and a plethora of other "theories" supported, sadly, only by inductive evidence....which reminds me....when did inductive reasoning start getting such a bad rap?
I make time, and take time, time-and-time again to help those few who make personal contact and ask that I check their ERADR exposure series, but I'm not up to the increase in submissions that surely would result were I to publicize the two camera models that have demonstrated, among their representatives, variations in sensor function as regards their individual allotments of ERADR. Those cameras are in fact, the reason for the recommendation that every individual camera be tested.
Anyone interested in this interesting phenomenon - whether considered extra raw-accessible dynamic range or "pseudo-dynamic range beyond JPEG clipping"...or "Erzatz DR" or...whatever... - should simply start by testing their own camera's ERADR allotment.
Try it. You'll either be convinced of the utility of EBTR or you won't. Take it or leave it. Makes me no never-mind. At the very worst you'll have wasted a bit of time and the cost of all those exposu....oh...wait....we're talkin' digital here....
Whatever you may call it, off to the right side- beyond the JPEG clipping warning - you will, if you actually look for it, find something that acts like more dynamic range accessible for raw capture. It may provide as little as 1/3 Stop (E.V.)., to as much as two and 2/3 stops (or more) of whatever you wish to call it..."ERADR"? "Pseudo stops"? "Headroom"? "Ersatz Dynamic Range"? And be not of faint heart upon seeing those washed-out with blend highlight thumbnails you see on your camera’s display. The ARE NOT...I REPEAT NOT OVEREXPOSED. it’s just that your camera thinks it's a JPEG! Just you wait until you tonally normalize it in your raw converter ...that same sensation you felt years ago,watching an image develop in the tray under that weak safelight will come back with a frisson of recollection!
And you can take heart in the observation that the majority of cameras tested by those of us using EBTR have been found to have at the very least one full stop of exposure "Beyond the Right".
In the spirit of full disclosure I must admit that, though I have recently acquired RawDigger and RapidRawViewer, I am still in the process of commencing to proceed beginning to prepare to climb that rather daunting learning curve, so any and all available examples of how their use incontrovertibly negates observations of the actual utility of EBTR would be humbly accepted and studied. In the meantime, I, and I suspect most of the others in my EBTR cohort will continue on our present course of pursuing the benefits of this exposure system that, we’ve used for a decade, but now learn....can't really work.
It is perplexing to me that folks who have paid through the nose for a mega-maxi-pixelated full frame, snazzy-assed camera are satisfied to have it deliver image files with a DR truncated by several stops and thereby, depending on file bit-depth, sacrifice several thousands to tens-of-thousands of tonal levels as well as dulled colors from similar loss of chromatic levels. It is all the more perplexing when the amount of un-used dynamic range can easily be determined and as easily used via EBTR...whether one believes that that stuff masquerading as DR (beyond the right of the JPEG histogram frame) is really DR or not...it sure as heck acts like it's DR, so why not use it as such?
And I guess I ought admit to another point of curiosity..
to wit:
who, among those who have corresponded here, have actually and seriously tested their own cameras in the manner described for the possible existence of extra raw-accessible dynamic range...or for something that seems to function as if we're ERADR?
Just curious.... just askin'....
Best regards,
Dave
