Pages: 1 2 3 [4]   Go Down

Author Topic: Using Spot Metering and Exposure  (Read 28634 times)

Guillermo Luijk

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2005
    • http://www.guillermoluijk.com
Re: Using Spot Metering and Exposure
« Reply #60 on: May 31, 2017, 05:51:27 am »

Your technique has been used for years with success by hundreds of RAW shooters, including me. It simply consists of correlating the camera JPEG blinking lights with what you find later at the RAW development stage, to try to figure out an estimation on how much you can allow your JPEGs to clip before the RAW file actually begins to clip as well.

The wrong part of the story is your assumption that this is an accurate calculation that will provide perfectly ETTR'd RAW files under any situation. No way, and you can check that with the RAW analysing tools you already have.

Regards

uuglypher

  • Newbie
  • *
  • Offline Offline
  • Posts: 34
Re: Using Spot Metering and Exposure
« Reply #61 on: May 31, 2017, 08:56:42 pm »

Hi Guillermo,
Thanks for commenting.  I understand, but  disagree with with your conclusion that users
of EBTR operate under the mistaken "...assumption that this is an accurate calculation that will provide perfectly ETTR'd RAW files under any situation."   Whether you are referring to the calculation of one's camera's allotment of extra raw-accessible DR (ERADR) or are referring to determination of the exposure at which jpeg highlight clipping will occur (when the determined ERADR must be added to the exposure to accomplish EBTR) is unclear. It is essential that both be accomplished accurately to assure utilization of all available raw-accessible dynamic range.

However, the explicit statement is that the EBTR process as described will fail to result in "...perfectly ETTR'd RAW files under any condition."   "Perfect" accomplishment of the entire available raw-accessible DR?    Well...O.K....I'll admit that we work within the slop factor of  something less than 1/3 stop ...which we find to  be adequate on a practical basis but which you may consider to be an error level of insufficiently "fine grain".  In fact, some practicing EBTR 
users actually ignore the last  intact 1/3 stop of determined ERADR and use it as "cushion" and a ( likely unnecessary)  guaran-damn-tee against any clipping of highlight detail in the raw image data.  There is no question that ACR does an amazing job ( at least to my mind...)  of recovering image data in the event of clipping in one of the  color channels and I am not unduly concerned when FRV informs me of that happenstance.  Too uncritical?  Perhaps guilty as charged, but it is so rarely (ever?)  of practical significance that I relegate it to inconsequentiality. 

Now, I have long suggested  that EBTR is really a simple procedure, but your statement:
"It simply consists of correlating the camera JPEG blinking lights with what you find later at the RAW development stage, to try to figure out an estimation on how much you can allow your JPEGs to clip before the RAW file actually begins to clip as well." is a touch too misleadingly simplistic. 

The key to successful use of EBTR is to realize that cameras vary not only in their maximum raw-accessible DR, but also in the accuracy of their histogram, correlation of the right end of the histogram frame with clipping, and/or correlation of the onset of the "clipping blinkies" with the actual onset of highlight clipping of a JPEG file. However, your salvation is that when you
identify  a given camera's   most accurate indicator of actual jpeg clipping, THAT remains consistently accurate, and it is at that point that one adds the determined ERADR of that camera to accomplish EBTR. .           

I have two cameras that both provide one and 2/3 stops of ERADR.
  In camera A the accurate indicator that highlight clipping will occur with the next additional 1/3 stop of exposure is the first definitive climbing of the right end of the histogram up the right side of the histogram frame. (at which point the onset of "blinkies" has not yet occurred). 

 In camera B the first indication that highlight clipping will occur with the next additional 1/3 stop of exposure is the  onset of the blinkies ( which occurs immediately after the histogram' right extreme barely "kisses" the right side of the histogram frame but before it begins its climb
up the right side.

With another camera that provides one full  stop of ERADR I have found that spot-metering
the brightest finely detailed highlights (subject Value VIII) and adding three full stops arrives at the point of jpeg clipping (which is past the point of the onset of blinkies).

A good friend spot-meters detailed highlights (value VII) , adds two stops, and then adds his camera's two stops of ERADR.

These indices of the onset of jpeg highlight clipping, although at great variance with each other, are consistently reliable in each camera.

The unavoidable axiom of success with EBTR?    "Know your camera !"

There are a number of benefits that accrue to use of EBTR as I've decribed the practice, and
one that is often forgotten is the decreased frequency  with which it's practitioners find the
need for  HDR, particularly those whose cameras have one and 2/3 stops or more of ERADR. Happily, my two work-a-day cameras come in just at that line!    Although some decry its presumed lack of precision, experienced users of EBTR routinely rely on its demonstrated accuracy and repeatable reliability...and were that accustomed precision found to be
unreliable, EBTR would quickly fall from use.

Is EBTR useful under all conditions? It offers complete reliability only when the scene DR
is exceeded by the camera's full, raw-accessible DR at the intended ISO.

Again, Guillermo, I offer explanation and attestation to the utility of EBTR not as a proselytizer, but simply to present a practical alternative to the presumed necessity of a more technical approach to assuring proper exposure of raw image data.  Seriously interested photographers will try both approaches and make their choice. and I suggest that use of one path rather than the other is a choice born of personal experience and is not, by any means (as I have heard) an act of faith nor a character flaw.

Regards,

Dave Graham
Logged

Guillermo Luijk

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2005
    • http://www.guillermoluijk.com
Re: Using Spot Metering and Exposure
« Reply #62 on: June 01, 2017, 08:23:57 am »

I get a bit lost in all those acronyms you invented Dave. Anyway what I mean is that figuring out an approximation of how much the JPEG can be overexposed to make sure you keep a relatively ETTR'ed RAW, is something most RAW shooters have already done for years through practice. If one wants to go a step beyond in accuracy, JPEG is useless and proper RAW histograms are needed (like those provided by Magic Lantern).

As you mention, nowadays with the increased DR of modern sensors, interest in ETTR and RAW histograms is vanishing. One can even play the game of introducing some deliberate underexposure just to make sure no higlights get clipped in the RAW file, and shadows will still retain a good SNR.

Regards

scyth

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 584
Re: Using Spot Metering and Exposure
« Reply #63 on: June 01, 2017, 08:39:31 am »

If one wants to go a step beyond in accuracy, JPEG is useless and proper RAW histograms are needed (like those provided by Magic Lantern).

my experience with M43 and Sony dSLM cameras is that with OOC JPG set to use UniWB (plus some other parameters to get as low constrast as possible) precision from "blinkies/zebra" in raw is within 1/3 EV to clipping in raw (as checked with rawdigger)... raw (or ooc jpg) histogram shown on your lcd is useless, in place clipping indicator (in whatever form - blinkies, zebra, some colors) where clipping happens in the frame is what people need.
Logged

uuglypher

  • Newbie
  • *
  • Offline Offline
  • Posts: 34
Re: Using Spot Metering and Exposure
« Reply #64 on: June 01, 2017, 09:59:55 am »

Guillermo states:
"One can even play the game of introducing some deliberate underexposure just to make sure no higlights get clipped in the RAW file, and shadows will still retain a good SNR."

Which, of course, abrogates the principle that maximal raw image data quality is sbsolutely dependent upon use of the greatest possible amount of raw-accessible DR.

And I'm sorry you were confused by acronyms in use by many who use EBTR (Expose Beyond the Right) by adding to the exposure that would cause clipping of a JPEG file the ERADR (Extra raw-accessible Dynamic Range) which must be accurately determined for each camera and added to the brightest possible (yet non-clipping)JPEG exposure to assure use of the full available dynamic range available for a capture of maximal quality raw image data.

Best regards,

Dave
Logged

uuglypher

  • Newbie
  • *
  • Offline Offline
  • Posts: 34
Re: Using Spot Metering and Exposure
« Reply #65 on: June 01, 2017, 10:06:07 am »

Scyth,
Thank you for amplifying my point that for appropriate use of EBTR the most accurate and replicable index of JPEG file clipping must be determined for each individual camera.

Best regards,

Dave
Logged

scyth

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 584
Re: Using Spot Metering and Exposure
« Reply #66 on: June 01, 2017, 10:18:34 am »

Scyth,
Thank you for amplifying my point that for appropriate use of EBTR the most accurate and replicable index of JPEG file clipping must be determined for each individual camera.

Best regards,

Dave

I 'd prefer naturally clipping indications done in place based on raw data... in some cameras using UniWB for OOC JPG settings in EVF/LV mode makes things difficult (like Sony dSLM and face detection - using UniWB makes it difficult... plus you need to swtich settings effects ON ... otherwise you are limited to post shot review of blinkies instead of using in in real time while looking in EVF... still way better than useless histograms in camera - raw or not)
Logged

uuglypher

  • Newbie
  • *
  • Offline Offline
  • Posts: 34
Re: Using Spot Metering and Exposure
« Reply #67 on: June 01, 2017, 12:03:27 pm »

Hi, scyth,

The methods I describe in an earlier response to Guillermo ( Reply #61 on: May 31, 2017, 08:56:42 PM)
are of the sorts that a number of EBTR practitioners have found to be of useful accuracy. Although useful indices vary from camera to camera, once determined, they have been found to be of repeatable useful accuracy. A bit of practice works wonders at making them second nature to use.

Obviously, subjects that are quickly moving between areas of differing lighting ( children at play and other wildlife, various sports)  militate against use of EBTR! But  conditions of consistent lighting seem made for EBTR. ( studio, landscapes, indoor sports "under the lights), zoos, museums, etc.)

Best regards,
Dave
Logged
Pages: 1 2 3 [4]   Go Up