Luminous Landscape Forum
Raw & Post Processing, Printing => Digital Image Processing => Topic started by: Mosccol on May 18, 2018, 04:37:37 am
-
ETTR has been documented and debated at length, including here (http://forum.luminous-landscape.com/index.php?topic=100261.0). However, with advances in both sensors and RAW developing applications do we still need to add the extra step of pulling exposure in in post/risking inadvertent burning of whites when shooting?
I am continuously amazed how far Lightroom has gone in retrieving shadow details since version 1 and I find that over-flat skies is a more difficult problem to solve than anything else. Also, as I blend proper camera and telephone pictures (shot in jpg) I then get a stream of mixed pictures, some exposed to the middle and others to the right...
Has your workflow/position changed over the last couple of years as a result of technical developments?
François
-
Working on a tripod with spare time I tend to use ETTR still. Why not.
Faster work where I am havinn to shoot quickly no I don’t. I protect highlights and will adjust in post as needed.
Modern sensors with the huge amount of dR make ETTR not as important as the t once was. That’s my opinion and that’s how I work.
-
ETTR is always worth it - when it makes practical sense...
The fact that DR has improved, and with it the issue of sensor-related noise generation, however, neither the essential principles of sensor light-gathering nor the physics of light itself has changed!
There are many scenarios where all one can realistically aim for is avoiding overall underexposure...
However, if one is shooting at a slow pace, particularly if one has the camera on a tripod, the scene is relatively still relative to shutter speed, and one can shoot at base ISO, then it is well worth it to get that histogram well over to the right!
Shot noise, related purely to the physics of light, is always going to be present, no matter how good a sensor gets, and so the more light hitting the sensor the better.
Any part of an image with the same hue (like a blue sky in a landscape shot) as well as any area with low exposure are where shot noise will become more apparent although it is present everywhere.
Frankly, for ultimate image quality, being able to sharpen without concern for exacerbating the effects of any noise that is present, and, conversely, not needing to bludgeon an image with noise reduction relying rather on very subtle noise reduction (some is always needed) can, potentially take image-making to another level...
(Obviously, pointing one's camera at a worthwhile scene is always primary and post-processing is always consequent to that!)
So, yes when I am shooting wildlife, my only concern is keeping my camera/lens still and having a fast enough shutter speed to freeze the action (aperture is wide open, and to hell with the ISO..) ETTR is the last thing on my mind!
However, if I am shooting on a tripod with sufficient light to allow me to capture what I want at the aperture I want and a sensible shutter speed for the subject then absolutely, ETTR will allow me to squeeze the last bit of goodness out of what are already excellent sensors...
Tony Jay
-
ETTR has been documented and debated at length, including here (http://forum.luminous-landscape.com/index.php?topic=100261.0). However, with advances in both sensors and RAW developing applications do we still need to add the extra step of pulling exposure in in post/risking inadvertent burning of whites when shooting?
I am continuously amazed how far Lightroom has gone in retrieving shadow details since version 1 and I find that over-flat skies is a more difficult problem to solve than anything else. Also, as I blend proper camera and telephone pictures (shot in jpg) I then get a stream of mixed pictures, some exposed to the middle and others to the right...
Has your workflow/position changed over the last couple of years as a result of technical developments?
François
Hi François,
When possible I always try to use the ETTR approach. Of course it is not always possible to do so to the limit, but I always have my camera set to add at least 1/3 stop to the meter reading. In traditional B&W photography the approach was always to expose for the shadows and develop for the highlights, the basis of the Zone System, perfected by Ansel Adams. And again, that approach often depended on the particular situation and the amount of time involved. The reason behind that approach was to be sure you had enough exposure in the shadows to accommodate a full tone B&W print, and not a lot of grain in the shadows. Of course we now refer to "grain" as digital noise, but the unwanted effect is very similar. If you haven't read the article on this site by Bob DiNitale "The Optimum Digital Exposure", I would suggest that you do so. It offers an in depth description of the process, along with the reasons for using it when possible. You can find it here - https://luminous-landscape.com/the-optimum-digital-exposure/ and I strongly suggest that you read it. His approach is a bit over the top, but after using it myself I can say that it definitely does work as advertised. And again, this approach is not always available in every photography situation, but when possible it can capture more tones than your camera meter can. The secret is in the processing, as usual.
Gary
-
Optimal exposure for ideal data is always something to strive for! Exposure is far, far from rocket science and if fact is a fundamental part of photography. Why not get it right IF you can?
-
Hi Francois,
You are raising a good question. The evolution of sensor technology has been so rapid that it's worthwhile reviewing past recipes insofar as technology is enabling, while the recipe at hand was meant to make the most of how linear response works and what the sensors of the time could deliver.
The logic underlying ETTR hasn't changed - every f/stop-worth reduction of aperture cuts the light reaching the sensor in half. The less light that reaches the sensor the greater the risk of seeing banding and noise especially in the quartertones. What has changed, however, are (1) the ways in which the sensors react to reduced exposure, and (2) the capabilities of post-capture software for adjusting tonality to reveal detail. The dynamic range of sensors has increased over the years and the software has improved dramatically. These two factors give us more latitude in how we place the exposure, insofar as compared with years ago the risk of banding and excessive noise in the quartertones has lessened with some departure from ETTR; this kind of flexibility is welcome.
The important operational consideration is how one exposes the scene with today's cameras, and I like to think of this in respect of two situations: a bright scene where protecting highlights is really important, and a darker scene where nothing all that bright causes concern. An example of the former includes skies where protecting clouds can be vital to the success of a photograph. As little as half a stop to the left of ETTR can work wonders for revealing excellent detail in clouds without a lot of gymnastics to pull it out in Lr or Ps. I usually use the camera's default scene metering with a bit of negative exposure compensation for achieving this, and it usually works well for both highlights and shadows when combined with appropriate post-capture adjustments in Lr. For darker scenes, no reason not to go ETTR.
-
Optimal exposure for ideal data is always something to strive for! Exposure is far, far from rocket science and if fact is a fundamental part of photography. Why not get it right IF you can?
Sure, but the question is "what's right" and does that change as a result of technological evolution? I'm suggesting that the basic principle hasn't, but in practice more variants can be "right" depending on the scene, because the technology enables better outcomes under more conditions than it would have years ago.
-
One of the downsides of practicing ETTR has been the sacrificing of JPEG quality and usability for the sake of essentially "overexposing" for the raw file. The general trend in increased sensor performance permits metering/shooting for raw and JPEG simultaneously - I see that as the ultimate benefit of less dependence on ETTR. If you need to shoot JPEG for expediency or other reasons, but expose for ETTR, your JPEGs were typically destroyed as a result. Nowadays, you can relax the extreme ETTR exposure shift and still get a good raw file to work with as well as acceptable, if not optimal, JPEGs. Tuning the in-camera JPEG controls (color styles, or whatever they are called on one's camera) will also help bring the raw v JPEG exposure closer together. Being able to relax the ETTR exposure also gives you some wiggle room when lighting in the frame is changing, ensuring that, even if the JPEG highlights are blown, the raw highlights will be preserved.
kirk
-
One of the downsides of practicing ETTR has been the sacrificing of JPEG quality and usability for the sake of essentially "overexposing" for the raw file.
Different data, different exposure. There is nothing about over exposing with ETTR! It's solely about optimal exposure for the data (raw) which requires a different exposure than a JPEG. Pick one, expose optimally for that data. Or not....
-
I use ETTR when I'm taking a single shot but.... if I'm shooting a scene with a wide range of light in it then I'm shooting for HDR and skipping the ETTR routine. Its so easy these days in the camera to bracket exposures and then process in any of a number of HDR tools to achieve great looking results. HDR is no longer synonymous with grunge.
-
Sure, but the question is "what's right"....
What’s ‘right’ is doing whatever it takes to get the the friggin’ shot.
If you are in a sutuation where you only get one chance to grab some frames bracket...if you only have one shot, preserve the highlights because you can pull up shadows but not easily replace burned out highlights.
But if you have the time, shoot till you got it rights-pixels are free (it’s the camera and lenses that cost real money)!
-
One of the downsides of practicing ETTR has been the sacrificing of JPEG quality and usability for the sake of essentially "overexposing" for the raw file. The general trend in increased sensor performance permits metering/shooting for raw and JPEG simultaneously - I see that as the ultimate benefit of less dependence on ETTR. If you need to shoot JPEG for expediency or other reasons, but expose for ETTR, your JPEGs were typically destroyed as a result. Nowadays, you can relax the extreme ETTR exposure shift and still get a good raw file to work with as well as acceptable, if not optimal, JPEGs. Tuning the in-camera JPEG controls (color styles, or whatever they are called on one's camera) will also help bring the raw v JPEG exposure closer together. Being able to relax the ETTR exposure also gives you some wiggle room when lighting in the frame is changing, ensuring that, even if the JPEG highlights are blown, the raw highlights will be preserved.
kirk
Some rather strange comments above...
ETTR is NOT about overexposing a raw file.
If one is shooting JPEG's no one in their right mind would expose those images in the same way as shooting raw, ETTR style, so...complete straw-man argument.
If one is shooting JPEG and raw together then the above point applies, one cannot employ ETTR.
You seem fixated on shooting JPEG - ETTR is never an option for you...
-
Modern sensors with the huge amount of dR make ETTR not as important as the t once was. That’s my opinion and that’s how I work.
Increased DR actually increases the ability to ETTR. However I get what you are saying - with increased DR generally comes decreased noise in shadows and less need to ETTR. While I totally get the ETTR thing, especially at base ISO, I always thought the original proponents were so enthusiastic because they shot Canon gear, which at the time, and indeed, until only recently, produced files with lower DR and ugly shadows.
-
ETTR has been documented and debated at length ... However, with advances in both sensors and RAW developing applications do we still need to add the extra step of pulling exposure in in post/risking inadvertent burning of whites when shooting?
I will put aside for a moment my skepticism about pursuing increases in SNR and such that are often measurable but not visible in the final displayed image, at the cost of both time and effort (and thus perhaps lost opportunities for other photographs) and also losing the convenience of having default JPEG conversions that are useful at least for preview and triage, and in many cases perfectly good for final use.
That leaves another question: one option for this goal of maximizing sensor exposure and thus SNR levels is to bracket exposure and then select the file with the best raw histogram(s) as viewed on the computer; so:
how often does one need instead to fiddle in the field with "overexposing" (relative to traditional light metering approaches and to what is needed for usable default JPEGs) and approaches like using "UniWB" white balance settings that mess up the live view and in-camera preview?
-
IMHO ETTR has some negative side effects; it changes the overall contrast of the image ( different contrast slope)
I get the better contrast envelope in my images by not to exposing as ETTR but just normal.
-
IMHO ETTR has some negative side effects; it changes the overall contrast of the image ( different contrast slope)
I get the better contrast envelope in my images by not to exposing as ETTR but just normal.
It's normal that if you expose to the right more of the image data is parked in the brighter zones of the histogram - that's the whole point, and it is expected that in post capture processing you will redistribute it, making contrast correct. But by doing this you have more data to redistribute, hence less risk of undesirable discontinuities.
-
It's normal that if you expose to the right more of the image data is parked in the brighter zones of the histogram - that's the whole point, and it is expected that in post capture processing you will redistribute it, making contrast correct. But by doing this you have more data to redistribute, hence less risk of undesirable discontinuities.
Exactly!
Tony Jay
-
It's normal that if you expose to the right more of the image data is parked in the brighter zones of the histogram - that's the whole point, and it is expected that in post capture processing you will redistribute it, making contrast correct. But by doing this you have more data to redistribute, hence less risk of undesirable discontinuities.
Yes, i know all that, i am only saying it is more difficult for me to get a photo with the overall contrast that i like.
Sometimes it seems even impossible to get the same contrast as with a normal exposure....
Still i would say my skills are good with photoshop and ACR. I work with photoshop since layers were introduced, so 20 years or so.
-
Exactly!
Tony Jay
+2 An ETTR raw will almost always be "brighter overall" on initial ingest into PP software than I find "ideal" - BUT as long as none of the data is clipped, I have more data to move about where I want it in terms of tone(s). Shadows w/ less/no noise and lots of detail is often one benefit. As Andrew says "also," all this can be very scene reference image dependent.
Rand
-
Once I had purchased and read "The Optimal Digital Exposure" from Bob DiNitale I immediately set about some experiments of my own. As he mentions in his article, if the raw unprocessed image does not look like it has been dipped in milk, it is probably suffering from underexposure. Therefore, his procedure is built around pulling the exposure down, instead of pushing it up as is usually the case when using the in-camera meter only. Once I had enough various subjects to satisfy my experiment I started opening them in ACR, and the one with the most exposure (I do bracket) was always visually very "milky", and the "Exsposure" scale would generally show approximately + 80 on average. However, once I started to pull the exposure "down" the details started to show immediately. Yes, there is often a small degree of clipping, but when I compare it to the one with the least exposure there is nothing there of any consequence, and may time pure white. Therefore, nothing to be concerned about in my opinion. I will say that until I ran my own experiments I was somewhat skeptical, but I was rather quickly convinced that this method does indeed work very well when applied properly. And of course not all situations lend themselves to this procedure, but I think that is quite obvious.
As the old adage goes - "The proof is in the pudding". Or perhaps in this case, "the milk". :)
Gary
-
IMHO ETTR has some negative side effects; it changes the overall contrast of the image ( different contrast slope)
I get the better contrast envelope in my images by not to exposing as ETTR but just normal.
The raw file is linear and the different contrast slope to which you refer is imposed by the tone curve of the raw converter and not by the position in the raw file of the ETTR higlights. In Adobe converters it is true that the calibration Versions 3 and 4 do remap the highlights, it is possible to obtain a roughly linear tone curve by using Version 2 (PV 2010) and setting the sliders to neutral and the tone curve to linear. Capture One (https://www.dpreview.com/forums/post/57237875) does offer a linear tone curve and its use in preserving the highlights is discussed here. (https://www.dpreview.com/forums/post/57237875)
Personally, I find the current Version 4 to be fine with reasonably ETTR images, but if you want a linear file there are means to obtain it and do your own tone mapping.
It's normal that if you expose to the right more of the image data is parked in the brighter zones of the histogram - that's the whole point, and it is expected that in post capture processing you will redistribute it, making contrast correct. But by doing this you have more data to redistribute, hence less risk of undesirable discontinuities.
While is true that a linear file contains more levels in the highlights, the real reason to use ETTR is to obtain a better signal to noise ratio as explained here (http://www.photonstophotos.net/Emil%20Martinec/noise-p3.html#ETTR). While Michael deserves credit fof introducing many of us to the concept of ETTR, his original rationale concerning the number of levels was incorrect. Raw data are never posterized and it would require herculean editing to introduce banding.
When shooting in dim light using higher ISOs, moving the histogram to the right by increasing ISO while leaving exposure the same (same aperture and shutter speed) serves mainly to reduce highlight headroom as Emil explains in the above link and the signal to noise ratio is not improved (assuming a reasonable ISO-less sensor). In this case using the histogram to represent exposure is misleading, and some prefer to use proper exposure rather than ETTR. Even with earlier Canon sensors, raising the ISO beyond 1600 has little effect on SNR.
Bill
-
As the old adage goes - "The proof is in the pudding".
No, it doesn't; there is no such adage, since the sentence has no meaning (unless you're hiding a test print inside a spotted dick, of course).
"The proof of the pudding is in the eating."
Jeremy
-
Of course ETTR is still relevant. You may not use it as frequently because sensors are so much better, but that doesn’t mean it becomes irrelevant. I still use it regularly whether I’m shooting with my D800E, my Sony RX-10iii or my iPhone.
Bottom line: Pay attention to your histogram no matter what you’re shooting, unless you enjoy working with blown out highlights.
-
Some rather strange comments above...
ETTR is NOT about overexposing a raw file.
If one is shooting JPEG's no one in their right mind would expose those images in the same way as shooting raw, ETTR style, so...complete straw-man argument.
If one is shooting JPEG and raw together then the above point applies, one cannot employ ETTR.
You seem fixated on shooting JPEG - ETTR is never an option for you...
I’ve been shooting raw for close to 15 years. With ETTR, we are purposely overexposing relative to the meter, not the raw histogram. If we had raw histograms on our cameras, ETTR would not be necessary, simply because you would have a true display of the raw data on your camera. Magic Lantern for some Canon cameras has this and you simply expose for the raw histogram. Using a tool like Raw Digger you can develop a relation between your meter and the raw data and infer the raw histogram’s highlight end from metering scenes and studying the raw data.
I am not fixated on shooting JPEG, but it is clear that if one can expose for both, I.e., because sensors are getting better, then that is optimal in terms of having an immediately useable image (the JPEG) and the high-quality raw data to use for further processing. I believe that the OPs point was exactly this- increased sensor quality may mean less of a need to ETTR. Clearly cameras are not there yet but way better than even a couple of camera generations ago.
Kirk
-
I love this forum! - Some topics are like touch paper ;)
Thank you everybody for the extremely useful contributions. My biggest take home is the realisation that it's the quantity of light hitting the sensor that counts, not the exposure (helped by ISO).
Talking about contrast curve, what would be a way to 'restore' a 'correct' contrast in LR using the current process (as suggested by Pieter Kers)?
-
Talking about contrast curve, what would be a way to 'restore' a 'correct' contrast in LR using the current process (as suggested by Pieter Kers)?
Typically we aren't "restoring" "correct" contrast in a raw file because there is no restoration and there is no standard for what is correct in much of the photography we do. We are using the software to create a usable visual appearance of the photo when we open it in Lr and we are editing tone and colour to taste when we start using the tools. What's correct is how we want our photos to look.
-
My biggest take home is the realisation that it's the quantity of light hitting the sensor that counts, not the exposure (helped by ISO).
That is the idea I asked about when I started yet another thread about ETTR that you might want to browse:
http://forum.luminous-landscape.com/index.php?topic=123749.0
Talking about contrast curve, what would be a way to 'restore' a 'correct' contrast in LR using the current process (as suggested by Pieter Kers)?
To clarify your question: are you looking for a way to first lower the levels from the "maximally exposed" ETTR file to get a correct mid-tone placement and then apply to that the default tone curve that you would have got if you had exposed for that mid-tone placement in the first place?
-
No, it doesn't; there is no such adage, since the sentence has no meaning (unless you're hiding a test print inside a spotted dick, of course).
"The proof of the pudding is in the eating."
Jeremy
Hi Jeremy,
I beg to differ. When I first read your reply to my usage of this phrase I decided not to take this any further, since it has no relevance to this thread. However --- I have decided to take you on this morning, since I have heard both versions of the aforementioned phrase, but mostly the version I used to fit the situation. Language is a fluid and ever evolving part of any society, and certainly not written in stone, as can be demonstrated in many ways and by many cultures. I used this version of the idiom (the proof is in the pudding) for two reasons. First, it's the one I am most familiar with. Secondly, it fit perfectly with the thought I was trying to convey.
If you have a minute to spare, perhaps you would like to Google(verb) the phrase I used. If so, you will find at least a full page committed to both versions of the phrase and the etymology of the one I used simply to fit my own personal needs. Here's one of the sites you might find interesting, or not - https://www.urbandictionary.com/define.php?term=the%20proof%20is%20in%20the%20pudding.
Also a snippet from another site - "The idiom is usually stated the proof is in the pudding and means that the end result is the mark of the success or failure of one’s efforts or planning. The phrase may also be used in the past and future tenses: the proof will be/was in the pudding".
Well Jeremy, it's the Victoria Day holiday weekend here in Canada, and I had nothing more pressing to do at this somewhat early time, so I hope this explanation will suffice. Just a small break from the initial thread, and no more to be said about I'm sure.
Have a great day,
Gary
-
My biggest take home is the realisation that it's the quantity of light hitting the sensor that counts, not the exposure (helped by ISO).
The light hitting the sensor is the exposure; not ISO.
-
With ETTR, we are purposely overexposing relative to the meter, not the raw histogram
We are not over exposing. We are ignoring the wrong meter reading for JPEG which isn't raw, nor correctly exposed like raw, hence, we ignore the Histogram because it isn't at all necessary to properly expose. OR we expose to the right of that incorrect Histogram because we can't expose without that lie told to us first, then compensate. But it's never over exposing!
What we need ideally is a raw Histogram or we just learn, like film, to expose without a Histogram, one that is true or one that is a lie.
-
We are not over exposing. We are ignoring the wrong meter reading for JPEG which isn't raw, nor correctly exposed like raw, hence, we ignore the Histogram because it isn't at all necessary to properly expose. OR we expose to the right of that incorrect Histogram because we can't expose without that lie told to us first, then compensate. But it's never over exposing!
What we need ideally is a raw Histogram or we just learn, like film, to expose without a Histogram, one that is true or one that is a lie.
Taken in context, the term "over-exposing" kirkt used here (my interpretation) may have meant "too bright an exposure relative to the tonality I want to end-up with for the photo", not over-exposure in the technical sense of clipping information.
Also, the alternatives you propose here may be a shade more dire than necessary. We can do tests to determine approximately the extent to which our histograms misrepresent the bright end of the scale and compensate accordingly in how we expose the capture. Admittedly a bit rough, but I think useful nonetheless.
-
Taken in context, the term "over-exposing" kirkt used here (my interpretation) may have meant "too bright an exposure relative to the tonality I want to end-up with for the photo", not over-exposure in the technical sense of clipping information.
Clipping is clipping (maybe over exposure, maybe not). Exposure is exposure and may subject some pixels to clipping, maybe not. Too bright is a subjective evaluation of rendered pixels. Over exposure of raw data isn't and we need tools to see such clipping not available on most of our camera systems to test exposure of that raw data.
-
Sure, but all I'm suggesting is a workaround until we have such tools.
-
Sure, but all I'm suggesting is a workaround until we have such tools.
We (I) don't need such tools; I exposed all kinds of film decades before histograms appeared on cameras. Also, such tools exist. So it's all a bit moot. Exposure (ideal exposure for the media you capture) is photography 101. It's not rocket science. ;D
-
When I was still shooting 4x5 B&W I always carried two exposure meters, a LunaSix 3 and a "vintage" Pentax Spot Meter. I very seldom used the LunaSix. I determined which part of the scene was most important and what exposure it required to maintain the detail necessary for what I wanted to see in the final print. I then metered the opposite end of the scene and tried to determine the necessary development necessary to hold the highlights as well. Again the "expose for the shadows and develop for the highlights" scenario, now referred to digitally as HDR. And of course I am referring to a "properly" used HDR procedure, not the "over the top" versions I see online, my opinion only. Personally, I do not yet use the HDR approach as such, since the "Optimal Digital Exposure" seems to produce what I need when the situation is such that it can be put into use. But of course similar restrictions apply to HDR exposures to some extent. The possibility of "overexposing" was not part of my vocabulary then, except for perhaps an occasional mistake on my part. The final product was a combination of exposure and development, which together produced a negative that could, with some degree of manipulation, produce the print I had envisioned when the shutter was fired. Did it always work to my expectations?. No, it did not, which meant a higher degree of manipulation during the printing cycle than I would prefer. However, it did usually produce a negative that contained enough information to provide a good print in the final outcome. I learned a lot from Ansel's writings, which are still relevant in the digital darkroom/Lightroom. It is after all the foundation of photography, and Ansel found a way to expand on that and make it more technical and indeed, controllable. I agree with the premise that there is no such thing as a "correct" exposure. It all depends on the vision of the photographer and the ability to get enough light hitting the sensor to capture the full range of detail. The rest is as it always has been, processing.
Gary
-
We (I) don't need such tools; I exposed all kinds of film decades before histograms appeared on cameras. Also, such tools exist. So it's all a bit moot. Exposure (ideal exposure for the media you capture) is photography 101. It's not rocket science. ;D
If that's the case why are you supporting Guillermo's petition? Now I don't understand where you're coming from. I exposed film successfully since the early 1950s Andrew and have been doing so successfully with DSLRs since 2004, so I could also argue I don't need any new tools, but I think we agree the idea of a raw histogram is a good one. But given that we don't have one, I was simply suggesting there are ways of compensating for that deficiency that can still include making use of the JPEG-based histogram. I don't see what should be so remarkable about that notion.
-
There is zero reason not to provide a raw Histogram! A JPEG based Histogram alone? No!
-
Alright, claro - raw histograms are preferred. That being the case I'd like to see the discussion move forward about why the camera manufacturers have not been providing raw histograms on their LCDs at the very least for the more professional models.
-
We are not over exposing. We are ignoring the wrong meter reading for JPEG which isn't raw, nor correctly exposed like raw, hence, we ignore the Histogram because it isn't at all necessary to properly expose. OR we expose to the right of that incorrect Histogram because we can't expose without that lie told to us first, then compensate. But it's never over exposing!
What we need ideally is a raw Histogram or we just learn, like film, to expose without a Histogram, one that is true or one that is a lie.
I know precisely how much raw headroom I have for my camera when I meter the reference in the scene, and I can compensate for the meter to get the desired, non-clipped but ETTRed raw exposure - 100 percent of the time the compensation involves adding light compared to the metered reading. That is what I mean by overexposing - "over" meaning positive exposure compensation or adding light. Not clipping a JPEG or a raw file, adding exposure compared to the meter. You don't ignore the meter, you simply accept that it is lying and use your experience and understanding of your camera's raw capability to convert its lie into a useful reading.
The camera histogram is practically irrelevant for raw files, unless pains are taken to manipulate it with crafty camera settings like UniWB, etc..
If you ignore the meter reading, how do you set your exposure?
kirk
-
Ah well us mere mortals will to keep ETTR just below the clipping of those imperfect jpeg screens! :o 8)
Interestingly I have tended to go +⅓ for all my quick shots as suggested above with some success... It's just the skies where I have to be more careful.
-
Ah well us mere mortals will to keep ETTR just below the clipping of those imperfect jpeg screens! :o 8)
Interestingly I have tended to go +⅓ for all my quick shots as suggested above with some success... It's just the skies where I have to be more careful.
Yes, I have the meter set at +1/3 as well, but if in doubt I do bracket. I've never run out of camera card space nor HDD space, so that part is always covered. The other alternative is to spot meter the scene when possible, but always, ALWAYS, check the Histogram and then wok from there.
Gary
-
I know precisely how much raw headroom I have for my camera when I meter the reference in the scene, and I can compensate for the meter to get the desired, non-clipped but ETTRed raw exposure - 100 percent of the time the compensation involves adding light compared to the metered reading. That is what I mean by overexposing - "over" meaning positive exposure compensation or adding light. Not clipping a JPEG or a raw file, adding exposure compared to the meter. You don't ignore the meter, you simply accept that it is lying and use your experience and understanding of your camera's raw capability to convert its lie into a useful reading.
The camera histogram is practically irrelevant for raw files, unless pains are taken to manipulate it with crafty camera settings like UniWB, etc..
If you ignore the meter reading, how do you set your exposure?
kirk
Thanks for confirming - this is what I thought you meant (my Reply #30 and the follow-ons). But I wouldn't go as far as to say the camera histogram is practically irrelevant for raw files. While flawed, I think it provides useful guidance provided we know its limitations and how to work around them.
-
Hi Jeremy,
I beg to differ. ... Here's one of the sites you might find interesting, or not - https://www.urbandictionary.com/define.php?term=the%20proof%20is%20in%20the%20pudding.
Also a snippet from another site - "The idiom is usually stated the proof is in the pudding and means that the end result is the mark of the success or failure of one’s efforts or planning. The phrase may also be used in the past and future tenses: the proof will be/was in the pudding".
I prefer definition #2 from that site. I realise the your formulation is in common usage, but it's still meaningless.
Well Jeremy, it's the Victoria Day holiday weekend here in Canada, and I had nothing more pressing to do at this somewhat early time, so I hope this explanation will suffice. Just a small break from the initial thread, and no more to be said about I'm sure.
Have a great day,
It isn't a holiday here (that's next Monday). I hope you had a good day too.
Jeremy
-
Made my day! :-)
-
... I'd like to see the discussion move forward about why the camera manufacturers have not been providing raw histograms on their LCDs at the very least for the more professional models.
A speculation, it is because they believe that only a tiny fraction of even high-end camera users are interested, even if it is popular in certain circles. Some possible reasons for that belief:
- ETTR is irrelevant in many sorts of photography, such as almost anything with moving subjects or low light/action photography where inevitably the sensor will be exposed "far to the left".
- There are easier strategies that many photographers judge to be good enough for all practical IQ purposes, such as using a fixed 1/3 or 1/2 stop of overexposure (relative to what the light meter recommends for default JPEG conversions) or pushing the JPEG histogram to the right and then an extra 1/2 stop or so beyond.
- In almost any situation where ETTR can be done, so can exposure bracketing, and with a burst of 7 at 1/3 stop increments (2 stop range), one will probably be darned close to the ETTR ideal.
- The fact that they got good results without ETTR when using film leads many good photographers to conclude that it is at best of a small marginal value with today's digital sensors, given that noise levels are far lower than their filmic equivalents.
Another question: would a raw clipping indicator be enough, or an exposure metering display mode that shows how far below clipping the brightest pixels are?
-
...............
Another question: would a raw clipping indicator be enough, or an exposure metering display mode that shows how far below clipping the brightest pixels are?
A raw clipping indicator could be useful but less so than a histogram which shows not only whether clipping would occur, but also what proportion of image pixels would be affected.
-
A raw clipping indicator could be useful but less so than a histogram which shows not only whether clipping would occur, but also what proportion of image pixels would be affected.
And it would be easier to implement since they already have code for the processed file histogram. It's also more familiar to photographers.
-
A raw clipping indicator could be useful but less so than a histogram which shows not only whether clipping would occur, but also what proportion of image pixels would be affected.
But the histogram does not show you where the clipping is occurring. For example, in harsh light, a face, hair and a light background may be clipped but you might only care about the face. Another example - a yellow bird on a branch. Both the yellow feathers and the branch may have clipping and the yellow feathers are only a small area in the image. They will often require a much larger exposure adjustment to eliminate clipping than the branches, but that will not be evident in the histogram.
-
But the histogram does not show you where the clipping is occurring.
Yes; for me, highlight blinkies/zebras are more useful than either a histogram or a simple yes/no overexposure warning. So maybe my wish is for a version of such indications that is based on raw data.
-
Yes; for me, highlight blinkies/zebras are more useful than either a histogram or a simple yes/no overexposure warning. So maybe my wish is for a version of such indications that is based on raw data.
Blinkies would be good. Also, something like a slow blink that was recognizably different that indicates where the next stop of more exposure would clip but isn't currently clipping. That way you could eliminate moving up to clipping then backing off.
-
But the histogram does not show you where the clipping is occurring. For example, in harsh light, a face, hair and a light background may be clipped but you might only care about the face. Another example - a yellow bird on a branch. Both the yellow feathers and the branch may have clipping and the yellow feathers are only a small area in the image. They will often require a much larger exposure adjustment to eliminate clipping than the branches, but that will not be evident in the histogram.
I would care about any clipping. Smart histograms can indicate what colour is being clipped. And there is no reason why these options should be mutually exclusive - one can have both blinkies and a raw histogram - why not?
-
To the OP:
Hello François.
Short answer:
You need to do your own testing on your own camera. My workflow has certainly changed.
Long answer:
Six years ago when I was using a Canon camera I would religiously ETTR because I had to. The noise in the shadows could get very ugly. And the OOC jpegs weren't that useful.
Times have changed.
Because it is not an optical viewfinder, with a Fuji camera I use the viewfinder to gauge the exposure. When I can see detail in the highlights I press the shutter and let the shadows fall where they may. I don't refer to the histogram that much.
The OOC jpegs are fine for medium size prints. This is a great time saver when there is rush to get some initial work out.
Most of the files I send to Photoshop have not had the exposure, highlight recovery, saturation and clarity sliders touched in the raw converter. And sharpening is set to minimum. With this sensor, mostly I want to remove sharpness, not add it.
To my eye the result is better tonality and colour. I don't find raw files to be as malleable as many make out. Oh, and I almost never need to use noise reduction on photos shot under 1600 ISO (Standard Output Sensitivity). The shadow detail is just fine.
I'm not talking about those who post here, but in general most people lose detail and fine tonality in the three quarter tones and above due to their desire to avoid shadow noise. Think white horses or wedding dresses. You see it regularly in photographs. Those highlights really matter to the eye in a full tonal range image.
The saturation and colour in the highlights is often off as well due to the exposure and highlight recovery sliders altering their balance. In PV 2012 Lightroom in particular can do unpleasant things to the three quarter tones and above, sometimes without even moving the highlight recovery slider.
Try this for yourself to see if it matters for your camera and your processing method. Render out a file in PV 2012 where the highlight recovery slider has been used. Now go to Camera Calibration _ Process and change it to 2010. Adjust the highlights using only the exposure slider and render.
Compare the highlight tonality in both images.
David
Edit: Some time ago when I tried the above I could see a marked improvement in the rendering of clouds. I tried it last night on an image of a white horse and the highlight recovery slider in the latest process version gave a better rendering. Who knows, maybe it's camera/image dependent or maybe I've gone mad. :)
Google brought up a link to Mike Johnston's article on tonal values in digital B&W here (http://theonlinephotographer.typepad.com/the_online_photographer/2017/07/how-to-cure-the-digital-bw-nasties.html). Although I wouldn't go so far as to agree that "almost all digital B&W is like drinking rotten pond scum".
-
Made my day! :-)
+1,
Gary
-
But the histogram does not show you where the clipping is occurring.
It most certainly can!
Everything you thought you wanted to know about Histograms
Another exhaustive 40 minute video examining:
What are histograms. In Photoshop, ACR, Lightroom.
Histograms: clipping color and tones, color spaces and color gamut.
Histogram and Photoshop’s Level’s command.
Histograms don’t tell us our images are good (examples).
Misconceptions about histograms. How they lie.
Histograms and Expose To The Right (ETTR).
Are histograms useful and if so, how?
Low rez (YouTube): http://www.youtube.com/watch?v=EjPsP4HhHhE (http://www.youtube.com/watch?v=EjPsP4HhHhE)
High rez: http://digitaldog.net/files/Histogram_Video.mov (http://digitaldog.net/files/Histogram_Video.mov)
-
It most certainly can!
The context is in-camera feedback. Can you explain how the in-camera histogram shows where in the image will be clipped (in under 40 minutes) for all cases. I provided examples where I do not think a histogram will reveal the clipping I am interested in, but where the blinkies work nicely.
-
. . . and no more to be said about I'm sure.
You ain’t gettin’ off that easy. As you say, language is fluid. So word meanings drift, phrases get truncated in some way or other. You know, like, OMG, “Him and me went to the mall.”
But on a more serious note, there’s something else often expressed in the shift/drift/fluidity of language over time. Philosophy. Let’s take the “proof” being “in the eating” in the first instance. The end product of some venture, the quality of it, is subject to evaluation and judgement by a value laden human agent. The “aptness/truth” of the saying hangs upon an assumption that good pudding is obvious to the “average person” as a value judgement, a consensus. This philosophical position sees the necessity for both process and agency as an adequate explanation for “the good ness (or not) of pudding.” There are much deeper philosophical/worldview overtones, but we can leave those for now.
Now, “the proof is in the pudding” is another kettle of fish altogether. The “proof” no longer needs any kind of agent to provide a value laden, and one might say ‘personal’ determination of pudding-goodness. No, this version says the proof is in the pudding itself, independent of any kind of agent making an evaluation. The goodness of the pudding is only determinable from the pudding itself, presumably the chemical makeup and characteristics of the pudding. This shift in the phrase is HUGE and reflects the shift toward a materialist/naturalist/reductionist philosophical frame.
Rand
(PS - This is an attempt at humor, in case it isn’t obvious at first glance.)
-
You ain’t gettin’ off that easy. As you say, language is fluid. So word meanings drift, phrases get truncated in some way or other. You know, like, OMG, “Him and me went to the mall.”
But on a more serious note, there’s something else often expressed in the shift/drift/fluidity of language over time. Philosophy. Let’s take the “proof” being “in the eating” in the first instance. The end product of some venture, the quality of it, is subject to evaluation and judgement by a value laden human agent. The “aptness/truth” of the saying hangs upon an assumption that good pudding is obvious to the “average person” as a value judgement, a consensus. This philosophical position sees the necessity for both process and agency as an adequate explanation for “the good ness (or not) of pudding.” There are much deeper philosophical/worldview overtones, but we can leave those for now.
Now, “the proof is in the pudding” is another kettle of fish altogether. The “proof” no longer needs any kind of agent to provide a value laden, and one might say ‘personal’ determination of pudding-goodness. No, this version says the proof is in the pudding itself, independent of any kind of agent making an evaluation. The goodness of the pudding is only determinable from the pudding itself, presumably the chemical makeup and characteristics of the pudding. This shift in the phrase is HUGE and reflects the shift toward a materialist/naturalist/reductionist philosophical frame.
Rand
(PS - This is an attempt at humor, in case it isn’t obvious at first glance.)
Hi Rand,
Well, unfortunately the only sentence that made any sense to a linguistic deprived fool such as I, was the last one - "PS - This is an attempt at humor, in case it isn’t obvious at first glance". If that isn't enough proof, cut into the pudding. You'll find more there I'm sure :)
Gary
-
Hi Rand,
Well, unfortunately the only sentence that made any sense to a linguistic deprived fool such as I, was the last one - "PS - This is an attempt at humor, in case it isn’t obvious at first glance". If that isn't enough proof, cut into the pudding. You'll find more there I'm sure :)
Gary
I guess you'll have to take my word for it that it's funny! ;D LOL (You're in good company, I get a lot of groans from my philosopher friends.)
Rand
-
I guess you'll have to take my word for it that it's funny! ;D LOL (You're in good company, I get a lot of groans from my philosopher friends.)
Rand
Hmmm... My philosopher friends are all resting comfortably in their graves, so we don't communicate very often. I think they're waiting for me to join them, but so far my stubborn demeanor has prevented me from doing so :)
Isn't it interesting that such an innocent use of a phrase can sidetrack a thread in this way. I do believe that we all need to take a break occasionally and back away from the technical aspects and try to manufacture a bit of a chuckle if possible. However, I must admit that it was not my intent to do so when I used that phrase initially, in the way I have always heard it. I totally agree with Jeremy that the version I used makes no sense whatsoever. Again, the liquidity of language is at play I suppose. Da proof is da proof.
I will now enjoy my morning java as I reread your ever so "humourous" former offering. Wish me luck ???
Gary
-
Hmmm... My philosopher friends are all resting comfortably in their graves, so we don't communicate very often. I think they're waiting for me to join them, but so far my stubborn demeanor has prevented me from doing so :)
Isn't it interesting that such an innocent use of a phrase can sidetrack a thread in this way. I do believe that we all need to take a break occasionally and back away from the technical aspects and try to manufacture a bit of a chuckle if possible. However, I must admit that it was not my intent to do so when I used that phrase initially, in the way I have always heard it. I totally agree with Jeremy that the version I used makes no sense whatsoever. Again, the liquidity of language is at play I suppose. Da proof is da proof.
I will now enjoy my morning java as I reread your ever so "humourous" former offering. Wish me luck ???
Gary
If a word or a phrase is used wrongly by enough people for long enough time, it becomes right.
-
If a word or a phrase is used wrongly by enough people for long enough time, it becomes right.
I couldn't have said it better. Well, perhaps differently, but only Rand would be able to decipher it ;)
Gary
-
If a word or a phrase is used wrongly by enough people for long enough time, it becomes right.
"If fifty million people say a foolish thing, it is still a foolish thing." -Bertrand Russel
-
"If fifty million people say a foolish thing, it is still a foolish thing." -Bertrand Russel
"If fifty million people say Bertrand Russel said something, it does not make it so." Anatole France. :)
-
"If fifty million people say Bertrand Russel said something, it does not make it so." Anatole France. :)
WOOOOOOW! What have I started? My sincere apologies. And if fifty million more people believe I'm not sincere, they are dreadfully mistaken! Therefore, I wash my hands of this from this moment forward. Take it where you wish.
Gary
-
"If fifty million people say Bertrand Russel said something, it does not make it so." Anatole France. :)
"Facts are stubborn things."* Unlike words where meanings can evolve over time.
This quote was from John Adams when he was defending the British soldiers who were on trial subsequent to the "Boston Massacre." That didn't exactly make him popular amongst the people at the time.
-
"Facts are stubborn things."* Unlike words where meanings can evolve over time.
This quote was from John Adams when he was defending the British soldiers who were on trial subsequent to the "Boston Massacre." That didn't exactly make him popular amongst the people at the time.
That was our 2nd President. We have come a long way to the 45th who knows a thing or two about of power of repeating "facts."
(My last OT post.)
-
I would care about any clipping. Smart histograms can indicate what colour is being clipped. And there is no reason why these options should be mutually exclusive - one can have both blinkies and a raw histogram - why not?
What I really miss from switching from my Canon to Sony it was that in the Canon I could install MagicLantern and have a real RAW histogram + Zebras. They were color zebras, so you could easily see which color channel or channels were overexposed and which areas of the image.
Camera manufacturers are just lazy... if a group of hackers was able to implement this in my old Canon 5D Mark II... they don’t do it in actual cameras because they don’t want... there is no other excuse...
http://dgpfotografia.com
-
"If fifty million people say a foolish thing, it is still a foolish thing." -Bertrand Russel
This is objectively true in my view, so “+1”.... but we are living in an age where nothing is seen as objectively true. Postmodernism has changed the landscape (mostly sloping downward and a bit slippery).
Rand
-
You ain’t gettin’ off that easy. As you say, language is fluid. So word meanings drift, phrases get truncated in some way or other. You know, like, OMG, “Him and me went to the mall.”
But on a more serious note, there’s something else often expressed in the shift/drift/fluidity of language over time. Philosophy. Let’s take the “proof” being “in the eating” in the first instance. The end product of some venture, the quality of it, is subject to evaluation and judgement by a value laden human agent. The “aptness/truth” of the saying hangs upon an assumption that good pudding is obvious to the “average person” as a value judgement, a consensus. This philosophical position sees the necessity for both process and agency as an adequate explanation for “the good ness (or not) of pudding.” There are much deeper philosophical/worldview overtones, but we can leave those for now.
Now, “the proof is in the pudding” is another kettle of fish altogether. The “proof” no longer needs any kind of agent to provide a value laden, and one might say ‘personal’ determination of pudding-goodness. No, this version says the proof is in the pudding itself, independent of any kind of agent making an evaluation. The goodness of the pudding is only determinable from the pudding itself, presumably the chemical makeup and characteristics of the pudding. This shift in the phrase is HUGE and reflects the shift toward a materialist/naturalist/reductionist philosophical frame.
Ah well, despite having lived in London for 30 years I still think that most traditional puddings (plum pudding, Christmas pudding and, of course, the unique Spotted Dick) are pretty revolting. So for me the proof of the pudding is in choosing some more refined patisserie... ;)
-
Ah well, despite having lived in London for 30 years I still think that most traditional puddings (plum pudding, Christmas pudding and, of course, the unique Spotted Dick) are pretty revolting. So for me the proof of the pudding is in choosing some more refined patisserie... ;)
Ah, us Colonists need to learn English :) In the States, when you ask a colleague if they want you to knock them up in the morning, you get sued.
-
Ah, us Colonists need to learn English :) In the States, when you ask a colleague if they want you to knock them up in the morning, you get sued.
...and not forgetting that in North America a "spotted dick" would be classified as a social disease.
-
...and not forgetting that in North America a "spotted dick" would be classified as a social disease.
Which of course reminds me of one of my favourite watering holes (https://loummorgan.wordpress.com/2010/11/23/the-curious-legend-of-dirty-dick/) in the City
-
What I would like to get:
- raw histogram only
- normal exposure mode attempting to render a neutral grey 18% card at the correct level of brightness in the scene
- configurable ETTR exposure mode attempting to expose to avoid clipping the highlights of either the worst channel of the average of the 3 channels with an average computed based on the ability of the actual raw converter I use to extrapolate highlight recovery
Cheers,
Bernard