Jack Hogan
Veteran Member
Overexposed
-
Welcome to the new forums! Please read this first. For known issues we are working to resolve, click here.
why no "blown pixel count" in (raw) viewfinder? Instead of histogram?
- Thread starter RussellInCincinnati
- Start date
Overexposed
GeorgianBay1939
Senior Member
Russell, while you were posting the (excellent) images above I was looking in my files for an example of where I might've upped my ISO. .... and ended up with the image at http://www.dpreview.com/forums/post/52512843 which ironically contains a lot of blown pixels!! (which I really didn't care about) I tend to ignore all of the advice not to shoot into the sun ... as I am not a Photographer, I am a snapshooter. As evidenced in my gallery !!!!Sure anything is possible, but just as sure there are many folks like myself who want almost no blown pixels in our image, hardly ever. This situation
and this one
where you probably want way less than 0.1% of the area in blown pixels, is so much more common than this one
...where you might be willing to tolerate 0.5% of the area pixels blown white.
Really in just about every one of my portrait photos, it would be a plenty enough Godsend to super quickly to read and unobtrusively in my viewfinder tell me how many blown raw pixels there are. There's such a huge class of photos where you'd want less than a thousand. A blown pixel count is plenty useful for the majority of nice images, just like there really are very few nice paintings in the world where the painter chose to leave more than 1% of the white canvas untouched.
Yes there are plenty of paintings that have untouched area of white canvas, but unless we're talking cartoons, percentage-wise there aren't. So sure it would be great to have an option for blinky raw blown pixels in our viewfinder, and/or histograms, but am claiming that usually a simple blown-pixel-counter in the corner is all the nudge I need to dial down my exposure. In fact, can't recall any static scene have photographed in the past few years, where I both allowed there to be a bunch of blown pixels, and didn't take a second image at much darker exposure settings (i.e. bracket) just to have available for merging in post-processing.
Maybe you could show us some nice images you have, with no regrets, where more than 5% of the frame is full of blanked-out white pixels?
We've come a long way from your original proposal. And ended up discussing techniques ... which I suppose influences what we want in the form of information and feedback from our cameras.
Up until very recently I used spot metering (and the histogram, and lots of chimping, and "calibrating" in raw digger.... to SURVEY the SCENE LUMINANCE, then use AELock to hold exposure, recompose etc and release. I found it necessary to spot meter as I did NOT have a photographer's eye to detect (subtle) changes in tone in the scene. Now I am a bit better at that and can get away with multiple (pattern) metering (and thumbing in +/- EC) to get ETTR when in A, S or P modes, thus being a bit quicker.
So as I am transitioning to more manual shooting with pre setting of exposure, ISO setting is becoming more significant to me. That is why I want to get it RIGHT!!!
Many thanks,
Tom
PS I find your third image to be difficult! What is more important to you ... the couple posing or the beautiful sunlit side of the skyscraper? I often wrestle with that issue when having a shadowed "subject" in a brighter "field". In many of those cases I often blow/blur the "field" . I am not sure that it would've worked with this photograph though. Difficult!
exdeejjjaaaa
Veteran Member
and in this situation blinkies (or their absence) will be perfectly fine...
and in this one too...
sure and I will see where the clipping is and dial in proper exposure to blow the sky if I want to (and may be I will blow some windows too - but I will see what I am clipping and you will be guessing)
whatever % I will be tolerating I will see where it is happening and you not
exdeejjjaaaa
Veteran Member
bracket, that's it... it is faster to bracket 3-5 shots for a natural light (even w/ a flash if you have a proper one and suitable powersource for it) than to spend time reviewing histograms during the shot (even that is a "A quick glance at your raw-calibrated histogram shows" - because then you might need to adjust and make another shot)Not if you are comfortably at ETTR already and all parameters are in their sweet spot.
But say for instance you are shooting one of your grandchildren indoors, no flash, for that great natural-light look. You have maxed out aperture at base ISO and you can't slow shutter speed further because otherwise the image will surely be blurry. A quick glance at your raw-calibrated histogram shows that you are two stops short of ETTRing the brightest desirable highlight. Do you increase ISO? With a GX7 I wouldBecause:
1) Better IQ (lower noise)
2) Brighter OOC image (easier to see if you nailed it on back screen)
3) If you nailed it, potentially no need to PP (especially for Raw+Jpeg shooters)
Jack
Last edited:
Jack Hogan
Veteran Member
Not to me, in the context of a non-ISOless camera like yours.
If you are in M mode you are in total control. So if IQ is important to you your first order of business should be to maximize Exposure (ss and aperture only) given your technical/artistic constraints - so choose the slowest shutter speed you think you can stand (say between 1/50th and 1/100th with 'posing' children) and the largest aperture (say f/5.6 in your example above - this is probably also your lens' sharpest f/#). Assuming nothing important is blown Exposure is now set and you can leave it alone.
Your second order of business in maximising IQ is to fine tune the in-camera processing. If the live histogram/blinkies of your GX7 show that you are far from The Right at base ISO, it behooves you to increase the ISO until you are fairly close To The Right. Anything important blinking? No, increase ISO*, check, iterate until clipping, back a little off to keep a safety margin, done. ISO set.
From that moment on, in that exact setting, forget about camera settings and shoot away with abandon concentrating on composition and capturing the moment - confident that you are capturing the best quality information possible from the scene. If the scene conditions change, re-evaluate in light of the new situation.
Jack
* With your GX7, should you be increasing ISO past 800 given the table I posted earlier? Probably not, because the advantage you gain in lower noise is minimal compared to the loss of DR. Should you stop at ISO400? In non DR critical situations I would. Note that this last paragraph has nothing to do with Exposure, which a Manual shooter always evaluates and sets independently of ISO.
Last edited:
Jack Hogan
Veteran Member
Good idea. On the other hand with a moving grandchild...bracket, that's it... it is faster to bracket 3-5 shots for a natural light (even w/ a flash if you have a proper one and suitable powersource for it) than to spend time reviewing histograms during the shot (even that is a "A quick glance at your raw-calibrated histogram shows" - because then you might need to adjust and make another shot)Not if you are comfortably at ETTR already and all parameters are in their sweet spot.
But say for instance you are shooting one of your grandchildren indoors, no flash, for that great natural-light look. You have maxed out aperture at base ISO and you can't slow shutter speed further because otherwise the image will surely be blurry. A quick glance at your raw-calibrated histogram shows that you are two stops short of ETTRing the brightest desirable highlight. Do you increase ISO? With a GX7 I would
1) Better IQ (lower noise)
2) Brighter OOC image (easier to see if you nailed it on back screen)
3) If you nailed it, potentially no need to PP (especially for Raw+Jpeg shooters)
Jack
Jack
PS This exchange is really about learning how to set the camera up to capture the best IQ possible, shot by shot.
GeorgianBay1939
Senior Member
Overexposed
"Overexposed" or overbrightened? And what short of indicator would you have liked to tell you just how much you "overexposed"? Olympus blinkies would've been blinking all over that ice (except for less reflective painted goal crease and the goalies shadows) and in the floodlights. My histogram would've had a high peak at the right and a smaller peak on the left, giving me the usual quandary ... how much blowing of highlights am I willing to tolerate, eh?
BTW I just looked at the RawDigger rendition of that shot of my grandkidlet that I sent earlier....
3 MP overexposed, 186 P underexposed. PERFECT .... using artistic license!!! LOL
Would something like the above ... toggled off of the RAW file, with better contouring of degree of blowing/clipping, be helpful? If RawDigger can do it why can't manufacturers do it ... simply?
I am beginning to conclude that maybe I should just bracket and hope that I get a good one. I know some commercial photog friends who shoot bracketed JPEGS and pick the best ones ... as a routine matter. We have stopped arguing about their beloved exposure triangle as it is of no use!!!!
One says, " Why would I study that? I only study on "a need to know" basis. If I need to know something about the mechanics of exposure I'll learn that. But so far after 40 years of shooting film and digital, I don't. Besides I am an Artist, not a fricking technician."
Ah, the fun of it all, eh?
Tom
PS I am starting to "delve" into GX7 data you provided. Need to use those complex equations/explanations on Wiki, though. Good exercise for the synapses !
GeorgianBay1939
Senior Member
Right. I have often done that when I had some prep time to set the camera up. As above with the ISO at BASE as that is where I usually leave it.
Ok, I got the * below. BUT we still have the following to deal with:
That is the issue!
Russell would say to leave it at base ISO. You say increase the ISO to move the histogram to the right. Russell would say that this introduces photon noise, I believe.
Russell says : The reason you would always reduce ISO gain to ISO 100 in this case, before resorting to reducing exposure, is that no photographer would ever want to increase photon shot noise significance and throw away big chunks of dynamic range, in search of some trivial gain in sensor read noise.
I am at his mercy because I do not have the background to challenge his statement above. Maybe I'll have to ask Russell to further explain his statement.
AND THAT IS THE BEST PART!!!!! (it brings out the lil bit of artiste in this old f@rt engineer's soul!
Makes sense!
I always use exposure to be affected by scene luminance, f/, shutter interval. INDEPENDENT OF ISO.
So this last paragraph has to do with GAIN, the amount of boost the ADC is giving to the signal from the sensels. The more in-camera gain, the less gain required in later post raw conversion brightening.
Right?
T
GeorgianBay1939
Senior Member
I bracket now, when I know that I need to. I used to have to spot meter to know the LR (luminance range) of the scene. Now my eyes are a bit better trained and so I can now (most of the time) use multiple (pattern) metering to get a good idea of exposure and then to use whatever appropriate exposure dial to ETTR looking at the histogram.bracket, that's it... it is faster to bracket 3-5 shots for a natural light (even w/ a flash if you have a proper one and suitable powersource for it) than to spend time reviewing histograms during the shot (even that is a "A quick glance at your raw-calibrated histogram shows" - because then you might need to adjust and make another shot)Not if you are comfortably at ETTR already and all parameters are in their sweet spot.
But say for instance you are shooting one of your grandchildren indoors, no flash, for that great natural-light look. You have maxed out aperture at base ISO and you can't slow shutter speed further because otherwise the image will surely be blurry. A quick glance at your raw-calibrated histogram shows that you are two stops short of ETTRing the brightest desirable highlight. Do you increase ISO? With a GX7 I would
1) Better IQ (lower noise)
2) Brighter OOC image (easier to see if you nailed it on back screen)
3) If you nailed it, potentially no need to PP (especially for Raw+Jpeg shooters)
Jack
I think that my motivation in making images is different than many photographers.
Many photographers want the best possible shot, in the most reliable way, with minimum fiddle time and minimum post processing.
Me? I treat each image as a challenge to learn more about the science and art of photography. Sure, I want good imagery, but I want to learn as much as possible from the experience. So I prefer to really study the scene, play with the dials, play with composition, pov etc .... and then to play in Lightroom, (which I now enjoy) to see what I can do ... stretch those sliders. I have lots of time!
As a result I get a lot of cr@p, but I also get a lot of enjoyment LEARNING how to make good images. When you get to be an old f@rt like me the Learning part of the experience is precious!
t
GeorgianBay1939
Senior Member
Yes, and I appreciate it.Good idea. On the other hand with a moving grandchild...bracket, that's it... it is faster to bracket 3-5 shots for a natural light (even w/ a flash if you have a proper one and suitable powersource for it) than to spend time reviewing histograms during the shot (even that is a "A quick glance at your raw-calibrated histogram shows" - because then you might need to adjust and make another shot)Not if you are comfortably at ETTR already and all parameters are in their sweet spot.
But say for instance you are shooting one of your grandchildren indoors, no flash, for that great natural-light look. You have maxed out aperture at base ISO and you can't slow shutter speed further because otherwise the image will surely be blurry. A quick glance at your raw-calibrated histogram shows that you are two stops short of ETTRing the brightest desirable highlight. Do you increase ISO? With a GX7 I would
1) Better IQ (lower noise)
2) Brighter OOC image (easier to see if you nailed it on back screen)
3) If you nailed it, potentially no need to PP (especially for Raw+Jpeg shooters)
Jack
Jack
PS This exchange is really about learning how to set the camera up to capture the best IQ possible, shot by shot.
Thank you.
t
GeorgianBay1939
Senior Member
I am not an Olympus user but have friends who swear by the blinkies. Both under and over exposure. I have seen them a couple of times and wish that I had them in my Pannys.
Do the blinkies come off of the RAW DATA or off of the (processed) jpeg data that is used to produce the image in the EVF/LCD?
Can you set the levels at which they become active? Or do they only indicate "overruns" or whatever in the shadows?
Tom
I have to leave for a bit. Be back later.
Jack Hogan
Veteran Member
C'mon Tom, you can do it
A few questions (about the case in this subthread):1) Does ISO have anything to do with Exposure?
2) If Exposure is unchanged, doesn't that mean that the signal (the mean number of arriving photons) is also unchanged?
3) If the mean number of arriving photons is unchanged, wouldn't photon noise also be unchanged?
4) If ISO has only to do with processing the captured photoelectrons from the given unchanged Exposure, what happens to the full well count in stops when raising ISO from base to 400? To 800? Refer to the chart of the GX7 above.
5) So you have lost that many stops at the top end. What has happened at the same time to your engineering eDR?
6) And if it has dropped so much less, doesn't that mean that your deep shadows have improved by the difference?
7) So if by raising ISO that way no desirable highlights were clipped, would your IQ maximizing good self feel compelled by his conscience to do it?
Right!
Jack
PS. N, Y, Y, -1.5, -2.5, -0.8, -1.6, +0.7, +0.9, Y, heck yeah
Last edited:
C
Christof21
Guest
No, Russell gave a detailed explanation of the process when you have blown pixels. He just says to start with decreasing ISO.
When you have place to go to the right in your histogram after maximizing the exposure, this is another context.
This is not at all contradictory. If the histogram is on the left, start maximizing the exposure and after the ISO. If this is on the right (blown pixels), start decreasing the ISO and after the exposure if there are still blown pixels. This is the reverse process.
Both processes give the same settings results.
You do not have to care about the cause of the blown pixels, whether it is over exposure or too high ISO. Just follow this simple process.
GeorgianBay1939
Senior Member
NoC'mon Tom, you can do it
1) Does ISO have anything to do with Exposure?
Yes
Yes. I (re)read Shot Noise and recalled that Shot Noise is significant with small signals, small numbers of photons being collected, therefore in the shadows. If shot noise is amplified (along with signal) by increasing gain, it becomes significant in the image.
Base ISO is said to be ISO 200 (I don't know why.)
ISO 200: FWC= 13501 electrons
ISO 400: FWC= 6663 electrons Diff= 6838
ISO 800: FWC= 3375 electrons Diff= 3288
Increasing the ISO by a stop decreases the FWC by ~1/2 .... which seems to be a pattern all of the way up the gain scale. Does that mean that since the capacity of the well is halved by each stop increase in ISO that a constant gain factor is applied at the sensel level as the ISO is increased stop by stop?
Stops?
Stops, f-stop conventions and exposure
In photography, stops are also a unit used to quantify ratios of light or exposure, with each added stop meaning a factor of two, and each subtracted stop meaning a factor of one-half. The one-stop unit is also known as the EV (exposure value) unit. On a camera, the aperture setting is usually adjusted in discrete steps, known as f-stops. Each "stop" is marked with its corresponding f-number, and represents a halving of the light intensity from the previous stop. This corresponds to a decrease of the pupil and aperture diameters by a factor of 1/sqrt2 or about 0.7071, and hence a halving of the area of the pupil.
So with each increase in stop of ISO the FWC is decreased by a stop. WRONG ANSWER!!
If we were to compensate for the decrease in the CAPACITY for electrons in the wells for EACH ISO we would have to open the f/ by 1/sqrt2 or 0.707 for each ISO stop. So if we were to increase ISO by 7 stops from ISO200 to ISO25600 we would have to close the relative aperture f/ by 7 stops.
In other words, a little trial:
Hold the shutter speed constant at say 1/1000 second:
and expose at f/2.8 at ISO 200 to get say 13501x electrons in the wells.
If you maintained the same exposure (f/, ss) and cranked ISO to 25600 you'd only get ~98x electrons before exceeding capacity.
These ~98x electrons in the wells would be equivalent to shooting at f/22 (you would have to stop down by 7 stops, to f/22) at ISO 200.
OR
Hold the relative aperture at f/2.8,
and expose at 1/15 at ISO 200 to get say 13501x electrons in the well.
If you maintained the same exposure (f/, ss) and cranked ISO to 25600 you'd only get ~98x electrons before exceeding capacity.
These ~98x electrons in the wells would be equivalent to shooting at 1/1000 (you would have to shorten exposure interval down by 7 stops, to 1/1000) at ISO 200.
_________________________________________________________________________
I have to break here. To be continued tomorrow, I hope.
Maybe, in the meantime you could tell me how you get your answers to #4. (I get -1 or -0.7 in f/ "equivalents" per change in ISO)
Thanks,
t
_____________________________________________________
f/2.8, f/4, f/5.6, f/8, f/11, f/16, f/22
1/1000, 1/500, 1/250, 1/125, 1/60, 1/30, 1/15
Jack Hogan
Veteran Member
Hi Tom,
The reference for noise in DSCs is this treatise by Professor Martinec, read it at your leasure. There you will find that light is statistical in nature and random noise is measured as the standard deviation of the mean signal. Noise inherent in natural light (photon/shot noise) follows poisson statistics, where the standard deviation is the square root of the mean (of the number of photons, the signal). So if the number of photons hitting a sensel (i.e. Exposure) is unchanged, so will its square root, shot noise.
Right. Here is the table again for easy reference. Recall that Exposure is linear with the number of photons hitting a photosite, which is linear with the number of photoelectrons generated by it, which is linear with the Raw value written to the Raw file. The relationship is:
Raw Value Written (ADU) = NumberofPhotonsArrivingatphotosite x EffectiveQE x CameraGain. So if 20k photons arrive at the photosite during the given Exposure (determined by scene, ss and aperture only), about 2780 electrons (20k*13.9%) will be captured by it. What value will be written to the Raw file?
Preliminary Approximate Values
That depends on the gain applied by the electronics downstream of the sensor, which is controlled by the ISO dial on your camera. If the camera was at base ISO (I believe 125 for the GX7) the inverse of gain would be 4.67 e-/ADU (see table, igain counterintuitive but so by convention). So the Raw value written to the file would be approximately 595 ADU (= 2780e- / 4.67 e-/ADU). If it was at ISO 800, 3390 ADU. If at ISO 1600 it would be clipping at this exposure, since the GX7 is a 12 bit camera and its maximum Raw value is 2^12-1.
The column mis-indicated as FWC represents the number of electrons that will result in clipping for each camera gain. It goes down because as you raise gain you reach clipping sooner, just like in the example above.
As you know in photography we tend to think in stops: 1 stop = a doubling/halving of the signal (Light/Exposure/number of photons/number of electrons aotbe). For example doubling exposure time from 1/100s to 1/50s will result in 1 additional stop of light aotbe [log2(2)]. Twice the aperture diameter = four times the area = 2 stops more light [log2(4)]. Hence the linear and quadratic relationship of EVs and time and f/# respectively.
So if at camera ISO125 19126e- result in clipping, but at ISO400 only 6663e- do, that means that by increasing ISO from 125 to 400 the maximum signal we can record was lowered by log2(19126/6663)= 1.52 stops. If the noise 'floor' has remained unchanged, it follows that DR has dropped by that much.
But has the noise floor remained unchanged when increasing camera ISO from 125 to 400? It hasn't, for reasons that are the subject of another post:
Read Noise (that is the noise added by the sensor and camera electronics during capture) has improved from 5e- to 2.7e-, or +0.89 stops at the higher ISO. So raising ISO has resulted in a lower potential maximum signal recordable, but also a lower noise added and hence a lower minimun acceptable signal. Therefore DR has not dropped by -1.52 stops (looking at clipping only), but by +0.89 stops less, or a total of -0.63 stops as you can see in the relative column*
The question is why you would not do it if there are no desirable highlights that would clip: you lose nothing and you have lower noise in the deep shadows. A no brainer. Of course the answer would be different if the maximum desirable signal were higher, and increasing gain/ISO would cause it to be clipped.
So the next question becomes: assuming we start at base ISO with the max possible Exposure given artistic constraints, how high should one raise the ISO as long as no desirable highlights are clipped?
The answer is as high as the read noise keeps dropping meaningfully (it's for you to decide what meaningful is according to the situation): no more and maybe a little less because you don't want to come home and realize that you should have kept an additional stop of highlights which instead you gave up for a measly 0.2 stops better SNR in the deep shadows - what I think Russel was driving at earlier ;-)
*My question in the previous post was about eDR which is more correct, but I realize that's an unneeded complication at this stage. In that case the correct answers are above.
Last edited:
RussellInCincinnati
Senior Member
Agreed this image is an example of where my proposal of a blown-raw-pixel counter would not give you enough exposure information. And I doubt a full histogram of raw pixel values would help much either. For this photo you would really need the blinkies, so that you could differentiate quickly between background blown pixels (which you find acceptable) and blown pixels on the model's cheeks (which you and most viewers probably don't want).
Not sure how many photos in the world are like your good counter-example. Could you at least agree that having a blown raw pixel counter is at quite useful, in comparison to having no raw pixel exposure info in most camera viewfinders at all?
Was looking at my many years old light meters (including an electronic flash meter) the other day, and thinking about the way that electronic viewfinders make not only light meters, but also autoexposure in general, pretty much archaic. It's way too easy with LCDs and electronic viewfinders to get a good exposure so quickly, with 100% confidence that things are not wildly overexposed. A confidence that no autoexposure system could ever give you in the field. And who cares about a flash meter any more, when you can take a test flash and infinitely scrutinize the sample image 1.5 seconds later.
Probably a good default or beginning strategy with any high-range-of-luminances scene, is to get a print where nothing is completely black or completely white. For the same reason that most painters of things besides cartoons, do not plan on leaving large blotches of white canvas or spilled black ink.
In general rule #1 is to risk underexposure more than overexposure. Lost highlight detail is so much more noticeable than lost dark detail.
It is so easy and fun, with raw files, in something like Lightroom or RawTherapee to move exposure and highlight recovery sliders around, and see what you can wrestle out of a difficult-scene image. The cameras that have a high dynamic range rating in DxOmark are in general best at this game. My Nex C3 is rated as 12.3 EV's of dynamic range, so 12.3 is apparently reasonably high dynamic range if my sample photo is any indication.
If it's a super-impossible-range-of-brightnesses scene, you simply must bracket exposures in the field so that you have a couple of photos from which you can piece/clone together something reasonable.
RussellInCincinnati
Senior Member
Well this indeed makes a fairly good case for wanting blinky blown raw pixels shown in your camera viewfinder. Because a simple blown pixel count would not tell you whether or not anything besides the ice is blown high.
Or maybe the raw counter would have been OK, with a bit of fiddling. Couldn't you have played with this exposure while watching a blown pixel counter? At way underexposure, the count is near zero. As you increased the exposure, the count would jump to a couple of thousand as the lights blew out. Then suddenly the count would go to 3 million as the ice blew out, and you would stop right there or perhaps back off the exposure just until the 3 million went away.
exdeejjjaaaa
Veteran Member
then bracket even more, those pesky kids...Good idea. On the other hand with a moving grandchild...
well, sometimes the situation does not allow you to make a good decision in a timely manner and then nothing prevents you from making a good decision and still bracket (better be safe than sorry).
RussellInCincinnati
Senior Member
Hey speaking from experience, you can make a lot of photos that make a lot of people happy without being an "artist." Am no artist myself, am a journalist...but even so, just clearly capturing what's in front of you, just showing up and doing the work, is in and of itself enough to make the world a slightly nicer place.
RussellInCincinnati
Senior Member
Nobody has said blinkies can't supply the same information as a single blown pixel counter. The question is, what is the simplest, easiest-to-implement-in-firmware and least obtrusive bit of overexposure information conceivable? My claim is that a blown raw counter is much easier and less obtrusive for camera manufacturers to implement than blinkies. I.e. if we can't have blown-raw blinkies, can we at least get a blown-raw counter?
You're acting like most photographers most of the time choose to blow out large parts of their photos. I think that's way off, most photographers most of the time don't hardly want to ever blow out anything but a few specular highlights.
Why don't you show us some great photos you've taken, where there's lots of blown out pixels and you have no regrets about them?
Here you are acting like a regular electronic viewfinder doesn't already give you most of that information. In fact, if you can't see instantly in a regular viewfinder where the blown pixels are, would further claim that the blown pixels aren't an important issue in that photo in any case.
Also haven't heard you acknowledge a single positive thing about the minimalist adequacy of a simple blown pixel counter. Seems like you're searching for negatives, more than mulling over everything looking for best practices.
Keyboard shortcuts
- f
- Forum