Optimize In Camera Jpeg White Balance RGB Histogram For Better ETTR.

[digitaldog]

Exposure simple?

"While intelligent people can often simplify the complex, a fool is more likely to complicate the simple."  -Gerald W. Grumet
Some of this discussion seems to back up that concept.
Those paying to subscribe and find this idea complex, this is a good start:
https://luminous-landscape.com/the-optimum-digital-exposure/
For Franz who may or may not be a subscriber (but should just for this):
Compare the noise from the camera’s “metered” exposure at 200 ISO (Section A) and the optimum exposure at 800 ISO (Section C). The noise from the optimum exposure at 800 ISO is the same or better than the 200 ISO at the metered exposure!   
As to ISO, depends on the camera design too:
https://photographylife.com/iso-invariance-explained

[faberryman]

I imagine Andrew might jump in and perhaps deflate my theory to some extent at least.  However, it's the way I see the situation and understand what's actually happening when I raise the ISO setting.  It does not increase the amount of light striking the sensor.  The sensor is then set to gobble up more of it, and for the most part that's a good thing.  I'm also certain that Andrew would offer a much more understandable way of putting it, which I look forward to.

Except that as you increase ISO your camera meter tells you to use a faster shutter speed or smaller aperture; i.e. less exposure and less light striking the sensor.

[digitaldog]

Except that as you increase ISO your camera meter tells you to use a faster shutter speed or smaller aperture; i.e. less exposure and less light striking the sensor.

Yes, IF you follow that 'recommendation' but as Bob's article and my example show, you don't have to do so.
It's akin to someone pointing a reflective meter at a white dog on snow or a black cat on coal and blindly accepting that exposure as being anything close to being correct or ideal. Meters (and resulting Histograms) can be easily fooled. Photographers who understand exposure, how differing meters work, what to meter on and how to test exposure for differing media don't get fooled. Going full circle as to why futzing with WB and JPEG Histograms along with easily fooled reflective meters is kind of a waste of time. Instead of simply testing the sensor/meter and viewing a raw Histogram to understand how to approach exposing in the field or otherwise, AFTER testing.
As many of us did with transparency film, decades before Histograms existed.

[faberryman]

Yes, IF you follow that 'recommendation' but as Bob's article and my example show, you don't have to do so.
It's akin to someone pointing a reflective meter at a white dog on snow or a black cat on coal and blindly accepting that exposure as being anything close to being correct or ideal. Meters (and resulting Histograms) can be easily fooled. Photographers who understand exposure, how differing meters work, what to meter on and how to test exposure for differing media don't get fooled. Going full circle as to why futzing with WB and JPEG Histograms along with easily fooled reflective meters is kind of a waste of time. Instead of simply testing the sensor/meter and viewing a raw Histogram to understand how to approach exposing in the field or otherwise, AFTER testing.
As many of us did with transparency film, decades before Histograms existed.

I have been a photographer for 45 years, shot a lot of black and while film and slides in that time. I learned early on how to meter for correct exposure. Served me well as I transitioned to digital. And I still shoot film as well. Never had to go through the gyrations outlined in the OP to get correct exposure. If people thought more about their images and less about their technique, we all would be better off.

[nirpat89]

Best explained in this post from way back in 2011 and the diagram:
https://forum.luminous-landscape.com/index.php?topic=56906.msg466484#msg466484

You are basically "pushing" the first image (or "pulling" the second) to compare them side-by-side.  If the noise was not more in the former, then the high iso would be redundant. 

[digitaldog]

You are basically "pushing" the first image (or "pulling" the second) to compare them side-by-side.  If the noise was not more in the former, then the high iso would be redundant.

The first is 'pushed' as it isn't an ideal exposure for raw as 'recommended' by the meter. The 2nd appears far too bright until (using Michaels term from the original ETTR article) normalized. Had the first been optimally exposed, it too would need to be normalized and wouldn't be as noisy. But the main take away is the incorrect concept that higher ISO always produces more noise as the OP implied.

[DP]

as the ISO is increased the sensor's sensitivity to light increases

increasing the nominal ISO in a best case scenario (there are exceptions) leads to a bigger analog gain pre ADC - and that is benefical in many implementations with off sensor ADC mostly (see old Canon cameras)... or in modern processors (based on Aptina's patent) at some point in switching off extra capacitance that also decreases readout related noise (pre ADC)... exceptions are numerous - for example sometimes, at some point, in some camera cases increasing nominal ISO will read just to writing a different tag - instructing converter to do multiplication behind the scenes... or firmware might actually do that multiplication when writing raw data (post ADC)... etc, etc... what does not happen in sensors so far used in consumer cameras is that no "sensor's sensitivity to light" is changing... albeit we can't exclude some technologies in the future that indeed will do just that... but not now, sorry

[nirpat89]

The first is 'pushed' as it isn't an ideal exposure for raw as 'recommended' by the meter. The 2nd appears far too bright until (using Michaels term from the original ETTR article) normalized. Had the first been optimally exposed, it too would need to be normalized and wouldn't be as noisy. But the main take away is the incorrect concept that higher ISO always produces more noise as the OP implied.

I think when most people think about noise and iso, it is in relation to 2 shots of the same scene, one with a slower shutter and low iso and the other with faster speed and higher iso (keeping the same "metered" exposure level) the former will have lower noise than the latter.  Under those conditions, the common perception is right.  Hopefully so, otherwise I have been lugging my heavy tripod for nothing...

[digitaldog]

I don't know what most people think about ISO, exposure, noise etc. I believe a fair number don't understand how they all work.

[Aram Hăvărneanu]

The only cameras for which this matters anymore are Canon cameras. Almost (all?) every other manufacturer switched to badly-called ISOless sensors.

You can see here exactly how much it matters for your camera: http://www.photonstophotos.net/Charts/PDR_Shadow.htm

In any case, I do not see the relevancy of this to the initial posting.

[digitaldog]

That maybe so. There a heckuva of Lotta Canon cameras out there. 

[DP]

The only cameras for which this matters anymore are Canon cameras. Almost (all?) every other manufacturer switched to badly-called ISOless sensors.

new Sony sensors are not ISO-less just because of the extra capacitance switch off at certain nominal ISO (hence a noticeable difference in readout noise upon the switch off) - hence truly truly ISO-less sensors are dying breed.... what you are referring to is the migration to multiple small simple ADCs on sensor vs few off sensor ADCs and hence the decrease of importance of analog pre ADC gain...

[Frans Waterlander]

Best explained in this post from way back in 2011 and the diagram:
https://forum.luminous-landscape.com/index.php?topic=56906.msg466484#msg466484

It seems to me that that thread, and hence this one, is misleading on the issue of noise without some clarification. Yes, noise will be lower (but by how much differs from camera to camera) if correctly exposed at a higher ISO than underexposed (with the same shutter speed and aperture) at a lower ISO. But, noise (and things like posterizsation) will be lowest when correctly exposed at the lowest ISO setting. The term "correctly exposed" is of course open to further explanation/interpretation, like ER, ETTR, etc.

[digitaldog]

It seems to me that that thread, and hence this one, is misleading on the issue of noise without some clarification.

Because you can't understand it, doesn't make it misleading.    You see no difference in noise between the two examples shown here or the multiple one's shown on the other thread? Another display issue?

Yes, noise will be lower (but by how much differs from camera to camera) if correctly exposed at a higher ISO than underexposed

Unless you love non-image forming data, (noise), lower is better. But then if you're OK under exposing your image data too, I suspect you're the kind of photographer (to be kind) that doesn't care much about noise either.
Some of us do. Some of us were professional photographers. Others are just serous about the craft and quality of their images and image data. Some of us believe in GIGO:Garbage In Garbage Out. Some here should stick to in camera JPEGs or cell phones for image capture.

But, noise (and things like posterizsation) will be lowest when correctly exposed at the lowest ISO setting.

Expect when it isn't with respect to ISO and noise as shown and apparently misunderstood by some. I'll refrain from commenting on whatever posterizsation is

The term "correctly exposed" is of course open to further explanation/interpretation, like ER, ETTR, etc.

Only for those unable to understand how to view a raw Histogram.
AS was outlined in the 2011 post, with the cameras that operate as shown, you've clearly missed the advantages of HIGHER ISO with less noise:
One may use the same aperture and shutter speed (because you are somehow restricted to those settings), it can be advantageous to increase ISO because it can reduce the noise as seen above!
Achieving so called "ETTR" through ISO only makes sense when exposure is inadequate. This can occur in many shooting conditions like low light conditions, large DOF requirement, quick action, etc.

[digitaldog]

new Sony sensors are not ISO-less just because of the extra capacitance switch off at certain nominal ISO (hence a noticeable difference in readout noise upon the switch off) - hence truly truly ISO-less sensors are dying breed.... what you are referring to is the migration to multiple small simple ADCs on sensor vs few off sensor ADCs and hence the decrease of importance of analog pre ADC gain...

There a heckuva of Lotta Sony cameras out there too.
Good to know, just got one coming from years shooting Canon, will have to play around with a similar test.

[Tim Lookingbill]

I have been a photographer for 45 years, shot a lot of black and while film and slides in that time. I learned early on how to meter for correct exposure. Served me well as I transitioned to digital. And I still shoot film as well. Never had to go through the gyrations outlined in the OP to get correct exposure. If people thought more about their images and less about their technique, we all would be better off.

Except in the digital world at least for me I set exposure to capture the entire tonal scale of a scene and to do this I have to adjust exposure to preserve highlights. They'ld rather mention UniWB which is not what I was talking about or asking them to indicate whether they've tried what I've indicated in the topic.

In camera Jpeg based histograms have been a quick guide for me to make this more easier to calculate exposure on the fly "chimping" the LCD but jpeg histograms don't indicate max head room on how far to let in more light that won't wipe out a lot of detail in the highlights as an example bilious white clouds on bright sunny day.

My point in posting this topic is whether other cameras that have more head room can now use the RGB histogram in pushing exposure. No one as expected would indicate if they've ever tried this since they're cameras have a wider dynamic range than my old Pentax.

[TonyW]

Except in the digital world at least for me I set exposure to capture the entire tonal scale of a scene and to do this I have to adjust exposure to preserve highlights. They'ld rather mention UniWB which is not what I was talking about or asking them to indicate whether they've tried what I've indicated in the topic.

In camera Jpeg based histograms have been a quick guide for me to make this more easier to calculate exposure on the fly "chimping" the LCD but jpeg histograms don't indicate max head room on how far to let in more light that won't wipe out a lot of detail in the highlights as an example bilious white clouds on bright sunny day.

My point in posting this topic is whether other cameras that have more head room can now use the RGB histogram in pushing exposure. No one as expected would indicate if they've ever tried this since they're cameras have a wider dynamic range than my old Pentax.

Good luck in capturing the entire tonal scale of a scene with the 200D: there is just not enough sensor DR for many scenes so you make a compromise sacrifice shadow detail to capture highlight detail or vice versa or add light to the scene or use HDR.

Its a shame when threads turn out to be argumentative and others PoV are denigrated.

Two mentions here of UniWB, dismissed as irrelevant when the purpose of the OP seemed to indicate looking for ways to make the JPEG histogram more useful.

UniWB is a "trick". A camera set up to use UniWB is instructed to produce a distorted JPEG image. That JPEG will typically look horribly green when viewed on a display. However, the distorted image is specifically designed to produce a histogram whose right hand extremity closely matches that of a real RAW histogram.

I do not have a 10 year old Pentax to make a comparison but AFAIK all DSLR offered some form of Custom WB and also probably some form of Picture mode e.g. Natural, Bright, Vibrant etc.
Any of which when implemented by the user alter the rendering of the raw WITH the obvious change to the histogram.  This does not change the exposure but merely the rendering. 

If you decide to use the camera LCD view including histogram as an indicator of raw exposure then you may be better selecting one of the seven scene modes  of the 200D, maybe Portrait or  Landscape bearing in mind that your raw converter of choice will have added its own values and results will be different e.g. Adobe to RT etc.

[bjanes]

And my ISO 800 images show LESS noise than ISO 100 in this case because ISO doesn't have anything to do with exposure; the amount of light striking the sensor. It does have something to do with dumb meters and not too intelligent photographers who don't understand how to expose:

This is a nice demonstration of optimizing shadow detail with a camera having less read noise as the ISO is raised, but this strategy works only with a short scale scene. With a full scale scene, the highlights wold be blown out.

Regards,

Bill

[nirpat89]

This is a nice demonstration of optimizing shadow detail with a camera having less read noise as the ISO is raised, but this strategy works only with a short scale scene. With a full scale scene, the highlights wold be blown out.

Regards,

Bill

Bingo!

[digitaldog]

This is a nice demonstration of optimizing shadow detail with a camera having less read noise as the ISO is raised, but this strategy works only with a short scale scene. With a full scale scene, the highlights wold be blown out.

Regards,

Bill

Exposure 101 for digital don’t blow out the highlights.
Again, the demonstration is to illustrate that higher ISO doesn’t necessarily equate to more noise.
Edit. Don’t blow out the highlights you the photographer in control of your exposure do not wish to blow out!