Need help with exposure question

To get a 'proper and usable image' we set many things including ISO, focus mode, meter mode, actual focus but these are not a part of the exposure apart from the fact that you cannot get an image without setting them.

It is the 'exposure triangle' that caused the op's original concerns.
 
Last edited:
No one has answered my question about how you achieve proper exposure without considering ISO to do it.
See post #29, above.

No, in no way. That just keeps on talking around it, intently trying to evade the specific question. He should be a politician (meaning, a politician of the worst kind). :) Along with his adding some really stupid stuff like: "ISO does not cause noise. If anything what ISO does suppresses noise. " Really laughable.

The question meaning was this: Suppose you want to take a picture of the crowd on Main Street, and you raise your camera to adjust for proper settings and do it. The Suns illumination does affect the brightness of the image you will see (as certainly also does the aperture, shutter speed and ISO selected), so there are the three settings necessary in an appropriate matched group for this proper exposure: aperture, shutter speed, and ISO. First Basics of photography. Each of the three requires selecting suitable choices for the situation (due to depth of field, motion blur, and sensor noise), or we could allow automation to chose something to do it (but other than exposure, it has no knowledge of the scenes actual requirements), but all three settings must absolutely be addressed. ISO is one of the settings we are required to make for a proper exposure (proper meaning so that we do get a usable image file with an acceptable histogram... there is no other way to get a proper image file out of the camera other than to select the three suitable settings). ISO is one of those extremely necessary settings to deal with, arguably one of the most important.

It is argued that digital ISO is just a gain amplification factor after actual instant of "exposure", which is just show-off clever in this context, but it is not useful in practice, or to newbies, since there is no other way to get the proper image file out of the camera without setting a suitable ISO too. Any image file and any histogram WILL have this ISO result in it. It is what the camera does. It is what ISO does. Setting ISO is one of the required ways we deal with situations, in order to get a proper and usable file out of the camera. The three basic essentials required for newbies (or automation) to get a proper image is aperture, shutter speed, and ISO. This was of course true of film too.

My question, still totally ignored, is: When standing on Main Street, how does one take any proper and usable image without concern for setting a usable ISO? It is not a question about semantics of terms, it is about when standing on Main Street, what does one do without setting ISO? Agreed, it is a stupid question (we all know the answer, we set ISO), but this has become a stupid thread.

I answered your question directly. If you're not going to take the time to read through a thread you shouldn't keep returning to the thread and re-posting the same errors that have already been addressed. Such arrogance set in ignorance makes you look very much the fool. And I'm still waiting for you to repost that definition of exposure with ISO underlined.

Joe

Oh and, what happened to this: "Either way, there is no point or any desire to continue this dumb stuff with you. Have fun."
 
I think a lot of people think that by upping the ISO they are increasing the sensitivity of the sensor by juicing it up. In fact, cranking up the ISO is more like turning up the output gain. The weaker the signal, the noisier it becomes. Image sensors are analog devices that have a noise floor. The best way to avoid noise is proper exposure at the base ISO and, exposure is determined by lens aperture and shutter speed as determined by the camera's light metering system or a handheld meter. Which means always shooting in good light, or shooting with a tripod, or shooting with flash or studio lights. If you are shooting moving subjects or shooting with a tripod is impractical, you will need to up the ISO to a point that will let you use the action stopping shutter speeds and smaller aperture for greater DoF. I look at exposure vs ISO as a compromise to get the action stopping and DoF I need with an acceptable amount of noise that can be hidden in post processing. dSLR's have gotten so good these days that I shoot in manual with auto ISO.
 
Last edited:
I think the main problem is the same as it always is, that too many use their knowledge of how a camera works to understand light rather than how light works to understand the camera. If you like, on the simplified diagram below, they read from the right rather than reading from the left:

ex-1.jpg


With digital you can add that the less light you use, (the lower the exposure), to create an image with the desired brightness then the more noise you will see. The calibrated brightness levels programmed into your camera (ISO) take this into account and so are also programmed to remove a little of that noise. User defined brightness, deliberate underexposure in relation to the calibrated brightness levels programmed into your camera (ISO), does not and so no noise reduction takes place.

Simple...
 
Last edited:
Wow, such debate of terminology :D
Just to clarify, I believe the following is correct:
  • Due to post processing or the exposure triangle; 99% of the photographers state exposure includes three aspects, shutter speed, aperture and ISO.
  • Technically, as a few have pointed out. Exposure is just a measure of the amount of photons which hit the sensor or film.
  • ISO settings in the age of film affected the sensitivity of the film to light. In digital, it affects how the raw sensor data is utilized. Note: How this is done is camera manufacturer specific, and often tied to the sensor itself.
  • In digital, noise is defined as invalid or unwanted data. As you increase ISO, you are actually making the sensor more sensitive, however this comes at a cost. The cost is you actually increase noise, thereby reducing the signal to noise ratio.
  • For digital ISO, there are three general methods, a straight voltage gain applied to the sensor data before analog is converted to digital, or the gain is applied to the digital values after conversion, or a hybrid which combines the first two methods. Based on reading on the net, most newer cameras are hybrids, and involve anti-aliasing and other features after the digital conversion in an attempt to reduce noise.
  • When you shot in RAW, ISO is baked into the data. You cannot adjust it afterwards with darktable, Lightroom, Photoshop, GIMP...
  • When you shot in JPEG, ISO is baked into the data. You cannot adjust it afterwards with darktable, Lightroom, Photoshop, GIMP...
  • darktable, Lightroom, and other products can adjust EV. EV is this concept is often a short hand for Exposure Value. This is done by formulas which amplify or decrease values in specific color channels across individual pixels. Each software has its own solution, and can often approximate each other; however they are not exact matches.
  • I have never seen the ability to take a picture with an ISO of 100, then keep the same shutter and aperture; and take a second shot with an ISO of 800, and be able to get them to match identically in view and histogram afterwards with post processing software.
The first point, and the last point is by far the most critical in terms of semantics. You can wail against the tide all you want, but at the end of the day, exposure outside of a technical discussion has come to mean the combination of ISO, Shutter Speed, and Aperture.

The inability to effectively adjust the ISO afterwards with software in post processing, means that you must include ISO as you take the picture. A long time ago, people may have differentiated between "exposure" and "camera exposure"/"proper exposure"; however with almost no photographer using or needing to technically define "exposure" as just the light photons hitting the sensor, the first adjective (camera or proper) has been dropped from common usage. Leaving us where we are today, exposure means the triangle.

Tim
 
Wow, such debate of terminology :D
Just to clarify, I believe the following is correct:
  • Due to post processing or the exposure triangle; 99% of the photographers state exposure includes three aspects, shutter speed, aperture and ISO.
  • Technically, as a few have pointed out. Exposure is just a measure of the amount of photons which hit the sensor or film.
  • ISO settings in the age of film affected the sensitivity of the film to light. In digital, it affects how the raw sensor data is utilized. Note: How this is done is camera manufacturer specific, and often tied to the sensor itself.
  • In digital, noise is defined as invalid or unwanted data. As you increase ISO, you are actually making the sensor more sensitive, however this comes at a cost. The cost is you actually increase noise, thereby reducing the signal to noise ratio.
No. You can't do anything to alter the sensitivity of the sensor. It is fixed in manufacture and ISO changes do not either raise or lower it. The noise most photographers see is not a function of anything ISO does to the sensor.
  • For digital ISO, there are three general methods, a straight voltage gain applied to the sensor data before analog is converted to digital, or the gain is applied to the digital values after conversion, or a hybrid which combines the first two methods. Based on reading on the net, most newer cameras are hybrids, and involve anti-aliasing and other features after the digital conversion in an attempt to reduce noise.
    [*]When you shot in RAW, ISO is baked into the data. You cannot adjust it afterwards with darktable, Lightroom, Photoshop, GIMP...

    [*]When you shot in JPEG, ISO is baked into the data. You cannot adjust it afterwards with darktable, Lightroom, Photoshop, GIMP...
    [*]darktable, Lightroom, and other products can adjust EV. EV is this concept is often a short hand for Exposure Value. This is done by formulas which amplify or decrease values in specific color channels across individual pixels. Each software has its own solution, and can often approximate each other; however they are not exact matches.
    [*]I have never seen the ability to take a picture with an ISO of 100, then keep the same shutter and aperture; and take a second shot with an ISO of 800, and be able to get them to match identically in view and histogram afterwards with post processing software.
I have to run right now for a 1:00 pm appointment. That's trivially easy to do and I'll be happy to show you an example this afternoon when I return.

The first point, and the last point is by far the most critical in terms of semantics. You can wail against the tide all you want, but at the end of the day, exposure outside of a technical discussion has come to mean the combination of ISO, Shutter Speed, and Aperture.

Given that I can thoroughly refute your last point (hang in there) you may want to reconsider this conclusion.

Joe

The inability to effectively adjust the ISO afterwards with software in post processing, means that you must include ISO as you take the picture. A long time ago, people may have differentiated between "exposure" and "camera exposure"/"proper exposure"; however with almost no photographer using or needing to technically define "exposure" as just the light photons hitting the sensor, the first adjective (camera or proper) has been dropped from common usage. Leaving us where we are today, exposure means the triangle.

Tim
 
ISO settings in the age of film affected the sensitivity of the film to light.

Camera settings don't affect the sensitivity of film one single jot... What the ASA/ISO setting did was adjust the exposure meter to a calibrated sensitivity (the sensitivity which is the ASA rating on the film) so it would indicate the *suggested* combination of aperture and shutter speed that would create the best exposure for film of that sensitivity.


The first point, and the last point is by far the most critical in terms of semantics. You can wail against the tide all you want, but at the end of the day, exposure outside of a technical discussion has come to mean the combination of ISO, Shutter Speed, and Aperture.

Exposure simply means the amount of light you allow to hit the sensor, or the range of point intensities recorded by the sensor.

In digital, noise is defined as invalid or unwanted data. As you increase ISO, you are actually making the sensor more sensitive, however this comes at a cost. The cost is you actually increase noise, thereby reducing the signal to noise ratio.

The sensitivity of the sensor remains unchanged, ISO only adjusts the output brightness and does a noise reduction. It is supposed to be transparent as in it does the calculation and post processing for you, but the downside to decreased exposure is a lower signal to noise. it's not a product of higher ISO, (higher ISO settings decrease noise because they are also programmed to remove some of it). All you do by setting a higher ISO is to recalibrate the meter so if reflects the *suggested* exposure for the brightness setting you've chosen to process the data at.
 
@Ysarex

Make sure to include the histogram for each before and after :)

Sent from my SM-J737T using Tapatalk
 
Wow, such debate of terminology :D
Just to clarify, I believe the following is correct:
  • Due to post processing or the exposure triangle; 99% of the photographers state exposure includes three aspects, shutter speed, aperture and ISO.
  • Technically, as a few have pointed out. Exposure is just a measure of the amount of photons which hit the sensor or film.
  • ISO settings in the age of film affected the sensitivity of the film to light. In digital, it affects how the raw sensor data is utilized. Note: How this is done is camera manufacturer specific, and often tied to the sensor itself.
  • In digital, noise is defined as invalid or unwanted data. As you increase ISO, you are actually making the sensor more sensitive, however this comes at a cost. The cost is you actually increase noise, thereby reducing the signal to noise ratio.
  • For digital ISO, there are three general methods, a straight voltage gain applied to the sensor data before analog is converted to digital, or the gain is applied to the digital values after conversion, or a hybrid which combines the first two methods. Based on reading on the net, most newer cameras are hybrids, and involve anti-aliasing and other features after the digital conversion in an attempt to reduce noise.
  • When you shot in RAW, ISO is baked into the data. You cannot adjust it afterwards with darktable, Lightroom, Photoshop, GIMP...
  • When you shot in JPEG, ISO is baked into the data. You cannot adjust it afterwards with darktable, Lightroom, Photoshop, GIMP...
  • darktable, Lightroom, and other products can adjust EV. EV is this concept is often a short hand for Exposure Value. This is done by formulas which amplify or decrease values in specific color channels across individual pixels. Each software has its own solution, and can often approximate each other; however they are not exact matches.
  • I have never seen the ability to take a picture with an ISO of 100, then keep the same shutter and aperture; and take a second shot with an ISO of 800, and be able to get them to match identically in view and histogram afterwards with post processing software.
The first point, and the last point is by far the most critical in terms of semantics. You can wail against the tide all you want, but at the end of the day, exposure outside of a technical discussion has come to mean the combination of ISO, Shutter Speed, and Aperture.

The inability to effectively adjust the ISO afterwards with software in post processing, means that you must include ISO as you take the picture. A long time ago, people may have differentiated between "exposure" and "camera exposure"/"proper exposure"; however with almost no photographer using or needing to technically define "exposure" as just the light photons hitting the sensor, the first adjective (camera or proper) has been dropped from common usage. Leaving us where we are today, exposure means the triangle.

Tim
1. Could we theorize that all of them are wrong?
2. There is no such thing as "photons" except in SciFi.
3. Film sensitivity is fixed by the manufacturer.
4. A digital sensor cannot be made more or less sensitive. It is fixed at the time of manufacture.
5. Your statement has proved that ISO is not involved in exposure.
6.
7.
8.
9. What you are looking at is the JPEG that your camera's firmware has generated in order to display a viewable image.

Since you are locked into the "exposure triangle", let me add another variable; the light. So now you have an "exposure square" to learn and pass on to the newbs. Try not to confuse them.
 
After you explain that exposure is shutter and aperture, to beginners, it's time to let them move on to understand the the triangle since they have to deal with ISO a third element. Beyond this, it's just word games we're playing.
 
After you explain that exposure is shutter and aperture, to beginners, it's time to let them move on to understand the the triangle since they have to deal with ISO a third element. Beyond this, it's just word games we're playing.

Unfortunately on a forum we have to use words to create a picture of a concept. It is not a *word game* but us trying to find a different set of words that will trigger a different understanding. That some see this as a word game suggests that they automatically dismiss, or will not consider, thinking of the relationships in any other way.

The biggest problem with exposure is that we use a fixed setting of say 125sec at f5.6 to supply a *value* but exposure is in fact a range of brightnesses that we try to record on a media that has a fixed range that it can record well. In film it was easier to understand in that film had a fairly fixed sensitivity so we could use a calibrated *standard* to align our *average light reading* of the scene to align with the *optimum* range of densities we produced on that film to form an image. Our variations from that calibrated *norm* was where experience and understanding of both light and the media came into effect. This is ASA or ISO.

Again the image is a range of densities produced by the different reactions of the film to different intensities of light, and again 125sec at f5.6 are only the settings we use to ensure that we capture the important parts of the scene well within the range that is optimum for the film to record. The *exposure* of 125sec at f5.6 does not equate to any quantity of light but is simply a measurement of the length of time we expose against the diameter of the hole in which we allow light to pass. It may produce either high or low key, it may produce and image with blocked shadows or blown highlights, the snow may be grey along with the cat... The way we align this reading with the sensitivity of our media is through ISO.

With digital we do not have that *floor* or the minimum amount of light we need to produce a reaction the way we did in film and therefore record a density that becomes the image. So with digital we use ISO as a calibration of *brightness* and is linked to a more complex and predictive version of the *average light reading*, but these are still pre-programmed settings on the camera. But it is important to understand that sensors have a fixed sensitivity and that because there is no minimum *floor* that we can *under-expose* and still record an image. ISO then is just a way of adjusting the brightness of the exposure that you made to produce the same range of brightnesses that you would see if you used the *standard* exposure for the base sensitivity of the sensor (which never changed). You do not adjust the sensitivity of the sensor but simply under-expose it and adjust the brightness of the output (post-process) and add a little noise reduction.

As in my original diagram, photography is based on an understanding that a camera collects and focuses light reflected from a scene on light sensitive media. A good understanding of both the light reflected and the way the media reacts to light is essential in understanding how to manipulate this process. All the exposure triangle does is to echo the way the camera is programmed to make this understanding transparent, that you don't need to understand it but merely *point and shoot an let the automation take care of it*. It explains nothing or gives you any real clue, it merely says *accept that the way we've programmed your camera is correct*.
 
Wow, such debate of terminology :D
Just to clarify, I believe the following is correct:
  • In digital, noise is defined as invalid or unwanted data. As you increase ISO, you are actually making the sensor more sensitive, however this comes at a cost. The cost is you actually increase noise, thereby reducing the signal to noise ratio.

Tim
No, as you increase ISO you are either boosting the analog signal from the sensor pre A/D or digitally adding brightness gain post A/D. Sensors have a noise floor that gets boosted as you increase brightness pre or post A/D.
 
@Ysarex

Make sure to include the histogram for each before and after :)

Sent from my SM-J737T using Tapatalk

Not sure what you mean or want in the way of a "before" histogram. I recorded 2 CR2 files with my Canon G7 -- camera on a tripod and the same exposure for both. One of those was taken with the camera ISO set to 1000 and the other with the ISO set to 125 (my G7 doesn't have a 100 ISO). Here are your two images with their histograms inset in the upper corner. Right now I'm going to leave it to you to pick out which is which as I have removed the EXIF data -- just for fun. I'll gladly post the photos with the EXIF data but first you sort them and explain what you see that makes it obvious which is which.

I did have to post process the files and some subtle differences remain but I believe these two photos do a good job of demonstrating that there was no advantage of any value that I got from that ISO 1000 setting. I can just as easily ignore it and get basically the same photo by leaving the ISO at 125.

I used to use a Fuji XE-2 which truly was ISO invariant and as such raising the ISO provided no value beyond allowing me to get an on the spot JPEG to look at. As far as processing the RAF files the only thing raising ISO did was harmful as it reduced DR. My G7 isn't quite as ISO invariant as the XE-2 but it's darn close and it represents the current trend. My new XT-2 is ISO invariant and most of the newest cameras from Nikon, Sony and Fuji are likewise. We're quickly moving to a point where the only point of raising the ISO is to get a JPEG you can chimp if that's what you need to do.

Sorry I didn't have more time earlier, here's the two photos:

Joe

photo_a.jpg


photo_b.jpg
 
The inability to effectively adjust the ISO afterwards with software in post processing, means that you must include ISO as you take the picture.

See photos posted above. I'm calling that effective.

A long time ago, people may have differentiated between "exposure" and "camera exposure"/"proper exposure";

2011 may be a long time ago for you but I still remember a whole lot of what I did that year (big flood on the Mississippi!). 10th edition The manual of Photography Focal Press 2011: https://www.amazon.com/Manual-Photography-Tenth-Elizabeth-Allen/dp/0240520378

And this is today: "In photography, exposure is the amount of light per unit area (the image plane illuminance times the exposure time) reaching a photographic film or electronic image sensor, as determined by shutter speed, lens aperture and scene luminance." Underline ISO in that definition and repost it. :)

however with almost no photographer using or needing to technically define "exposure" as just the light photons hitting the sensor,

Speak for yourself. It helps me to understand how things work and I do a better job when I understand how things work. It's important to my students too because they don't like seeing this:
f-.jpg


the first adjective (camera or proper) has been dropped from common usage.

It was dropped from common usage long before I picked up a camera and many many many and many more decades before the ET raised it's stupid head. Ilford didn't just recently add "camera exposure" to The Manual of Photography. They did that before I was born as they were dealing with the same issue. And they made sure to add an accurate definition of exposure. Colloquial usage and industry jargon have been with us since the invention of language.

Leaving us where we are today, exposure means the triangle.

See industry standard definition above. Who are you that you think you get to change it?

Joe

 
Last edited:
As an Amazon Associate we earn from qualifying purchases.
I never even HEARD of the so-called exposure triangle until well after the year 2000. Seriously. I learned that exposure was Intensity x Duration, with Intensity including the lens f/stop and the brightness level of the light, and duration being shutter-open time. I. Never. Ever. Heard. Of. The. Exposure. Triangle. Until.The Web.Era.

I consider the so-called Exposure Triangle to be bull$h!+. The way it is commonly shown/illustrated is WRONG. Wrong,wrong,wrong,wrong! Why is it wrong,wrong,wrong? Because as shown so often, the so-called Exposure Triangle is illustrated so that ISO 100 gives a clean,crisp picture, with no visible noise, while ISO 1,600 and higher are illustrated as being "awful", and creating "noisy" images. That is WRONG. Wrong,wrong,wrong!!! Period. Wrong!

This thread is filled with half-truths, quarter-truths, outright misinformation, and outdated concepts. And some actual truths. We will have to leave it to every person to pick his or her "truth" from among the misinformation.

If ISO is part of "exposure", what does one make of the new, ISO-invariant sensors? How does one explain the beautiful images made by shooting very "dark" images, and then brightening the images up, in post software, with an ISO-invariant sensor-equipped digital camera? How does one explain the nice images Ysarex shows above, made at ISO 125 and ISO 1,000, in Post #58 above? Why does the ISO 1,000 image not look like crap--as the so-called exposure triangle would predict? If the "triangle" were valid, then the ISO 1,000 image would look like rubbish, and it would be VERY obviously different from the ISO 100 image--yet, the two images are visually, for all intents, identical. Now, to me, if a theoretical model does not produce the results predicted by the model, then that hteoretical model is bull$h!+.

I for one do NOT consider that "exposure" means any type of triangle....the definition of "exposure" is well over 100 years old. This triangle concept is internet-era nonsense.

Equivalent camera exposure settings, for either one ISO in use, or for different ISO values...now that concept, the concept of equivalent camera exposure settings, that has been around for a long,long time. How a basic, true, proven idea came to be bastardized into this so-called exposure triangle idea...well...that has happened during the Internet era.

Anyway, this thread has drifted very far from the OP's query about why two photos look different...
 
Last edited:

Most reactions

New Topics

Back
Top