First lens purchase for a D3300 user

Started Oct 19, 2016 | Discussions thread
FingerPainter Senior Member • Posts: 5,806
WARNING! Video needs a re-do

PhotographyTV wrote:

As the others mentioned, the issue with noise is due to the lack of light,


which means your ISO had to be high in order to get a good exposure.

No. The lack of a good exposure meant both that the image was noisy and the ISO had to be high in order to get a good image brightness. You are confusing exposure with image brightness. Increasing ISO while holding aperture and shutter constant doesn't increase exposure. It increases the brightness of the image.

6400 ISO is pretty high, which is what causes the noise..

No. The high ISO is not causing the noise. It is the low exposure that is causing the noise. If he took the same picture in the same light with the same shutter and aperture settings, but set the ISO to 800 instead of 6400 the image would have the same exposure and about the same noise (probably slightly more) but would be less bright.

I made a video to explain ISO,

And repeated in it a whole bunch of very common misconceptions.

The video is very nicely presented and produced but most of its factual statements are wrong.

so I hope this helps you.

I understand that you are trying to help. But because you have an incorrect understanding of the subject matter, you are inadvertently misleading people. This could lead your viewers to make make mistakes in complex photographic situations.

Your video is essentially a re-presentation of a badly flawed model called "The Exposure Triangle". The internet has made this conceptual model very widely known. Because it is repeated over and over on many sites, a lot of unsuspecting people believe it is true. It isn't.

Here is a summary of the mistakes in your video:

What you refer to as “exposure” isn’t exposure at all. It is image brightness. This confusion of terms is very common, but leads to misunderstanding. Exposure has had a precise definition in photography and sensitometry for over 125 years. Exposure is the amount of light falling on the sensitive medium per unit area. Exposure is not the same as image brightness. Image brightness is how light or dark an image appears. Exposure is an important concept, in part because exposure is the prime variable affecting image noise and Dynamic Range (DR) for a given camera system. Exposure and ISO setting together determine the image brightness.

Exposure does indeed have three parameters, and shutter speed and aperture are two of them. But the third parameter of exposure is scene luminance, not ISO. Image brightness has two parameters: exposure and ISO.

Changing the ISO setting doesn’t change the sensitivity of the sensor to light. It does not change the amount of light the sensor is able to absorb. The ISO doesn’t control the amount of light the camera is able to capture. Rather, it changes the output sensitivity of the camera system. That is to say, it changes the brightness of the image that the camera produces from a given amount of captured light. So, higher ISO, brighter image.

Increasing the ISO doesn’t allow or cause the sensor to capture more light. If you increase the ISO, the sensor will still capture the same amount of light for a given exposure. What increasing the ISO will do is increase the digital number that is stored in a pixel in the digital image file for a given amount of light captured for that pixel by the sensor. The higher number will result in the pixel being brighter when the image is displayed or printed.

A lower ISO does not cause a higher quality image. Lower ISO is often found associated with higher quality images (and higher ISO with lower-quality images), but it is a non-causal correlation and the correlation is not universal.

Higher ISO doesn’t introduce noise. Higher ISO doesn’t cause higher noise. In fact increasing ISO can sometimes result in a slight decrease in noise but never an increase.

Higher ISO doesn’t equal more light. Higher ISO leads to brighter images from the same light. You did correctly state that increasing the ISO increases the brightness of the photograph, but you mistakenly conflate brighter images from higher ISO with brighter images having more light. Only images which are brighter due to a higher exposure (more light falling on the sensor) are brighter because of more light.

Your first series of pictures of the Betamax(?) figure doesn’t show you increasing the amount of light being captured. It shows the camera making progressively brighter images from the same amount of captured light.

As you sum up the first series, you again mistakenly claim that ISO controls the sensitivity and amount of light the photograph can capture. It does neither. It just controls how bright will be the image produced from a given amount of light. In fact, when using an autoexposure mode like P, A or S, increasing the ISO will act to limit the amount of light that can be captured, because the higher you set the ISO, the lower the camera will set the exposure.

You claim that increasing ISO (as you did in this first series) increases noise. Noise is a calculable property. If you use the appropriate software tools to calculate the noise in the raw image for the ISO 100 version and for the ISO 800 version, you will find that the noise in the ISO 800 version is very slightly lower then the noise in the ISO 100 version.

In your second series of images, you purport to show how ISO impacts the quality of your images. The problem with this series is that you have two variables in each photo: shutter speed and ISO. You progressively increase shutter speed and progressively increase ISO, and as you do, the noise gets worse. The mistake you are making is to attribute the increase in noise to the increase in ISO. This is the common error of confusing correlation with causation. In fact, it is the increase in shutter speed, and consequent decrease in exposure, that is causing the increase in noise. The decrease in exposure is also prompting you to increase the ISO, in order to maintain image brightness. The decrease in exposure is the common cause of both the increase in noise and the increase in ISO. This results in a non-causal correlation between ISO and noise.

A proper experiment seeks to isolate variables. If you were to repeat the series, in each case changing only one of the factors, you could properly determine the causal factor.

For instance if you took a similar series of photos but only increased ISO you would find that the noise in the images declines slightly with each increase in ISO. From this you should determine that increasing ISO does not increase noise, but actually has a slight effect in the other direction.

If you took a series of images with shutter speed increasing but ISO and apertue remaining constant, you would find that noise increased by a factor of about 1.4 with each stop increase in shutter speed. Doing the same experiment but changing only aperture, you will find that noise increases by a factor of about 1.4 with each stop reduction in aperture diameter. Finally you could do a series in which you keep the camera settings constant but progressively reduce the output of your studio lights by a stop for each photo. Again you will see about a 1.4 factor increase in noise for each stop reduction in light. This should lead you to correctly conclude that a reduction in exposure results in an increase in noise equal to about the square root of the amount of decrease in exposure.

Post (hide subjects) Posted by
Keyboard shortcuts:
FForum PPrevious NNext WNext unread UUpvote SSubscribe RReply QQuote BBookmark MMy threads
Color scheme? Blue / Yellow