Learning perfect exposure - Expose to the right or WRONG? 

Hope this video helps you understand the whole ETTR concept!

From what I understand, exposing to the right is never about clipping the highlights/details to the right. Its about exposing it to the right as long as the image is not clipping. Some photographers do some tests images to learn how much they can safely push the historgram, so that next time they shoot, they already know what is the safe limit to expose to the right. This will result into a much cleaner and smoother image while at the same time giving the photographer more flexibility to adjust the shadows without amplifying the noise.

ETTR but don't blow the highlights.Then modify the exposure and look to your taste and what you want to achieve.This means anyone wanting to ETTR should have good post processing skills.

Thank you, I like how you are not afraid to put your opinion out into the public arena. I can have a live histogram in my viewfinder. I was taught to expose to the histogram, nothing touches the extreme left nor the extreme right. After I meet this requirement, I try to get as close as possible to a balanced exposure. Normally, this results in my images being around -0.7 to -0.3EV

You left the Exposure slider alone on the ETTR image. If you slid that to the left, you almost certainly could have duplicated the "properly" exposed image. And you proved that exposing to the right causes less noise in the shadows. The difference is minimal at ISO 100, but that difference is more pronounced at higher ISOs.

Once again another great video very helpful.

I think your test was different than others by you using the incident light meter. Which is an accurate meter resding. Most people use the cameras light meter that can be fooled depending on the scene. Obviously you would not expose to the right if it is clipping the highlights.

Very interesting video. One point I would like you clarify. At 4:00, you say the picture in slightly underexposed. But you have used a light meter which is supposed to calculate correctly the light. Does the LM underexposed the calculation ? OR is the calculation of the LM correct but the amount of light was a little weak ? Thanks for reply

than q sir good lesson for ettr, i expect next video for raw file how to remove noise.

I don't understand this video and its conclusion. ETTR means, you increase the exposure such that you use the full range of highlights, but don't clip the highlights. Since you used a motive with high contrast, your normal exposure already used the full range of the histogram. As a consequence the correct exposure was already exposed to the right (and also to the left!). The third picture clearly was overexposed as the highlights were clipped. The same also holds for the example for the exposure to the left which you say "is significantly underexpose".

I think that moving the histogram to increase low-level shadow signal is solid. However, I feel the only way the signal t noise ratio (SNR) can be increased is by allowing more light to reach the sensor, be that by opening the aperture or increasing the light hitting the shadows. I feel that amplifying the signal by any other method (be that ISO/speed/compensation adjustments) will cause a similar amplification of inherent noise and will not result in any SNR/quality improvements. Can you provide any arguments to dispel that theory?

It has never made sense to me in any way that artificially pushing our tones to the right in camera, thereby changing overall brightness and saturation can be superior to capturing the tones where the they belong natively. (This is not the same as correctly capturing very bright tones without clipping that belong on the right of the histogram) When we have pushed tones to the right, we are then depending on a software program to correctly establish the overall tonality and saturation as we then move the histogram to a more normal rendering of the image. Why not render the tones correctly with a proper exposure...where they belong natively when we shoot? Oh...the shadow noise. Yeah...if we're shooting a Rebel XT! I understand that the right side of the histogram stores substantially more data, but if we are artificially pushing tonal data there where those tones don't naturally belong in relation to the lower tones in our images and then expecting our software to give us a linear representation of tonality and saturation as we reconstruct our images by moving parts of the histogram to the left, are we doing more harm than good? If we are confronted with an extreme wide dynamic range in an scene, wouldn't it be better to capture the shadow detail in a separate frame for blending while leaving the natural tonality of the rest of the image in tact? It would be nice for instance to get a read from Adobe, let's say, on this. Until then, I'll just meter for intent and shoot the native tonality in my scenes. BTW, for those who think that hand held meters are a thing of the past in digital photography...this is a part of why I think they're still relevant and important today.

You are confused. You can't ettr and blow the highlights. That's just wrong exposure. Also, you can't ettr for high contrast scenes that exceeds the dynamic range of your sensor

You blew the highlights and didn't change the exposure slide when you were processing the image. Since this is not a good comparison, it's not possible to assess your main argument.

So wrong then? lol