Rob Trek Forum

On my journey searching for new information I came across this astrophotographer. He has some interesting M43 videos...


Link:-


He did say something that I do not agree with. You can not use the ISO to expose to the right (ETTR). Why?

Question:- For example, do you think in terms of reflected light at the image sensor when setting up your exposure?

For more on reflected light, study the link below...

Link:- https://myolympusomd.blogspot.com/2022/ ... ns-at.html

Best

Siegfried

I know Ben personally and have great respect for his work as I do yours. I only mention this because I'm not sure he knows you. I will correct this.

I missed the part in his video on not using iso for ettr. As you stated in your articles, it's basically about maximizing reflected light data and balancing the exposure mix. The way I see it, once you've hit your limits on aperture and/or shutter, either physically or creatively, pushing the luminance via iso to maximize the data points before the raw file is created is valid.

In your article, you stated setting the iso first. Do you mean to set it to base iso first, then adjust shutter/aperture? If not, can you give me an example of why you'd set iso to say "1600" first?

I teach a class where I tell my students to set the aperture and/or shutter first, then adjust the iso to get the brightness they want, or just use auto-iso. This is in the context of using ETTR.
Thanks,
-Rob

I use ETTR and with a G9 and EM1II I tested using the ETTR with ISOs. With those cameras, there was no real difference I could see pixel peeping. The actual light reaching the sensor is the same set using the aperture and shutter. I thought ISO increasing would be less noise because the signal was being amped at the sensor instead later after the electronics add noise. I just could not see it. I also tried underexposing up to 2 stops and raising it in post. I could find no difference and I really tried so I would have the best photo. Robin Wong did a video on this where he found ETTR raising ISO was not as good as a proper exposure and recommended not doing ETTR in low light. I use linear profiles (free profiles by Tony Kuyper) which do not have an profile added to the sensor so there are no highlights to recover and or shadows to raise. These give the best detail and a much richer file to post process.

Everything I did was with RAW. For jpegs I found it better to not raise the ISO as with the EM1II the jpegs had more noise reduction and lost details at the higher ISO. This was just how the camera processed the jpeg apparently using a different algorithm based on ISO??? So for my mentee subscribers who do not process. Which of course without post processing ETTR does not work although some people process jpegs.

Hi Rob

Thank you for your support

The shutter speed and Aperture determine how much light is reaching the sensor. The ISO only amplifies the image signal coming from the sensor.

We want to get as much as possible light onto the sensor because that increases sensor saturation and SNR. That means more tonal data and less visible noise.

It's a balancing act between exposing the sensor and amplifying the image signal.

Thinking in terms of illuminating the subject, and reflected light (luminance) exposing the sensor. This changes everything.

When I say set the ISO first, it's simply a process of carefully selecting the right ISO (manually) and then managing the performance of the sensor with the aperture and shutter speed. You will find that thinking in terms of light reaching the sensor, you will be much more considerate with what your starting ISO is. Your starting ISO can be anything...

Finally, everything is based on having a good understanding of the image sensor.

Best

Hi Morgan

This is not to make you look bad my friend. You only amplified the image signal, adjusting the ISO is not ETTR.

The easiest TEST is to use the same settings on the EM1 II and the G9 and to make any one of them look terrible in terms of noise. How is that possible you would say? Simply by regulating the external light and correcting it in PS. How do you think some proved the OM-1 has no so-called noise improvement and then they continue saying OMDS lied?

It's all about how much light is reaching the sensor and understanding the image sensor...

A better way to test is to take a normal image with highlights, mid-tones, and shadows. The sensor will be undersaturated in the shadows and that is where you study the performance of the sensor. It is also impossible to regulate the light of a normal landscape scene. Why did reviewers never think of this? Why do they always use a piece of white paper?

Siegfried

If you adjust the exposure with compensation for ETTR, the cameras both raised ISO. Which gave a lighter photo to process (which is what ETTR does) but did not improve the image quality over a correctly exposed photo. Many people use auto ISO which changes way too soon (for me) and do ETTR which just changes ISO. With dual gain sensor there is a ISO where the noise will lower. I use it when I have a camera with a dual gain sensor. In this case I know raising the ISO a 1 or 2 thirds of a stop will give a better photo. The dual gain sensors change the ADC rather than boosting gain like most ISO adjustments. The video said the EM1III was dual gain, but maybe the difference showed up under astrophotography conditions and not normal low light photography. People who did not ever use film cameras use ISO as part of an exposure camera treating it equally with aperture and shutter. Of course coming from film, I had no triangle.

The golden rule is, what can we as photographers do (adjust) to take better images, especially on older cameras.

We have 3 settings, Aperture, Shutter Speed, and ISO to work with (I explained what they do)
We know we can "Manage" the performance of the sensor (IQ and noise) by how much light is reaching the sensor
Anything like dual ISOs, where components are mounted on the sensor or the camera is improvements and ONLY interesting info

I explained the Image Taking Process in the camera - light goes via the lens to the sensor - ISO amplifies the image signal

Master these basics, study my articles, and you will see a difference

Adjusting the ISO (amplify) is not ETTR (use only manual ISO and get more light on that sensor)

The only M43 camera that allows us to use Dual ISO is the GH5S - I tried it and...

Astrophotographers like Ben will consider variables like heat because it's critical to them

Our aim is to improve the performance of the sensor by increasing sensor saturation and the SNR

Whatever is coming off the sensor (image signal) is amplified with the ISO

That means we only amplify what is happening on the sensor - if we have a desaturated sensor with low SNR then...

The solution is to get more reflected light onto the sensor

Your challenge is not only DOF, blur or bokeh, and freezing sports, you also want maximum light (information) on the sensor

Why did no one tell you this? The commercial size and capture strategy like you to think the only option is a bigger sensor

Hope this helps