NOTICE! I just recently deleted a comment that had some nonconstructive negativity towards my view on AI. Please keep in mind the following: if you find my views on AI "tiresome" or you are tired of hearing me rail against AI, I couldn't care less and I invite you to unsubscribe. You are not welcome here.
Really like you tutorials, they are practical and funny. I incorporated a lot of that in my processing workflow. However, as someone who works with AI for scientific images i do not want to admit to myself that it will kill us all🙂
Thanks for the support and encouragement. Regarding your AI comment... Like all technologies, AI for scientific imaging isn't necssarily bad in itself. In fact, I am sure there are some instances where AI is quite useful and could even be used to reverse some damage done to climate, maybe to research some new technologies. Even so, I believe that the whole of it is still dangerous because of its overwhelming psychological effects on humanity and its economic effects as well. Science and mathematics (speaking as a mathematician who worked in pure research for over 12 years) is the type of pursuit that rarely sees the true societal impact of its work until much later, and scientists usually only see the impact of their work directly in other humans by interacting with other scientists. Thus, it is sometimes hard to understand the overarching effects of technology. So even though I myself recognize that AI can be incredibly useful, I have taken a position not to use AI even when it can be beneficial to me, such as in image denoising or other similar areas. Again, the specific uses of AI such as denoising are not necessarily harmful, but the overall societal effects over decades are either harmful or too risky to support without extreme caution and wisdom, which is something we just don't have.
Very useful tutorial. Personally, in some cases, when the effect is too aggressive, I change the module's blend mode to "Average" or, quite simply in "Normal" blend mode, I reduce the opacity of the effect.
That is a good trick. The non-local means can be quite helpful if you use it as you describe. It does produce a bit of an oil-painting effect but at low levels, it's not notieable.
Thanks for that tip. I just tried it on a bunch of pics I took in a very dark bar with ISO's at 25,000. Your method did an awesome job and then following up with @WildEarthPhoto's contrast equalizer tips, I got some really nice results.
Denoise and contrast equalizer don't add grain, though increasing the contrast will enhance the appearance of grain. However, I explained how to actually add grain to your shots at the end of the video.
It is similar to this style. We do not go into the details of the scene-referred workflow but that may be an upcoming topic when we further update the course. For now, we do not talk about filmic.
What is better for simplicity and ease of use for somebody that has to denoise bunches of photos in a short time? This method or using a specific program like Topaz denoise?
I think it depends on your computer. For slower computers, Topaz takes some time to denoise and with my method, I can denoise faster than Topaz if the computer is slow, but if the computer is sufficiently fast, then not. However, I think you should also ask yourself whether the photos are worth denoising if Topaz is truly trequired. Personally, I feel that if there are so many images are in need of such extreme denoising, then it might be better to go out and just get better images. Personally, I don't use AI denoising in any of my work.
@@WildEarthPhoto it is for 360° photos for Virtual Tours taken with 360 camera. With bad light the noise is noticeable and convenient to denoise to get better overall quality.
