Enderman,
Generally, generally I don't mind educating those who are less knowledgable when they are interested in learning.
There is a concept in psychology called the Dunning-Kruger effect. In the interest of saving time, I have copied the first paragraph from the Wikipedia article:
"
The Dunning–Kruger effect is a cognitive bias in which low-ability individuals suffer from illusory superiority, mistakenly assessing their ability as much higher than it really is. Dunning and Kruger attributed this bias to a metacognitive incapacity, on the part of those with low ability, to recognize their ineptitude and evaluate their competence accurately."
Where CRI, light "quality", color interpretation, even the recording of movies is concerned, this effect applies to you.
This statement, "
Higher CRI, the more complete the spectrum is and the closer to the real colours in sunlight it will look." shows gross ineptitude on this topic. A high CRI 5500K (or similar) source will cause colors to be rendered like they will be under sunlight. This is completely different from the rendering of colors at <5000K. 5K and over CRI is based on a simulated sunlight spectrum that is somewhat even, but not "flat" and not blackbody unlike CRI under 5000K.
Do you know why they want to simulate sunlight as perfect as possible for movies? It's because everything in their processing stream, including film if they are using it, the filters on the sensors in digital cameras, the color space conversion algorithm in the digital camera, etc. are all tailored for accurate recording of color IF they use spectrum like the sun and since they often use natural sun, this is the best choice. When film was the only choice, you were stuck. With digital recording, cheap storage, and new techniques, there are explorations into superior illumination methods and image capture methods.
Do you even know what the CRI is of an HMI lamp? Hint, it's not 100. Its 90+. HMI is a trademark of OSRAM. Don't believe me that it's not 100? Here is a link for everyone to check out:
Family data sheet The only way to get to 100CRI with a HMI lamp would be with a lot of filtering (which is done). That is something that can be done without filtering with fluorescents as you just need to mix a lot of phosphors to get the right output that you want. Hint, you can filter a LED too (or mix phosphors or LEDs, etc.)
Your repeated comments about the "spiky" nature of LEDs for also shows your lack of knowledge w.r.t. to human perception of spectral power density. Don't worry, it's common in people who are not typically translating spectral density into something meaningful in terms of what humans will see. That is probably why you posted a bunch of irrelevant spectral plots I assume. Note that your HMI plot had "spikes" ... some actually in areas where the eye is quite sensitive which is why it takes a bit of a hit in CRI .... sort of like the red deficiency in plasma lighting.
It's all understandable. I can tell that you don't have formal training in this area and don't have the application experience of say chillin.
I have a master's specifically in electro-optics. I (yes me! -- the salty guy), have co-authored published papers on color space conversion with multi-spectral illumination (image capture) and adapting color space conversion algorithms to variable spectra to improve perceived image quality. By a twist of fate (okay they made me an offer I could not refuse), I have done a lot of work in illumination and lighting the last many years and have been fortunate to converse with AND LEARN FROM some of the recognized experts in the area of color perception and quality. I don't consider myself an "expert" in this area. They are. I trust their expertise in this area, not some guy who send links to spectrum that I can probably draw from memory with pretty good accuracy.
No LED is 100CRI. It's almost impossible to obtain in any artificial light source that is not a true black body. Fortunately, this is a case where "close enough" really is good enough. It's not like a 2700 kelvin black body is common in nature (it's not), and sunlight is highly variable and if you measured it, would rarely be "100 CRI".
How good can an LED get? This good:
http://www.xicato.com/technology/color-rendering Check out the CRI (not just Ra, but all values out to 15) of the Artist series. They are 98+ across all the color targets. What's also cools is the V95 series with the 120 gamut area index. Drop the CRI a bit ... makes a light that is preferred by HUMANs .... that's real quality. Your lack of knowledge will likely consider the spectrum in this data sheet "spiky"
http://www.xicato.com/sites/default/files/documents/DDS XTM Artist Series 160520.pdf but to someone who knows how SPD relates to color spectrum, they will see this as a very smooth graph that accurately replicates a tungsten source, Especially in the critical areas where the sensitivity of the eye is high and the ability to discern tone differences is highest.
If you are truly interested in this topic, I suggest doing some research. The Lighting Resource Center (
Rensselaer) and the solid state lighting metrology section at NIST are great places to start. Once you have digested those sites in detail, I can suggest some deeper reading that will provide more insight into human visual perception and spectrum.
I apologize to the OP for letting this get off track, but I think you have been provided with some good options for your usage.
Enderman, unless your post is related to a specific aspect of fulfilling the OPs requirements and/or links to professional research papers that may be of interests to the OPs requirements, I will not be replying to any more of your posts.
