A rather interesting rebuttal to our current project was recently printed on Petapixel.
Our response to this rebuttal is as follows:
First of all, I completely agree with you in regards to the infrared spectrum as most people think of it as being out of the reach of vision in general. We have been quite careful to distinguish that our project is seeking a visual response in the near infrared spectrum which has been demonstrated time and time again in a number of different ways. While I wasn’t working with those exact values for absorption, I am familiar with cone and rod absorption curves. The Rhodopsin curve tapers off near 700nm assuming a certain level of light intensity. The ISO 20473 scheme defines near infrared as beginning at 780nm. Because of the slow taper at the high end of the rhodopsin absorption curve even rhodopsin is sensitive to a 780nm wavelength of light given a high enough magnitude. I believe I know where you got this “absolute limit of long wavelength photoreceptors” being 650nm but I’ve never seen this number in a journal. It matches none of the research or data that I’ve seen and trust me.. I’ve seen quite a lot. First, I’d like to argue against this 650nm limit you describe. As far back as 1955, Brindley discussed the “yellowing” of light that occurs when wavelengths longer than 700nm and of sufficient magnitude are subjectively described. For example, subjects described an 850nm light of sufficient brightness as being orange rather than red. Of course, this study was subjective but electrophysiology studies have since confirmed this. T.D. Lamb’s article in Vision Research Volume 35, Issue 22 proposes an equation attempting to match these findings. Your claim of an “absolute limit of long wavelength light” without once mentioning lumens is an erroneous statement. If you’d like, I can certainly provide you with links to a few good textbooks regarding photometry that can help you understand what I’m talking about. Now, I am perfectly willing to acknowledge that I may be misinterpreting some of the data and fine points of these journal articles. I do not have a PhD in your field; however, I take issue with some of the values you presented.
In short, there are 4 kinds of photoreceptors in our eyes, rod photoreceptors which detect electromagnetic spectral energy from ~400nm to ~580nm with maximal sensitivity at 496nm (blue/green light)
The long wavelength sensitive, L-cones (“red cones”) are known to be maximally sensitive to wavelengths peaking around ~560nm, medium wavelength sensitive, M-cones (“green cones”) peak around ~530nm and short wavelength cones, S-cones (“blue cones”) peak at ~420nm respectively.
I found it strange that you chose these values which are rather outdated. I have the Bowmaker and Dartnell article which dates back to 1980 proposing these values. The 2004 Hunt (Didn’t you coauthor a work with Hunt?) work is far more recent and likely its sources more accurate. You also chose 496nm as the Rhodopsin maximum sensitivity. The most commonly reported value is 498nm. While trying to backtrack your data… I found your source. Your single and sorely misinterpreted source was Helga Kolb’s Webvision page from which you word for word took the majority of your explanation. The worst aspect of all is that you interpret image figure 14a as somehow indicative of a 650nm cutoff that you’re promoting as an “absolute limit of long wavelength photoreceptors.” This type of occurrence is exactly why independent researchers are so vital to the advancement of science. I’m sure some people will believe your criticisms based on the fallacy of appeal to authority regardless of your blatant plagiarism from a single source.
The bottom line here is that even without dehydroretinal which contemporary studies have shown to peak closer to 620nm rather than the 522nm or so reported by Wald, humans can already see into the near infrared spectrum assuming a light of appropriate intensity. By replacing the retinol in the human eye with dehydroretinol it’s very likely that the observable spectra will shift further into the NIR range and become observable at a lower intensity such as normally present around us at all times. I say this is likely as it has been demonstrated time and time again in murine models with electrophysiological measures which if you’d taken the time to read, is described and cited in detail in our original study proposals. I agree with you that we have a serious problem with ignorance towards science in this country but I’d like to add that quote mining a source without so much as fully understanding the material isn’t any better than blind faith.
Furthermore, I can’t emphasize how much you’ve jumped to conclusions without taking any time to read our protocols. Sure, the images posted state “850nm Flashes” and “950nm Flashes.” This is referring to the wavelength of peak light emission of the LED, not some pure wavelength of light. Had you read any of our design posts, you’d have seen the section explaining why our control light which is clearly visible by any person with normal and unaugmented vision is labeled 850nm. An 850nm LED bleeds a swath of light with wavelengths in the normal range of human vision. We have all of the documentation detailing the curve of different wavelengths being emitted by the 950 and higher wavelength LEDs as well. Of course, you didn’t take the time to read anything we’ve written or worked on and instead pasted together a rather remedial introduction to the physiology of the eye. Yes, I am aware of the work you’ve done and the articles you’ve written. I have no doubt that your understanding of the eye is far greater than mine; however, you haven’t taken the time to actually read what we’re doing. Your rebuttal is little more than a leap to a conclusion heavily seasoned with condescension.