From: Jonathan Kirwan
Subject: Re: Make infrared goggles inexpensively (like $10!!!)
References: <firstname.lastname@example.org> <3D848F5E.91EC2A2F@xympatico.ca> <email@example.com>
X-Newsreader: Forte Agent 1.92/32.572
NNTP-Posting-Date: Tue, 17 Sep 2002 11:06:17 GMT
Organization: AT&T Broadband
Date: Tue, 17 Sep 2002 11:06:17 GMT
On 17 Sep 2002 00:59:18 -0700, firstname.lastname@example.org (William J. Beaty)
>Huh? Convolving the filters will smooth out the sharp passband
>edges in the graphs, while actually stacking up some bandpass
>filters will sharpen the edge of the passband. You don't want to
>convolve, instead you just want to multiply each filter's absorbtion
>at a particular frequency by the absorbtion of all the other filters
>in the stack.
>If you want to do convolution too, you could assume that the original
>instruments which produced those absorbtion curves had 10nM windows
>of accuracy, then use convolution to include that smoothing effect
>(but what if the spectrometer slits were actually 1nM?)
My mistake in using the term. I meant "multiply."
>> curves shown are in linear scale response, not logarithmic, and there
>> is probably substantial flux remaining almost across the spectrum,
>> though most of it will be in the 650+nm area. Certainly, there will
>> be plenty of 680nm light available, given the curves I saw. And that
>> is visible.
>If a particular filter response is down 99.0% at a certain frequency,
>and you stack three of those filters up, then it's down six orders of
>magnitude at that frequency. (If I'm wrong, can you please explain
Of course, if you assume that's the figure you can get what final
figure you want to argue with. But my point was that the curves do
NOT eliminate 99% at 680nm. At least, not the two of them I looked
at, which you mentioned. I went directly to the web pages you mention
So I've no idea why you are arguing a strawman, when it's not what I
was talking about.
>> There is no need to pull out some incorrect analogy to explain being
>> able to see some odd looking things. It's enough to use traditional
>Declaring my analogy to be incorrect, yet not saying WHY, is a
>typical debating trick.
I explained why in another post. It's no trick as I know what I'm
talking about, at least enough for this.
>It's wise to withhold judgement before investigation. But simply
>declaring something to be wrong is just as irrational as declaring
>something to be right.
The analogy is without value, that's all. There may be an interesting
bit of fun in changing the mix of spectra your eyes receive. I think
it's a great idea, putting different filters together. I don't mean
to take any of that away. And changing the mix can definitely bring
out different aspects you might not notice as well, otherwise.
Frankly, I think it's a great idea. But the analogy is still flawed.
>How is the rolloff in the human eye sensitivity function at the
>IR end fundamentally different than the rolloff of human ear
>sensitivity at the ultrasonic end? If you think my analogy is
>misguided, please show this. Are you that confident that you
>can't hear a pure 30KHz tone no matter how loud the signal?
>Are you confident that you can't see a narrow 800nM emission line
>no matter how bright the light?
Did you read my other post? I'll wait.
>> If anyone has the exact filter response data, over wavelength, for
>> these filters, I'd be more than happy to convolve them with the human
>> photopic vision curves to find the resulting response. I think we
>> will find, from that, where the dominant is at. And I rather doubt it
>> will be in the longer than 800nm area.
>We agree. For eyes wearing "congo blue" goggles, the sensitivity
>probably peaks out at around 710 to 720 nanometers. I also made up
>some goggles with four layers of Lee #120 "deep blue" which has a
>much deeper IR cutoff. They make the sky look even blacker, but the
>light is so dim that it takes about 20 seconds before you can see
>anything at all, even in full sunlight. I figure that these
>"deep blue" goggles peak out your eye's response at around 720 to 730nM.
>To see an outdoor scene at 800nM, the sun would have to be many orders
>of magnitude brigher, and I'd have to use much deeper IR filters
>(like Wratten 87B for example.)
I applaud doing all these things, Bill. Please don't get me wrong. I
also agree that some things can become visible that aren't, otherwise.
For example, there was a child's game which used a red bit of plastic
to remove out some confusing colors so that an image became visible.
I just believe that there is nothing quite the same about this process
and, say, phosphor up-conversion or electronically enhanced
instrumentation to achieve wavelength translation.
Just wanted to make that clear.