Re: CCFT versus LED [was: Spectros better than colorimeters for monitor profiling?]
Re: CCFT versus LED [was: Spectros better than colorimeters for monitor profiling?]
- Subject: Re: CCFT versus LED [was: Spectros better than colorimeters for monitor profiling?]
- From: "edmund ronald" <email@hidden>
- Date: Sun, 26 Oct 2008 11:01:46 +0100
Robin,
I see there is no way to escape my engineering school courses, the
nightmare formulae come back to haunt me.
Usually, when people invoke the Nyquist theorem I start running for
cover, signal processing is really not my thing - kilai desu as our
friend Eric would say :)
However, it would seem to me that the limit sampling frequency one
should be looking at is the one used for computing the observer
responses. I think the CIE sets this at something like 5 or 3 nm these
days. So by your invocation of Nyquist, one should ideally be sampling
at 2.5 or 1.5 nm, by your reasoning.
Another interesting question which I know nothing about is whether in
the i1Pro design light cannot "fall between the floorboards" so to
speak.
Edmund
On Sun, Oct 26, 2008 at 1:34 AM, Robin Myers <email@hidden> wrote:
> For a variety of reasons LEDs (not LED lasers) do not have really narrow
> spectral emissions. Their peaks are generally from 20 to 50 nm wide,
> sometimes wider. One measure of how wide the peak is the Full Width Half Max
> (FWHM) value which is taken as the distance between the point on the left
> side of the peak at half the peak's maximum and the corresponding point on
> the right side of the peak. The FWHM is often reported in the LED
> manufacturer's specification sheets.
>
> A spectrometer such as the i1 Pro, which measures the spectrum in
> approximately 3.5 nm bands will get at least 5 or more samples in a 20 nm
> peak. It combines these 3.5 nm measurements into 10 nm values reported to
> the user. Since Nyquist showed that a signal must be sampled at twice the
> frequency of the signal one is looking for, even the 10 nm reporting
> interval of the i1 Pro is adequate for LED measurement (even more so when
> considering the 3.5 nm measurements physically measured by the instrument's
> sensor).
>
> Perhaps the literature nomenclature is partly at fault here for the
> confusion. Most of the LED literature describes LED emissions as
> "narrowband" which can lead one to think of almost monochromatic emission.
> However, the "narrowband" is a relative term, usually used when comparing
> LED output to incandescent sources which are often described as "broadband".
>
> By these definitions a white LED is an example of a narrowband light source
> being converted into a broadband source. White LEDs are usually a blue or
> violet narrowband LED coated with one or more phosphors. The most common
> white LEDs use a yellow phosphor which emits light in the green and red
> portion of the spectrum. Along with unconverted blue light the result is a
> very broadband spectral emission, usually almost across the entire visible
> spectrum.
>
> One advantage of the diffraction grating spectrometer (e.g. i1 Pro) is that
> it can measure the backlights and filters of monitors as technology changes.
> Filters for colorimeters must be matched to specific technologies, thus
> rendering a given colorimeter filter design less accurate as the technology
> changes. The spectrometer is therefore more versatile, removing the
> difficulty and expense of changing colorimeters for every monitor
> technology.
>
> One disadvantage of the diffraction grating spectrometer is that it can be
> sensitive to polarization angle. This affects LCD monitor measurements since
> LCDs produce images using polarized light. Depending on the spectrometer
> design, rotating the instrument in the plane of the LCD monitor's surface
> will produce slightly different results for the same point on the surface.
>
> Robin Myers
>
>
> On Oct 25, 2008, at 15:59 , Marco Ugolini wrote:
>
>> In a message dated 10/25/08 2:53 PM, edmund ronald wrote:
>>
>>> ROTFL :)
>>> LEDs are quantum devices. The simples ones just emit one single
>>> spectral ray. That's why we even have LED lasers.
>>> It's harder to be spikier than that.
>>> Some LEDs have a fluorescent shell encapsulating the actual diode, I
>>> think this can turn them into white light LEDs, but the light here is
>>> re-emitted.
>>
>> Edmund,
>>
>> If the following graph accurately reflects the typical emissions of
>> phosphor-based white LEDs, their spectral output is nowhere as spiky as
>> that
>> of CCFTs:
>>
>> <http://en.wikipedia.org/wiki/Image:White_LED.png>
>>
>> More on the subject (with emphasis on the ending statement in the quote):
>>
>> <http://en.wikipedia.org/wiki/LED>
>>
>>> Light quality: Most white LEDs have spectra that differ significantly
>>> from a
>>> black body radiator like the sun or an incandescent light. The spike at
>>> 460 nm
>>> and dip at 500 nm can cause the color of objects to be perceived
>>> differently
>>> under LED illumination than sunlight or incandescent sources, due to
>>> metamerism, red surfaces being rendered particularly badly by typical
>>> phosphor based LEDs white LEDs. However, the color rendering properties
>>> of
>>> common fluorescent lamps are often inferior to what is now available in
>>> state-of-art white LEDs.
>>
>> A quick review of spectral graphs shown at the following URL makes the
>> point
>> quite well about the superior smoothness of LED light when compared to
>> CCFTs
>> (yes, there is some spikiness in LEDs, but *comparatively* they seem to
>> represent a marked step towards improved smoothness):
>>
>> <http://ledmuseum.home.att.net/spectra7.htm>
>>
>> Also review the fluorescent spectra on this page (scroll halfway down):
>>
>> <http://en.wikipedia.org/wiki/Fluorescent_lamp>
>>
>> Marco Ugolini
>
> _______________________________________________
> Do not post admin requests to the list. They will be ignored.
> Colorsync-users mailing list (email@hidden)
> Help/Unsubscribe/Update your Subscription:
>
> This email sent to email@hidden
>
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Colorsync-users mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
This email sent to email@hidden