LEDs : Colour Rendering
LEDs : Colour Rendering
Some of the recent discussions have turned rather confrontational on this matter; so I thought it best to start afresh ...
The specification of this SCHOTT Spectra light source might be of interest:
https://www.lambdaphoto.co.uk/schott-sp ... -100w.html
https://www.lambdaphoto.co.uk/pdfs/Scho ... source.pdf
MichaelG.
.
See also: https://en.m.wikipedia.org/wiki/Color_rendering_index
The specification of this SCHOTT Spectra light source might be of interest:
https://www.lambdaphoto.co.uk/schott-sp ... -100w.html
https://www.lambdaphoto.co.uk/pdfs/Scho ... source.pdf
MichaelG.
.
See also: https://en.m.wikipedia.org/wiki/Color_rendering_index
Too many 'projects'
Re: LEDs : Colour Rendering
LED'S ARE PERFECT FOR HOBBY USE.. THEY ONLY BECOME AN ISSUE WITH SNOBBISH, CRANKY FOLKS THAT WANT TO CLING TO THE OLD FASHIONED LIGHTING.. THEY ARE CHEAP, LOW VOLTAGE, ALMOST HEATLESS, AND LAST FOR YEARS....
BILLT
BILLT
Re: LEDs : Colour Rendering
I neither know nor careapochronaut wrote:price?
... The information was posted for its technical merit.
i.e. an indication of what can be achieved, with appropriate LEDs, as an approximation to 'black body radiation'
MichaelG.
.
Edit: I see that the fight has started already
Too many 'projects'
Re: LEDs : Colour Rendering
Michael,MichaelG. wrote:I neither know nor careapochronaut wrote:price?
... The information was posted for its technical merit.
i.e. an indication of what can be achieved, with appropriate LEDs, as an approximation to 'black body radiation'
MichaelG.
.
Edit: I see that the fight has started already
Unless someone adds logwood, the fire will soon extinguish itself...
To the point:
Paging the Cree site, I found they produce LEDs of 95-98 CRI, at 2700-4000K.
Under the title:
XLamp CXA2 LED Arrays: Premium color
What I could not find is LED components (say, a single 6mm/9mm LED) for sale, where the specifications are these very high CRIs. The component tables specify LEDs with CRI of 90-91 instead - or I lost my way among the data...
Because, given that there are LEDs that simulate daylight or warm light, I guess that the straightforward and cheapest way to evaluate them for the microscope is to buy a single chip, connect a 9/12V source and On!
Re: LEDs : Colour Rendering
Why so large? The filament size of the Zeiss 60w incandescent bulb is about 4mm x 3mm (as near as I could measure it without breaking the glass).What I could not find is LED components (say, a single 6mm/9mm LED) for sale, where the specifications are these very high CRIs.
Zeiss Standard WL (somewhat fashion challenged) & Wild M8
Olympus E-P2 (Micro Four Thirds Camera)
Olympus E-P2 (Micro Four Thirds Camera)
Re: LEDs : Colour Rendering
Here's an enthusiastic young man demonstrating the merit of high CRI
https://m.youtube.com/watch?v=L3LWXznJx_0
... and a couple of pages from the yuji site:
https://store.yujiintl.com/collections/vtc-series
https://store.yujiintl.com/collections/vtc-d50-series
All rather impressive.
MichaelG.
https://m.youtube.com/watch?v=L3LWXznJx_0
... and a couple of pages from the yuji site:
https://store.yujiintl.com/collections/vtc-series
https://store.yujiintl.com/collections/vtc-d50-series
All rather impressive.
MichaelG.
Too many 'projects'
Re: LEDs : Colour Rendering
Worth keeping an eye on them. The CRI is impressive indeed.
I see that there are roughly two sizes: the tiny 2835 or 5730, that fit on a flexible strip and provide ca. 20-30 lm, and the large 11mm diameter COB LEDs, of a ten-fold intensity or more, and operate on 18-21V supply.
They are targeting museums and galleries, and indeed should illuminate paintings very well.
I see that there are roughly two sizes: the tiny 2835 or 5730, that fit on a flexible strip and provide ca. 20-30 lm, and the large 11mm diameter COB LEDs, of a ten-fold intensity or more, and operate on 18-21V supply.
They are targeting museums and galleries, and indeed should illuminate paintings very well.
Yes, I was seeking a high CRI small LED of high luminosity, maybe a 3-5W LED, at least.75RR wrote:Why so large? The filament size of the Zeiss 60w incandescent bulb is about 4mm x 3mm (as near as I could measure it without breaking the glass)....LED components (say, a single 6mm/9mm LED) ... very high CRIs.
Last edited by Hobbyst46 on Wed May 22, 2019 8:45 pm, edited 1 time in total.
-
- Posts: 6327
- Joined: Fri May 15, 2015 12:15 am
Re: LEDs : Colour Rendering
I was merely asking if you knew the price. thanks for the respectful answer.MichaelG. wrote:I neither know nor careapochronaut wrote:price?
... The information was posted for its technical merit.
i.e. an indication of what can be achieved, with appropriate LEDs, as an approximation to 'black body radiation'
MichaelG.
.
Edit: I see that the fight has started already
Re: LEDs : Colour Rendering
You're welcomeapochronaut wrote:I was merely asking if you knew the price. thanks for the respectful answer.
... If I do happen to find a price, I will post it for reference; but it is honestly of no interest to me.
MichaelG.
Too many 'projects'
Re: LEDs : Colour Rendering
Interesting YouTube video of an led conversion on an Zeiss OPMI1 microscope.
https://www.youtube.com/watch?v=jfKX8ZNqXRs
https://www.youtube.com/watch?v=jfKX8ZNqXRs
Re: LEDs : Colour Rendering
A nice video. I like the mechanical-electrical approach.Roldorf wrote:Interesting YouTube video of an led conversion on an Zeiss OPMI1 microscope.
https://www.youtube.com/watch?v=jfKX8ZNqXRs
Some details about the LED itself and the results are missing, IMO:
1. It seems that the LED is 1.6 W (500mA x 3.2V). Does it provide the same illumination as a (new) 30W halogen ? perhaps if the LED is of a narrow beam type. What LED brand and model was it ? Is the final result sufficient ?
2. An OMPI 1 manual that I looked at (on a web site) states that the appropriate bulbs are 12V 100W halogen, not 30W.
3. Can the LED be driven directly from the original power supply of the microscope ? just a drop-in replacement ? no driver or current-limiting resistor ?
And comments:
4. The original bulb, the one that the guy replaces, looks like an incandescent and not a halogen bulb. I could be wrong.
5. Are 12V halogen bulbs that expensive ($50) ? such bulbs are vehicle parts, I think not at that cost.
Re: LEDs : Colour Rendering
Hi together,
I knew a guy who had a small electronics company and supplied the NDR (NordDeutscherRundfunk = public north german broadcasting) with LED light panels for TV studios and ring lights for TV cameras. This was about 10 years ago and he used a mixture of different coloured LEDs to acheive the desired (not necessarily neutral) light effect. Especially women liked his ring lights very much because they look younger under the multidirectional frontal light that reches into every wrinkle.
For microscopy purposes a multi-LED-lamp would be less than ideal, but for other lighting it might be a good path to acheive a nice colour curve.
Bob
I knew a guy who had a small electronics company and supplied the NDR (NordDeutscherRundfunk = public north german broadcasting) with LED light panels for TV studios and ring lights for TV cameras. This was about 10 years ago and he used a mixture of different coloured LEDs to acheive the desired (not necessarily neutral) light effect. Especially women liked his ring lights very much because they look younger under the multidirectional frontal light that reches into every wrinkle.
For microscopy purposes a multi-LED-lamp would be less than ideal, but for other lighting it might be a good path to acheive a nice colour curve.
Bob
Re: LEDs : Colour Rendering
As promised:MichaelG. wrote: ... If I do happen to find a price, I will post it for reference
https://www.villalighting.com/2310304/P ... ND_1114078
MichaelG.
Too many 'projects'
-
- Posts: 2790
- Joined: Sat Mar 03, 2018 9:09 pm
Re: LEDs : Colour Rendering
Absolutely. A friend makes booth lighting systems for jewelry/gem trade shows (http://www.showofflighting.com) with alternating light temperatures. The mix is wonderful for gems of most sorts. I imagine it could make for interesting stereo microscope lighting anyway.MicroBob wrote: For microscopy purposes a multi-LED-lamp would be less than ideal, but for other lighting it might be a good path to acheive a nice colour curve.
Bob
Re: LEDs : Colour Rendering
They say 1750 Lumens from 96W - this seems a faily low ratio! This is about what a halogen lamp offers.MichaelG. wrote:As promised:MichaelG. wrote: ... If I do happen to find a price, I will post it for reference
https://www.villalighting.com/2310304/P ... ND_1114078
MichaelG.
Re: LEDs : Colour Rendering
I think the main 'selling point' is the 50,000 hour lamp life ... The cost of swapping-out halogen bulbs in a display environment can be very significant.MicroBob wrote:They say 1750 Lumens from 96W - this seems a faily low ratio! This is about what a halogen lamp offers.
MichaelG.
Too many 'projects'
Re: LEDs : Colour Rendering
From his youtube post:-
""The LED is a Lumiled Luxeon Rebel, 1W, 2700K warm white.
I used an old PC power supply (5V) and a power resistor to limit the current. Multiple leds will not make an even illumination. You will probably see a pattern of light instead of an uniform brightness. But the 1W LED is bright enough.
The light bulb is only 30W 6V. At full power it is way too bright to work with. So I normally had it on the dimmed switch position. The 1W LED has the perfect brighness. I'm happy with that.
No fan needed. The brass strip takes away enough heat. I added a small piece of brass to act as a heat sink, but that's all.""
As regards multiple led's:-
""That may work... you must try it. The problem is that you have not much space for a big LED light. And I'm not sure if it makes a good light pattern when you use multiple light sources. Try it...!""
Hope that answers a few of your questions
""The LED is a Lumiled Luxeon Rebel, 1W, 2700K warm white.
I used an old PC power supply (5V) and a power resistor to limit the current. Multiple leds will not make an even illumination. You will probably see a pattern of light instead of an uniform brightness. But the 1W LED is bright enough.
The light bulb is only 30W 6V. At full power it is way too bright to work with. So I normally had it on the dimmed switch position. The 1W LED has the perfect brighness. I'm happy with that.
No fan needed. The brass strip takes away enough heat. I added a small piece of brass to act as a heat sink, but that's all.""
As regards multiple led's:-
""That may work... you must try it. The problem is that you have not much space for a big LED light. And I'm not sure if it makes a good light pattern when you use multiple light sources. Try it...!""
Hope that answers a few of your questions
Re: LEDs : Colour Rendering
Thanks Roldorf. Do I correctly understand that the Youtube video was actually done and posted by yourself ? if so, my compliment for a very nice job.Roldorf wrote:From his youtube post:-
""The LED is a Lumiled Luxeon Rebel, 1W, 2700K warm white.
I used an old PC power supply (5V) and a power resistor to limit the current. Multiple leds will not make an even illumination. You will probably see a pattern of light instead of an uniform brightness. But the 1W LED is bright enough.
The light bulb is only 30W 6V. At full power it is way too bright to work with. So I normally had it on the dimmed switch position. The 1W LED has the perfect brighness. I'm happy with that.
No fan needed. The brass strip takes away enough heat. I added a small piece of brass to act as a heat sink, but that's all.""
As regards multiple led's:-
""That may work... you must try it. The problem is that you have not much space for a big LED light. And I'm not sure if it makes a good light pattern when you use multiple light sources. Try it...!""
Hope that answers a few of your questions
Perhaps the 100W halogen is only required for very demanding operations...
Re: LEDs : Colour Rendering
Sorry the answer is 'NO', but it is a very well made video and compliments to the author for a very informative post. As regards the power supply I would think that a standard 1watt usb phone charger would avoid the use of resistors for the LED you could possibly even supply power from a laptop PC 500ma USB port (I am not an electrical engineer )
Re: LEDs : Colour Rendering
Colour photo enlarger used a white box into which light was projected through three adjustable filters. The mixed light in the box was used to project the negative on the paper. This would work for a microscope too, the question is how far from a perfect illumination this will be. With a diffused acting as the new light source this probably work acceptably. In the end it would be a choice more with its own advantages and disadvantages over the common light sources.
I haven't compared the wave spectra of the different LED types. It would be interesting to know how much closer to a nice spectrum one gets by use of 2,3,4,etc LEDs.
Bob
I haven't compared the wave spectra of the different LED types. It would be interesting to know how much closer to a nice spectrum one gets by use of 2,3,4,etc LEDs.
Bob
Re: LEDs : Colour Rendering
Michael,MichaelG. wrote:Here's an enthusiastic young man demonstrating the merit of high CRI
https://m.youtube.com/watch?v=L3LWXznJx_0
... and a couple of pages from the yuji site:
https://store.yujiintl.com/collections/vtc-series
https://store.yujiintl.com/collections/vtc-d50-series
All rather impressive.
MichaelG.
Thanks again for the info about high CRI LEDs. It is really interesting how good their photomicro performance will be.
Personally, I am using two small cheap 5mm warm white (2700-2900K), high luminosity LEDs (either 7 or 18 cd, do not remember the lm data) for epi-illumination near the objective front lens. They are about 0.09W each. Have no idea about their CRI - probably not better than 80-85.
It would be nice to acquire similar LEDs, but of a high CRI. Say, 5mm, narrow viewing angle, ~30mA current, ~3 VDC forward voltage, CRI 95 or better.
The color temperature is not critical - say, 3000-5000K. i would then be able to compare.
So far I searched the YUJI and CREE sites (as well as other LED supply houses), but there are no such components for sale - at least, not just a few or single pieces.
Re: LEDs : Colour Rendering
Hi Doron,Hobbyst46 wrote:... for epi-illumination ...
do you want to use this setup with higher powered objectives, offering light low down in a flat angle?
Bob
Re: LEDs : Colour Rendering
Hi Bob,
It is already in use in my ongoing project: please see
viewtopic.php?f=6&t=7504&p=65881#p65881
the last two responses. There is a scheme of the setup. The objectives are Plan 2.5X, Plan 6.3X; mostly the 6.3X. Sometimes the Neofluar 10X, but not higher. The 16X is already too much for the Forams. Moreover, the "chamber" is thick (or deep, you may call it) so the working distance of the objective becomes a limitation.
It is already in use in my ongoing project: please see
viewtopic.php?f=6&t=7504&p=65881#p65881
the last two responses. There is a scheme of the setup. The objectives are Plan 2.5X, Plan 6.3X; mostly the 6.3X. Sometimes the Neofluar 10X, but not higher. The 16X is already too much for the Forams. Moreover, the "chamber" is thick (or deep, you may call it) so the working distance of the objective becomes a limitation.
Re: LEDs : Colour Rendering
What I was thinking about was an incident illuminator that works with a transparent plastic part as a light transmitter. This could be quite flat in the middle area to fit under higher powered objectives and be thicker are the outer area to fit to the size of the LED. With polished surfaces all around and a matte surface towards the object in the middle this might allow illumination up to the 25:1 or 40:1 objective. It could be made from exoxy or silicone in a mould made from childrens dough, the waxy kind. A planconcave lens might also offer a good piece of material.
- Attachments
-
- Auflicht Illuminator.jpg (57.12 KiB) Viewed 16522 times
Re: LEDs : Colour Rendering
I have just searched YUJI for 5mm ...Hobbyst46 wrote:It would be nice to acquire similar LEDs, but of a high CRI. Say, 5mm, narrow viewing angle, ~30mA current, ~3 VDC forward voltage, CRI 95 or better.
The color temperature is not critical - say, 3000-5000K. i would then be able to compare.
So far I searched the YUJI and CREE sites (as well as other LED supply houses), but there are no such components for sale - at least, not just a few or single pieces.
https://store.yujiintl.com/search?q=5mm ... ge,product
... Sadly, it looks like this might be a 'not inexpensive' exercise.
Hopefully someone, somewhere, sells-them-on in smaller quantities.
MichaelG.
.
Edit: I've found this posting, with some useful-looking links:
https://community.bulb.co.uk/discussion ... ri-r9-leds
Too many 'projects'
Re: LEDs : Colour Rendering
So I looked into them, and they only sell LED strips or LED lamps for home lamp fixtures, like E27 bulbs, 220V, etc. It seems that the high CRI LED market has erroneously developed for residential rather than amateur miroscopy clients... After all, who on earth needs 5mm 95CRI white-light LEDs ? But, when someone will discover they are useful for hydro growers, then...MichaelG. wrote:Edit: I've found this posting, with some useful-looking links:
https://community.bulb.co.uk/discussion ... ri-r9-leds
Re: LEDs : Colour Rendering
Eureka !!MichaelG. wrote:Hopefully someone, somewhere, sells-them-on in smaller quantities.
https://www.ledsales.com.au/index.php?m ... 48_160_291
Noting that 'somewhere' would, almost inevitably, be the Antipodes !!
MichaelG.
Too many 'projects'
Re: LEDs : Colour Rendering
On the 'Chinese Microscope' thread, Alan [Roldorph] recently wrote:
BUT there are other applications where reliable quantitative colour assessment is important.
e.g. How would you distinguish a Red herring from an Orange herring if you did not have a light source with known characteristics ?
Different 'commonly available' White LEDs have significantly different spectra, so the cross-calibration of results from different microscopes might be near impossible.
The merit of using 'High CRI' LEDs will vary from 'who cares?' through 'nice to use' to 'important' according to the user and the work being done.
Please note: My interest in this is not elitist, it is technical.
MichaelG.
.
P.S. I should also mention that actual 'daylight' is itself not particularly consistent, and that the reference curve for CRI is 'Black Body Radiation ...
I think the important part of that statement is "for our application" ... and I certainly wouldn't try to persuade him that high CRI is required for their application.Oh one last thing. Led lighting and whether or not it is true to natural daylight or not, is a bit of a red herring certainly for our application as mostly the samples we look at have to be stained to be able to see them as with most things that are microscopic, so who cares what colour the light is as long as the heat from the lamp doesn't kill the specimen. Cool ha.
BUT there are other applications where reliable quantitative colour assessment is important.
e.g. How would you distinguish a Red herring from an Orange herring if you did not have a light source with known characteristics ?
Different 'commonly available' White LEDs have significantly different spectra, so the cross-calibration of results from different microscopes might be near impossible.
The merit of using 'High CRI' LEDs will vary from 'who cares?' through 'nice to use' to 'important' according to the user and the work being done.
Please note: My interest in this is not elitist, it is technical.
MichaelG.
.
P.S. I should also mention that actual 'daylight' is itself not particularly consistent, and that the reference curve for CRI is 'Black Body Radiation ...
Too many 'projects'
Re: LEDs : Colour Rendering
Thanks, you found them ! Specs: current - 20-30mA, luminosity - ~8lm, angle - 45deg. I estimate 8*360/(4*pi*45) = ~ 5cd. About X3 weaker than my plain LEDs, yet acceptable. The price is fair (less than 1USD apiece) but transport brings it to $30-40. I will wait until they diffuse nearer.MichaelG. wrote:Eureka !!MichaelG. wrote:Hopefully someone, somewhere, sells-them-on in smaller quantities.
https://www.ledsales.com.au/index.php?m ... 48_160_291
Noting that 'somewhere' would, almost inevitably, be the Antipodes !!
MichaelG.