Making tri LED fixtures 'deeper' in colour

[Tomo]

We have used a mixture of different RGB and RGBW lanterns in the last year or so at our tiny theatre and they have proved immensly useful for giving a punch of deep colour, especially where the precise shade does not matter. We certainly don't try to light faces with them but have found we can balance different makes pretty well to each other. We do almost exclusively plays here, so they have worked despite being in an unpromising subject area!

 

We are now looking into extending this experience with some of the Selecon PL1 range (with which I have no self interest) which are deliberately designed to cover a much wider spectrum so will mimic a lot more favourite gel colours than a typical RGB lantern (and obviously cost a lot more).

The Selecon PL1 are RGBW fixtures, so you'll get a very similar colour result from them as your existing RGBWs - optics are another matter of course!

Their output wavelengths are on page four of this PDF: http://www.seleconlight.com/images/stories/files/english/productspecs/PL1/PL1_Frequently_asked_questions_11_10.pdf

 

This is typical of RGBW, it shows the extremely narrow bands for RGB and the wide-band white blue-pumped & phosphor very clearly.

 

RGBW is much better at making white light than an RGB or W alone, but has no effect at all on saturated colours - you don't use the white diode for them.

That means their colour gamut as drawn on a Hue/Sat circle or CIE is almost the same as an RGB - but they do have a much better CQS near white.

 

The differentiation between different RGBWs is generally the phosphor mix on the White diode - warmer or cooler with a higher or lower CRI*, thus affecting the fixture's CQS.

Selecon will have chosen their White phosphor mix to offer a good range of whites using the RGBs to 'tint' it, but you can see that it's a very different shape to the tungsten curve shown in the background below.

 

The actual light output is the sum of all active diodes, as seen is these graphs of the light output of an ETC Selador fixture and a typical RGB fixture:

http://www.layersoflight.com/Layers/media/MediaLibrary/Images/4-Source%20Four%20LED%20-%20non-Homepage/Detail%20Images/detail-qol-1-color-lg.jpg?width=200&height=200&ext=.jpg

 

The control you have in the fixture is the height of the peaks in these graphs.

A Gel (colour filter) cuts/reduces out some colours of light from the final mix - eg overlay the filter diagrams from Lee.

 

I just thought of a mental image that might help:

Think of the LEDs as being fresnels focussed on stage a distance apart with a little bit of overlapping spill. You cannot refocus them, only adjust their brightness.

 

The 'Blue' one is stage right, 'Green' centre stage and 'Red' stage left.

Your 'White is a single circuit of multiple floodlights covering the whole stage, with a bright spot far over on stage right, a dim patch just to stage left of the centre of the 'Blue' beam.

 

You want to light the 'Yellow' area between the Green and Red ones.

 

You have to turn on Red and Green to get any light in the 'yellow' area - but that means lighting Green and Red areas much brighter!

You can add White, but then you just flooded the whole stage and that's not right either.

 

In theory Barndoors or blackwrap (gel) could block the green and red areas - but then there's hardly any light left, and you can't light the Green area at all.

 

The solution is of course the same as in theatre: Add more lights to cover individual areas, then you will have much more controllable lighting.

 

So with LED, more colours gives you better light.

 

* CRI was designed for the wideband phosphors in HID and fluorescent, so can be a misleading measurement of luminaires that use narrowband emitters like Red, Green and Blue.

That means some LED fixtures get higher and lower CRIs than they really 'deserve'. The CQS (Colour Quality Score) is a better method for these types of fixture.

[bigclive]

Since the LEDs themselves are single wavelength you're not going to get a different wavelength of blue from the unit. You may be able to cheat a deeper looking blue by adding a tiny amount of red to it, but it's also worth experimenting with turning the intensity of the blue channel down to make a "darker" blue. The brain seems to interpret the lower intensity as a deepening of the colour to a degree.

 

Likewise the bright red can have a tiny touch of blue added to "deepen" the colour.

 

Other than that, for such wavelength specific colour requirements you may need to go to tungsten and gels.

[Jamtastic3]

Interesting info guys cheers.

 

Yeah, the blue and the red are OK, but nothing exactly special. It's especially annoying since the two together don't make a very nice magenta/purple. If I could bring the blue a bit deeper then this would help. Green I'm not too fussed about. The red, again, it's a primary red. It's OK, but as an LED output, it looks a bit 'wishy washy'.

 

I'll maybe try the different ideas some of you have suggested. I just thought a tinted gel of some description (blues or reds spectrum) may be able to help.

[alistermorton]

that on that basis it is not going to pass risk assessment here.

 

But ... but ... the director saw exactly the same effect in the West End last week and wants to duplicate it for the Christmas show.

 

 

 

[gyro_gearloose]

Other than that, for such wavelength specific colour requirements you may need to go to tungsten and gels.

Or discharge lamps and CMY colour mixing

[bigclive]

Before red LEDs had competition from blue and green the "high output" ones were typically 660nm which is a very deep red. They weren't really high output though since the human eye is less sensitive as the red goes closer to the infrared end of the spectrum, but the manufacturers could quote the intensity figures to make them appear brighter on paper.

 

Now many red are around the 635nm mark due to the eyes greater sensitivity to the area where it is starting to border on orange. It follows that the LED industry chooses the actual "brightest" looking to compete with the blues and greens and that happens to be bright orangey red.

[indyld]

Isn't it about time we had a debate about CQS here on the Blue Room? Am not much of tick box and numbers person in practical performance lighting design, but it's certainly a topic that I'm sure will start a few arguments. Didn't find one here when I Googled.

 

It is right that CQS is "better" for a particular type of light source (er, oh yes, LED...) although what they "deserve" is open to question.

[Brian]

Isn't it about time we had a debate about CQS here on the Blue Room?

 

Well, I'm all for something that makes buying a house easier.

 

The Conveyancing Quality Scheme (CQS) provides a recognised quality standard for residential conveyancing practices. Membership achievement establishes a level of credibility for member firms with stakeholders (regulators, lenders, insurers and consumers) based upon:

 

the integrity of the Senior Responsible Officer and other key conveyancing staff

the firm's adherence to good practice management standards

adherence to prudent and efficient conveyancing procedures through the scheme protocol