Page 2 of 2

two colour wave files

PostPosted: Tue May 17, 2011 8:52 am
by grahamhunt
have found an excellent orange filter medium repair tape for car indicator
lenses :D

PostPosted: Tue May 17, 2011 11:53 am
by gary
Hi Holtzman,

The level is set thus because that is the club standard, i.e. it represents 1V p-p on a standard (as if there was one) CD player.

Keep in mind you only need 8 bits to represent full contrast (i.e. black = 0, white = 255), so the level could be much lower than that even and still be displayed with full contrast.

I think the perceived lack of contrast is more to do with the 2 colour scheme, after all how do you reproduce white when you only have 2 colours?.

Now, as I have previously mentioned, I am certainly no expert in colourisation, so if you see how to recover the RGB more accurately than I am doing (I am just reversing the process of going from RGB to 2 colour, but, of course, it is not a fully reversible process) then I would be only too happy to try it out.

If you would like to try the Baird system I could implement that easily enough.

EDIT: Of course I realise you can always come up with some scheme that calculates to white = 255,255,255 - but as far as I can tell that only shifts the problem to somewhere else in the colour space, but maybe would be a perceptual improvement after all white is such an important colour.

PostPosted: Tue May 17, 2011 9:31 pm
by Lawnboy
Excellent work gary! i have been contemplating building a 2 color system like this for demo purposes for months and now i have no reason not to. i have converted a few videos so far and this system delivers pleasing results.

i am planning on using standard orange leds and some aqua leds for the monitor. i did a search for an appropriate cyan/aqua led a little while back and found a good one here (USA company): ... LED/1.html

two colour wave files

PostPosted: Wed May 18, 2011 7:20 am
by grahamhunt
On the subject of cyan L E Ds MUTR list cyan+aqua leds only problem
£10 min order Re colours on play back filters only white perfect blue ok
if vivid red a bit pinky will report when rgbs + cyan leds arrive 8)

PostPosted: Wed May 18, 2011 8:57 pm
by gary
Lawnboy wrote:i did a search for an appropriate cyan/aqua led a little while back and found a good one here (USA company): ... LED/1.html

That looks perfect!

PostPosted: Wed May 18, 2011 9:57 pm
by Klaas Robers
The use of colourfilters and white LED's might cause a problem, because white LED's aren't white. Whit LED's are normally blue LED's with some stuff that converts part of the blue light into yellow light. So the output is only yellow and blue. If you try to get other colours out of it it might not work, just like seeing other colours in the light from low pressure sodiumlamps than yellow is not possible.

It is due to the working of our eyes that we can simulate quite some colours while mixing red, green and blue, but even those colours are different from the original colours, e.g. it is impossible to simulate the light coming from the said sodiumlamps while mixing red, green and blue.

2 colour wave files

PostPosted: Wed Jun 15, 2011 8:13 am
by grahamhunt
"Ideas please" I have been kicking around the idea of wiring a set of rgb LEDs so that all the blues on one channel all the reds on the other and then mix some signal from each to go to all the green leds .I am no expert but supposed that a diode in each lead would prevent back feeding any thoughts :?: :roll:

first thought

PostPosted: Thu Jun 16, 2011 12:56 am
by holtzman
It seems interesting, I made some sketch. You'll need R1 almost surely. I am not so sure about R2, where it should be. The green component must be weaker than red, but red led has already lower resistance. Of course, diodes can affect gamma. But ,hopefully, tuning may give good results.

PostPosted: Thu Jun 16, 2011 9:02 am
by holtzman
Sorry, I think the circuit published above by me is not good. The green channel won't be this way a truly mixed one. Either orange or aqua will "lock" each other, depending which of them gets stronger signal.
As I see it by now, 3 amps required for driving rgb leds. The green channel amp should be connected via audio mixer circuit to both inputs.

Mixer for NBTV

PostPosted: Tue Jul 19, 2011 12:12 am
by holtzman
Hi all,
I got some powerful RGB led and want to experiment with 2-channel colored NBTV. There will be 3 amplifiers, R, G, and B. The G amp should get mixed signal, the sum of Aqua and Orange signals. This way I hope to avoid the mess with finding leds with exact color.
Now the question is, what kind of cirquitry can do the mixing job? I googled the subject, but all I got were or mixers wich invert, or complicated ones, or those wich seem won't work. Anybody here knows how to mix 2 channels correctly, without attenuation/inversion/dc offset problems etc.?

PostPosted: Wed Sep 19, 2012 10:47 pm
by Harry Dalek
I never really looked into colour NBTV as it was a bit beyond me .
But i have been reading what others have been doing past few years on the forum ,Garys 2 colour system software makes and seems the easiest way into it apart from if you want a closed system .You have a easy way to make test colour sound files and a viewer to test on .
Is there a reason its not the club standard ?
Am i right in saying the a percentage of the luminance governs the 2 colours
and the 3rd is recovered by the percentage left over ....
Thinking this thread was worth another shot i found it very interesting.

PostPosted: Thu Sep 20, 2012 7:43 pm
by gary
Graham (Hunt), I heard through Jeremy that this two colour format was demonstrated at the last NBTV convention. I assume it was you who did the demonstration - I was wondering if you could give us a run down on it?

two colour wave files

PostPosted: Wed Sep 26, 2012 7:15 am
by grahamhunt
The monitor I displayed at the convention used my monochrome disc televisor, the original white LED being replaced with a blue one.Three more LEDs two green and one red made up the new display.One blue and one green being fed from one channel and the red and other green from the other.the reason Iused two greens is my inability to make a suitable mixer,but as it turned out for a good selection of colours more green was rquired than one Led would produce.Skin tones being faithfully reproduced reds and blues ok as long as you avoid very dark shades these show as black.This is eaven worse with greens only the lighter shades showing but the poorest colour is yellow usualy shows as pink, eaven though blond hair appears ok this may be down to seeing what you expect.
If anybody is interested in trying to colour I would recommend you start with filters as very good results can be had in this way
Graham Hunt :oops:

PostPosted: Wed Sep 26, 2012 10:15 am
by gary
Thanks Graham, did you use a current splitter to provide the correct proportions between the LEDs? (obviously with some sort of tweaking to allow for slightly different characteristics in the LEDs). I wonder what the best method of calibration would be?

BTW I wonder if the holy grail of an acceptable 2 colour system warrants my writing an application that allows the three primary colours to be mixed in any combination into 2 colours to enable someone (it's pointless it being me as I am a little red-green colour blind) to determine the best arrangement for the widest range of "perception" 2 colourisation? Or do you think we are flogging a dead horse?

The idea would be to be able to see on the screen what the colours should look like but also to be able to output it through the sound card to an LED cluster that is being calibrated.

two colour wave files

PostPosted: Wed Sep 26, 2012 11:46 pm
by grahamhunt
I used four separate amps so that all the leds were seperatly adjustable.The other amps were single transistor amps modified for 12 volts