Literibbon clones

[andrew ward]

anyone found a reasonable clone for literibbon?

Id like to make up some panels but dont want to pay thousands.

Is there any more reasonably priced hybrid led strip kits that arent too green?

[Phil Rhodes]

Two that I'm aware of. A company called Yuji LED sells stuff at about $200 for a 5-metre reel of their most powerful type, which is somewhat cheaper than literibbon and comes in specified colour temperatures. They also specify a high R9 reading, where R9 is the highly saturated red colour patch which is excluded from the average used when calculating CRI. This colour is often badly underexposed under LED lighting and this deficiency in red is the main reason why skintones often look so bad; the Yuji stuff should avoid this problem. I have a reel of it but I haven't used it on anything yet.

 

There are also companies selling stuff on eBay, which are claiming 90 CRI but simply describing the output as "cool white" or "warm white" and don't give R9 numbers. I have not seen or used this stuff, although someone did post that they'd bought some recently.

 

While googling around for the links in this post, I also happened across the Neonica company in Poland, who claim CRI >90 for their LED strip. I know nothing more about them, but it's half the price of Yuji again. No R9 numbers.

 

The problem with most LEDs is not that they look green - that's a fluorescent tube thing, and an increasingly dated problem in any case - but that they lack red. In general, for obvious reasons, the "warm white" types tend to be better, but the "cool white" types may give you a cleaner, more sparkling look if you're trying to create a sci-fi environment or something like that. It would mainly be useful as an accent light on sets and I wouldn't on any account use the basic types as a key light, although the high CRI stuff may be OK.

 

After R9, the thing we're interested in is R12, which is a very deep blue. Despite the fact that white LEDs are blue LEDs driving a yellow phosphor, there is often a lack of very deep blue because the LED output is somewhat sapphire - a royal blue - and the physics behind phosphor conversion mean you can't ever shorten the wavelength, only increase it. Even more recent developments use UV-emitting LEDs to drive phosphors that create all of the white light, something like a fluorescent tube, and while this costs us in terms of efficiency it is the best available technology as regards colorimetry. I believe the Kino Celeb LED uses this approach, but I'm not completely sure.

 

There's also the factor that in a few years, having LED ribbon all over the place is going to look terribly mid-2010s.

 

P

[JD Hartman]

Phil, what LED ribbon lights are you trying out?

[Phil Rhodes]

I have some of the Yuji stuff, but I don't have a colour meter at the moment.

 

P

[Albion Hockney]

I was curious about this same thing, was looking at the Yuji stuff yesterday. Phil do you know anything about dimming/powering them properly is there anything you have to worry about? a big cost from the LiteGear stuff is their power supplies/dimmers, is there anything special to note and to make sure they don't flicker? LiteGear claims you can shoot up to 1000fps with theirs and have no flickr.

 

I think the way to go is to make panels out of them, you can make super thin bendable panels ....super useful.

 

I keep thinking about the future when this stuff will be a lot cheaper and have them in the source maker blanket style in 8x8 and 12x12 varieties. that will make for some beautiful soft wrapping light.

[Mark Dunn]

I have some of the Yuji stuff, but I don't have a colour meter at the moment.

 

P

I suspect you could force a DSLR with a custom WB function to behave rather like a colour meter.

Illuminate a white card, photograph it at a given WB and work out the offset needed to bring it back to neutral. Probably.

But I might be talking out of my hat.

[Phil Rhodes]

Not... really.

 

There are two kinds of colour meters. The first is the cheaper option, in which there are several (often four but sometimes more than ten) photodetectors behind filters. Obviously, that can't actually produce a complete representation of the spectrum of the light source, instead using mathematical modelling to produce the desired data. That's presumably how something like a Fuji C700 colour meter works, and how monitor probes like an X-Rite i1 work. It can be a bit approximate. Any sort of monitor probe that asks you what sort of backlight the monitor is using is one of these - it uses the information to improve its guesswork.

 

The proper way to do it is with a spectroradiometer, in which the light passes through a prism (or more usually a diffraction grating, but same principle) which is directed toward a narrow slot, beyond which is a single photodetector, often a photomultiplier tube for very high sensitivity. The width of the slot controls sampling range (often a few nanometres' range of wavelengths, where the whole visible spectrum is a few hundred nanometres wide). Rotating the prism on a servoactuated mount allows the photodetector to read a narrow part of the resulting spectrum, so you can get a fully comprehensive sampling of the actual spectrum. Instruments called "monochromators" work in a similar way. This interesting YouTube video shows a monochromator being worked on. The long tubular monitor probes you see in high-end circumstances are generally spectroradiometers and they cost five figures. It's an analytical laboratory instrument.

 

 

Your DSLR is most similar to the former arrangement, a bunch of photodetectors behind filters, and it is designed, in concert with the software, to have a response similar to your eyes. The filters on Bayer sensors cross over quite a bit - the red sees a bit of green, which sees a bit of blue, which is necessary for reasonable performance from the demosaic process. This is why the colours on single chip cameras are never as good as on 3-chip blocks; it has to be matrixed back out again, or in everyday language the saturation gets wound up. A lot. But I digress.

 

You could figure out very roughly what the white point of an approximately-white light was, which is the combination of white balance and green-magenta balance, if you knew exactly what the response of the filters in the camera was, and what the response was of the combined photodetector array and accompanying electronics was (you could establish this under test). But you can't know what its CRI is, let alone any lumps and holes in its spectrum, just by observing a non-coloured object. This is why we shoot macbeth charts - because we know that some of the chips have a fairly narrow reflect band.

 

P

[Phil Rhodes]

Phil do you know anything about dimming/powering them properly is there anything you have to worry about?

 

LEDs are DC devices and can, in theory, be driven completely flicker free at any frame rate. However, many LED dimmers use pulse width modulation, which is generally unsuitable for film/tv work.

 

I built dimmers using off the shelf parts from eBay. Essentially it's a step-up converter chained to a step-down converter, which takes whatever the battery is doing, boosts it up to about 30V, and then drops that back down to a variable zero to 12V DC. This provides full nothing-to-100% dimming, avoids severely overvolting the LED strip from "hot" 14.4V batteries straight off the charger, and ensures the light output stays constant as the batteries decay. This could be done more efficiently using a SEPIC converter topology, but I built these a few years ago. The amount-of-rotation-to-apparent brightness relationship is horribly nonlinear unless you mess about with a lot of serial and parallel resistances. The flicker characteristics of this depends largely on how good a DC converter you're using, and how heavily loaded it is. When using the super-cheap stuff from eBay, overspecify everything two or three times and you're probably fine. You can read the output with even a cheap oscilloscope and observe how much it's rippling and at what frequency. Most of the small cheap DC converters run at hundreds of kilohertz (hundreds of thousands of cycles per second) so even if they do ripple you're fine at conventional framerates.

 

 

Note that at 8.8V (blue digits), 5M of Yuji's LED strip doesn't draw enough current to register on the (red) ammeter!

 

I have made panels out of the super-cheap LED strip and it's perfectly effective, but I'm not sure it's the most cost effective way to an LED light if you want to use the high-CRI options.

 

P

[Albion Hockney]

so theoretically to power the LED's properly all I need to a 12v power supply of any kind?I don't need battery power of any kind so just looking for a nice affordable way to get them wall power and then dim without flicker.

 

so is your dimmer also a power supply? in general would these be a two step thing where I get a dimmer and a power supply?

 

I wonder what the best way to make a dimmer in this kinda setup would be or if there is any off the shelf sorta thing I could get without spending $300.00 on the LiteGear one.

 

Sorry little lost there as Ohms law is about where my electrical knowledge starts and end haha...

Edited October 3, 2015 by Albion Hockney

[Phil Rhodes]

so theoretically to power the LED's properly all I need to a 12v power supply of any kind?

 

Yes. They can be quite hungry; assuming you want to be safe, I'd specify something capable of 6A to run a full 5M run. It won't be a wall wart.

 

 

 

so is your dimmer also a power supply?

 

The dimmers run from 12V.

 

 

 

$300

 

Christ, I think I built mine for less than $30 in parts apiece.

 

P

[andrew ward]

Phil!

LEGEND.

Ill look into that.

 

I might have to ger Litegear but its soooooo expensive.

[Phil Rhodes]

Perhaps I should start manufacturing these things...

[Mark Dunn]

Not... really.

 

Thanks for the explanation. Now I actually know how they work.

[JD Hartman]

Perhaps I should start manufacturing these things...

 

Yes you should.

As discussed in another thread, the proceeds could subsidize your C-stand/grip gear purchases.

[Phil Rhodes]

Ehhh. It's one of those things. Everyone says "yeah, make some!", then you make some, and everyone goes "Yeah, great!" and you say, "How many d'you want, then?" and they say "Me? Oh, not me, but I think it's a really good idea..."

 

Also everyone would want DMX control. Although I could probably do that.

[JD Hartman]

Make one and put it up on UK Fleabay. Test the waters.

[andrew ward]

Make it for me!

[Albion Hockney]

As far as I know there is no LED dimming solution other then the LiteGear ...you should really do it! huge market on the horizon

 

I'm your first buyer!

[Stuart Allman]

Well...LEDs really respond semi-linearly to current, not to voltage. They are, after all, diodes with a threshold voltage and a steep voltage slope from small current to large current. So you have to be careful when applying a voltage power supply to a string of LEDs. The diodes could be on the low or high end of the voltage threshold spectrum, based on LED manufacturing process variation. What you really want is a current power supply to keep the design within safely controlled operating parameters. Yes, if you don't go crazy on brightness then you can use a voltage power supply - but that's simply not me. I want something blinding most of the time - and I want it to be rock solid reliable.

 

The other problem is that LEDs shift wavelength based on how much current you drive into them - sort of like a tungsten light on a dimmer. That's why a lot of vendors use DC drive at higher currents and switch to PWM drive at lower currents. That gives them a somewhat consistent spectrum.

 

In my spare time I'm designing an LED ballast for 300W+ applications. I have the design work done, but the board layout is going incredibly slow (just my time availability). Driving lots of LEDs in a consistent and reliable manner can be a challenge when done to maximum power efficiency. Who knows if I'll ever finish the board and build prototypes, but it was a good learning experience.

 

Stuart

---------------------------

illuma.blogspot.com

[Adrian Sierkowski]

If the price is right I'd be happy to grab some. Ideally something 4' by 6' or so in a coop light thing.

[Phil Rhodes]

 

 

So you have to be careful when applying a voltage power supply to a string of LEDs.

 

These strips of LEDs we're discussing here already have resistive current limiting (they're a long string of three-in-series strings), so dimming them is just a voltage driver.

 

P

[Satsuki Murashige]

Phil, can we pay you in c-stands? It's the new bitcoin.

 

I'd be interested in buying a few for my homebrew Varicon idea.

[Phil Rhodes]

I don't know. What ratio of dimmers to stands?

[Satsuki Murashige]

Given that the stands are $200 US or thereabouts, you tell me :)

[Adrian Sierkowski]

I'll go LED-Standzy for you.