[Music]

This episode is basically just a nerd–out session to kind of look at the state of LED lights in 2023, relative to tungsten lights. Now, in this entire episode, when I say tungsten, I'm referring specifically to tungsten lights or halogen lights; anything that has a filament that heats up and emits light, and that might not even be the best definition. But in any case,

I think you know what I’m talking about: tungsten lights, halogen lights, we'll just refer to them as tungsten, versus LED lights.

Now, LED lights have come a long way.

My first LED panel I bought in… 2014, maybe?

And it was awful in terms of the color that it produced.

That's an important thing to understand, is that light has color to it and it falls across an entire spectrum and we'll talk about that in a little bit more detail in just a moment.

Now, who needs to care about this? Who needs to care about color quality when it comes to choosing what light you use for video?

And if you're just filming YouTube videos, or stuff for fun, and you're getting good results with what you have, then that's fine. Don't– I'm not here to convince you that you need to upgrade your lights necessarily.

If you're getting the results you're happy with, stick with it.

But for those that are working for clients, especially doing things like product shots, where you have to get the colors right, this is where this may be a more interesting and more relevant discussion and your choice of lights, whether you're renting or buying, becomes a lot more important.

Now, when you set the white balance on your camera, your camera is expecting the light to come in in a very specific way.

That is to say, it's expecting so much light at various points along the color– the light color spectrum, in certain ratios to each other.

So, here, for example, is a spectral power distribution chart and it shows how much light at various parts of the visible spectrum the light is generating, and this is a tungsten light or a halogen light.

And you can see it's very, very continuous and very smooth in terms of its overall light output.

And the camera can expect that; a camera can be programmed, and its white balance can be programmed, very specifically to know exactly what to expect when you use a tungsten light and you set your white balance to tungsten.

That means that the color is going to be rendered in a predictable way on your camera.

Now back to spectral power distribution charts.

Here's one, of course, with a tungsten or a halogen light.

And again, notice how distinct and clear the pattern is.

Now, here's a modern LED's spectral power distribution.

Notice how the pattern is different. Here's another, and here's another, and note how they're all different from each other.

So this creates a potential problem.

And the problem is that when you set your camera to a specific white balance and then you use one of these LEDs, the LED is not actually producing light in the different colors along the spectrum at the same in the same ratio that the camera is expecting.

And so you will get what are called color rendering errors.

Is it going to be a problem? Again, if you're just doing talking head videos for YouTube like I'm doing right now… probably doesn't matter. And you can probably do a little bit of color correction if you have to.

Although, if you really want to get things looking exactly how you want, you might have to do secondary corrections, which are pretty… a little bit more intensive in terms of effort and time. Nevertheless, that's the state. Now, the interesting thing about these halogen and tungsten lights is that they're all pretty much the same.

They're very, very consistent from light to light to light to light, and it's just a– it's just because of the technology.

Now I realize halogen lights in particular just, they have a lot of downsides to them.

So while they have this massive upside where they're very predictable in terms of their spectral output, they're… they get super hot.

They use a ton of electricity for the amount of light output that you have.

If you want to color match daylight light, that you have light coming in a window and you want to match that, you have to put a gel on and you lose a ton of power when you put a gel on.

When you dim them, they change color.

All these different things that make them not the easiest to work with.

So I'm not here to have a debate about whether you should use halogen lights or not versus LED lights. I think that in my experience nowadays,

LEDs are so much more convenient from the standpoint that you can dim them without them substantively changing color.

They are much more efficient, so you're not going to be tripping circuits all over the building where you're trying to film.

They don't get nearly as hot, so they're a lot safer to use.

But they have the downside of their color is not generally as predictable.

So what do we do? Again, if you're getting decent results with your existing lights,

I'm not here to convince you to change or to upgrade, but here are some thoughts. First, humans can generally see between 380 nanometers and 750 nanometers in terms of color radiation; evidently for cameras, whether it's film or digital, they're most sensitive in the 380 through 670 nanometer range.

So that's the part that the cameras are most sensitive to.

Now, the Academy of Motion Pictures has created a measurement to measure the predictability of a lighting source called Spectral Similarity Index.

Now, it is not a perfect measure, but it's pretty useful, and let me explain what it does. It essentially measures the difference in color power in 10 nanometer bands between 380, that's violet, all the way up to 670 nm, which are your reds. So what SSI does is it compares a light that you're measuring– you have to have a device that can measure the color– and it compares that to a known light source.

So for example, if I'm measuring this LED light,

I can get a reading on this and then compare it to a tungsten or halogen LED source.

And then it calculates the difference between the two, and that's an SSI score.

Now, an SSI score above 90 generally means that you're unlikely to experience color rendering errors that, at least, are detectable by the camera.

So that's how SSI works.

Now, there are lots of problems with this metric; if you just look at the number, you don't know where it differs.

So for example, if I get a score of 75 out of 100 on a particular LED light relative to a tungsten light, or a daylight source, I don't know which colors are kind of out of whack, if you will.

I don't know where I'm going to experience color rendering errors, but if you do look at the spectral power distribution chart where you do compare the two, you can get some ideas there. So again, not a perfect measure, but it actually, I think, is quite useful.

Now in my experience, and I've reviewed quite a few LED lights, at least in the, I would say, in the prosumer range, and the highest score up until just recently, the highest score I ever saw on any of these lights was an SSI score of about 85 or 86. And generally, the daylight balanced lights almost always came in around 73 to 75.

So when you get down into the seventies, you're going to start to have some color rendering errors. If you get down into the sixties, you're definitely going to probably start to actually notice things by eye very easily.

So again, I'm looking mainly for those that are working on doing color critical type work.

I would say you'd probably want to have something in the ninety and above range.

Let's just take a look at a few things here and show you what I mean.

So here is a comparison of a shot made with tungsten and look mostly at the skin tones.

I've also put a color chart in there, but look primarily at the skin tones, which is going to be the most important thing for any sort of talking head or interview type videos, which is mainly what I do. So here is our first LED light compared to a tungsten light.

And in this case, this is a– this is actually the only light I've seen that actually achieves a score of 90 SSI.

And this is the Amaran 200x S.

It's a bicolor LED light and we're using it in the tungsten color right now.

So this is probably the best example I've seen of any LED in all of my experience, reviewing both prosumer and pro level lights.

Now, I have– there are a lot of pro level lights I haven't looked at.

Any of the Arri lights I haven't used.

So lots of great lights there, I don't know how they perform in terms of SSI overall, but in any case, this is– and this, interestingly, is not the most expensive light by a long stretch.

In fact, Amaran is Aputure's sort of budget line for creators.

So they've done something interesting here.

What they've done is they've created an LED light that has three different colored chips.

It has two blues and one tungsten colored… combination of chips, anyway.

So what it's able to do is it's able to compensate and reach farther down into the blues and closer to the violets.

And therefore it's not a… it doesn't have to push this massive blue spike that you see on almost every LED light out there.

So that's how it's achieving the score. And the results, if you look at the test shots here, from my point of view, the skin tones, especially, are rendered very, very close to how tungsten lights render skin tones.

Now out of curiosity, here are some other lights. Here's a pro level light, a Lupo. This is the 60 Full Color, it’s a big 60 centimeter panel, and you'll notice that the colors are actually substantially different.

Here's another one. This is the Nanlite Forza 60.

It's an RGB ACL light, and this is really made as an effects light, to be honest. But nevertheless, from my point of view, the skin tones look really quite yellow here.

You know, I need to test higher end RGB ACL lights, but here's just one example. Now,

I think a lot of people think that– and actually, the RGB ACL lights that I've looked at– and for those that aren't familiar,

RGB ACL lights produce– they have six different colore chips in them: red, green, blue, amber, cyan, and lime.

And the thinking is that with those six different LED colors, you can create a larger color space, which indeed you can. Is it more accurate?

We'll have to do some more testing, but again, that's where I think SSI becomes a useful tool.

So what to take away from all this rambling?

I think if you're doing color critical work, you're going to want to pay attention to those SSI scores.

You'll probably want to get something like a Sekonic C800, so you can test those. We'll do more of that on this channel here, of course, like we always do. But overall, just now we're starting to see LED lights that are able to achieve 90 SSI and above.

And to me that's a pretty exciting new world, where you get all the benefits of an LED light and you don't sacrifice in terms of overall color quality– at least not to a great extent. So, I hope that was helpful.

If you have any questions, go ahead and leave those down below. If you've not already subscribed, make sure you do that. And we'll be sure to get you more great videos on how to improve your lighting and sound for video. Talk to you soon.