[Review] SST-20 in FW3A is divine!

That just added some pie slices. What do they tell us?

Hell if I know, but if this is set to replace Ra/R9 we should probably figure it out :slight_smile:

AFAIK you want a big wide circle like this for even, full spectrum. If it jets in or out too much you will have a bias towards that

The black circle is the reference, the red circle is the SST20. When the arrow goes to the interior of the circle, the colors of the region are less saturated than the reference ( = negative chroma shift), when it goes to the exterior = more saturated than the reference ( = positive chroma shift). When it goes to the left or the right, there is a hue shift.
The SST20 colors rendition are very close to the reference but with slight hue shift, overall saturation matches the reference except for colors of region 14 which are a bit more saturated.
This sample is very good for colors rendition with high fidelity (compared to the reference), better than a 219B SW45K.

Is it a filtered SST20 ?

The SST-20 actually renders colors better vs a 219B SW45K, since it doesn’t have a tendency to saturate on the magenta part of the spectrum due to its more balanced tint, and actually having less tint shift.

TLDR: The SST-20 is better than the 219B in every aspect, even in terms of tint since the FD2 is now more easily available, and the FA3 and FA1 color tint bins will hopefully be available as binning gets better.

Tint though different amperages is its biggest weakness, but I’m okay with it being on bbl at lowest outputs and rosy at highest.

Yea, but that was all part of the TM-30 before the update. I was wondering what difference the update made.

+1
I think it’s just going to take some time for the general opinion on the forum to accept the 219B is obsolete by all perceivable measures.

I think a lot of folks just tend to subjectively prefer the rosier tint of the 219B SW45K. Under normal circumstances, I don’t find the tint of the SST-20s in my FW3x’s to be particularly green, but when I put them directly next to some SW45K on a white wall, the SST-20s look dreadfully green (at low outputs) and the SW45Ks are decidedly pink.

I prefer the SW45K tint but its efficiency drawbacks and sensitivity to high currents make it impractical for a lot of applications.

An FA1 or FA4 binned SST-20 might be perfection.

Tint is quite perceivable, and multiple studies have clearly demonstrated most people prefer a tint no available bin of SST-20 can achieve even at unsustainably high currents. I don’t want a light that’s 0.0016 at a level I use regularly when I could just use a 219B and get a much more pleasing neutral or rosy tint that becomes even rosier at higher outputs. There is still strong demand for 219B years after it was discontinued precisely because no other LED can replicate what it does, except perhaps E21A, which is impractical to use in most lights due to the lack of mounting options.

I may not have explained completely either. I’m not aiming for a meter-verifiable value of 0.000 duv. I’m aiming for something which I think looks good in person… like ~4500K with negative duv.

Adding a minus-green filter is a pretty common modification when using SST-20 emitters, LH351D, and sometimes others. The 89% transmittance value given earlier is a bit generous though… the range I generally see is 70% to 86% transmission, and I’m expecting to need roughly an 80% model to get most of my greenish lights down to a nice-looking tint.

As is, when I use this SST-20 FW3A outdoors, the grass looks nice and green… but so does the sidewalk. And when I point it at purple-colored objects, they turn blue. When aimed at objects with several shades of green, I find it difficult to tell the shades apart. It’s all lemongrass. So I’m hoping a filter can fix that.

I recently ended up with a D4 SST-20, which demonstrates the low-mode tint issue pretty vividly. If I put it in stepped ramp mode, steps 1 to 4 are green while steps 5 - 7 are white. If I make it go back and forth between steps 4 and 5, it looks like two different lights… green, white, green, white, green, white. Because of this, I’m tempted to convert it from a FET+1 light to FET-only, to make the tint usable below 150 lm.

… except for hitting the tint sweet spot which studies have found agreeable to the most people.

It can be tint-corrected with a minus green filter, but after filter losses it ends up making fewer lumens per amp than the aging 219B.

It’s definitely old, but there are good reasons why people still prefer 219B over 219C, why people prefer XP-G2 over XP-G3, and why SST-20 isn’t universally loved. In the quest for ever-increasing efficiency, several different LED manufacturers over the past few years sacrificed optical quality for higher total output.

The changes, like in Cree’s current-gen emitters, make a lot of sense for use in lighting fixtures where the LED is meant to run near full power all the time behind a diffuser. In that sort of device, the Cree rainbow doesn’t matter and neither does current-related tint shift. It only needs to look good when integrated and running at the recommended power level.

That sucks a bit for flashlight purposes, but flashlights aren’t a big enough market for LED companies to really care. So we’ve been using a lot of old stuff lately instead of always adopting new tech.

In any case, the point I’ve been getting at this whole time is: There is no One True LED. Different people prefer different lighting, and that’s okay.

Not everything, but an ANSI white chart should answer a few questions…

CCT goes lower left to upper right along the dotted line. Tint goes up and down along the more vertical lines. “Duv” is the distance from the dotted line. The shades most people prefer are below the dotted line between 4000K and 5000K.

The SST-20 4000K emitter this thread is about would fall into the 5A3 square, according to its official specs and the measurements it gets on turbo. However, at lower levels it shifts up and to the right, to about 5C4.

The “FA1” bin mentioned a couple comments ago is “5A1” on this chart. It might be a decent tint for SST-20, since at low levels it should only shift as far as 5D2. However, as far as I’m aware, no one on this site has ever acquired any in the FA1 bin. They don’t seem to exist. Most SST-20 around here are the 5B4 bin, which is roughly 5C3 at low levels.

My 2c:

Some time ago I owned two D4S, XP-L HI 3D 4000K and SST-20 4000K because I wanted to check which I prefer and leave just one for myself.
The throw was nearly the same, I had problems to determine which throws further.
But with both lit up at the same time, the SST-20 tint was obviously biassed with green. And I’d guess SST-20 boils up 20% sooner.
For these two reasons I sacrified high CRI for better tint and less heat. XP-L version is still with me.

(After all I still consider SST-20 to be a great LED. But similary to 219B - it just should not be pushed hard. A perfect diodes for keychain or other small lights)

I agree…. Love my 219B’s. And there are many out there who would agree.

Thank you! This was very informative!

Do you have a link to the filter products you are referring to? I would like to have a look.

I’m surprised nobody seems to complain about rosy tints. I have an equal dislike for rosy tints and green tints.

It seems green-tint phobia is a thing?

Yes.

I don’t actually like rosy tints, which is why I dislike the 219B, which I’ve actually seen in real life.

I prefer the FD2 SST-20’s tint.

Of course, we can’t deny that the current high power LEDs that have the best tint are the E21As, and in the mid-power range, the Optsolis.

My 5000K E21’s have major changes at powers above 700ma. Granted the worst in this CIE are just bare emitter.