LED test / review – Osram OSTAR Projection Compact LE UW Q8WP – Black Flat's big brother, very high luminance at higher flux!

LED test / review


Osram Ostar Projection Compact LE UW Q8WP

UPDATE --- 06/21/18 --- diagram Vf / light flux depending on Tj added!

UPDATE --- 06/23/18 --- raw data added! (link)

A longer time ago I got an Osram Ostar Projection Compact LE UQ Q8WP. This LED looks very interesting for luminance and maximum light flux, which could possibly beat also the good old XP-G2 S4 2B in luminance.

Technical data

Tj 25 °C / If 1,400 mA unless noted otherwise

Order code: unknown

Type: single die (lateral)
Binning: unknown
Rated voltage: min. 2.80 V, max. 3.50 V
Forward current: 40 to 5,000 mA
Peak forward current: 40 to 6,000 mA
Viewing angle: typ. 130° (at 50 % Iv)
Thermal resistance: typ. 2.6 °C/W (max 3.6)
Junction temp.: max. 150 °C

The official datasheet can be found here (download starting by click on link, offficial Osram website, pdf)

The Q8WP was developed for projection /stage applications, mostly for beamer and stage or entertainment lighting. I got this sample from a member of the german forum a longer time ago.

It features a glass window on top, against many other LEDs which are equipped with silicon domes or protective coatings. This makes the Q8WP very sensitive against damage in case of falls or mishandling. For that the LED can be easily cleaned and is also less sensistive to soldering flux.

Because of the not symmetric shape the use of center rings made with lathe is not recommended, also the die is not centered on the LED package which could make the precise centering of the reflector more difficult.

The Q8WP fits on a XP Noctigon MCPCB you can see in the previous pictures.

The die is connected with four bonding wires. At the bonding pad edges there are yellow-colored surfaces, which is known from many newer LEDs, especially that one in Flip Chip design. So I decided to look further to check if light is emitted sideways.

And indeed it is true, the sides are also lit up which can be seen in the both pictures above. From top view the side light isn't reflected by the substrate really well, but it is visible. But this means also that the real LES is a lot bigger than stated in the package outline shown in the datasheet in which the die is 1.55 x 1.23 mm in size, but only in top view.

In the official datasheet the ‚radiating surface‘ is given as 1.8 mm², which is true. But apparently this is valid for the one only visible from above, not for the light emitting sides.


25 °C Tsp, unless noted otherwise.

raw data here

The Q8WP (red graph) is not the most efficient LED. The XP-L HI is a lot more efficient and delivers higher light flux. For this the Vf is fairly low, at the current maximum of 9,600 mA it reaches only 3.53 Volts, at 1607.4 Lumens (33.9 Watts).

It delivers lot more light than Black Flat and dedomed XP-G2 S4 2B (old version, the good one).

At binning conditions (25 °C Tsp, 1,400 mA) it delivers 451.2 Lumens, so it equals to the Osram flux bin PA (450 to 560 lm).

The higher the Tj and therefore Tsp, the higher is the loss in light flux. From 25 to 125 °C, the difference is at least -18 %.

The Vf decreases in this range by about 130 mV.


I established a new method for determining luminance, especially to ensure more realistic values for 'real-life' conditions (flashlight use). The measurements are taken with a new original Convoy C8 reflector, but with same heatsink / setup as previously used in flux measurement.

Reflectivity for reflector 85 %

Transmission UCL glass 97,8 %

Values at 25 °C Tsp, for 85 °C Tsp values are 6 to 14 percent lower, depending on LED

LEDs marked with Warning sign uses old values are still taken with previous method (determining die size) due to problematic light distribution (donut holes etc).

The luminance of the Q8WP is extremely high. In general the Q8WP is the 'Black flat with higher light flux', I suppose the Q8WP is built on the same LED chip technology used in Black Flat and maybe SYNIOS. The side LES do not seem to influence the results. The unsymmetrical LES still remains.

The light color is a little bit blueish white, but without green or yellow parts in it. I estimate the CCT at approx. 7500 to 8000 K, the color rendition is low which is normal for cool white light sources in this color (rendering) class.


Again I had to revoke my conclusion for this LED. After the new luminance measurements it is clear that the Q8WP is very powerful, even in luminance. It delivers a lot more flux, at same luminance.

The efficiency is not as high as of other emitters, but there might some higher binned samples available on the market.

I like the low Vf, which allows a good regulation with linear drivers. Also the beam pattern is not bad, there are no miscolored coronas if used in conjunction with SMO or OP reflectors. The rectangular area could be a problem for aspherical optics / lights, though.


  • Low Vf
  • High max current
  • Extremely high luminance, like Black Flat


  • Tint a bit blueish
  • Glass window on the which makes the LED more sensitive but also makes cleaning more easy
  • Bigger (total) LES than stated in official datasheet because of side emitting areas


  • Relatively inefficient
  • Unsymmetric shape makes centering more difficult

Thanks a lot for reading! :)

Greetings, Dominik (aka BLF member koef3)

Mistakes, suggestions or offers / sponsoring of LEDs which I should test next are best sent via PM.

Thanks for the testing. This looks like an interesting LED.

To calculate the luminance you just divide the measured intensity above the LED (in cd) by the light emitting surface area, correct? In your picture from above the lit LED it doesn’t look like the area around the die is significantly lit from the side-escaping light, though this could be hard to tell in the photo. So if you are not measuring the light coming from the side areas you don’t need to include those areas in your calculation.

In lots of the new flip chip design LEDs the phosphor to the side is excited and emits significant light upwards which increases the LES for the purposes of your calculation, but that doesn’t seem to be happening with this LED. So maybe the luminance is actually higher.

Examining the LES is harder than on other LEDs.
I wasn’t sure if I should add the side areas to the total LES. And yeah, I divide the intensity with the total LES (in mm²).

For almost all Flip Chip based emitters the LES is much bigger than of the actual die itself. If this Q8WP is used in reflectors, there would be also the light coming from the side used? I understand that the reflector ‘catches’ not only the light from above (head on) and also this side areas affects the luminance. I am not sure about this but I can’t know everything about this topic. :wink:

Edit: I guess that the radiating surface is stated in the datasheet because this LED is used in projection systems. Maybe only the light from top surface is needed for this purpose and the side light emissions doesn’t matter. Maybe the glass window is also there to ensure precisely delimited edges for such applications.

A reflector might catch the side light, but for your luminance calculation that might not be relevant. If that side light doesn’t light up a surface for your luxmeter to see during your measurement, you shouldn’t include that area in the calculation. For some LEDs the LES is not a simple plane and the luminance is not a constant over the surface which makes your method of calculating the luminance prone to errors.

Really the most direct measure of the thing we care about is the measured intensity in a flashlight with known reflector area. The flashlight beam center is sampling the luminance of just the center of the die and so all of the complications associated with the other LESs to the side of the die are avoided.

Okay, I should have listen to my feeling. (Which says that the side LES doesn’t matter in this case.)
Maybe I will correct the luminance chart. Just take some further measurements to be on the safe side. But the yellow-colored area on the bonding pad belongs to the LES, because it emits significant amounts of light, right?

Thanks for the test! I believe I was the one who found this LED (by clicking through all of Osrams datasheets at the time). The sample here was imported from China (Alibaba), maybe they get rejected LEDs?

Regarding the luminance: measuring Candela with a reflector might help to get ballpark figures.

It was always important for me to measure LEDs without optics, especially these of flashlights.There are also many errors that could occur and I know that some LED manufacturers doesn't use secondary optics for luminance measurements.

I take some measurements with a Revtronic (Nitefighter) F30B, with SMO reflector, same focus, same lens, same centering:

Q8WP: 744 lm @ 57600 cd (77 cd per lm)

XP-G2 R5 dd: 420 lm @ 43100 cd (103 cd per lm)

XP-L HI V3: 802 lm @ 45200 cd (56 cd per lm)

You can see that the Q8WP is not the best choice for throw, like shown in the first luminance chart. I'm not sure to change my results, because the direct comparison in same optics apparently confirms my luminance values...

By the way - nice beamshot of Q8WP. :D

The eye of the sky...

Great, thanks! The old XP-G2 is much better.

The oval Corona was to be expected.

Yeah, and also the Q8WP has a lower luminance. It is not bad and for most thrower it will be enough.
I will not change the results until there are really good reasons to do so. If the luminance results would be much higher (like in case of smaller calculated LES) the shown ‘real life’ results (in flashlight) wouldn’t fit together.

What was the current during these measurements? It looks like the Q8WP has the highest throw?
Edit: I guess the Q8WP should have close to double the throw of the XPL HI if the LES was near 2mm^2.

The draw current was in all three measurements 4.00 Amperes (lab power supply and measured with clamp meter). In general I look closely to get the absolutely same environment and measurement conditions also in case of repetitions. The 105c limits the max current.
The XP-G2 seems to be underdriven, but originally (in domed state) it was only bin R5. Without dome the flux is another measured 27 percent less so it equals more to Cree flux bin Q5 or R2. Against it the Q8WP is a lot more efficient!

Thanks for the test.
EasyB is correct that if your luxmeter is right in front of the LED, the area that should be used for the calculation is the front area, which would be 1.55*1.23 = 1.9065mm^2

You can see in the real world test in your flashlight that the Q8WP still gets higher cd than the G2.
However, since it also emits light to the sides (which then get collected by the reflector but don’t add to the intensity) you end up with a worse cd/lumen ratio.
If all you care about is throw, cd/lumen doesn’t really matter as long as it still has higher cd.
Taking the front area of 1.9065mm^2, we can do 153.9/1.9065*3.23 = 260.73cd/mm^2 which is just slightly below the Black Flat but still better than any cree LED.

You can see this reflected in your real world test, the Q8WP gets 1.336 times higher intensity than the XP-G2.
At 260.7cd/mm^2 it is 1.321 times higher than the 197.3cd/mm^2 of the XP-G2 (as seen in your table).

I’m not exactly sure why your XP-L HI did better than the G2 in your real world test but worse in the cd/mm^2 table, probably because you used two different XP-G2s, but it is still the case that the Q8WP outperforms the XP-L HI in both the real world test and in the cd/mm^2 measurements, suggesting that the 260cd/mm^2 is correct.

Yeah, but only because of the bigger (side) LES and higher efficiency than the old shabby R5 XP-G2.
If I had used a dedomed XP-G2 S4 2B the intensity would be much higher than of the Q8WP, at approx 73 kcd or so - despite smaller LES because the XP-G2 doesn’t emit any light directly to the side.

Gradually I have the impression that the (also my) method to get the luminance is wrong. As far I understood the luminance depends on solid angle (Steradian), the light flux in this angle and the LES. If the brightness is measured and divided by the LES, we get such result that not depends on the angle of emitted light.

The next days I test another XP-G2, maybe in S2 bin. The dedomed XP-G2 S4 2B of my chart is broken and I am not sure about the by Hank newly discovered old XP-G2 S4 that they are as good as the old ones…

You’re right, the side light does seem to affect the emission profile a lot, especially at about 45 degrees, since it is very different from the no-dome black flat or synios LEDs.

I am still unsure of how putting this LED in the flashlight affected your lux readings though, since the reflector on a flashlight collects the side emission more than the forward emission of an LED.

I would suggest using a lens instead, to just take the front view and emission of the LED, maybe that will give more accurate comparison between LEDs.
It will also be easier to get a consistent lux reading since using a lens will give a nice uniform die projection rather than a spot with a corona.

Thanks for testing this led, it was on my list but you keep relieving my BLF-workload :smiley:

Some thoughts.The light emitted to the side does hit the reflector but predominantly the deepest part that gives a much worse hotspot projection than near the edge of the reflector, and does not add much to the cd of the hotspot.
It is a matter of choice if you use a reflector or lens for testing throw, I assume the optimal emission profile for both is slightly different.

Imho this is nicest looking emitter I ever seen. To bad about performance :frowning:
Thank you very much koef3.

Important notice:

I updated the luminance values for all LEDs shown in test chart including Q8WP. After several 'real-life condition' measurements in flashlight optics (SMO reflector) this LED has a very high luminance at much higher light flux than Black Flat which makes it very interesting for ultra-long-range flashlights with more output!

To ensure more realistic values for 'real-life' conditions (flashlight use / secondary optics) I established a new testing method to determine the luminance with a SMO reflector. All upcoming and already existing tests will use this method!

Updating of my other tests will be done in the next hours / days.

At this point big thanks to everyone who helped me out to establish this new test method and answered my questions about this! :)

Interesting, so my assumptions when I found this LED were correct after all. This is a special LED. Thanks Koef3 ;).

Thanks Koef3,

The Driver and Enderman will have to change their emitters now :))

So we have excellent results here but with higher current draw… No I don’t wan’t it it in single cell light…

But if there would be good and reliable buck driver with 9.6A draw I would think about adding additional 18650 tube to my lets say special edition lights.

20% lux performance with larger die than old g2s42b is not a joke… And it is best looking emitter ever imho.

Well, my big thrower was designed for this LED, but it’s just too much work. I don’t want to wait for the light again, currently. Maybe, when higher Bins become available. I think Enderman told me about a slightly updated version of this LED a while ago which doesn’t have the lowest bin anymore.

There are good buck drivers for 9.6A, but they have a large diameter.

Because of the low Vf you could just use a LD4 linear driver. It’s compact, works very good and not too expensive!