Texas_Ace integrating PVC sphere with no math involved

I have really enjoyed the reading, thanks for sharing your knowledge.

Just wanted to say thank you TA for creating this thread. I've been playing around with improving my light tube using the ideas you published in the OP. Really has helped a lot. It has also got me trying other materials and designs. So far, nothing has done better than your model in the OP, but it's interesting to explore this aspect of the hobby. In the past, I've been frustrated and discouraged by how much my measurements could vary by just the slightest change in flashlight depth into the tube and such.

I call mine a PVC light box . I'm a little surprised the round spheres haven't taken off more so than the PVC light box designs. I think I read of some issues with the spheres, but haven't followed it closely. I haven't had time to look in adapting mine to how yours is set up here, but looks like it has definite benefits. For me, I'd probably want to build up a 2nd one, rather than losing my only means of lumens measurement while I build, tweak and test, then also I'd have something to compare with.

This is great work though TA!!

Thanks, glad others are finding it useful. It is by no means a perfect sphere but it does work far better then my prior designs without the “integrating layer”.

I do think a thinner layer would be better though. Maybe something like DC fix. 2 separate layers would be even better I think.

The other major change I would make if I did it again would be to make the inside of the PVC a matte finish. Either by sanding it (quick and simple) or with a high quality flat paint. The flatter the finish the better.

Yeah, it is more of a light box really but no one searches for light box. :wink:

The issue with the round ones is mostly availability and cost. The styrofoam versions work great for low powered lights but once you get to ~1000 lumens or so they start to divagate due to so much light making it through the foam. Plus they are big and bulky which is the biggest reason I prefer the PVC setup, it is much easier to deal with on the desk.

Yeah, build a second one for sure so that you can use the first one to calibrate the second. Calibration is the hardest part in any sphere build. I am still not that confident in mine, it is simply the best I have been able to come up with using the flashlights and LED’s I have access to.

In my LED tests my readings are generally within the margin of error for the datasheet specs, so that would say I am not too far off.

Nice. The crazy angles matter though, not just "realistic" conditions. The crazy angles matter because the whole point of an integrating sphere is not how consistently it can give the same value for the same light, you don't need any kind of sphere or box for that, just a meter stick. The point is how well it can account for angular distributions of different lights.

The main point here is many bounces average out. I wonder about using a jar of some granular cloudy material. Beads? Salt? etc. I also though like EASYB's method, just measure brightness at several angles, and... integrate. Of course it's not as comfortable.

This pipe seems simple and effective though.

This is not true, loss of integration from light loss through the wall is with the average wall thickness of styrofoam balls (over here 2cm) negligable, and is completely independent of the light power of the measured source. The limit of my integrating spheres are solely determined by the range of the luxmeter that is used, which can be extended to wish by either a neutral density filter or the use of a satellite sphere. My most used sphere measures up to 15,000 lumen, if I need to measure higher, a small adaptation can extend that range.

Is the bad availability of suitable styrofoam balls the only reason for people to not build integrating spheres? (which despite the love of flashoholics for tube-systems are according to all literature of light measuring and measurements in practical use the way to go for good integrated light flux measuring). It is not that it is difficult, via my sigline people will find 4 IS build threads, of which one is simpler, cheaper, faster to build, and better integrating than any tube-system that is in use by CPF and BLF members.

I don’t know what your sphere consists of but I know that the JoshK sphere that I have will deviate by 5% or 10% (forgot which, been a year since I messed with it) per 1000 lumens due to losses.

So basically I have to adjust the multiplier by adding an additional 5-10% per 1000 lumens as the lumens go up in order to get an accurate result.

I tested this by testing three S2+’s separately and then putting them all together and taking a reading and the readings did not match up like they should. I used a few other lights in the same way with the same results. I also compared it to my PVC sphere. The PCV did not deviate to a statistically significant extent where the foam ball did.

I ran these tests due to over time the readings I was getting on high output lights was just not adding up compared to the low output.

All of the data is posted in the joshK thread.

I was 100% in favor of the sphere when I started all of this, which is why I was so excited to get the josh sphere and give it a try. The move away from it was not easy and only happened because I got significantly better and more consistent results on the PVC sphere.

I have positively nothing to prove in this case, I would love a proper sphere but the cost to build one would be way too much to justify the result based on my experiences so far. It would also be a lot of work as it would need to be coated with something to keep the light from making it through the foam.

Which is a point, light makes it through the foam super easy, even moon can be seen just making it glow in a dark room. It is quite possible that denser foam would not have this issue.

Measuring light has little to do with light loss but all with keeping the loss constant, the fact that a small fraction (in the order of a few %) of the light goes through the wall of the styrofoam is of no concern for either integration or accuracy.

About Joshk’s spheres:
I stopped reading and posting in Joshk’s thread after he refused to take on any advice from me (two years of building spheres and numerous tests on the performance), and was being rude to Texlite who was giving very good input at the time. Then he went completely his own way with some clear misconceptions about measuring light and produced and sold a number of spheres without any proper test data on their performance.
He would not even know what kind of tests to perform, not even sure if he really understands what light integration is. Which is all perfectly alright but not if you are into integration sphere making and distributing.
Not having followed his thread any further I would not know why his spheres do not perform, it is certainly not the concept of the integrating sphere, which has been proven worldwide by every light lab.

Just measured two flashlights (that I know have a good constant output) in my integrating sphere, first separately, then together next to each other in the entrance hole. I did not correct for entrance hole effects on the multiplier (but that effect is limited in my large sphere)

First light: 71 lumen
Second light: 115 lumen
Together: 185 lumen

For the record I never doubted your sphere works or that the sphere design works better.

I doubt that cheap spheres using cheap styrofoam balls work better then a PVC setup like I have now.

The only thing with the josh sphere I could not isolate and test was the sphere itself, which is why I came to, and still hold, the conclusion that the foam sphere itself is the cause of the non-linear readings.

With a better and/or coated sphere I have no doubt it would be superior. The question is, at what cost?

I am very much a results minded guy. I am not overly worried about how the results are achieved, as long as they are the desired results.

Thing is that we try to achieve some accuracy (don’t we all?), but we are actually pretty satisfied with a ballpark figure. So if our numbers are not consistent or match up, we just call it a day. And why not?

Btw, in integrating sphere making (and measuring light in general) I found that achieving a good performance has little to do with costs, it has to do with having your theory in order, thinking a lot about how to avoid possible errors, and testing, testing, testing.

Solving an error can be as cheap as taking a piece of sandpaper and carefully sand the inside of a sphere to a nice matt finish (just an example).

I'm guessing most folks would prefer a proper sphere. Whenever I have read a build thread on one, I would get a bit intimidated. Not so much by the physical building, but by understanding what is going on in one. I've always trusted djozz and others that take measuring light seriously, but it's difficult for me to picture in my mind how a sphere can measure both extremes in beam profiles accurately. Couple that with having to purchase or construct a large rigid structure, the barrier to entry was just to high.

I did, though, already have adequately sized PVC parts on hand. With those, I tried my hand at it years ago, but didn't like the fiddly, inconsistent results. I pretty much wrote it off until this thread was created. Someday, I would like a proper sphere that is big enough to measure my big lights. It may have to do double duty as a table top stand or something to overcome the SAF though.

I would agree, it took dozens if not hundreds of hours worth of testing to get my sphere where it is now. It is now consistent enough that I will at least post the numbers although I still have no idea exactly how accurate it is, I just know it is in the ballpark.

I appreciate the heroic efforts to measure total light flux as reliably as possible. However
only total lumens and OTF lux are insufficient to characterize the beam.
I think it’d be better to measure “intensity” (in the photometric definition), as a function of angle instead of only along the flashlight axis. If we had a table of candela X angle we’d have a much more complete picture of the beam and could answer questions like how big/intense is the hotspot, how much spill or corona there is, etc. This is unusual because it’s too much for the general public but for this community it’d be very useful and it seems that measuring lux reliably is a lot easier than total emission.

In principle we need measurements in the whole plane but almost all flashlights have a beam with cylindrical symmetry so readings along a single line are enough. It can be along a circle or a perperdicular to axis.

If we know the distance to the flashlight we can also easily integrate and determine total lumens emitted in the measured angular range, and going far enough even total lumens. This way we wouldn’t even need an integrating sphere and if we have such measurements we can compare with the angular result.

These readings take some time, so it’s better to do them at low/mid levels to avoid output drop. Since the beam profile won’t change with intensity level, doing 1 or 2 readings at max is enough to scale the complete graph.

Didn't follow everything but I have to 100% agree with djozz that the performance of any sphere/box, whatever, has nothing to do with the brightness of light. Any two photons on the same path act the same and have the same chances of getting through regardless of how many other photons there are around it. If something is non-linear, it's the meter, not the box (or maybe light leaking in from other sources skewing the low-end results?).

While that is all well and good, 2 seconds to take a lumen reading > doing math.

Put simply we are messing with flashlights, the actual light output simply doesn’t matter. It is purely for comparison purposes on the internet and bragging. All that matters is that it lights up what you want it to light up.

There is no need for anything more then a basic light box / sphere setup in the flashlight modding world.

In an engineering world, sure more accurate numbers can be important but even then I would rather spend that time and money on simply overbuilding the setup instead of trying to plan everything out perfectly on paper.

To each their own. As long as the end result is good, that is all I care about.

I thought this same thing, which is why it took me so long to track down the issue.

I still can’t fully explain it but I know for 100% sure it is not the meter or any other aspect of the sphere except for the sphere itself.

I tried 3 separate meters on the sphere, they all read the same with a deviation as lumens increased.

I tried the same 3 meters on my PVC setup, they all read the same with no statistically significant deviation regardless of lumen output.

light leaking in is also not an issue as the sphere is contained inside a box plus the light in the room is nothing compared to a 5000+ lumen blast from a flashlight. Plus that would cause the readings to go up, not down.

The readings do deviate, the sphere itself is the only thing left that it could be as I have tested everything else in the PVC sphere. I still do not understand why or how it is happening, only that it is.

About the only thing I can see as remotely possible is if the light is heating the surface and this affects its reflectivity. I guess it goes in that category of mysteries along with many other things.

I agree that to see mod results made by the same person it’s enough but my post aims a more general situation. First, we’d have more info on the beam. Second, reliable measurements are important to compare different flashlights, measured by different people.