My favorite by far is XP-L HI. I think I’ve been fairly clear about that, but just in case… it’s the one I consider the best for this light. I would ideally only make the light with XP-L HI emitters, and offer it in 7D, 5D, 4D, 3D, and 1D tints. Or maybe 6D, 5A, 3D, 2A. Perhaps 7A, 5D, 4A, 2D. Something like that.
I’ve tried both LH351D and SST-20, but my experience with both so far has been “icky green tint”, even at high CRI. But maybe that would be better with the correct tint bin.
However, this project isn’t just for me. Or just for Fritz. We already have one. This project is for the people. And a lot of people really like LH351D or SST-20. So we’ll use those… or at least one of the two. If either one can be obtained in a rosy tint bin, I think it would turn out well.
I use XP-L 4000k 5D in everything. Always very nice tint and usually goes a little pink.
My favorite by far is XP-L HI. I think I’ve been fairly clear about that, but just in case… it’s the one I consider the best for this light. I would ideally only make the light with XP-L HI emitters, and offer it in 7D, 5D, 4D, 3D, and 1D tints. Or maybe 6D, 5A, 3D, 2A. Perhaps 7A, 5D, 4A, 2D. Something like that.
I’ve tried both LH351D and SST-20, but my experience with both so far has been “icky green tint”, even at high CRI. But maybe that would be better with the correct tint bin.
However, this project isn’t just for me. Or just for Fritz. We already have one. This project is for the people. And a lot of people really like LH351D or SST-20. So we’ll use those… or at least one of the two. If either one can be obtained in a rosy tint bin, I think it would turn out well.
cabfrank wrote:
My favorite, and many other’s as well, is also XPL-HI. I think some people want to just try another emitter, or keep the cost slightly lower. I’m sure I’d be very happy with just two XPL versions.
Yep, the XP-L HI is my choice too. I am usually good with 2D to 4A… with 3A – 3D being the sweet spot.
Your probably right too about some people wanting to try another emitter. In fact, i had been hearing so much about the SST-20; I ordered a S2+ to try the 4000K 95CRI version. But after reading THIS that may well have been a mistake unless I just happen to get a good tint. Oh well, it was on sale for $10.99.
Will probably try a LH351D sometime too, if I know for sure it is a good tint.
Choosing an emitter solely based on saving money is a mistake from the start in my book.
For me, and this applies to most anything in life; sometimes the latest is not the greatest.
There is always a lot to be said for the ‘tried and true’.
—
You never know how a horse will pull until you hook him up to a heavy load./"Bear" Bryant
.................................. "Slow is Smooth, Smooth is Fast" ...................................
teacher, that $10.99 is less than the cost of an S2+ host plus it’s got a biscotti driver. I say you came out way ahead. Just swap in your favorite led. I have one from Hank and ordered 4 more from banggood for giveaways.
Yeah, great deal, and a great way to try the SST-20. Our standards have changed quickly (meaning we are getting spoiled by more great lights at great prices), but this should be a very solid and useful flashlight.
Your exactly right shirnask… “cheaper than the host”. . That was the deciding factor on getting it, I figured it was a win – win on my end; whether I liked the SST-20 or not.
And, as cabfrank noted; it was a cheap way to try the SST-20 & get a decent light (Convoy S2+) also.
Just a win-win all the way around….
—
You never know how a horse will pull until you hook him up to a heavy load./"Bear" Bryant
.................................. "Slow is Smooth, Smooth is Fast" ...................................
Wow, the gray anodization has a bit more blue in it than I realized.
It's only slightly blue-tinted. If I take an average color value from each of three people's pictures, and normalize them, here's what I found: * R/G/B = 77.7% / 88.3% / 100.0% Bluzie * R/G/B = 89.0% / 93.8% / 100.0% TK * R/G/B = 94.0% / 97.0% / 100.0% Maukka Maukka's pictures are color-calibrated, so I'd trust those values the most. If I understand correctly, the others used auto white balance so their colors may be skewed.
Does R/G/B = 94.0% / 97.0% / 100.0% (for Manuka) equate to R=240, G=247, B=255? I'm trying to duplicate the FW3A color, out of curiosity, in HTML or Photoshop. Thanks!
It’s only slightly blue-tinted. If I take an average color value from each of three people’s pictures, and normalize them, here’s what I found:
R/G/B = 77.7% / 88.3% / 100.0% Bluzie
R/G/B = 89.0% / 93.8% / 100.0% TK
R/G/B = 94.0% / 97.0% / 100.0% Maukka
Maukka’s pictures are color-calibrated, so I’d trust those values the most. If I understand correctly, the others used auto white balance so their colors may be skewed.
Since ActiveAl brought it up, I too would be interested to know what this means, how you got “an average color value”, & how you arrived at those percentages.
It’s like Greek to me & I am very curious too.
Thanks in advance….
—
You never know how a horse will pull until you hook him up to a heavy load./"Bear" Bryant
.................................. "Slow is Smooth, Smooth is Fast" ...................................
how you got “an average color value”, & how you arrived at those percentages.
It’s fairly straightforward.
Blur the image. Something like a Gaussian blur filter, with an appropriately-wide sample size for the given image. Maybe 20 to 50 pixels across, depending on the image. This makes each pixel an average of itself and every neighbor within N pixels.
Use the eyedropper tool to read the R/G/B values from a region near the middle of the light.
Note the highest of the three values. Multiply all three numbers by 100, divided by the noted highest value.
Also, it’s important to make sure beforehand that the original image isn’t clipping (overexposed) anywhere near the sampled region. I mean, make sure none of the values are maxed out, 255, in the original image… or at least not anywhere within N pixels of the sampled point, where N is the size of the Gaussian filter. Clipping distorts the data.
how you got “an average color value”, & how you arrived at those percentages.
It’s fairly straightforward.
Blur the image. Something like a Gaussian blur filter, with an appropriately-wide sample size for the given image. Maybe 20 to 50 pixels across, depending on the image. This makes each pixel an average of itself and every neighbor within N pixels.
Use the eyedropper tool to read the R/G/B values from a region near the middle of the light.
Note the highest of the three values. Multiply all three numbers by 100, divided by the noted highest value.
Also, it’s important to make sure beforehand that the original image isn’t clipping (overexposed) anywhere near the sampled region. I mean, make sure none of the values are maxed out, 255, in the original image… or at least not anywhere within N pixels of the sampled point, where N is the size of the Gaussian filter. Clipping distorts the data.
Thank you TK for that clear explanation.
—
You never know how a horse will pull until you hook him up to a heavy load./"Bear" Bryant
.................................. "Slow is Smooth, Smooth is Fast" ...................................
The values I get in Photoshop depend on where I “pick” the color
That’s why it’s important to blur the image first, to get an average. After blurring, the values will still vary from pixel to pixel, but they pretty much only vary in brightness, not hue. The ratio of red to green to blue doesn’t change much.
I wonder if GIMP has a filter to normalize each pixel individually. This would make it much easier to see changes in color without having to pay attention to brightness.
Do let us know what you think when you get it, teacher. I think it is quite tempting, even at around fifteen bucks.
I certainly will my friend.
Also, I just checked; the $10.99 Coupon Code for the Convoy S2+ SST20 4000K 95CRI is still active. Get the details HERE …
It is, but a word of warning, i ordered one on the 20th Feb using that link and code and got an e-mail on 1st March saying that the order had been delayed due to manufacturing process. I asked Simon and he confirmed here that there were no such issues. Could be a fake reason for a genuine delay but who knows.
Do let us know what you think when you get it, teacher. I think it is quite tempting, even at around fifteen bucks.
I certainly will my friend.
Also, I just checked; the $10.99 Coupon Code for the Convoy S2+ SST20 4000K 95CRI is still active. Get the details HERE …
It is, but a word of warning, i ordered one on the 20th Feb using that link and code and got an e-mail on 1st March saying that the order had been delayed due to manufacturing process. I asked Simon and he confirmed here that there were no such issues. Could be a fake reason for a genuine delay but who knows.
Yeah, I ordered 18 February & I am still waiting. My order still says “Processing”. But I figure what the heck, it’s only $10.99 & a good deal. It’ll come sooner or later.
I’m thinking about ordering another to use as a host.
—
You never know how a horse will pull until you hook him up to a heavy load./"Bear" Bryant
.................................. "Slow is Smooth, Smooth is Fast" ...................................
The values I get in Photoshop depend on where I “pick” the color
That’s why it’s important to blur the image first, to get an average. After blurring, the values will still vary from pixel to pixel, but they pretty much only vary in brightness, not hue. The ratio of red to green to blue doesn’t change much.
I wonder if GIMP has a filter to normalize each pixel individually. This would make it much easier to see changes in color without having to pay attention to brightness.
GIMP does have a lot of useful filters, but I don’t know if that specific ability is available. I do know it has a “posterize” filter that will reduce the image to any number of “colors” chosen by the user. I suppose that could be used effectively to normalize the shape of the light into one hue and one brightness level, if you choose the right number of “colors”.
EDIT: Nope. I was wrong. I just tried it. It reduces to a user-selected number of “levels” not “colors” and the difference means it won’t work for this purpose.
—
The Cycle of Goodness: “No one prospers without rendering benefit to others”
- The YKK Philosophy
I wonder if GIMP has a filter to normalize each pixel individually.
It doesn’t, but it didn’t take long to write a script to repeat what I did to the entire image, so we can see how that looks. To make these, what I did is:
Gaussian blur, with the sample size set to about 1/3rd of the width of the FW3A in each image.
Run the images through a quick per-pixel brightness normalizer script. This scales up the brightness of each pixel so at least one of the red, green, or blue channels will be exactly 255. Basically, it maxes out the brightness of each pixel as far as possible without changing the hue.
Mouse-over to compare before/after/after:
Initial pic: Original image.
Mouse-over to see it fully normalized.
Mouse-out to see it blurred and normalized. This is how I measured the “normalized average color” of each FW3A photo.
This visualization shows only the hue of each pixel, without regard for the original brightness. Using this, a perfectly grey surface would end up looking pure white. That’s why the older light grey FW3A prototype disappears almost completely… and why the FW3A in maukka’s image almost disappears. It’s virtually indistinguishable from the background color. Any deviation from pure grey shows up as an off-white color.
Please please please, yes I would like 1. Dark grey XPL-HI.
“You need the dark in order to show the light.” Bob Ross
I’m changing my mind on emitter, I’ll want to go with an XP-L instead of the LH351D. Edited my original post.
It doesn’t matter. The list is just for how many total lights you want.
Texas Ace Lumen Tube and JoshK Sphere calibrated with Maukka lights
Click this to go to signature links. I'm still around, just not reading many new threads.
I’m in.
Dark grey LH351D 4000K 90CRI
Thanks
I’m already down for 2 so maybe I should take a 3rd for my father in law
Forgot to say- Also XPL for the 3rd
I’m in for for one xpl-hi
I use XP-L 4000k 5D in everything. Always very nice tint and usually goes a little pink.
Your probably right too about some people wanting to try another emitter. In fact, i had been hearing so much about the SST-20; I ordered a S2+ to try the 4000K 95CRI version. But after reading THIS that may well have been a mistake unless I just happen to get a good tint. Oh well, it was on sale for $10.99.
Will probably try a LH351D sometime too, if I know for sure it is a good tint.
Choosing an emitter solely based on saving money is a mistake from the start in my book.
For me, and this applies to most anything in life; sometimes the latest is not the greatest.
There is always a lot to be said for the ‘tried and true’.
You never know how a horse will pull until you hook him up to a heavy load./"Bear" Bryant
.................................. "Slow is Smooth, Smooth is Fast" ...................................
Texas Lumens Flashlights / M4D M4X Deals : sign up - save $$$$
Rudeness Level _ mΩ _ {width:70%} _ LightWiki _ LED Tint Chart
Xlamp size chart _ BatteryU _ Flashaholic? Need Professional Help??? TheOriginal _ TAB _ LightSearch _ BatterySearch _ 14500's _ DiCal
teacher, that $10.99 is less than the cost of an S2+ host plus it’s got a biscotti driver. I say you came out way ahead. Just swap in your favorite led. I have one from Hank and ordered 4 more from banggood for giveaways.
Yeah, great deal, and a great way to try the SST-20. Our standards have changed quickly (meaning we are getting spoiled by more great lights at great prices), but this should be a very solid and useful flashlight.
Your exactly right shirnask… “cheaper than the host”.
. That was the deciding factor on getting it, I figured it was a win – win on my end; whether I liked the SST-20 or not. 
And, as cabfrank noted; it was a cheap way to try the SST-20 & get a decent light (Convoy S2+) also.
Just a win-win all the way around….
You never know how a horse will pull until you hook him up to a heavy load./"Bear" Bryant
.................................. "Slow is Smooth, Smooth is Fast" ...................................
Texas Lumens Flashlights / M4D M4X Deals : sign up - save $$$$
Rudeness Level _ mΩ _ {width:70%} _ LightWiki _ LED Tint Chart
Xlamp size chart _ BatteryU _ Flashaholic? Need Professional Help??? TheOriginal _ TAB _ LightSearch _ BatterySearch _ 14500's _ DiCal
Do let us know what you think when you get it, teacher. I think it is quite tempting, even at around fifteen bucks.
i would like to buy one, with XP-L HI please
Also, I just checked; the $10.99 Coupon Code for the Convoy S2+ SST20 4000K 95CRI is still active. Get the details HERE …
You never know how a horse will pull until you hook him up to a heavy load./"Bear" Bryant
.................................. "Slow is Smooth, Smooth is Fast" ...................................
Texas Lumens Flashlights / M4D M4X Deals : sign up - save $$$$
Rudeness Level _ mΩ _ {width:70%} _ LightWiki _ LED Tint Chart
Xlamp size chart _ BatteryU _ Flashaholic? Need Professional Help??? TheOriginal _ TAB _ LightSearch _ BatterySearch _ 14500's _ DiCal
I’ll take 1 too thanks!!!
I prefer dark gray. It looks nothing but perfect on the calibrated photos.
Does R/G/B = 94.0% / 97.0% / 100.0% (for Manuka) equate to R=240, G=247, B=255? I'm trying to duplicate the FW3A color, out of curiosity, in HTML or Photoshop. Thanks!
It’s like Greek to me & I am very curious too.
Thanks in advance….
You never know how a horse will pull until you hook him up to a heavy load./"Bear" Bryant
.................................. "Slow is Smooth, Smooth is Fast" ...................................
Texas Lumens Flashlights / M4D M4X Deals : sign up - save $$$$
Rudeness Level _ mΩ _ {width:70%} _ LightWiki _ LED Tint Chart
Xlamp size chart _ BatteryU _ Flashaholic? Need Professional Help??? TheOriginal _ TAB _ LightSearch _ BatterySearch _ 14500's _ DiCal
You just open the photo in Photoshop, use the dropper tool to pick up the color from a pixel of the picture, and read the RGB values.
OH….. Thank you EnlightenMe!! You have enlightened me….
. 
You never know how a horse will pull until you hook him up to a heavy load./"Bear" Bryant
.................................. "Slow is Smooth, Smooth is Fast" ...................................
Texas Lumens Flashlights / M4D M4X Deals : sign up - save $$$$
Rudeness Level _ mΩ _ {width:70%} _ LightWiki _ LED Tint Chart
Xlamp size chart _ BatteryU _ Flashaholic? Need Professional Help??? TheOriginal _ TAB _ LightSearch _ BatterySearch _ 14500's _ DiCal
It’s fairly straightforward.
Also, it’s important to make sure beforehand that the original image isn’t clipping (overexposed) anywhere near the sampled region. I mean, make sure none of the values are maxed out, 255, in the original image… or at least not anywhere within N pixels of the sampled point, where N is the size of the Gaussian filter. Clipping distorts the data.
You never know how a horse will pull until you hook him up to a heavy load./"Bear" Bryant
.................................. "Slow is Smooth, Smooth is Fast" ...................................
Texas Lumens Flashlights / M4D M4X Deals : sign up - save $$$$
Rudeness Level _ mΩ _ {width:70%} _ LightWiki _ LED Tint Chart
Xlamp size chart _ BatteryU _ Flashaholic? Need Professional Help??? TheOriginal _ TAB _ LightSearch _ BatterySearch _ 14500's _ DiCal
Great, what values did you get?
The values I get in Photoshop depend on where I "pick" the color - and they vary widely between the light and dark areas of the FW3A photos.
That’s why it’s important to blur the image first, to get an average. After blurring, the values will still vary from pixel to pixel, but they pretty much only vary in brightness, not hue. The ratio of red to green to blue doesn’t change much.
I wonder if GIMP has a filter to normalize each pixel individually. This would make it much easier to see changes in color without having to pay attention to brightness.
It is, but a word of warning, i ordered one on the 20th Feb using that link and code and got an e-mail on 1st March saying that the order had been delayed due to manufacturing process. I asked Simon and he confirmed here that there were no such issues. Could be a fake reason for a genuine delay but who knows.
I think Banggood threw Simon under the bus.
They’re basically saying: we did not order enough units.
My very own high current Beryllium Copper springs Gen 3:
http://budgetlightforum.com/node/67401
Liitokala Aliexpress Stores Battery Fraud: http://budgetlightforum.com/node/60547
I’m thinking about ordering another to use as a host.
You never know how a horse will pull until you hook him up to a heavy load./"Bear" Bryant
.................................. "Slow is Smooth, Smooth is Fast" ...................................
Texas Lumens Flashlights / M4D M4X Deals : sign up - save $$$$
Rudeness Level _ mΩ _ {width:70%} _ LightWiki _ LED Tint Chart
Xlamp size chart _ BatteryU _ Flashaholic? Need Professional Help??? TheOriginal _ TAB _ LightSearch _ BatterySearch _ 14500's _ DiCal
GIMP does have a lot of useful filters, but I don’t know if that specific ability is available. I do know it has a “posterize” filter that will reduce the image to any number of “colors” chosen by the user. I suppose that could be used effectively to normalize the shape of the light into one hue and one brightness level, if you choose the right number of “colors”.
EDIT: Nope. I was wrong. I just tried it. It reduces to a user-selected number of “levels” not “colors” and the difference means it won’t work for this purpose.
The Cycle of Goodness: “No one prospers without rendering benefit to others”
- The YKK Philosophy
It doesn’t, but it didn’t take long to write a script to repeat what I did to the entire image, so we can see how that looks. To make these, what I did is:
Mouse-over to compare before/after/after:
Bluzie: orig, norm, blur+norm
TK: orig, norm, blur+norm
Maukka: orig, norm, blur+norm
This visualization shows only the hue of each pixel, without regard for the original brightness. Using this, a perfectly grey surface would end up looking pure white. That’s why the older light grey FW3A prototype disappears almost completely… and why the FW3A in maukka’s image almost disappears. It’s virtually indistinguishable from the background color. Any deviation from pure grey shows up as an off-white color.
Is it too late to still get one of these?
Pages