CRI / tint
- Thread starter Brasso
- Start date
Help Support Candle Power Flashlight Forum
kaichu dento
Flashaholic
Re: HDS Systems EDC # 15
No ones preferences are important enough to negate the validity of other peoples preferences.
This is one of the most stupid statements made in this thread over this issue. You're assuming that what you like is progress, and what you don't like is a step backwards. Cold tints have always been available from LED's and now that warm and neutral tint fans are getting something to get excited about you want to consider it regression.Same here.In retrospect, how silly it was of me to hope that the new high-CRI emitter would be a 5000K 92 CRI Nichia or something equally compelling.
I would have bought such a Clicky without hesitation, but since it seems that they've actually regressed instead...oh well.... :shrug:
No ones preferences are important enough to negate the validity of other peoples preferences.
Re: HDS Systems EDC # 15
I'd like to humbly thank you for taking the high road and going easy on me. :bow: I realize that what you must be thinking is far, far worse, but in the interest of respectful discourse you've shown remarkable, nay, extraordinary restraint, and this, good people, is a shining example of why lovecpf:kiss:.
No, I just gave my own point of view on the subject, much like everybody else does around here. To me the change HDS made is a step backwards, away from natural lighting and toward artificial lighting similar to what we've had for more than a century, although I understand that others have different points of view. You assumed that I consider my way the only "right" way, but you're wrong--not stupid, mind you, just plain wrong in this instance.
That's true enough, but not cool white LEDs that render colors with a high degree of accuracy. Such LEDs exist, I believe (I'd have to use them myself to confirm), but I've yet to see one used in a flashlight.
As regrettably uncommon as these tints still are, given how much people like them, they've been around and there are more high-CRI versions on the way from several manufacturers. It's nothing we haven't seen before with incandescent lighting (which incidentally has a higher CRI), and I was hoping for something new, which I would view as progress.
This may seem a bit nitpicky, but it's not that I WANT to consider it a regression, it's just that in some ways I DO consider it a regression, and I have every right to hold and express my point of view.
Did I ever say otherwise? :thinking: That's very imaginative--I'm impressed!
I'd like to humbly thank you for taking the high road and going easy on me. :bow: I realize that what you must be thinking is far, far worse, but in the interest of respectful discourse you've shown remarkable, nay, extraordinary restraint, and this, good people, is a shining example of why lovecpf:kiss:.
No, I just gave my own point of view on the subject, much like everybody else does around here. To me the change HDS made is a step backwards, away from natural lighting and toward artificial lighting similar to what we've had for more than a century, although I understand that others have different points of view. You assumed that I consider my way the only "right" way, but you're wrong--not stupid, mind you
, just plain wrong in this instance.That's true enough, but not cool white LEDs that render colors with a high degree of accuracy. Such LEDs exist, I believe (I'd have to use them myself to confirm), but I've yet to see one used in a flashlight.
As regrettably uncommon as these tints still are, given how much people like them, they've been around and there are more high-CRI versions on the way from several manufacturers. It's nothing we haven't seen before with incandescent lighting (which incidentally has a higher CRI), and I was hoping for something new, which I would view as progress.
This may seem a bit nitpicky, but it's not that I WANT to consider it a regression, it's just that in some ways I DO consider it a regression, and I have every right to hold and express my point of view.
Did I ever say otherwise? :thinking: That's very imaginative--I'm impressed!
Last edited:
the.Mtn.Man
Flashlight Enthusiast
- Joined
- Oct 3, 2008
- Messages
- 2,516
Re: HDS Systems EDC # 15
Were people really expecting the high CRI emitter to not have a warm tint?
Were people really expecting the high CRI emitter to not have a warm tint?
Re: HDS Systems EDC # 15
I was thinking the same thing. This disappointment and claiming the XPG HighCRI is a regression of some sort makes no sense. The HighCRI XPG is one of the most efficient High CRI emitters that is available currently. HighCRI lights tend to be warm, this is normal, not the exception. The Nichia emitter, which is more neutral supposedly, hasn't been proven reliable and we know Henry doesn't use an emitter unless it has a proven track record. I tried to find information about the HighCRI Nichia emitter but didn't have much luck (specifically with CRI rating and efficiency). I doubt high CRI alternatives to the XPG can match the XPG availability, efficiency, etc.
JWRitchie76
Flashlight Enthusiast
Re: HDS Systems EDC # 15
All high CRI's are significantly warmer than that. Your specs fall into the neutral category and we ain't getting those.
All high CRI's are significantly warmer than that. Your specs fall into the neutral category and we ain't getting those.
calipsoii
Flashlight Enthusiast
- Joined
- Apr 21, 2010
- Messages
- 1,412
Re: HDS Systems EDC # 15
Not all, most. The aforementioned Nichia 183A's fall firmly in the cool end of neutral. That said, I can't say I'm shocked by it being an XP-G either. I assumed that was a given the moment Henry announced the cool whites would be using XP-G's. All the other emitters other than the Nichia 119 have such a different footprint and dome design that he'd need a different reflector to accommodate them.
Not all, most. The aforementioned Nichia 183A's fall firmly in the cool end of neutral. That said, I can't say I'm shocked by it being an XP-G either. I assumed that was a given the moment Henry announced the cool whites would be using XP-G's. All the other emitters other than the Nichia 119 have such a different footprint and dome design that he'd need a different reflector to accommodate them.
Re: HDS Systems EDC # 15
Since the emitter is a Cree, then no, because they're not much into high CRI, and it's easier to achieve that spec with warm white LEDs. I was hoping for a different emitter from a different manufacturer. In the big picture, I didn't expect the flashlight to have a cooler tint, since that is not the (niche) market trend I've been seeing with high-CRI flashlights, but I had hoped that HDS would go against the trend and give us a new alternative to the other warm high-CRI flashlights that are going to be released in the near future.
Previously I had shied away from the SSC P4-based high-CRI HDS flashlights because I wasn't satisfied with certain aspects of the emitter. I did, however, take notice of its neutral tint, as opposed to warm, which is what got my hopes up for an improved neutral white emitter or even a cool white high-CRI emitter. But it didn't happen, boo-hoo, so I still can't justify (speaking only for myself, which should be implicit for what everybody says) paying that much for a flashlight that in my eyes can't match even my ZebraLight H51c in terms of color rendering. I really like the HDS EDC design and interface, and I wish that the new version would have met my requirements regarding beam tint. No big deal, though, as we all can't have everything we want in life--it's just a statement that I had hoped for something different, and I'm not the only one. It's just for general knowledge, but maybe somebody will listen and bring something new to the market.
That would have been fine, but I was hoping for 5000K-6000K, since such high-CRI LEDs actually exist, even today.
It's a regression from the point of view of those who wish for flashlights (as well as other light sources) to emulate daylight as closely as possible. Those who prefer alternatives to the light that we primarily evolved under are perfectly within their right to do so, of course (if somebody prefers deep purple light, that's OK too), but the preference for daylight tint and color rendering does make some sense (to me, everything else makes colors look wrong).
Yes, it's normal for LEDs, and it's going to become more common in the near future, but there is nothing wrong with being exceptional, either (as long as there is a market for it, anyway).
Admittedly, it's hard to find data for some of the LEDs I have in mind, but here is one that I might be interested in:
http://www.nichia.co.jp/specification/en/product/led/NS6W183A-H1-E.pdf
It is multi-die, which is not ideal for all purposes, but other flashlights have used such LEDs successfully. Its efficiency at 350 mA is not bad (slightly above that of the high-CRI warm XP-G), although its maximum drive current is much lower (should be sufficient for the HDS EDC).
Here is another that looks interesting (specifically the LXW8-PW50):
http://www.philipslumileds.com/pdfs/DS61.pdf
This one has a single die, and while 85 CRI may not seem very high to some, due to the peculiarities of white phosphor-based LEDs (which I can explain if anybody cares), the actual accuracy of color rendering is understated by the CRIs of higher-CCT LEDs. In other words, an LED can have more accurate color rendering (relative to sunlight) and a more balanced spectrum than another even when its CRI is lower.
Not all, as my example above (Nichia NS6W183A-H1, 92 CRI @ 5000K CCT) shows, and CRI is rather flawed as an indicator of color rendering accuracy in any case.
Since the emitter is a Cree, then no, because they're not much into high CRI, and it's easier to achieve that spec with warm white LEDs. I was hoping for a different emitter from a different manufacturer. In the big picture, I didn't expect the flashlight to have a cooler tint, since that is not the (niche) market trend I've been seeing with high-CRI flashlights, but I had hoped that HDS would go against the trend and give us a new alternative to the other warm high-CRI flashlights that are going to be released in the near future.
Previously I had shied away from the SSC P4-based high-CRI HDS flashlights because I wasn't satisfied with certain aspects of the emitter. I did, however, take notice of its neutral tint, as opposed to warm, which is what got my hopes up for an improved neutral white emitter or even a cool white high-CRI emitter. But it didn't happen, boo-hoo, so I still can't justify (speaking only for myself, which should be implicit for what everybody says) paying that much for a flashlight that in my eyes can't match even my ZebraLight H51c in terms of color rendering. I really like the HDS EDC design and interface, and I wish that the new version would have met my requirements regarding beam tint. No big deal, though, as we all can't have everything we want in life--it's just a statement that I had hoped for something different, and I'm not the only one. It's just for general knowledge, but maybe somebody will listen and bring something new to the market.
That would have been fine, but I was hoping for 5000K-6000K, since such high-CRI LEDs actually exist, even today.
It's a regression from the point of view of those who wish for flashlights (as well as other light sources) to emulate daylight as closely as possible. Those who prefer alternatives to the light that we primarily evolved under are perfectly within their right to do so, of course (if somebody prefers deep purple light, that's OK too), but the preference for daylight tint and color rendering does make some sense (to me, everything else makes colors look wrong).
Yes, it's normal for LEDs, and it's going to become more common in the near future, but there is nothing wrong with being exceptional, either (as long as there is a market for it, anyway).
Admittedly, it's hard to find data for some of the LEDs I have in mind, but here is one that I might be interested in:
http://www.nichia.co.jp/specification/en/product/led/NS6W183A-H1-E.pdf
It is multi-die, which is not ideal for all purposes, but other flashlights have used such LEDs successfully. Its efficiency at 350 mA is not bad (slightly above that of the high-CRI warm XP-G), although its maximum drive current is much lower (should be sufficient for the HDS EDC).
Here is another that looks interesting (specifically the LXW8-PW50):
http://www.philipslumileds.com/pdfs/DS61.pdf
This one has a single die, and while 85 CRI may not seem very high to some, due to the peculiarities of white phosphor-based LEDs (which I can explain if anybody cares), the actual accuracy of color rendering is understated by the CRIs of higher-CCT LEDs. In other words, an LED can have more accurate color rendering (relative to sunlight) and a more balanced spectrum than another even when its CRI is lower.
Not all, as my example above (Nichia NS6W183A-H1, 92 CRI @ 5000K CCT) shows, and CRI is rather flawed as an indicator of color rendering accuracy in any case.
Re: HDS Systems EDC # 15
I am interested. It would help explain why I find all my Zebralight 'c' models to be so satisfying to use. Thanks!
I am interested. It would help explain why I find all my Zebralight 'c' models to be so satisfying to use. Thanks!
Re: HDS Systems EDC # 15
No surprise that the CRI emitter is an XP-G, it was never likely to be anything else.
I'm not personally a big fan of the CRI XP-G, it's too warm and makes everything look a muddy brown but this is just my personal opinion. Many people will like it and it has one big plus in that you can drive it much harder than all the other High CRI emitters (1.5 amps) so the output is higher (mute point in the HDS lights as they aren't driven that hard). In lights that do (ArmyTek) it has a real advantage over the other High CRI LED's.
I'm with GaAslamp, it would be nice to see more lights using the Nichias and Luxeon Rebel High CRI emitters. I'm tempted to get a Zebralight just to try the Rebel emitter. These Rebels have great potential in 'mainstream' lights.
As for the Nichia 183A-H1, it looks even better. With a 5000K temperature and the potential to be driven at 800mA it should be an amazing LED. The only way you will ever see these in a light is from Don (McGizmo) I would guess. I can't wait to see if he has an offering using these, they would be perfect in a Sundrop head.
No surprise that the CRI emitter is an XP-G, it was never likely to be anything else.
I'm not personally a big fan of the CRI XP-G, it's too warm and makes everything look a muddy brown but this is just my personal opinion. Many people will like it and it has one big plus in that you can drive it much harder than all the other High CRI emitters (1.5 amps) so the output is higher (mute point in the HDS lights as they aren't driven that hard). In lights that do (ArmyTek) it has a real advantage over the other High CRI LED's.
I'm with GaAslamp, it would be nice to see more lights using the Nichias and Luxeon Rebel High CRI emitters. I'm tempted to get a Zebralight just to try the Rebel emitter. These Rebels have great potential in 'mainstream' lights.
As for the Nichia 183A-H1, it looks even better. With a 5000K temperature and the potential to be driven at 800mA it should be an amazing LED. The only way you will ever see these in a light is from Don (McGizmo) I would guess. I can't wait to see if he has an offering using these, they would be perfect in a Sundrop head.
Re: HDS Systems EDC # 15
They ARE satisfying to use, aren't they? :twothumbs I think that's because their specific Rebel emitter's color balance (despite its 4000K CCT) where it counts the most is more like that of sunlight at around 5800K than a closer match to an ideal 4000K source would be. The latter would have given this emitter a higher CRI than 85 because CRI is measured against an ideal source of the same CCT, but then it would be less like sunlight than it already is. Does that make sense? I know I keep repeating this mantra, but CRI doesn't mean all that much.
As for the "peculiarities" of LEDs in regard to measuring CRI that I offered to explain, it's simple yet not that easy to explain, although I'll try. Basically the spectra of phosphor-based white LEDs all have a large, narrow spike in the blue range, all have a significant deficiency in the cyan range (between blue and green), and most have a deficiency in the red range (unless they're warm white, in which case they probably have way too much red and orange relative to blue). Because CRI is measured against an ideal incandescent source of the same CCT, as mentioned above, the cyan deficiency that all LEDs have (some more than others, but it's always major)--being close to the blue end of the visible spectrum--has a more significant negative impact at higher CCTs than at lower CCTs. This means that just by adding more of the same phosphor blend to convert more of the underlying blue light (these LEDs are all actually blue with a phosphor blend on top, hence the narrow blue spike) to other wavelengths, for example, the CRI automatically goes up as the CCT goes down, all else being equal. The resulting LED's color rendering may be superior or inferior, but either way it will have a higher CRI because the cyan deficiency has by definition become less significant in the calculation. This is why most high-CRI LEDs tend to be warm white--it's relatively easy to achieve and doesn't necessarily mean a whole lot (such an LED could have superior color rendering, but it wouldn't be because of its CRI alone). This is a major flaw in using CRI to judge the color rendering accuracy of phosphor-based LEDs, especially for those who consider sunlight an ideal light source.
There are other issues that come into play that seriously call into question the usefulness of CRI even when comparing LEDs of the same CCT. I mentioned color balance above, and went into greater depth in the following posts, complete with spectral graphs and sample photos referenced from other posts of mine:
http://www.candlepowerforums.com/vb...-CRI-Release&p=3730719&viewfull=1#post3730719
http://www.candlepowerforums.com/vb...-CRI-Release&p=3730790&viewfull=1#post3730790
That's what I've been hearing for a long time, but I just kept wondering "Why does it have to be a Cree?" Not that I have anything against Cree, but they're mostly about maximum output rather than accurate color rendering, which gave me some doubt until now.
I haven't seen its spectrum, but if it's similar to that of other warm white high-CRI LEDs, then it probably just has a bit more blue to render cool colors a little better. At such a low CCT, everything is bound to look orangish or brownish in general, though, as there is still way more red and orange than blue.
That's exactly what I did, and I was surprised by how accurately it renders colors (perhaps in part because I used to put blind faith in CRI, and 85 didn't seem all that high). It's not the sort of thing that would immediately hit most people in the face, so to speak, but when virtually every color you view with it looks practically as natural as you'd expect from viewing it under sunlight, the cumulative effect is kind of a revelation as LEDs go (or even incandescent flashlights--this one is more accurate to my eyes). While it has a yellowish cast, colors still look extremely well balanced otherwise. I can't guarantee that you'd like it as much, as we all see a bit differently, but I'm very impressed.
It certainly could be, I think. While I'm a bit concerned about its color balance from looking at its spectrum, its very high CRI notwithstanding, I'd take a chance on an HDS EDC equipped with this LED. With the SunDrop, I don't know--it's a bit out of my price range and also rather specialized. In addition, I'm pretty satisfied with my H51c right now, even if I'm always shooting for something that is potentially better.
They ARE satisfying to use, aren't they? :twothumbs I think that's because their specific Rebel emitter's color balance (despite its 4000K CCT) where it counts the most is more like that of sunlight at around 5800K than a closer match to an ideal 4000K source would be. The latter would have given this emitter a higher CRI than 85 because CRI is measured against an ideal source of the same CCT, but then it would be less like sunlight than it already is. Does that make sense? I know I keep repeating this mantra, but CRI doesn't mean all that much.
As for the "peculiarities" of LEDs in regard to measuring CRI that I offered to explain, it's simple yet not that easy to explain, although I'll try.
Basically the spectra of phosphor-based white LEDs all have a large, narrow spike in the blue range, all have a significant deficiency in the cyan range (between blue and green), and most have a deficiency in the red range (unless they're warm white, in which case they probably have way too much red and orange relative to blue). Because CRI is measured against an ideal incandescent source of the same CCT, as mentioned above, the cyan deficiency that all LEDs have (some more than others, but it's always major)--being close to the blue end of the visible spectrum--has a more significant negative impact at higher CCTs than at lower CCTs. This means that just by adding more of the same phosphor blend to convert more of the underlying blue light (these LEDs are all actually blue with a phosphor blend on top, hence the narrow blue spike) to other wavelengths, for example, the CRI automatically goes up as the CCT goes down, all else being equal. The resulting LED's color rendering may be superior or inferior, but either way it will have a higher CRI because the cyan deficiency has by definition become less significant in the calculation. This is why most high-CRI LEDs tend to be warm white--it's relatively easy to achieve and doesn't necessarily mean a whole lot (such an LED could have superior color rendering, but it wouldn't be because of its CRI alone). This is a major flaw in using CRI to judge the color rendering accuracy of phosphor-based LEDs, especially for those who consider sunlight an ideal light source.There are other issues that come into play that seriously call into question the usefulness of CRI even when comparing LEDs of the same CCT. I mentioned color balance above, and went into greater depth in the following posts, complete with spectral graphs and sample photos referenced from other posts of mine:
http://www.candlepowerforums.com/vb...-CRI-Release&p=3730719&viewfull=1#post3730719
http://www.candlepowerforums.com/vb...-CRI-Release&p=3730790&viewfull=1#post3730790
That's what I've been hearing for a long time, but I just kept wondering "Why does it have to be a Cree?" Not that I have anything against Cree, but they're mostly about maximum output rather than accurate color rendering, which gave me some doubt until now.
I haven't seen its spectrum, but if it's similar to that of other warm white high-CRI LEDs, then it probably just has a bit more blue to render cool colors a little better. At such a low CCT, everything is bound to look orangish or brownish in general, though, as there is still way more red and orange than blue.
That's exactly what I did, and I was surprised by how accurately it renders colors (perhaps in part because I used to put blind faith in CRI, and 85 didn't seem all that high). It's not the sort of thing that would immediately hit most people in the face, so to speak, but when virtually every color you view with it looks practically as natural as you'd expect from viewing it under sunlight, the cumulative effect is kind of a revelation as LEDs go (or even incandescent flashlights--this one is more accurate to my eyes). While it has a yellowish cast, colors still look extremely well balanced otherwise. I can't guarantee that you'd like it as much, as we all see a bit differently, but I'm very impressed.
It certainly could be, I think. While I'm a bit concerned about its color balance from looking at its spectrum, its very high CRI notwithstanding, I'd take a chance on an HDS EDC equipped with this LED. With the SunDrop, I don't know--it's a bit out of my price range and also rather specialized. In addition, I'm pretty satisfied with my H51c right now, even if I'm always shooting for something that is potentially better.
Last edited:
the.Mtn.Man
Flashlight Enthusiast
- Joined
- Oct 3, 2008
- Messages
- 2,516
Re: HDS Systems EDC # 15
P4 is what was used in previous HDS lights, correct? Because if my high CRI Clicky is any indication it's much closer to warm than neutral.
Well, if it's color rendering you value then you should know that the H51c at 83 CRI lags behind the HDS offerings at 90+ CRI. Reading some of the early reports of the H51c, many users were disappointed that it didn't make colors "pop" like they were expecting from an emitter marketed as "high CRI" with some reporting that it was little better than a standard neutral white (read bluish) emitter.
I appreciate the tone of your posts, but I think you're trying to make a subjective judgement sound like an objective one. A CRI of 90 will render colors more accurately than 80 at any color temperature. This is undeniable. What it sounds like is that you prefer cooler tints over warm and judge those colors to be more "accurate" even at a lower CRI and therefore technically less accurately rendered than a warm emitter with a higher CRI. That's fine, and your preference isn't wrong, but I think it's somewhat misleading to downplay the importance of a high CRI for accurate color rendering.
To put it another way, your "ideal source" is sunlight; my "ideal source" is incandescent (or late afternoon sun, what photographers call "the golden hour"). Thus we are each going to judge colors to be more "accurate" at our preferred color temperature even if the CRI is actually low relative to a high CRI source.
As far as I'm concerned, LED technology will have reached its zenith once we have emitters capable producing 3200K at a CRI of 100.
P4 is what was used in previous HDS lights, correct? Because if my high CRI Clicky is any indication it's much closer to warm than neutral.
Well, if it's color rendering you value then you should know that the H51c at 83 CRI lags behind the HDS offerings at 90+ CRI. Reading some of the early reports of the H51c, many users were disappointed that it didn't make colors "pop" like they were expecting from an emitter marketed as "high CRI" with some reporting that it was little better than a standard neutral white (read bluish) emitter.
I appreciate the tone of your posts, but I think you're trying to make a subjective judgement sound like an objective one. A CRI of 90 will render colors more accurately than 80 at any color temperature. This is undeniable. What it sounds like is that you prefer cooler tints over warm and judge those colors to be more "accurate" even at a lower CRI and therefore technically less accurately rendered than a warm emitter with a higher CRI. That's fine, and your preference isn't wrong, but I think it's somewhat misleading to downplay the importance of a high CRI for accurate color rendering.
To put it another way, your "ideal source" is sunlight; my "ideal source" is incandescent (or late afternoon sun, what photographers call "the golden hour"). Thus we are each going to judge colors to be more "accurate" at our preferred color temperature even if the CRI is actually low relative to a high CRI source.
As far as I'm concerned, LED technology will have reached its zenith once we have emitters capable producing 3200K at a CRI of 100.
Last edited:
Re: HDS Systems EDC # 15
+100! I prefer warmer tints than cool ones. Although what GaAslamp said makes sense but it is like what you have said, a matter of personal preference. Honestly, after reading all the lengthy replies from GaAslamp, I can't help feeling that we are being "force-fed" into accepting that a high CRI of 85 with a cooler CCT is the way to go. I am sorry GaAslamp but this is what I felt after reading your posts. We all have our preferences so let's just leave it as that.
I appreciate the tone of your posts, but I think you're trying to make a subjective judgement sound like an objective one. A CRI of 90 will render colors more accurately than 80 at any color temperature. This is undeniable. What it sounds like is that you prefer cooler tints over warm and judge those colors to be more "accurate" even at a lower CRI and therefore technically less accurately rendered than a warm emitter with a higher CRI. That's fine, and your preference isn't wrong, but I think it's somewhat misleading to downplay the importance of a high CRI for accurate color rendering.
To put it another way, your "ideal source" is sunlight; my "ideal source" is incandescent (or late afternoon sun, what photographers call "the golden hour"). Thus we are each going to judge colors to be more "accurate" at our preferred color temperature even if the CRI is actually low relative to a high CRI source.
As far as I'm concerned, LED technology will have reached its zenith once we have emitters capable producing 3200K at a CRI of 100.
+100! I prefer warmer tints than cool ones. Although what GaAslamp said makes sense but it is like what you have said, a matter of personal preference. Honestly, after reading all the lengthy replies from GaAslamp, I can't help feeling that we are being "force-fed" into accepting that a high CRI of 85 with a cooler CCT is the way to go. I am sorry GaAslamp but this is what I felt after reading your posts. We all have our preferences so let's just leave it as that.
kaichu dento
Flashaholic
Re: HDS Systems EDC # 15
Favorite light sources, now that we can be so picky, are really like preferences in cuisine, and we should accept that the guy that likes pizza is no more wrong than the one preferring sushi.
There is no superior light source, only a wider range of preferences, and to constantly try to rate LED's in comparison to mid-day sunlight is only helpful for those who prefer it over early/late sun, as many of us do.
Once again the straw man of daylight comes up when in nature there are many presentations of light - all the way from a clear dawn to a cloudy day, sunny noon to blazing orange sunset, not to mention light of the moon, or even on a moonless night, the aurora and other favorites.
Favorite light sources, now that we can be so picky, are really like preferences in cuisine, and we should accept that the guy that likes pizza is no more wrong than the one preferring sushi.
There is no superior light source, only a wider range of preferences, and to constantly try to rate LED's in comparison to mid-day sunlight is only helpful for those who prefer it over early/late sun, as many of us do.
Great post and it was pretty nice to read after having been waiting all night to write a response with some of the same points made. I guess we'll be in the same state of bliss when we get that 3200k high CRI 100. :thumbsup:To put it another way, your "ideal source" is sunlight; my "ideal source" is incandescent (or late afternoon sun, what photographers call "the golden hour"). Thus we are each going to judge colors to be more "accurate" at our preferred color temperature even if the CRI is actually low relative to a high CRI source.
As far as I'm concerned, LED technology will have reached its zenith once we have emitters capable producing 3200K at a CRI of 100.
bondr006
Flashlight Enthusiast
Re: HDS Systems EDC # 15
Although I don't like blueish purple cool tints, I don't like the other end of the spectrum either. I do not believe that overly yellow brown renders colors any better than blueish purple does. Being a retired Naval photographer I would not use either of those tints to render true color, but something in the middle. White balance is an amazing thing for true color rendering. I realize that tint is a subjective preference, but extremes either way do not truly render colors accurately.
Although I don't like blueish purple cool tints, I don't like the other end of the spectrum either. I do not believe that overly yellow brown renders colors any better than blueish purple does. Being a retired Naval photographer I would not use either of those tints to render true color, but something in the middle. White balance is an amazing thing for true color rendering. I realize that tint is a subjective preference, but extremes either way do not truly render colors accurately.
Last edited:
Re: HDS Systems EDC # 15
Correct, and more specifically it's either the S42180 (4000K CCT) or N42180 (3000K CCT). I've always been told that its CCT is 4000K, although some photos taken under its illumination appear orangish to the point where I can't be sure. The S42180 has a large peak in the orange range of the visible spectrum, just like the N42180 has, which means that orange is going to be emphasized either way, making it hard to tell its CCT due to the many unknown variables involved in examining photos taken by others.
It may be "warm," it may be on the warm side of "neutral" due to manufacturing variation, it may be "neutral" (~4000K) with a peak in the orange range (as its spectral power distribution graph indicates), or a combination these possibilities. It is for this reason that I chose not to purchase this version of the flashlight, as impressed as I was--at least at the time--with its 93 CRI. In fact, this was what first made me doubt the usefulness of CRI in evaluating color rendering accuracy, and everything I've learned about CRI and from my experiences with using flashlights with varied spectra since then has confirmed my doubts.
I'm not sure where you got 83, but it is rated by the manufacturer as having a minimum of 80 and a typical value of 85. A few months ago that would have meant something to me, but not anymore.
The purpose of color rendering accuracy is not to make colors "pop" (more than they should) but to render them accurately. To make them really "pop" requires exaggerating certain wavelengths, usually in the red range of the visible spectrum. I find that my H51c makes many colors "pop" a little bit more than the typical neutral white LED due to its higher red output, but at the same time it's not overly red to the point where it adds any more "pop" or "richness" than objects should have like most (not all) other high-CRI LEDs do (sort of like putting on BluBlockers).
Although I fully acknowledge that we all see a little differently (or maybe even a lot in some cases), I wouldn't be surprised if somebody expressed disappointment in the H51c/SC51c (same LED) even if their eyes and visual perception were exactly like mine. That's because they may be looking for something different, such as more "pop" or a specific tint rather than what I personally define as accuracy. I'm quite pleased that it gives me the accuracy--relative to sunlight--that I'm seeking, or something pretty darn close to it. I've always found artificial light sources lacking in this respect, whether they "pop" or not.
Although what I've said does include both subjective and objective elements, I assure you that there was no deliberate attempt to confuse the two. Critical readers should be able to sort them out, I think.
It is neither undeniable nor self-evident. I think that my arguments and examples are sufficient to introduce doubt with regard to placing blind faith (as I once did, effectively) in CRI as an indicator of color rendering accuracy. I tried to do it in a way that laymen (myself included) could understand, but there are people with much better credentials who are thinking the same thing, as evidenced in the following links:
http://en.wikipedia.org/wiki/Color_rendering_index#Criticism_and_resolution
http://www.knt.vein.hu/staff/schandaj/SJCV-Publ-2005/521.pdf
My personal preferences are no secret, and I don't hide the fact that they form the perspective from which I discuss these issues. This is perfectly natural since this particular discussion started when somebody accused those who have such a perspective of making no sense (which I suppose implies that those who have other perspectives are making sense :ironic. In response, I tried my best to make sense out of my (and others') perspective, without invalidating other perspectives (we all have different preferences and goals).
As for CRI, there are many good reasons to rightfully downplay its importance--particularly where phosphor-based LEDs are concerned. I've given a few of my own, as well as references to the harsh criticisms of professionals for everybody here to peruse. If anything that I or they said is wrong, then by all means refute it--I'd rather be corrected than mistaken while still believing that I'm right (and misleading others).
Generally speaking, CRI is crude and only useful for certain cases. Because of this and the fact that by definition it only measures accuracy against an ideal reference of the same CCT, it is virtually useless as an indicator of the color rendering accuracy of, say, 4000K LEDs against an ideal 5800K source such as sunlight. Does this make sense? I'm saying that if a specific 4000K LED renders colors very much like my ideal of 5800K sunlight does (with some flaws, of course), then its CRI--due to the very definition of CRI--will in all likelihood be lower than that of other 4000K LEDs that come closer to the theoretical ideal 4000K source. The problem is that I don't care about the ideal 4000K source while CRI--by definition--does, therefore the higher CRIs of these other LEDs means nothing to me. This would be true even if CRI were a PERFECT method of determining accuracy, but in reality it is far, FAR from that.
Correct, and more specifically it's either the S42180 (4000K CCT) or N42180 (3000K CCT). I've always been told that its CCT is 4000K, although some photos taken under its illumination appear orangish to the point where I can't be sure. The S42180 has a large peak in the orange range of the visible spectrum, just like the N42180 has, which means that orange is going to be emphasized either way, making it hard to tell its CCT due to the many unknown variables involved in examining photos taken by others.
It may be "warm," it may be on the warm side of "neutral" due to manufacturing variation, it may be "neutral" (~4000K) with a peak in the orange range (as its spectral power distribution graph indicates), or a combination these possibilities. It is for this reason that I chose not to purchase this version of the flashlight, as impressed as I was--at least at the time--with its 93 CRI. In fact, this was what first made me doubt the usefulness of CRI in evaluating color rendering accuracy, and everything I've learned about CRI and from my experiences with using flashlights with varied spectra since then has confirmed my doubts.
I'm not sure where you got 83, but it is rated by the manufacturer as having a minimum of 80 and a typical value of 85. A few months ago that would have meant something to me, but not anymore.
The purpose of color rendering accuracy is not to make colors "pop" (more than they should) but to render them accurately. To make them really "pop" requires exaggerating certain wavelengths, usually in the red range of the visible spectrum. I find that my H51c makes many colors "pop" a little bit more than the typical neutral white LED due to its higher red output, but at the same time it's not overly red to the point where it adds any more "pop" or "richness" than objects should have like most (not all) other high-CRI LEDs do (sort of like putting on BluBlockers
).Although I fully acknowledge that we all see a little differently (or maybe even a lot in some cases), I wouldn't be surprised if somebody expressed disappointment in the H51c/SC51c (same LED) even if their eyes and visual perception were exactly like mine. That's because they may be looking for something different, such as more "pop" or a specific tint rather than what I personally define as accuracy. I'm quite pleased that it gives me the accuracy--relative to sunlight--that I'm seeking, or something pretty darn close to it. I've always found artificial light sources lacking in this respect, whether they "pop" or not.
Although what I've said does include both subjective and objective elements, I assure you that there was no deliberate attempt to confuse the two. Critical readers should be able to sort them out, I think.
It is neither undeniable nor self-evident. I think that my arguments and examples are sufficient to introduce doubt with regard to placing blind faith (as I once did, effectively) in CRI as an indicator of color rendering accuracy. I tried to do it in a way that laymen (myself included) could understand, but there are people with much better credentials who are thinking the same thing, as evidenced in the following links:
http://en.wikipedia.org/wiki/Color_rendering_index#Criticism_and_resolution
http://www.knt.vein.hu/staff/schandaj/SJCV-Publ-2005/521.pdf
My personal preferences are no secret, and I don't hide the fact that they form the perspective from which I discuss these issues. This is perfectly natural since this particular discussion started when somebody accused those who have such a perspective of making no sense (which I suppose implies that those who have other perspectives are making sense :ironic
. In response, I tried my best to make sense out of my (and others') perspective, without invalidating other perspectives (we all have different preferences and goals).As for CRI, there are many good reasons to rightfully downplay its importance--particularly where phosphor-based LEDs are concerned. I've given a few of my own, as well as references to the harsh criticisms of professionals for everybody here to peruse. If anything that I or they said is wrong, then by all means refute it--I'd rather be corrected than mistaken while still believing that I'm right (and misleading others).
Generally speaking, CRI is crude and only useful for certain cases. Because of this and the fact that by definition it only measures accuracy against an ideal reference of the same CCT, it is virtually useless as an indicator of the color rendering accuracy of, say, 4000K LEDs against an ideal 5800K source such as sunlight. Does this make sense? I'm saying that if a specific 4000K LED renders colors very much like my ideal of 5800K sunlight does (with some flaws, of course), then its CRI--due to the very definition of CRI--will in all likelihood be lower than that of other 4000K LEDs that come closer to the theoretical ideal 4000K source. The problem is that I don't care about the ideal 4000K source while CRI--by definition--does, therefore the higher CRIs of these other LEDs means nothing to me. This would be true even if CRI were a PERFECT method of determining accuracy, but in reality it is far, FAR from that.
the.Mtn.Man
Flashlight Enthusiast
- Joined
- Oct 3, 2008
- Messages
- 2,516
Re: HDS Systems EDC # 15
It depends on what you mean by "accurate". A bluish/purplish tint will not render colors "accurately" if your reference is a 3200K incandescent.
I think where this discussion is getting lost is that people are using "accurate" to mean "preferred".
It depends on what you mean by "accurate". A bluish/purplish tint will not render colors "accurately" if your reference is a 3200K incandescent.
I think where this discussion is getting lost is that people are using "accurate" to mean "preferred".
Last edited:
the.Mtn.Man
Flashlight Enthusiast
- Joined
- Oct 3, 2008
- Messages
- 2,516
Re: HDS Systems EDC # 15
Then I guess it was inadvertent.
But CRI is an indicator of color rendering accuracy. That's what "CRI" means: color rendering index as measured on a 100 point scale. A CRI of 100 is as accurate as it gets at any color temperature.
That said, I think you do have a valid point that someone who prefers cool white isn't going to necessarily like how colors look with a warm high CRI emitter even if the latter offers more accurate color rendering (relative to its reference).
Then I guess it was inadvertent.
But CRI is an indicator of color rendering accuracy. That's what "CRI" means: color rendering index as measured on a 100 point scale. A CRI of 100 is as accurate as it gets at any color temperature.
That said, I think you do have a valid point that someone who prefers cool white isn't going to necessarily like how colors look with a warm high CRI emitter even if the latter offers more accurate color rendering (relative to its reference).
Re: HDS Systems EDC # 15
I'm sorry that you got this impression because it was not my intention. All I'm saying in this regard is that I've found a specific LED--not just any LED of a certain CRI and CCT, but one specific model--that appears to my eyes to render the vast majority of colors in a manner very similar to that of 5800K sunlight (despite having too much yellow that serves to bring its CCT down to 4000K). This only applies to those who are looking for the same thing that I am.
However, in a broader sense I used this specific example (and others) to show why CRI is not the most meaningful and useful indicator of color rendering accuracy (especially for LEDs). Two LEDs that have the exact same CRI and CCT can still have extremely different spectra and therefore render colors very differently. If one could be judged as subjectively more accurate than the other, which I think is pretty likely, then CRI didn't really tell us much. With different CCTs, the result is totally meaningless for the reasons I gave earlier.
Straw man? Anyway, I think you're reading things into what I said. I never said that my ideal light source was the ONLY correct one. You're right that there are many forms of natural light, although I must point out that sunlight (as an average that excludes the extremes) is what I've personally spent more time under (even when inside) during the course of my life. What others have experienced and may prefer for whatever reasons (if any) is just as valid, and talking about my preferences in no way implies otherwise.
That is true, but my point was that CRI does not do this in most cases. Some have tried to tell others that a higher CRI always means superior color rendering accuracy, but that is NOT what it means--it means something very specific and narrow in scope, and on top of that it's pretty crude and inherently ill-suited for use with LEDs (all backed up earlier with solid reasoning that nobody has bothered to refute).
Additionally, there is also the point about looking at the whole picture of color balance (relative spectral power distribution) rather than just CRI and CCT, which captures only a tiny fraction of the information about a light source. I have to say that it has always been difficult to broaden the perspectives of others, but it's even more difficult, as well as ironic, when everybody pretty much ignores all of that and focuses on a tiny aspect of the discussion where they think (mistakenly in this case) that my perspective is overly narrow. :laughing:
I'm sorry that you got this impression because it was not my intention. All I'm saying in this regard is that I've found a specific LED--not just any LED of a certain CRI and CCT, but one specific model--that appears to my eyes to render the vast majority of colors in a manner very similar to that of 5800K sunlight (despite having too much yellow that serves to bring its CCT down to 4000K). This only applies to those who are looking for the same thing that I am.
However, in a broader sense I used this specific example (and others) to show why CRI is not the most meaningful and useful indicator of color rendering accuracy (especially for LEDs). Two LEDs that have the exact same CRI and CCT can still have extremely different spectra and therefore render colors very differently. If one could be judged as subjectively more accurate than the other, which I think is pretty likely, then CRI didn't really tell us much. With different CCTs, the result is totally meaningless for the reasons I gave earlier.
Straw man?
Anyway, I think you're reading things into what I said. I never said that my ideal light source was the ONLY correct one. You're right that there are many forms of natural light, although I must point out that sunlight (as an average that excludes the extremes) is what I've personally spent more time under (even when inside) during the course of my life. What others have experienced and may prefer for whatever reasons (if any) is just as valid, and talking about my preferences in no way implies otherwise.That is true, but my point was that CRI does not do this in most cases. Some have tried to tell others that a higher CRI always means superior color rendering accuracy, but that is NOT what it means--it means something very specific and narrow in scope, and on top of that it's pretty crude and inherently ill-suited for use with LEDs (all backed up earlier with solid reasoning that nobody has bothered to refute).
Additionally, there is also the point about looking at the whole picture of color balance (relative spectral power distribution) rather than just CRI and CCT, which captures only a tiny fraction of the information about a light source. I have to say that it has always been difficult to broaden the perspectives of others, but it's even more difficult, as well as ironic, when everybody pretty much ignores all of that and focuses on a tiny aspect of the discussion where they think (mistakenly in this case) that my perspective is overly narrow. :laughing:
