6 ms·

I bought one, as I sell a product with a similar use case [0]. My impressions: The overall concept is great. The UX is cool, but some of the UX choices are str

I bought one, as I sell a product with a similar use case [0]. My impressions: The overall concept is great. The UX is cool, but some of the UX choices are strange to me. For example, automatically dimming down when you turn the light on (you have to press the button to pause it) vs. the other way around. It works well as a wake-up light, although I would have preferred something higher-blue for the morning. There's really nothing "low-blue" about this light compared to other 2700K LED lights, other than the fact that it has a dimmer built in. The 2700K diodes they chose don't have better color quality than normal LED bulbs. It has "amber" (somewhere around 2000-2200K) LEDs at the very bottom end of the range that have great color quality, but they are too dim for anything but navigating to the bathroom at night. They are modulating the LEDs with PWM at a very high frequency, almost 8 kHz. That's better than most of today's smart lighting by far. There's still debate about 100% flicker actually being healthy, even at higher frequencies. For the lighting nerds, I captured photometric and flicker measurements, which you can review at [1]. Overall, I think Casper's lamp is a good effort. I hope future versions can improve the color quality, some of the interactions, and the overall healthiness of the light compared to regular lighting. [0] Bedtime Bulb: https://bedtimebulb.com/ https://bedtimebulb.com/ [1] Casper Glow Photometrics: https://www.dropbox.com/s/u8n0way6xqwbqcd/Casper%20Glow%20Photometry.pdf?dl=0 https://www.dropbox.com/s/u8n0way6xqwbqcd/Casper%20Glow%20Ph...

pags 8y ago
I don't know anything about modulating LEDs, but could that high freq. have anything to do with the quiet but audible whine I hear when the light is on? Seems like it could be 8kish.
- Duhck 8y ago
  That's likely a singing capacitor. Lots of electronics will "sing" like this FWIW. Could also be a transformer but since this isn't a high voltage product it's more likely some disk capacitor in the circuitry.
  - samstave 8y ago
    Isnt it usually the ballasts that sing in most, if not all lights?
    - dfox 8y ago
      There isn't any ballast to speak of in essentially any LED driver topology, certainly there isn't any inductive component that is used for its reactance as part of the working principle (which is what "ballast" means). LED driver might have some kind of LC filters on its power input for EMC reasons, but these are too small to meaningfully "sing". On the other hand it may make sense to follow the driver with some kind of reconstruction filter to remove the "high frequency" flicker, but AFAIK nobody does that because it involves costly and large-ish components with significant losses for somewhat questionable benefit.
      - sokoloff 8y ago
        Wouldn’t a boost converter from LiPo to voltage needed to run a string of series LEDs have an inductor? That seems a common design pattern. It’s arguably before the LED driver chip but a meaningful part of the overall design. Indeed, the Casper uses the Ti TPS63021 boost converter, which needs an inductor (L1/L2 on the datasheet sample application). http://www.ti.com/product/TPS63021?qgpn=tps63021 http://www.ti.com/product/TPS63021?qgpn=tps63021
  - Bjartr 8y ago
    It could also be PWM, which is frequently used to control LED brightness.
degenerate 8y ago
I use flux and I know the importance of eliminating blue wavelength before bed. What I don't understand is the need to eliminate green, and this graph on your site doesn't help much in understanding that: https://i.imgur.com/fHJKxWA.png https://i.imgur.com/fHJKxWA.png (rehosting on imgur for short URL) I don't know what the grey area is, and there are no axis so I am unsure what this graph is even showing me. Also, wouldn't eliminating "orange" light be just as important (according to the graph you provided), since it seems to follow the same path as green? Everything else makes sense on the page, it's the graph that threw me.
- yeutterg 8y ago
  In simple terms, both blue and green light contribute equally to circadian interruption. The graph itself is a comparison of spectral power distributions among several light sources. It's basically the signature of the light source, showing relatively how much of each color is produced in comparison to all the other visible wavelengths. The gray area is melanopic sensitivity. This is the light that the ipRGCs[0] in the eyes are sensitive to, which provide input to the circadian rhythm. The peak is approximately 480-490 nm, but the overall sensitivity covers both blue and green wavelengths. There's a great article on the blue/green concept by Ian Ashdown [1]. If you use the f.luxometer tool [2], developed by the f.lux team, you will see that same curve. EDIT (for clarity): In the graph you linked, ignore the color of the lines. That is just to distinguish the different light sources. Just pay attention to the intensity of the wavelengths on the x-axis. The y-axis scale is 0 to 1, as this graph is normalized, so hardly necessary to show. Let me know if I can help clarify this further! [0] ipRGCs: https://en.wikipedia.org/wiki/Intrinsically_photosensitive_retinal_ganglion_cells https://en.wikipedia.org/wiki/Intrinsically_photosensitive_r... [1] Melanopic Green: The Other Side of Blue: https://www.ies.org/fires/melanopic-green-the-other-side-of-blue/ https://www.ies.org/fires/melanopic-green-the-other-side-of-... [2] f.luxometer: https://fluxometer.com/rainbow/#!id=iPad%20Pro/6500K-iPad%20Pro https://fluxometer.com/rainbow/#!id=iPad%20Pro/6500K-iPad%20...
- geofft 8y ago
  "Orange," "green," etc. in those graphs are just different models of lightbulbs. (Maybe using patterns - dotted, dashed, etc. lines - would have been clearer than overloading color.) For colors, look at the frequencies on the X-axis. Anyway, the trick is that photoreceptor cells are sensitive to a range of frequencies, with weights centered on a particular color, and return a single floating-point output. So your blue cones cannot distinguish bright green and dim blue - that's only calculated by taking the input of the green cones into account, too. See https://i.stack.imgur.com/5snTb.png https://i.stack.imgur.com/5snTb.png for what each cone's sensitivity looks like (in this graph, lines are light received by each cone, and the colors are semi-meaningful). If the goal is to not stimulate the blue cones at all, you need to emit as close to pure red light as possible. The wavelength of your light as measured by the single average or dominant color isn't as interesting as the spectrum of all light it emits and how much that overlaps with blue sensitivity.
GuiA 8y ago
> The UX is cool, but some of the UX choices are strange to me. For example, automatically dimming down when you turn the light on (you have to press the button to pause it) vs. the other way around I own one. If the brightness is anything above the middle mark or so (which you’d want to light up a room), then the device consumes more power than the charger provides, and it dies after 2-3 hours. I believe that’s why the default behavior is to start dimming the moment it is turned on. It bothers me a bit, but I do really like the product otherwise.
- yeutterg 8y ago
  That's odd—I haven't left it on long enough to notice. So they really intended this light to be used for short time intervals.
  - trhway 8y ago
    especially odd considering that those sub-$20 charging pads for phones are 10-15w which is more than plenty for a very light LED.
adrianmonk 8y ago
I suppose there are two types of people: those who would consider it fine if a night light is on continuously all night, and those who would not. I'm firmly in the second category. I want my room to be as dark as possible while I'm sleeping. I don't have a night light right now, but if I did, I would get something that errs on the side of not being lit. For example, a motion-activated one. If that is what this dim by default behavior is intended to accomplish, then it sounds like a wonderful idea to me. A night light is a hard sell for me in the first place, but this is the sort of thing that might convince me to consider it.
- yeutterg 8y ago
  I totally agree the dim to dark is an amazing feature before bed. I just wish they would use something better than a 2700K LED if it's supposed to be used in the 15-45 minutes before sleep. In my opinion, if people are able to sleep in darkness, they should, and it's not a good habit to start kids out sleeping with the lights on.
  - ticmasta 8y ago
    >> and it's not a good habit to start kids out sleeping with the lights on. Why? I usually see issues going the other way, where kids from birth get absolute quiet and dark, then can't get to sleep without those conditions for the rest of their lives (i.e. adults who need blackout curtains and ear plugs to sleep).
    - asteli 8y ago
      I've wondered about that -- my partner and her cousin both grew up in the Midwest and need white noise and eyeshades to sleep. For reference we live in the SF Bay. I grew up here and can sleep through damn near anything. I'm curious how much of the difference is due to childhood locality vs genetics vs child-rearing practices.
      - elbear 8y ago
        Have you ever tried sleeping with eyeshades? It has been shown[1] that even small sources of light can inhibit the secretion of melatonin (the hormone that regulates the circadian rhythm and consequently sleep). If you use eyeshades, you might feel more rested in the morning (if your bedrooms is not already pitch black dark, that is). As a personal anecdote, I always feel less rested when I don't use eyeshades. [1] https://www.ncbi.nlm.nih.gov/pubmed/3567262 https://www.ncbi.nlm.nih.gov/pubmed/3567262
alias_neo 8y ago
I don't know what these cost, the article didn't say,but I run Hue throughout my home. If I was to take just the absolute minimum you'd require for sleep/wake, and getting to the bathroom at night; 1 bulb in a lamp in your bedroom, 1 on the landing and 1 in the bathroom, add 2 Hue motion sensors (or just one if your can live with it turning both the landing and bathroom lights on) and the hub, you're looking at 1 Hue Starter kit and 2 sensors, in an Amazon sale you could do that for around £150, or closer to £200 if not. This is for the colour and white ambiance bulbs, which do full colour and a broad spectrum of white. Now you have your colour temps for sleep wake, as well as any colours you desire for the landing/hall and bathroom. You have hands free lighting from room to bathroom at night, and can pre configure different light temps, colours and brightness for day/night and can set sleep/wake schedules, sleep timers and various other bits (all without connecting the hub or your phone to their cloud service); if you want some physical control the battery powered dimmer switch can be had for ~£20. I've been running this for a couple of years now since buying my new home, being able to get up in the middle of the night, hall and bathroom lights automatically turn on to a dim red I've configured for night mode, then wonder back to bed and have them turn them selves off a couple of minutes later has been priceless. I also don't use an alarm because the bedside lamps are set for "sunrise" at my desired wake time. The sensors also sense ambient light which you can configure the level of to not activate the lights at all if the natural light level is sufficient in the day. The sensors also hide other features accessible by the (local) API that Hue don't use yet, such as temperature and Lux amongst other things. The Hue can also be extended almost limitlessly over time to (as I've done) cover your whole home (and outside it with their outdoor lighting). The question now is, how close in price is that minimal scenario to what these cost Casper lights cost?
- bosie 8y ago
  seems to be USD 99 plus tax https://casper.com/glow-light/buy/ https://casper.com/glow-light/buy/
  - alias_neo 8y ago
    At that link it says £90 for the one lamp. I'd say, that price, coupled with the above features, Hue is by far the better deal. It doesn't have fancy gyro features, but there's no reason someone can't make a ZLL (ZigBee Light Link) accessory with those if they were the killer feature.
balfirevic 8y ago
Is your product available in Europe?
- yeutterg 8y ago
  It will be in a few months. Shoot us an email and we'll make sure you're aware of the launch: hey AT bedtimebulb DOT com Thanks!
Freak_NL 8y ago
> They are modulating the LEDs with PWM at a very high frequency, almost 8 kHz. That's better than most of today's smart lighting by far. There's still debate about 100% flicker actually being healthy, even at higher frequencies. That's really quite high. 300 Hz makes stationary lights seem constant for the majority of people. Does that frequency have an impact on the total lifespan of the LEDs?
- et2o 8y ago
  You can likely still detect much higher frequencies due to saccadic eye movements. Up to around 2-3Khz.
fouc 8y ago
I'd prefer something that's too dim over something that is too bright, so maybe that's why it automatically dims when you first turn it on. (So it makes sense to me UX-wise) When navigating to the bathroom at night the light certainly should be dim or red light only.
neop1x 8y ago
what about placing a capacitor after PWM...

pags 8y ago

I don't know anything about modulating LEDs, but could that high freq. have anything to do with the quiet but audible whine I hear when the light is on? Seems like it could be 8kish.

Duhck 8y ago
That's likely a singing capacitor. Lots of electronics will "sing" like this FWIW. Could also be a transformer but since this isn't a high voltage product it's more likely some disk capacitor in the circuitry.
- samstave 8y ago
  Isnt it usually the ballasts that sing in most, if not all lights?
  - dfox 8y ago
    There isn't any ballast to speak of in essentially any LED driver topology, certainly there isn't any inductive component that is used for its reactance as part of the working principle (which is what "ballast" means). LED driver might have some kind of LC filters on its power input for EMC reasons, but these are too small to meaningfully "sing". On the other hand it may make sense to follow the driver with some kind of reconstruction filter to remove the "high frequency" flicker, but AFAIK nobody does that because it involves costly and large-ish components with significant losses for somewhat questionable benefit.
    - sokoloff 8y ago
      Wouldn’t a boost converter from LiPo to voltage needed to run a string of series LEDs have an inductor? That seems a common design pattern. It’s arguably before the LED driver chip but a meaningful part of the overall design. Indeed, the Casper uses the Ti TPS63021 boost converter, which needs an inductor (L1/L2 on the datasheet sample application). http://www.ti.com/product/TPS63021?qgpn=tps63021 http://www.ti.com/product/TPS63021?qgpn=tps63021
- Bjartr 8y ago
  It could also be PWM, which is frequently used to control LED brightness.

degenerate 8y ago

I use flux and I know the importance of eliminating blue wavelength before bed. What I don't understand is the need to eliminate green, and this graph on your site doesn't help much in understanding that: https://i.imgur.com/fHJKxWA.png https://i.imgur.com/fHJKxWA.png (rehosting on imgur for short URL) I don't know what the grey area is, and there are no axis so I am unsure what this graph is even showing me. Also, wouldn't eliminating "orange" light be just as important (according to the graph you provided), since it seems to follow the same path as green? Everything else makes sense on the page, it's the graph that threw me.

yeutterg 8y ago
In simple terms, both blue and green light contribute equally to circadian interruption. The graph itself is a comparison of spectral power distributions among several light sources. It's basically the signature of the light source, showing relatively how much of each color is produced in comparison to all the other visible wavelengths. The gray area is melanopic sensitivity. This is the light that the ipRGCs[0] in the eyes are sensitive to, which provide input to the circadian rhythm. The peak is approximately 480-490 nm, but the overall sensitivity covers both blue and green wavelengths. There's a great article on the blue/green concept by Ian Ashdown [1]. If you use the f.luxometer tool [2], developed by the f.lux team, you will see that same curve. EDIT (for clarity): In the graph you linked, ignore the color of the lines. That is just to distinguish the different light sources. Just pay attention to the intensity of the wavelengths on the x-axis. The y-axis scale is 0 to 1, as this graph is normalized, so hardly necessary to show. Let me know if I can help clarify this further! [0] ipRGCs: https://en.wikipedia.org/wiki/Intrinsically_photosensitive_retinal_ganglion_cells https://en.wikipedia.org/wiki/Intrinsically_photosensitive_r... [1] Melanopic Green: The Other Side of Blue: https://www.ies.org/fires/melanopic-green-the-other-side-of-blue/ https://www.ies.org/fires/melanopic-green-the-other-side-of-... [2] f.luxometer: https://fluxometer.com/rainbow/#!id=iPad%20Pro/6500K-iPad%20Pro https://fluxometer.com/rainbow/#!id=iPad%20Pro/6500K-iPad%20...
geofft 8y ago
"Orange," "green," etc. in those graphs are just different models of lightbulbs. (Maybe using patterns - dotted, dashed, etc. lines - would have been clearer than overloading color.) For colors, look at the frequencies on the X-axis. Anyway, the trick is that photoreceptor cells are sensitive to a range of frequencies, with weights centered on a particular color, and return a single floating-point output. So your blue cones cannot distinguish bright green and dim blue - that's only calculated by taking the input of the green cones into account, too. See https://i.stack.imgur.com/5snTb.png https://i.stack.imgur.com/5snTb.png for what each cone's sensitivity looks like (in this graph, lines are light received by each cone, and the colors are semi-meaningful). If the goal is to not stimulate the blue cones at all, you need to emit as close to pure red light as possible. The wavelength of your light as measured by the single average or dominant color isn't as interesting as the spectrum of all light it emits and how much that overlaps with blue sensitivity.

GuiA 8y ago

> The UX is cool, but some of the UX choices are strange to me. For example, automatically dimming down when you turn the light on (you have to press the button to pause it) vs. the other way around I own one. If the brightness is anything above the middle mark or so (which you’d want to light up a room), then the device consumes more power than the charger provides, and it dies after 2-3 hours. I believe that’s why the default behavior is to start dimming the moment it is turned on. It bothers me a bit, but I do really like the product otherwise.

yeutterg 8y ago
That's odd—I haven't left it on long enough to notice. So they really intended this light to be used for short time intervals.
- trhway 8y ago
  especially odd considering that those sub-$20 charging pads for phones are 10-15w which is more than plenty for a very light LED.

adrianmonk 8y ago

I suppose there are two types of people: those who would consider it fine if a night light is on continuously all night, and those who would not. I'm firmly in the second category. I want my room to be as dark as possible while I'm sleeping. I don't have a night light right now, but if I did, I would get something that errs on the side of not being lit. For example, a motion-activated one. If that is what this dim by default behavior is intended to accomplish, then it sounds like a wonderful idea to me. A night light is a hard sell for me in the first place, but this is the sort of thing that might convince me to consider it.

yeutterg 8y ago
I totally agree the dim to dark is an amazing feature before bed. I just wish they would use something better than a 2700K LED if it's supposed to be used in the 15-45 minutes before sleep. In my opinion, if people are able to sleep in darkness, they should, and it's not a good habit to start kids out sleeping with the lights on.
- ticmasta 8y ago
  >> and it's not a good habit to start kids out sleeping with the lights on. Why? I usually see issues going the other way, where kids from birth get absolute quiet and dark, then can't get to sleep without those conditions for the rest of their lives (i.e. adults who need blackout curtains and ear plugs to sleep).
  - asteli 8y ago
    I've wondered about that -- my partner and her cousin both grew up in the Midwest and need white noise and eyeshades to sleep. For reference we live in the SF Bay. I grew up here and can sleep through damn near anything. I'm curious how much of the difference is due to childhood locality vs genetics vs child-rearing practices.
    - elbear 8y ago
      Have you ever tried sleeping with eyeshades? It has been shown[1] that even small sources of light can inhibit the secretion of melatonin (the hormone that regulates the circadian rhythm and consequently sleep). If you use eyeshades, you might feel more rested in the morning (if your bedrooms is not already pitch black dark, that is). As a personal anecdote, I always feel less rested when I don't use eyeshades. [1] https://www.ncbi.nlm.nih.gov/pubmed/3567262 https://www.ncbi.nlm.nih.gov/pubmed/3567262

alias_neo 8y ago

I don't know what these cost, the article didn't say,but I run Hue throughout my home. If I was to take just the absolute minimum you'd require for sleep/wake, and getting to the bathroom at night; 1 bulb in a lamp in your bedroom, 1 on the landing and 1 in the bathroom, add 2 Hue motion sensors (or just one if your can live with it turning both the landing and bathroom lights on) and the hub, you're looking at 1 Hue Starter kit and 2 sensors, in an Amazon sale you could do that for around £150, or closer to £200 if not. This is for the colour and white ambiance bulbs, which do full colour and a broad spectrum of white. Now you have your colour temps for sleep wake, as well as any colours you desire for the landing/hall and bathroom. You have hands free lighting from room to bathroom at night, and can pre configure different light temps, colours and brightness for day/night and can set sleep/wake schedules, sleep timers and various other bits (all without connecting the hub or your phone to their cloud service); if you want some physical control the battery powered dimmer switch can be had for ~£20. I've been running this for a couple of years now since buying my new home, being able to get up in the middle of the night, hall and bathroom lights automatically turn on to a dim red I've configured for night mode, then wonder back to bed and have them turn them selves off a couple of minutes later has been priceless. I also don't use an alarm because the bedside lamps are set for "sunrise" at my desired wake time. The sensors also sense ambient light which you can configure the level of to not activate the lights at all if the natural light level is sufficient in the day. The sensors also hide other features accessible by the (local) API that Hue don't use yet, such as temperature and Lux amongst other things. The Hue can also be extended almost limitlessly over time to (as I've done) cover your whole home (and outside it with their outdoor lighting). The question now is, how close in price is that minimal scenario to what these cost Casper lights cost?

bosie 8y ago
seems to be USD 99 plus tax https://casper.com/glow-light/buy/ https://casper.com/glow-light/buy/
- alias_neo 8y ago
  At that link it says £90 for the one lamp. I'd say, that price, coupled with the above features, Hue is by far the better deal. It doesn't have fancy gyro features, but there's no reason someone can't make a ZLL (ZigBee Light Link) accessory with those if they were the killer feature.

balfirevic 8y ago

Is your product available in Europe?

yeutterg 8y ago
It will be in a few months. Shoot us an email and we'll make sure you're aware of the launch: hey AT bedtimebulb DOT com Thanks!

Freak_NL 8y ago

> They are modulating the LEDs with PWM at a very high frequency, almost 8 kHz. That's better than most of today's smart lighting by far. There's still debate about 100% flicker actually being healthy, even at higher frequencies. That's really quite high. 300 Hz makes stationary lights seem constant for the majority of people. Does that frequency have an impact on the total lifespan of the LEDs?

et2o 8y ago
You can likely still detect much higher frequencies due to saccadic eye movements. Up to around 2-3Khz.

fouc 8y ago

I'd prefer something that's too dim over something that is too bright, so maybe that's why it automatically dims when you first turn it on. (So it makes sense to me UX-wise) When navigating to the bathroom at night the light certainly should be dim or red light only.

neop1x 8y ago

what about placing a capacitor after PWM...