Hacker Newsnew | past | comments | ask | show | jobs | submitlogin
Average color of the NYC sky every 5 minutes (nskyc.com)
540 points by sethbannon on June 9, 2023 | hide | past | favorite | 109 comments


#50a7f1 from this morning is an incredibly beautiful blue

https://storage.googleapis.com/nskyc-3727d.appspot.com/nyc/1...

Makes me miss NYC.


Such a great little website. you can see it was more orange than Hackernews header just 2 days back.


What concentration of fine particulates would be necessary to produce a sky the exact color of HN's header?


YC116 aims to find out. Now hiring. Our mission: Making the world a better place by making the sky HN orange.


With occasional low hanging black clouds for when famous people die.


Do I have to relocate to nightvale in order to get a job?


The Tubbs fire did a fine job of this, but it’s always dimmer than the luminosity of HN orange on a typical screen, so it depends on how you define “color”.


Not being exact, but at least 150+ AQI seems to generate a significant impact on sunlight coming through creating that "smoky"/"faded" look. I can't comment on how to get the orange exactly as I think it also requires specific particulate matter to get that.


Not particulate matter, but photochemical smog, so nitrogen and sulfur oxides, volatile organic compounds and of course ozone.

Add sodium lamps for lighting and you get a rich, orange hue.


what a funny/scary thought :)


I wonder how the camera was calibrated in terms of white balance (if at all).

I have this idea to implement in my house: a sensor that would measure temperature of the light outside and set the lights inside to the same temperature. Getting this measurement right seems to be non-trivial. Though maybe it wouldn't have to be exact in terms of physical units, a pair of sensors calibrated together (one for outside, one for inside) might achieve the desired result too.


I think it must have a fixed white balance, or the average hue captured would hardly change over time at all.

Adjusting the inside lights until the inside sensor is satisfied, rather than based on specifying in absolute units, would also have the benefit of taking into account any influence on indoor overall color contributed by your walls, window coverings, and anything else. Might need to average multiple sensors for best results, unless a single sensor does that through a wide angle diffuser.

This is a bit like how automatic audio equalizers (I think Sonos offers this, for example) don't work just by ensuring that the speaker delivers the desired frequency response curve (or lack thereof), but by actually using a measurement mic, since your experience will be influenced by the combination of speaker+room. The difference is that you only need to have the mic around when making changes to the room/furniture (which is rare) whereas you'd be making changes to the indoor light bulbs continuously (based on outdoor light), so you need your indoor sensor running continuously.


Your idea is pretty straightforward. Set up a camera outside, point it at the sky, capture a screenshot, average the color, set smart lights to that color.

Presto, yellow sky makes yellow light.

You’ll quickly grow tired of how it looks. And not because it isn’t calibrated correctly.


What the high end variable-temperature lighting systems seem to do is just use your clock and latitude to simulate what color temperature the sky would be at a given time of day, and set the lights to that.

I've never seen it in action, I only have boring old single-color lights. I am not really annoyed by 2700K lamps during the day, and mixing daylight and 2700K doesn't really look as awful as you'd think during zoom calls.


The main point is that those high end systems won’t look any better to you just because they calculate out a color curve based on your lat,long. It’d look just as good if you point a camera at the sky. Your high end systems also won’t capture changes in weather, so that’s less cool. But then you’d get bored of it looking rainy whenever it rains.


Could be mistaken but I think my Mac already does this -- it has "true tone" and when I open the lid, my main external monitor changes color temperature.

I wonder if there's a way to scrape that data and then send it to Home Assistant for light control.


If only color calibration were that easy.

People should really get comfortable with the idea that color is relative, so choose whatever looks nice to you and stick with it. True Tone certainly isn’t scientific.

In this scenario, all that matters is that the instrument doesn’t try to change its white balance. You can actually calculate the spectral energy of the sky based on these color changes alone, and it doesn’t matter if you start at a white balance of fluorescent or incandescent.


I find it fascinating how Toronto is no where near as bad as New York. The fight to protect the environment has to be a global one


https://firesmoke.ca is an excellent resource for seeing the current smoke levels across the continent.


At the latitudes of the US and Canada, the prevailing winds blow from west to east.


Ahh, my geography was completely off, I was picturing NY where Buffalo currently is.. I can't believe I forgot it's on the Atlantic.


Has the developer of this posted any blog or documentation of how it works? What hardware/software is used for it?


The author is a designer who do many creative coding projects, he mentioned in an interview[1] back in 2011 that:

> I wrote this program that hooks up to a webcam. It takes a photo out the window every five minutes, and it will upload that to a server. The server then reads the sky portion of the photo, and it goes pixel by pixel. What it does is it takes all those values, the RGB values, and it averages them. So what you are seeing is not the dominant color in the sky it’s actually just the average color.

Mike Bodge's personal website:https://bod.ge/

Recent tweet by author: https://twitter.com/mikebodge/status/1666507978663690240

Here's also a submission from a decade ago (2011): https://news.ycombinator.com/item?id=2685621

Shameless plug: Assisted by my HNRelevant browser extension

[1] https://www.villagevoice.com/2011/06/23/mike-bodge-nskyc-cre...


Average how? In RBG? HSL? LAB?

Average raw RGB gives you bad results.


Well, I'm afraid I can't help you with that

If you're enthusiastic about the project, maybe consider DMing the author on Twitter to discuss it with him


The first two night photos have logos for EarthCam, Live, and some location data. Seems they found a way to remove those in later images.


It appears that the camera provider was switched around 1 day ago [0].

[0] https://twitter.com/mikebodge/status/1666541169273372672


probably a webcam


i presume you're on mobile and it's not visible there. on desktop you can hover any section and you will see the picture that is used as the source.


On mobile (i use firefox an android) you can tap on the sections


Tap on mobile


Nice, however "a day ago" isn't particularly helpful. For this local EDT time would be fine, but I'd like to know it it's 8am or 11am


Zooming out is kinda neat...

But I would like to be able to see the background image for some of the shots bythemselves - as there are some cool pics of that view.

https://i.imgur.com/S9I3arV.png


It would be interesting if he specified how he came up with RGB values for sky color, like what is his neutral grey point etc.


Yeah you can't just do this with a consumer webcam out of the box because they're constantly trying to auto-adjust the white point. I mean not if you want anything "accurate". (Some webcams allow you to lock the white point in software though.)

But indeed, it is very non-obvious what you'd select for the constant exposure and white point (or combined as a gray point). You might also want to apply a strong curve to the brightness so nighttime skies are quite visible.


Scroll down to the end of the page to 2 days ago! It's insane!



This is such a cool idea. Some people are so cleaver.


#000 I didn't think it'd be that dark even at local midnight


It's going to depend on the camera and processing, since the dynamic range of human vision extends well beyond the RGB color range.


I agree and if you want to refer to a color in an absolutely correct, objective, device-independent manner you need to define it in terms of L*a*b*

The CIELAB coordinate space represents the entire gamut of human photopic (daylight) vision and far exceeds the gamut for sRGB or CMYK.

You might initially think that the three axes of what is "RGB" defines a perfect color cube which accurately describes all perceptible colors but this is not true. The typical RGB we encounter when encoded with an sRGB gamut (almost always on smartphones, TVs, and PCs as of 2023) forms a bizzare polyhedron in the objective L*a*b* color space instead of filling the whole thing with itself because it is just a small subset of what we can actually see: https://en.wikipedia.org/wiki/File:SRGB_gamut_within_CIELAB_...

https://en.wikipedia.org/wiki/CIELAB_color_space


1WTC and all the other buildings are pretty bright at night. I assume if they were physically masked out, the camera would have better chance to pick up some light reflected off the atmosphere



My heart goes out to new yorkers. This cannot be healthy. If there are long-standing health repercussions I doubt there will be any recourse they can take.


Over a decade ago my wife was thru hiking and got caught in the smoke of a California wildfire. The ash scratched her corneas and her vision is still impacted to this day.

It's definitely not something to take lightly, even if you wear a mask.


It's actually back in the healthy range as of this morning! I could immediately tell the difference when I woke up before sunrise, I went for a long walk outside and it felt great.


It’s not great, but for comparison many cities in Asia have a much worse AQI much more often.


That's not the equality I wanted.


> My heart goes out to new yorkers. This cannot be healthy. If there are long-standing health repercussions I doubt there will be any recourse they can take.

It's not good, but it's far from the worst around. The peak AQI here in NYC was about 350, which is actually much lower than the average AQI in Delhi between November and January, for example. (Air pollution is highly seasonal).

Effects of air pollution are cumulative, so the people who really suffer the most from it are the ones who experience this regularly, as opposed to historical anomalies like this.


Or you can live in the Colorado with fresh air, but way more UV exposure. There are always trade-offs.


We've (Coloradoans)been dealing with wildfire smoke pretty much forever. On average our air quality is quite good, but fires in any of the PNW/SW states regularly impact our air quality during the spring, summer, and fall. We held the dubious title of worst air quality in the world for at least a day earlier this spring.

It's been a little comical having this NYC air quality situation keep finding it's way into my news sources -- it's been a part of my life as long as I've been alive. I check the FireNow site as much as I check the weather during the warm month.


On the Front Range, where most people live in Colorado (~Denver + north and south 60 miles), the air quality can actually get pretty bad due to inversion, and this has been noted since pre-statehood. We also are lucky enough to get a lot of wild fire smoke from California, Canada, the PNW, as well as our own home sourced wildfires. Maybe 2 years ago it felt like the whole summer was smoky, and we had stretches where the AQI was 500+ here in Boulder.

https://denvergazette.com/outtherecolorado/news/the-brown-cl...


People will be fine and no one is going to develop long term health issues from a couple days of smokey air. I feel bad for people with lung conditions as breathing was made more difficult for a bit but no one is dying from this.

And what recourse? Who?


> People will be fine and no one is going to develop long term health issues from a couple days of smokey air. I feel bad for people with lung conditions as breathing was made more difficult for a bit but no one is dying from this.

That's simply not true. People can, and do, develop long-term issues from acute exposure to bad air (that's literally why the range is called "Hazardous" on the AQI).

Not to mention that one in ten New Yorkers has asthma, which means that yes, people can literally die from acute exposure to air pollution.


Not to mention that one in ten New Yorkers has asthma,

One in 10 people do NOT have asthma.


> One in 10 people do NOT have asthma.

You are incorrect. Current prevalence state-wide is 8%[0] and lifetime prevalence is 14%, indicating that many people are expected to develop asthma but have not yet been diagnosed.

Incidence is more concentrated in the city due to a strong causative relationship between asthma and certain measurable factors: poverty, childhood exposure to vehicular-generated air pollution, and the tendency to have highways located near poorer neighborhoods.

If you look at a map of asthma incidence in the city by neighborhood, this heterogeneous distribution is even clearer.

[0] https://www.cdc.gov/asthma/stateprofiles/asthma_in_nys.pdf


Well, all I can say, is these numbers seem crazy to me. It appears that the condition is all a mishmash of "lungs in bad state", predicated by anything from "that dude was breathing deadly toxic air", to "always like this".

I think there should be more nuance here, but concede you are correct as the subject is discussed.


> Current prevalence state-wide is 8%[0] and lifetime prevalence is 14%, indicating that many people are expected to develop asthma but have not yet been diagnosed

Perhaps a bathtub curve? Many people have childhood asthma which resolves in adolescence. So they'd count for lifetime but not current.


> People can, and do, develop long-term issues from acute exposure to bad air

Not for a day, no. At least not for what was going happening in the East Coast. It wasn't like ash and debris was flying through the air. It was like sitting around a campfire. People aren't that fragile. Now, hypochondriacs might think they are (and we have a growing population of those), but they simply aren't. Is it bad for someone with COPD? Yes, but the damage had already been done for years.


> It wasn’t like ash and debris was flying through the air.

Isn’t that exactly what smoke is? It’s worse than visibly large particles, the small particles are more likely to get inside your lungs and do damage.

> Not for a day, no.

You might want to familiarize yourself with the evidence-based WHO air quality guidelines which sets standards based on short-term 8 & 24 hour exposure periods, and separate long-term recommendations for annual exposure. https://www.who.int/news-room/feature-stories/detail/what-ar...

The recommended short-term 24 hour exposure levels are far below what NYC saw (as well as what many cities in the west have had for months at a time over the last ten years. These numbers are published based on the WHO’s ability to demonstrate that adverse health effects appear in the population when exposure exceeds these levels. The recommendations were lowered recently compared to their 2005 numbers because the studies and data and evidence have grown in the last 20 years and it shows that even mild levels of exposure result in more doctor’s visits, more lung conditions, more athsma, more damage and more risk.


Junk “science”. It’s extrapolations that make no sense and that don’t translate to reality. Ever heard the saying “lies, damn lies, and statistics?” Well that’s what this interpretation is.

You can’t generalize from a 1-off event like we saw (where most nobody was being exposed for much of it) and then average across the population and say it causes things.

Repeated exposure, yes. But not this. Is it good for you? No. But it’s not like it’s really hurting anyone. Walking past a person smoking a cigarette isnt going to give you cancer no matter how hysterical sone people get about it. It’s ridiculous to even entertain it.


Crazy. Why do you believe that? Have you read the WHO’s methodology? What is your expertise in air quality studies and/or policy? Do you know how many scientists and organizations outside the WHO agree with their assessment?

Nobody said anything about walking past a single smoker, that’s pure straw man in this context. The stats also aren’t generalized from a single event. You’re arguing armchair FUD logic without any facts, against real-world evidence from a global organization with a many decades history of monitoring all available science on this topic.

The full methodology and 24 pages of scientific references are available online: https://www.who.int/publications/i/item/9789240034228


> Crazy. Why do you believe that?

Because no one is dying from some smoke blowing across the region for a day, lol.

Chronic pollution is a problem. Breathing bad air daily is a problem. working around smoky environments day in and day out is a problem Having some smoke in the area for a day is not.

I'm tired of the neurotics and hypochondriacs making big deals out of things that aren't. Like, don't get fat and don't smoke, and use bad things in moderation and get regular checkups and don't stress about everything and you'll live to 80 in most cases. Some people get unlucky.

People dramatizing this recent event like it was HARMFUL to them and "scary" or some type of public health crisis are insane. Posts like yours that for whatever reason want to reference something that isn't really science but is used like it's irrefutable truth to say that the smoke event was somehow dangerous - I mean, just c'mon. It's fine.


> Because no one is dying from smoke blowing across the region for a day, lol.

Why do you think that? How do you know, exactly?

I would totally recommend reading the WHO report! The guidelines are based on mortality statistics. They have in fact studied how often people are dying from a 24 hour exposure to bad air. Statistically, a few people are actually dying from one “bad day” exposure. It’s not many people, but it’s still a measurable number greater than zero, and they are demonstrating the number is greater than if they didn’t have the one bad day of smoke. Yes, for a single event it affects the most prone population. What we are talking about is risk factors. Most people will not die from one day of smoke, but that doesn’t mean no one will.

I could see this topic being irritating to hear about if you don’t believe in science and don’t trust the WHO, and we certainly have a political climate with people intentionally trying to reduce public trust in science and organizations like the WHO. But it’s worth keeping an open mind and studying a little bit about what they’re actually saying, what they’re not saying, and how they arrived at their conclusions.


> They have in fact studied how often people are dying from a 24 hour exposure to bad air. Statistically, a few people are actually dying from one “bad day” exposure.

Yeah it wasn’t the million other things they did in the last 80 years of their lives. Perfectly healthy people keeling over because of a bad air quality day lol. Please take a moment to use a small bit of logic here. This is a make believe idea like people who died in car accidents but had Covid were victims of Covid. This is aggregate nonsense and to suggest a single, mild event (hundreds of miles from danger) was the bullet to the head is so far fetched as to be hilarious to me.

I’m not a science renter but as a person highly educated in that field I can spit bullshit quickly. I know what the report is saying because stats and math are fun to tell a story. But just because you’re using the tools of science doesn’t mean you’re doing meaningful science.

Your report is a political statement to justify power/action when it isn’t warranted. Much like people used to reference the word of god as authorities truth.

Maybe all the alarmists and doom sayers caused unhealthy rubes undue stress that that’s what did it. They never even breathed a breath.

Your science is laughable at best in the context of this weeks event.


Go directly inhale camp fire smoke for 24 hours straight and tell me how you feel.

Do you live in NY? I dont know anyone that lives here that feels the way you do. I can only reason that someone who is so far removed from the situation can say something so callous.

My dad who has 65 years in NYC has never seen it this bad, and that was with pollution before EPA was created.


I do in fact life in the area. From Tuesday afternoon to Wednesday afternoon it smelled a bit smokey and it was hazy and the sun reflected off the particles. Was eerie.

And I kept my windows closed and we did not go to the playground. But the dog was still walked and I went about my business walking around or driving when needed. It was an exceptional scenario but it wasn’t dangerous.

People need to get over themselves I think. It was fine for the short duration it was.


Haven't the Americans looked for a reason to invade Canada, for like ... forever?


No, we’ve got a good thing going obviously. We have a nice big peaceful border, they send us celebrities and artists, and we will absolutely demolish any country that fucks with them.

We generate plenty of wildfire smoke internally anyway.


Because that worked out so well for us the last time we tried... [0]

[0] https://en.wikipedia.org/wiki/War_of_1812


According to that, the entire war cost ~$105 million dollars, which is about $2 billion dollars today. So 3 years of war equals around one day of the current US military budget.

I realize your post was mostly silly, but it made me curious so I ran the numbers.


The US kicked the British out of Canada in 1867.


I presume this is a joke (though I don’t get it), because AFAIK the US had nothing at all to do with Canadian federation.


Post civil war, there was concern that Canada would be manifest destinied. Britain was somewhat supportive of the CSA and Alaska was purchased in 1867 so there were reasons to think they were next.


Well, TIL. Interesting that I had so many hours of Canadian history yet this particular angle didn’t stick with me. I recall a framing of the constitution act as a quintessentially Canadian “we asked politely for independence, and Britain said okay”, but after some reading it does seem your “let’s get out of here, the neighbors are psycho” framing also holds water ;) I suppose both are true.


That was invading Britain more like, it's time for a rematch.

Of course, we may re"match" by lighting California on fire again when the prevailing winds are from the Southwest.


It mostly seems to have moved down to Pennsylvania at this point.

Smoke happens, it is bad, but I’m sure everybody in the country gets it a couple times in their life (unless you live on Martha’s Vineyard or something).


Is this data not polluted by the skyline? Or does it somewhat take buildings/the skyline into account, and correct for the shades of these objects?


Based on the color selected out of the night shots, I'm assuming the average is of just the plain sky, with the water and buildings cropped out.


From some fiddling in an image editor, it looks like it's a fairly small rectangle (I'd guess top left corner?)


Is the urban legend that it started out of controlled burn at WTREX conference true? Couldn't verify, but there are also no outlets debunking it.


Yes, but that was in Alberta. The smoke in New York is coming from fires in Ontario and Quebec, which were probably started by lightning strikes.


Hot take: Canadians may need to learn controlled burns.


That fire did happen, but was extinguished about a month ago: https://www.albertaprimetimes.com/beyond-local/banffs-out-of...

The 'urban legend' remains because it is constantly mentioned on right-wing 'news' sites as proof that DEI programs are dangerous to society.


I read that article. What is the possible connection to DEI programs?


It was a prescribed fire event called WTREX, which stands for Women-in-Fire [0] Prescribed Fire Training Exchange. Right wing sites haven't been shy about their viewpoint [1].

[0] https://womeninfire.org/

[1] https://freebeacon.com/politics/burning-down-barriers/


Very cute, although I'd like to see what the results are like when choosing the dominant rather than the average color.


Such a simple idea, yet so well executed.


Things start to get interesting if you zoom out:

https://files.littlebird.com.au/Shared-Image-2023-06-10-08-5...


It's unfortunate you can't bookmark dates but really cool none the less.


I stiched the last two days of this into a quick Timelapse - very cool to watch. https://youtu.be/KTv6tqyr8YU



The vague timestamps like "an hour ago" are an odd choice, since they're very non-unique.


With some kind of API to retrieve the colors the site would be really useful.


So simple yet beautiful!




airnow is almost useless in some areas

The only single sensor near my city is 50 miles away, it says air quality is fine, yet the smoke tracker map has several wildfires nearby

We need a huge, reliable pm2.5 etc network, start with every single public school


A significant number of the sensors show up as "LOW COST PM2.5 SENSOR (PurpleAir)"

https://www.purpleair.com sells those sensors (you can see their coverage on their map).

It would be something to assist/help fund and public schools would be an excellent place to have them be situated.


Does this go back further than 2 days?


gosh darn I love this.


great domain name!


This is going to sound brutal - but citizens of New York... if there is anyone in the world to blame for this, it is you...

Sorry. I know that's horrid. But it's also true.


Without really supporting or elaborating on your thesis here, this comment it basically content-free.


Can you explain that?


Sure, I will explain. New York is one of the epicenters of consumption that has driven climate change and precipitated the wild fires in Canada. The USA has been one of the most sustained producers of carbon in the world for the last century and has exported this tendency world wide.

Boys and girls - you did this. The sky is orange because of things you did. Face it, front up and deal with it.


My guess is they are referring to NYC as center of capital markets.


No - personal carbon consumption. New Yorkers have lived an unsustainable life style for the last 100 years, for the last 50 years you've all known about it - but you've done nothing about it.

Yes - the capital markets are also huge enablers.


And where do you live? the carbon footprint of the typical NYC resident is going to be lower than that of the Canadians more proximate to the fire as well as residents of much of the rest of North America, for that matter.




Consider applying for YC's Winter 2026 batch! Applications are open till Nov 10

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: