7 ms·

A barometer is present in new smartphones as a part of the GPS because GPS only provides accurate X/Y positioning, and using barometric data (which varies predi

by criley 14y ago

A barometer is present in new smartphones as a part of the GPS because GPS only provides accurate X/Y positioning, and using barometric data (which varies predictably as elevation changes) allows the GPS to have Z data as well, giving it a better representation of your location. (This is the best explanation I've seen, please correct me if I'm wrong). If you want to donate your barometric data for study, I recommend checking out PressureNet https://play.google.com/store/apps/details?id=ca.cumulonimbus.barometernetwork&hl=en https://play.google.com/store/apps/details?id=ca.cumulonimbu... The PressureNet dev(s) are active on HackerNews and have made a couple of posts about the app. For the record, my Galaxy Note II has a barometer as does the Galaxy S3 -- meaning the S4 didn't change anything in this regard!

cjg_ 14y ago
GPS provides accurate 3D positions. However if you know roughly your position the GPS fix can be acquired more quickly. So the barometer only helps you with an estimate Z pos, which helps you get the GPS fix faster.
- diroussel 14y ago
  If you have line-of-sight to only three GPS satellites then you don't have enough information to compute your altitude [1]. If you can see four satelites, the altitude can be computed. edit: spelling [1] http://gis.stackexchange.com/questions/12866/why-does-gps-positioning-require-four-satellites http://gis.stackexchange.com/questions/12866/why-does-gps-po...
  - archivator 14y ago
    Just tried it out and my phone doesn't report any position (however inaccurate) if it's only seeing 3 satellites. Also, although it's possible in principle, I doubt the timing information on a consumer phone is accurate enough to get a possible fix from 3 data points.
  - cjg_ 14y ago
    Ah, yes of course :)
  - StavrosK 14y ago
    If you have line-of-sight to three GPS satellites, you can compute all three of your coordinates (one will be impossible and can be ruled out), assuming you have an atomic clock locally. If you're like most of us and don't, then you need a fourth (and fifth, and sixth, and seventh) satellite for timing adjustments. Until then, you'll get no fix.
    - Retric 14y ago
      You don't need a local atomic clock for fairly accurate positioning using 3 satellites it just takes longer. Basically, what you need to know is distance from each satellite and an atomic clock let's you find that directly. However, GPS satellites are moving so much faster than you that at over time you narrow down the possible distances involved fairly quickly. Though this does take a lot of processing and is not on any consumer device I know of. Also of note, phones get a reasonable fix using base towers which can speed up GPS calculations though I don't know of any phone that does this.
      - correctifier 14y ago
        Almost all phones do use the base tower to speed up getting a fix with Assisted GPS: http://en.wikipedia.org/wiki/List_of_devices_with_Assisted_GPS http://en.wikipedia.org/wiki/List_of_devices_with_Assisted_G...
- jeza 14y ago
  The barometer gives you better resolution on the Z pos. So it's useful for measuring changes in altitude and the like, provided the weather conditions don't change (resulting in a change in atmospheric pressure). Though without some kind of calibration, you can't determine your altitude from the barometer alone. It can also be useful to calibrate when enough GPS satellites are visible, then continue providing the altitude, even when there aren't enough satellites in view.
- TheAnimus 14y ago
  The question to me is why would someone want a faster Z. Normally people only view their GPS in 2D, granted elevation could be useful for GPS speedos and the like, but when inside a car the barometer will lack precision. The Private Pilot (ie n00b of t3h pilots!111!) in me knows that its incredibly unlikely to have issues seeing many GPS sats when flying, and I wouldn't ever want a static source that wasn't outside. So I think this will be only a speed optomisation at most, however combined with GPS posistion and tempurature, such barometric data could no doubt be of use to weather forecasters who are trying to model incredibly localised patterns maybe?
  - pavel_lishin 14y ago
    No matter what your screen is showing you, you are currently occupying three dimensions of space, and if you want an accurate representation of that without already knowing at least your rough location, you need to lock on to more satellites than you do if you already know approximately how high you are. At least, such is my limited understanding.
    - XorNot 14y ago
      Also if I'm plummeting out of plane in the middle the night with no sense of position, then I'm not saying relying on my phone is a good idea but knowing when to pull my chute so I don't spend too much time in the stratosphere would be pretty handy. I mean I'm not saying it's likely but...luck favors the prepared right?
    - TheAnimus 14y ago
      As I said I support faster GPS (my last line in the post). I think it might be a faster, data light alternative to the A-GPS thing, but I can't really see it narrowing it down that much? I'd have thought dual tuning GLONAS and GPS was better? After-all the error for these sensors will be vast depending on where you are. My understanding (may well have changed, I don't know) is that currently most work on one or the other, not both. A reason I am saying this is I remember the first time I flew a plane in Asia after learning in mild weathered UK I was getting challanging readings from my static source. As such I'm not sure how well it would work as an optomisation. I'm not saying its bad, I'm not saying it's not fun for geeks considering the no doubt low cost of this scale. But I still don't see the need for faster Z, as most smartphone users are on the terraferma?
maxerickson 14y ago
I think it only speeds up solving for position. Without an up to date local reference, the barometer can't tell you absolute altitude with any sort of precision. I think the basic idea is that instead of dealing with the entire circle that you get from intersecting two spheres, you only have to deal with the parts of the circle that are remotely sensible according to the barometer.
mikeash 14y ago
GPS provides far better altitude data than a barometer. GPS altitude fixes are typically accurate to about 20m. An uncalibrated barometer can be off by 300m or more, depending on what the weather happens to be like at that particular moment. I'm sure the barometer is useful for something, but providing better altitude data than GPS gives you is definitely Not It. Edit: I should note that if you calibrate the barometer to the local atmospheric pressure, it can be extremely accurate, down to perhaps 3m or so, if it's a good barometer (like the one usually called an "altimeter" in an airplane). But it's not likely that a cheap phone barometer would achieve that, and of course you need some way to get the local base atmospheric pressure to begin with.

by criley 14y ago

cjg_ 14y ago

GPS provides accurate 3D positions. However if you know roughly your position the GPS fix can be acquired more quickly. So the barometer only helps you with an estimate Z pos, which helps you get the GPS fix faster.

diroussel 14y ago
If you have line-of-sight to only three GPS satellites then you don't have enough information to compute your altitude [1]. If you can see four satelites, the altitude can be computed. edit: spelling [1] http://gis.stackexchange.com/questions/12866/why-does-gps-positioning-require-four-satellites http://gis.stackexchange.com/questions/12866/why-does-gps-po...
- archivator 14y ago
  Just tried it out and my phone doesn't report any position (however inaccurate) if it's only seeing 3 satellites. Also, although it's possible in principle, I doubt the timing information on a consumer phone is accurate enough to get a possible fix from 3 data points.
- cjg_ 14y ago
  Ah, yes of course :)
- StavrosK 14y ago
  If you have line-of-sight to three GPS satellites, you can compute all three of your coordinates (one will be impossible and can be ruled out), assuming you have an atomic clock locally. If you're like most of us and don't, then you need a fourth (and fifth, and sixth, and seventh) satellite for timing adjustments. Until then, you'll get no fix.
  - Retric 14y ago
    You don't need a local atomic clock for fairly accurate positioning using 3 satellites it just takes longer. Basically, what you need to know is distance from each satellite and an atomic clock let's you find that directly. However, GPS satellites are moving so much faster than you that at over time you narrow down the possible distances involved fairly quickly. Though this does take a lot of processing and is not on any consumer device I know of. Also of note, phones get a reasonable fix using base towers which can speed up GPS calculations though I don't know of any phone that does this.
    - correctifier 14y ago
      Almost all phones do use the base tower to speed up getting a fix with Assisted GPS: http://en.wikipedia.org/wiki/List_of_devices_with_Assisted_GPS http://en.wikipedia.org/wiki/List_of_devices_with_Assisted_G...
jeza 14y ago
The barometer gives you better resolution on the Z pos. So it's useful for measuring changes in altitude and the like, provided the weather conditions don't change (resulting in a change in atmospheric pressure). Though without some kind of calibration, you can't determine your altitude from the barometer alone. It can also be useful to calibrate when enough GPS satellites are visible, then continue providing the altitude, even when there aren't enough satellites in view.
TheAnimus 14y ago
The question to me is why would someone want a faster Z. Normally people only view their GPS in 2D, granted elevation could be useful for GPS speedos and the like, but when inside a car the barometer will lack precision. The Private Pilot (ie n00b of t3h pilots!111!) in me knows that its incredibly unlikely to have issues seeing many GPS sats when flying, and I wouldn't ever want a static source that wasn't outside. So I think this will be only a speed optomisation at most, however combined with GPS posistion and tempurature, such barometric data could no doubt be of use to weather forecasters who are trying to model incredibly localised patterns maybe?
- pavel_lishin 14y ago
  No matter what your screen is showing you, you are currently occupying three dimensions of space, and if you want an accurate representation of that without already knowing at least your rough location, you need to lock on to more satellites than you do if you already know approximately how high you are. At least, such is my limited understanding.
  - XorNot 14y ago
    Also if I'm plummeting out of plane in the middle the night with no sense of position, then I'm not saying relying on my phone is a good idea but knowing when to pull my chute so I don't spend too much time in the stratosphere would be pretty handy. I mean I'm not saying it's likely but...luck favors the prepared right?
  - TheAnimus 14y ago
    As I said I support faster GPS (my last line in the post). I think it might be a faster, data light alternative to the A-GPS thing, but I can't really see it narrowing it down that much? I'd have thought dual tuning GLONAS and GPS was better? After-all the error for these sensors will be vast depending on where you are. My understanding (may well have changed, I don't know) is that currently most work on one or the other, not both. A reason I am saying this is I remember the first time I flew a plane in Asia after learning in mild weathered UK I was getting challanging readings from my static source. As such I'm not sure how well it would work as an optomisation. I'm not saying its bad, I'm not saying it's not fun for geeks considering the no doubt low cost of this scale. But I still don't see the need for faster Z, as most smartphone users are on the terraferma?

maxerickson 14y ago

I think it only speeds up solving for position. Without an up to date local reference, the barometer can't tell you absolute altitude with any sort of precision. I think the basic idea is that instead of dealing with the entire circle that you get from intersecting two spheres, you only have to deal with the parts of the circle that are remotely sensible according to the barometer.

mikeash 14y ago

GPS provides far better altitude data than a barometer. GPS altitude fixes are typically accurate to about 20m. An uncalibrated barometer can be off by 300m or more, depending on what the weather happens to be like at that particular moment. I'm sure the barometer is useful for something, but providing better altitude data than GPS gives you is definitely Not It. Edit: I should note that if you calibrate the barometer to the local atmospheric pressure, it can be extremely accurate, down to perhaps 3m or so, if it's a good barometer (like the one usually called an "altimeter" in an airplane). But it's not likely that a cheap phone barometer would achieve that, and of course you need some way to get the local base atmospheric pressure to begin with.