The worst part is, if a new antibiotics is discovered, it might help you right now, but after a couple of year, because of over use(*), the bacteria will eventually evolve some resistance against it. So the next patient with the same kind of infection will be again in the same situation...
Maybe time to dust off alternative therapies, like phage therapy ? (**) Cue in citation of your favorite strategist (Churchill, Sun-Tzu, Machiavelli, etc.) commenting about the millennia-old proverb that the enemy of your enemy is your friend.
---
(*) : over-prescription, industrial/agricultural use, etc.
(**) : phage are like viruses but specialize in infecting bacteria. So phage therapy is basically curing your sickness, by making your sickness itself sick, with its own sickness, in a kind of pathogen-ception.
Essentially you can either use the same improvements to make the cars more efficient in terms of gas usage or you can make them have more total horsepower.
And according to TFA is that cars are getting overall more efficient exactly because of that. All the time that the driver isn't spending with the foot stomping on the gaz pedal (which isn't the majority of a normal comute) is actually time during which the car has better efficiency than it's fore runners.
Or, in other words : this efficiency vs. power has been made dynamic and varies according to driver's needs moment by moment.
Technology is gotten that smarter that the driver thinks he's buying a more powerful car, but 99% of time the car will function as an efficient one.
you will read that Tesla's workplace safety is actually 32% better than the industry average. From TFA: [Tesla's] record of safety incidents went from slightly above the industry average in late 2016, to a performance in the first few months of 2017 that was 32% better than average.
That was exactly what I was going to ask : how do these numbers compare with the rest of the industry ?
It's good that they are both : - making progress (as mentioned in the summary) - already better than the average.
i.e.: they are making good efforts.
(Compare the situation with Apple's Foxconn reports. They were within the industry average. So on one hand, this wasn't as dramatic as the news wanted to make it seem, on the other hand Apple wasn't putting any effort back then to make things better)
Not OP but I can easily discern the harmonic beat between ultrasonic and sonic combinations.
In practice : - actually not at all.
the "beat" frequency is calculated by the delta of the 2 frequencies. to be able to hear the "beat" you'd need a frequency that is clearly in the hearing domain. to make it beat with an inaudible ultra-sound you would need a frequency that is clearly above the hearing domain. (i.e.: We're not speaking 10'000 Hz vs 10'001 Hz)
a delta between such two frequencies is sure to gives a beat frequencies that is beyond the response time of the ear. i.e.: your inner ear labyrinth receptor won't be able to notice a sound going on and off that fast.
not to mention that the delta might end up being higher than the lower frequency at which point there's no really beat to be heard at all. (a beat frequency is usuabl <<sound frequency)
in theory : a beat is just a sound oscillating in volume over time (a type of tremolo if you wish). in this case you're not actually hearing utlra sound (you don't have receptors for that), you're hearing distorsions in the audible domain, for which you DO have a receptor. and similarly a sampling rate of 48kHz won't be able to code the ultra sound frequency (beyond nyquist frequency), but can clearly code an audible frequency whose amplitude varies over time.
but that's just about the theory of perceiving or recording a beat a.k.a. an osciliation of volume a.k.a. a tremolo. for the practicality of a setup with a beat between an audible and inaudible frequency, see above.
However, if you understand Fourier's theorm, then you understand you can only capture a 22kHz sine wave. The harmonics get cut off. Even 22kHz is rather high, but its there to get the second harmonic of 11kHz.
I'm not saying that higher frequencies than 22kHz to not exist. I'm just saying that we lack the receptor to even detect it. 33kHz could be emitted as a harmonic of a source tuned for 11kHz base note (though a 11kHz base note is quite high already and doesn't make much sens musically). But there's no sane reason to record it and reproduce it. You're just wasting storage space on something that you'll never be able to perceive.
Translated into the light domain : I'm not saying that X-Rays do not exist. I'm just saying that the human eye lack any receptor to detect them. If your purpose is only to store a photo, you're wasting storage space on storing the x-rays value of image pixels and then trying to reproduce it : your watchers will never be able to see it, they only see the visual domain.
(If you're a radiologist looking at X-ray picture, it's an entire different story. the sound domain equivalent is : Similar to ultra-sound sonars used by the parking assistance of cars)
but I'm going to be that pedantic twat and point out that some people can reportedly perceive UV light, particularly after cataract surgery. Claude Monet is an oft-cited example.
And I'm going to out-predant you be making you notice I didn't mention which band of UV (UV-A, UV-B, UV-C, etc.) the rods and cones can perceive UV down to 280nm (but they are blocked out by the len, so if you still have your natural born len, you don't see anything below 390nm).
You can't see 240-260nm UV, even with an artificial eye len, that's why you need a spectrometer to measure DNA content in a sample, you can't do it by naked eye.
UV extends down to 10nm.
Also UV wavelengths are dangerously damaging to the retina (that's why lens evolved to filter them). Modern artificial lens used in cataract operation do filter UV light for this exact reason. (Unlike back in Monet's era)
Browser support? Only relevant to web developers (users don't know or care what format the file is).
...they don't care. But most of the web apps they are using are switching or have already switched.
(a different exemple: Spotify doesn't use MP3. Spotify actually uses OGG/Vorbis. meaning that tons of people use OGG/Vorbis for their music collection instead of MP3 or AAC. Phasing the system's MP3 codec won't even affect them)
Android support? Might be nice, but without iOS support it's irrelevant as a mobile format.
OPUS is available as a library under a permissive license (same as Vorbis and FLAC, BTW).
Means that for the iOS version, most App can simply package "libopus.dynlib" (or "libvorbis.dynlib") even if Apple doesn't botters. (Luckily, means that now that the patent expired, a "liblame.dynlib" could be similarly possible)
Meanwhile, there is nothing to indicate OPUS will get wider traction than VORBIS did.
Nothing, except all the application currently installed on your smartphone that already support it and use it. (most of the voice chat applications, most of the media player, etc.)
If you've talked to someone or left a voice message in WhatsApp, Skype, Facebook Messenger (and countless others), your voice was transmitted using OPUS codec.
Face it, MP3 is still the superior format because no other format can replace it.
MP3 is currently widely available and supported. OPUS is currently widely available and supported, and has much better quality than anything available for internet use today. (The exception are non-internet use for 4kbps)
If you are starting an application today that has audio, guess which you are going to pick ? Current trends in apps show that OPUS is picked.
Too bad Apple refuses to even acknowledge the existence of AV-1,
Again, the other side has Netflix, Google and Amazon on board. (Pretty much nearly all content viewed on-line). Apple isn't even in a dominant position on this market.
So if they want to make themselves irrelevant, let them...
let alone talk about supporting it on Apple gear or in Apple software.
That's probably due to the fact that they *DO HAVE* H265 licenses, and they produce their own hardware (which is one of the few that keeps supporting H265 despite the current mayhem).
on the other hand, given the track record of Xiph (Vorbis, FLAC, Speex, OPUS, Theora, etc.), and of such IETF-aproved standards in general (Matroska), there's going to be beer/free libraries supporting AV-1 under permissive license (short BSD and similar) that can be included easily even in commercial apps.
In other words : - No, iPhone 8 & 9 will still not support AV-1 because Apple will still be trying to push H265/HEVC using their license into the GPU core of their "A" processors. - But yes, Netflix and Youtube apps on iPhones, iPad and AppleTV will stream AOMedia AV-1 none the less, simply by embeding their own copy of "libavone.dynlib" or whatever. (Same for Google Chrome, Mozilla Firefox and other browser apps following IETF standards).
MP3 will probably be supported on pretty every device for the foreseeable. There's literally no reason not to.
There's a very clear reason not to: Force users to rebuy the song they are used to listen to.
Armor piercing question: How is this a concern for the makers of MP3 PLAYERS and software that plays/encodes MP3s? Hint: It isn't...
Yes it *DOES*. I'm quoting 1 sentence further into my post:
If you happen to be in a vertically integrated market, where you both produce the playing device and content shop,
If the manufacturer of the music player and writer of software, happens to *ALSO* be the owner of the shop where the content is sold (and *USUALLY* also produce part of the content), they have a strong incentive to phase out support for the older formats which are not featured in the shop.
If said manufacturer has a strong market share and brand loyalty (specially if users of their shop are already locked-in by their already-bought music collection, e.g. due to DRM), they can leverage it to try to pull the trick.
Currently Apple, could be in a position where they could realistically try to phase out MP3, and pressure their users to re-buy their old tracks as AAC from iTunes. I'm not say that they *will*, just that they are in a position where *they could risk trying* even if they probably won't. (Probably because they are already busy phasing out analog 3mm jacks because... huh... "reasons ?"... in order to pressure their users to re-buy new Apple approved earphones with lightning port or bluetooth wireless).
Historically, Microsoft has been in a slightly different position, a tried to pull one reminiscent of this : Tried to build a music market - or at least a microsoft standard for music markets to use (remember the "Play for sure" era, before they tried it again with Zune ?), and indirectly tried to pressure hardware player manufacturer : the certification for "Play for sure" specified an explicit exhaustive list of the format that a device should support. They could not unlist MP3 - it was way too popular, would have been a commercial suicide. But OGG/Vorbis was suspiciously missing (and the certification term worded in such an ambiguous way, that supporting OGG/Vorbis might exclude from being "Plays for sure" certified) - probably because it was a format competing favorably in terms of quality back then.
(On other hand, no name Asian manufacturer kept support for Vorbis, they just didn't advertise it, or sometime even advertised both "Plays for sure" and support for Vorbis).
Where the show was designed by the actor's race and sex instead of a plot and a casting call.
On the other hand, even the original serie, from the beginning has tried hard to be inclusive (the communication officier was a african american woman, the navigator comes from the other side of the iron curtain, etc.)
So trying to feature under-represented minory is absolutely nothing new in Star Trek.
The only key question is : are these characters otherwise well written, and are the actors portraying them good ? (or are their "under represented minority" the only noticeable thing about them). but that's hard to judge without watching 1-2 episodes of the serie. (Which isn't available here around, at least not to me. So I can't judge)
AOMedia's AV-1 is attempting to be exactly that but for video.
A codec that is either patent free or whose patent are free. Free code implementation. supported by nearly anyone involved in video, including content providers (includes both Google and Netflix, so a sizeable portion of all video played), software makes (Mozilla, VLC, etc.) and hardware manufacturer (AMD, Intel, Broadcom, ARM, etc.)
And there are quite a few big developpers involved : - Xiph (makers of Daala), Google (VP10) and Cisco (thor)
It's going to do exactly what OPUS did to audio. probably within a year.
Until then it's either H264 / AVC if you can afford the patents or Google's VP-9. Stay away from H265 / HEVC, it's a trap (that was the actual incentive to start AOMedia)
Until something better takes the world by storm, the patent expiration will only help the format become more widely available.
Let me intoduce your to this thing called OPUS. (It's also by Xiph, the people behind Vorbis, but this time in collaboration with Skype).
It's patent-free, it's free. it's accepted as a IETF web-standard, it's supported by web browser. it's already used by lots of voice chat application (Skype - obviously - but also e.g.: WhatsApp) your smartphone probably already supports it (if it's a recent enough version of android). there are even informal standards to use it in Digital Radio Mondial (the digital cousin of AM radio).
And it has the best audio quality ever. Beats nearly anything else in ABX tests. Except maybe ultra small bandwidths ( 4 kbits ) which are beyond its intended usage anyway.
so it is taking the world by storm (chances are if an app on your smartphone deals with audio, it supports it) it's just a very silent storm.
No, every piece of music is lossy because analog cannot be encoded into digital without an infinite amount of loss.
Usually, we talk in terms of music.
we don't try to record every possible vibration in existance in the universe, we try to record *sound*. and the human body, due to limitation caused by laws of physics, has a very narrow set of vibration that it can hear and interpret as sound.
you can't hear ultrasound. there's no physiological way for you to hear them. thus there's no point in storing them.
There's a range of frequencies (tactile can feel up to dozens/hundrer of hertz, ears can feel up to somewhere between 10kHz and 20kHz). There's a range of intensities (between impossible to hear, and causes pain/hearing loss). By virtue of mathematics of information theory, every possible sound that you could ever hear can nicely fit within a 44kHz to 48kHz samplerate and 16bits to 24bits sample size.
Everything beyond that is just overkill, you're not physically equipped to percieve it. (That would be like trying to see UV, X-Ray, etc.)
---
BTW: A piano is also bound by the laws of physics and the amount of different vibrations it can produce isn't infinite either.
To get anything reasonable, I get a maximum 4:1 compression ratio comparing lossless to lossy, 11:1 would be like listening to a concert through a cellular connection.
Try using OPUS at 96kbps (Also by Xiph, the same guys who made Vorbis - but in collaboration with Skype this time) Resulting quality is incredibly close to lossless.
(It's also patent-free, consieded a IETF web standard and probably already supported by your current smartphone)
Why not ? Because the same developpers - Xiph - this time working in tandem with Skype - brought OPUS to you. And that one busts nearly everyone else in ABX tests, is considered a IETF web standrad, and thus supported by most browsers and used by several on-line voice-chat apps (Skype, WhatsApp, etc.), there are even informal tools to support it inside Digital Radio Mondial (the digital cousin of AM radio), etc.
It's thus very likely to be supported by your smartphone (e.g.: recent versions of android do play OPUS) and will progressively see wider usage over time.
In other words, unless you're interested in very small bit rates ( 4 kbits , meaningless for the typical design of OPUS ) there is just no argument for not using it.
Surprisingly, more of the in-vehicle-infotainment computers support Ogg/Vorbis that you would have though. (Software comes for free, requirement to add it as yet another supported format are minimal and it adds a nice additionnal checkbox on the feature list).
It's not widespread among standard DIN-format radios, but the big screens with Satnav support it. (And stand-alone satnav too. It used to be the default audio format for Tomtom).
Also, most modern car IVI also feature an AUX-In jack and/or bluetooth audio link. So basically cars play anything that your smartphone supports.
Most mp3 are drm free. Especially if you ripped them yourself.
Yup. Exactly. And when they look at your collection, the RIAA feels deep disgust.
They would *definitely* prefer no to have to keep the support for your DRM-free MP3, but have manufacturers only produce players that exclusively only play some modern DRMed music files (probably AAC based).
MP3 will probably be supported on pretty every device for the foreseeable. There's literally no reason not to.
There's a very clear reason not to : Force users to rebuy the song they are used to listen to.
If you happen to be in a vertically integrated market, where you both produce the playing device and content shop, you a have a strong incentive to deprecate MP3 on some made-up reason (like low quality per bitrate in such an old format) (like all the modern bells and whistle that supposedly come with newer format), and produce devices that only play the modern format that is also used on your online shop. Thus lots of users won't have any easy way to play all their old collection of song (they'll need to convert their MP3 into whatever new format is the current one. Of course, you'll try to make the converter specially NOT user-friendly). The simplest for them would be to re-buy their song from your shop.
If you're only a small player in the field and you're the only one doing it, you're going to lose customers as they move to another manufacturer who won't deprecate their entire collections of old songs. If you're a major player, holding a huge chunk of the market, the "running away" is going to be a tiny bit more complicated. You have your custommers locked in, let's proceed to make sure that they only play stuff that comes in your DRM'ed format from your store, not whatever old MP3 file they happened to still have laying around.
In other words : Apple could be in a great position to deprecate MP3 files support, forcing all iGadgets users to buy their songs from their store.
Plus no-one who is going to see this in the theater is going to not do so now that they can see the movie 20 minutes at a time on their laptop.
And also, people who are interested in getting pirated copies of the movies will get one at the thousands other leaks/bootleg/screen cap/DVD-screener/Web-DL/whatever that will pop up in the following months. Even if Disney *did* decide to pay, and the blackmailer were "honest" and destroyed their copy, that wouldn't stop the countless others. It's just basically about one more extra download link at your usual download/dtreaming spot. It's just not worth for Disney to pay.
Really what happens is that a NAT router drops any packet that it can't figure out a destination for.
Nope. They drop any packet, because that's the default rule in the iptables (sidenote: anyone with a modern modem that uses netfilter ?) loaded into the linux kernel that runs on the MIPS (mostlikely) inside your modem/router. The rest are exceptions. On a NAT router the rules will be in the form 'if destination port is "6992", then replace destination ip with "192.168.2.13" and keep the packet'. On a regular IPv6 router the rules will be in the fromo 'if destination IP is ":81a6:3d0f:5025:9243:5660" and destition port is "6692" then keep the packet'. But the rest, on any sane modem implementation doesn't even leave the iptables rules.
It's kind of like the post office... they don't deliver mail for which they can't figure out the destination address.
Nope. To keep the metaphore: - security works because you have a post-office to begin with (filtering capabilities, thanks to iptables). - the default for that post-office is to burn with a flamethrower anything that doesn't match known names and/or street numbers. - the question of whether the destination address exists or not will never come, because most of the mail will have already been burnt before hand, on the grounds on not being on the list of allowed name or allowed street numbers. They never reach the postman's backpack / they're never scheduled for delivery.
It seems like a pedantic point, but it becomes important when you talk about IPv6. Computers behind NAT are protected because they don't actually exist on the Internet.
Nope. They are protected because the metaphorical post office is trained to burn mail my default. Any sane router drops incoming traffic by default. Even if they had public addresses, that could be reached from anywhere on the internet, they would be still protected because the default rules is to drop any packet that didn't get authorised by another rule. The fact that the address needs to be rewritten is just icing on the cake.
They can only be reached via special tricks, and those tricks have to be implemented for the thing to work.
Those "special tricks" are just "yet another entry in the iptables". In the very special sub-case of IPv4 behind a NAT, it happens that the rule also needs to rewrite the destination address. But that's about it. Aside from that smal detail, everything is the same, including with IPv4 and public addresses. Including IPv6 and public addresses. You could even imagine including IPv6 and private addresses, but it's not worth the hassle.
On the other hand, a device with an IPv6 address DOES exist on the internet.
Again : so what ? The question of whether the IP is publicly addressable or not is completely perpendicular to the quesiton of whether the current ruleset will allow the packet through, or not. (by default : it's not)
Unfortunately, protecting the IPv6 devices isn't as simple as just building routers that use the same filter as NAT routers do,
Yes, it is, as proven by nearly any modern sane router (random example : AVM's Fritz routers).
because NAT routers don't use a filter.
Wha.... ? Here's an incredible surprise for you : nearly every modern modem/router runs a Linux kernel. iptables are default feature of the linux kernel iptables are necessarily present in modem/routers. iptables ARE USED because you need it to put the default "masquarade out-going traffic rule". iptables ARE USED because you need them to put additional port-forwarding rules.
You could build an IPv6 router that imitates what a NAT router does,
Manufacturer DO ALREADY build IPv6 routers that do what NAT router do and drop by default anything incoming, exce
It is now common practice to do a full memory reset just before any standardized examination.
I'm wondering : Nowadays with extremely powerful (relatively to calculators) CPUs available in very small form factors - has anyone attempted to built one (say a RaspPi Zero) inside a calculator sell, programmed to mimick the "exam mode memory reset" in a believable way, but then unleash all its potential to the end user (full blown programmability, ability to use modern math language like R, Octave, Maxima, etc. Scripting with Python/Perl)?
NAT routers don't filter.* Any incoming traffic * some also filter, but that's not really part of NAT
(Note: I was using "filter" in a very liberal way. Basically: they don't just pass blindly ethernet packets around as a hub/switch would. Technically, yes, NAT router don't pay as much attention to the source IP as they pay to the destination port, so the applied rules are a bit unusual).
But most modem with NAT I've seen have their router set to drop most their inbound connection, unless addressed to a port that was white-listed : -...manually by the modem webinterface (forward port "6992" to the machine running bittorrent) -...ask by a machine over UPnP (skype running on a laptop asks router to have a port forwarded to the laptop) -...answering the port of an out-bound UDP request (so either a live video chat, or as part of a STUN firewall whole punching instead of classic UPnP) -...as part of an out-going TCP request (the answer of a HTTP request) -...as requested by a special protocol (as part of the in-bound TCP data channel in an FTP session, as specified in the port command)
Such modem systematically check any incoming packet for the destination port, and will block or forward it depending on the destination port. (= so they look pretty close to the filtering work done by a classic firewall, except for the "not having to care the destination IP" part) and in addition will remap the IP addresses (that's the extra part that NAT adds to the top of a regular firewall).
But again, the security brought by modern modems comes from the fact that they decide to drop or allow inbound traffic based on rules. The IP remapping is additional mumbo-jumbo necessary to circumvent the limited amound of public IPs.
TL;DR: it's secure because the Modem's embed Linux' iptables. The fact that you can use a private IP range is just bonus.
Otherwise it responds, or doesn't, to traffic addressed to it, just like any other machine would.
I haven't seen a modern modem that responds to traffic it self by default. All I've seen have: drop or forward rules (some forward rules built on the fly other manually set as mentioned above) It usually takes special settings to actually have the router itself respond to external traffic. (usually to allow the ISP to administer it).
Or in other words : all the inbound traffic gets into the modem's embed Linux' iptables ruleset, from that point onward it's either sent to other machine or dropped, but under normal circumstances (i.e.: default settings) it never reaches the embed Linux' network socket layer. The only exceptions are stuff like DHCPv4, RA for IPv6 and much more seldom telnet)
Slashdot Mirror
The worst part is, if a new antibiotics is discovered, it might help you right now, but after a couple of year, because of over use(*), the bacteria will eventually evolve some resistance against it. So the next patient with the same kind of infection will be again in the same situation...
Maybe time to dust off alternative therapies, like phage therapy ? (**)
Cue in citation of your favorite strategist (Churchill, Sun-Tzu, Machiavelli, etc.) commenting about the millennia-old proverb that the enemy of your enemy is your friend.
---
(*) : over-prescription, industrial/agricultural use, etc.
(**) : phage are like viruses but specialize in infecting bacteria. So phage therapy is basically curing your sickness, by making your sickness itself sick, with its own sickness, in a kind of pathogen-ception.
Newer cars likely weigh a lot more, because of all the safety & environmental regulations.
Actually *NOT*, according to TFA.
The opposite is actually observed : newer material used nowadays means cars are lighter.
Steel got replaced by aluminum and modern fibers.
Essentially you can either use the same improvements to make the cars more efficient in terms of gas usage or you can make them have more total horsepower.
And according to TFA is that cars are getting overall more efficient exactly because of that.
All the time that the driver isn't spending with the foot stomping on the gaz pedal (which isn't the majority of a normal comute) is actually time during which the car has better efficiency than it's fore runners.
Or, in other words : this efficiency vs. power has been made dynamic and varies according to driver's needs moment by moment.
Technology is gotten that smarter that the driver thinks he's buying a more powerful car, but 99% of time the car will function as an efficient one.
Woker union / Database Unions / C union :
Too bad I've already commented and can't "+1 Funny".
you will read that Tesla's workplace safety is actually 32% better than the industry average.
From TFA: [Tesla's] record of safety incidents went from slightly above the industry average in late 2016, to a performance in the first few months of 2017 that was 32% better than average.
That was exactly what I was going to ask :
how do these numbers compare with the rest of the industry ?
It's good that they are both :
- making progress (as mentioned in the summary)
- already better than the average.
i.e.: they are making good efforts.
(Compare the situation with Apple's Foxconn reports. They were within the industry average. So on one hand, this wasn't as dramatic as the news wanted to make it seem, on the other hand Apple wasn't putting any effort back then to make things better)
Not OP but I can easily discern the harmonic beat between ultrasonic and sonic combinations.
In practice :
- actually not at all.
the "beat" frequency is calculated by the delta of the 2 frequencies.
to be able to hear the "beat" you'd need a frequency that is clearly in the hearing domain.
to make it beat with an inaudible ultra-sound you would need a frequency that is clearly above the hearing domain.
(i.e.: We're not speaking 10'000 Hz vs 10'001 Hz)
a delta between such two frequencies is sure to gives a beat frequencies that is beyond the response time of the ear.
i.e.: your inner ear labyrinth receptor won't be able to notice a sound going on and off that fast.
not to mention that the delta might end up being higher than the lower frequency at which point there's no really beat to be heard at all.
(a beat frequency is usuabl <<sound frequency)
in theory :
a beat is just a sound oscillating in volume over time (a type of tremolo if you wish).
in this case you're not actually hearing utlra sound (you don't have receptors for that), you're hearing distorsions in the audible domain, for which you DO have a receptor.
and similarly a sampling rate of 48kHz won't be able to code the ultra sound frequency (beyond nyquist frequency), but can clearly code an audible frequency whose amplitude varies over time.
but that's just about the theory of perceiving or recording a beat a.k.a. an osciliation of volume a.k.a. a tremolo.
for the practicality of a setup with a beat between an audible and inaudible frequency, see above.
for more informations :
- watch this video
- read this wiki
However, if you understand Fourier's theorm, then you understand you can only capture a 22kHz sine wave. The harmonics get cut off. Even 22kHz is rather high, but its there to get the second harmonic of 11kHz.
I'm not saying that higher frequencies than 22kHz to not exist.
I'm just saying that we lack the receptor to even detect it.
33kHz could be emitted as a harmonic of a source tuned for 11kHz base note (though a 11kHz base note is quite high already and doesn't make much sens musically).
But there's no sane reason to record it and reproduce it.
You're just wasting storage space on something that you'll never be able to perceive.
Translated into the light domain :
I'm not saying that X-Rays do not exist.
I'm just saying that the human eye lack any receptor to detect them.
If your purpose is only to store a photo, you're wasting storage space on storing the x-rays value of image pixels and then trying to reproduce it :
your watchers will never be able to see it, they only see the visual domain.
(If you're a radiologist looking at X-ray picture, it's an entire different story.
the sound domain equivalent is :
Similar to ultra-sound sonars used by the parking assistance of cars)
but I'm going to be that pedantic twat and point out that some people can reportedly perceive UV light, particularly after cataract surgery. Claude Monet is an oft-cited example.
And I'm going to out-predant you be making you notice I didn't mention which band of UV (UV-A, UV-B, UV-C, etc.)
the rods and cones can perceive UV down to 280nm (but they are blocked out by the len, so if you still have your natural born len, you don't see anything below 390nm).
You can't see 240-260nm UV, even with an artificial eye len, that's why you need a spectrometer to measure DNA content in a sample, you can't do it by naked eye.
UV extends down to 10nm.
Also UV wavelengths are dangerously damaging to the retina (that's why lens evolved to filter them).
Modern artificial lens used in cataract operation do filter UV light for this exact reason.
(Unlike back in Monet's era)
Browser support? Only relevant to web developers (users don't know or care what format the file is).
...they don't care. But most of the web apps they are using are switching or have already switched.
(a different exemple: Spotify doesn't use MP3. Spotify actually uses OGG/Vorbis.
meaning that tons of people use OGG/Vorbis for their music collection instead of MP3 or AAC.
Phasing the system's MP3 codec won't even affect them)
Android support? Might be nice, but without iOS support it's irrelevant as a mobile format.
OPUS is available as a library under a permissive license (same as Vorbis and FLAC, BTW).
Means that for the iOS version, most App can simply package "libopus.dynlib" (or "libvorbis.dynlib") even if Apple doesn't botters.
(Luckily, means that now that the patent expired, a "liblame.dynlib" could be similarly possible)
Meanwhile, there is nothing to indicate OPUS will get wider traction than VORBIS did.
Nothing, except all the application currently installed on your smartphone that already support it and use it.
(most of the voice chat applications, most of the media player, etc.)
If you've talked to someone or left a voice message in WhatsApp, Skype, Facebook Messenger (and countless others), your voice was transmitted using OPUS codec.
Face it, MP3 is still the superior format because no other format can replace it.
MP3 is currently widely available and supported.
OPUS is currently widely available and supported, and has much better quality than anything available for internet use today.
(The exception are non-internet use for 4kbps)
If you are starting an application today that has audio, guess which you are going to pick ?
Current trends in apps show that OPUS is picked.
Too bad Apple refuses to even acknowledge the existence of AV-1,
Again, the other side has Netflix, Google and Amazon on board. (Pretty much nearly all content viewed on-line).
Apple isn't even in a dominant position on this market.
So if they want to make themselves irrelevant, let them...
let alone talk about supporting it on Apple gear or in Apple software.
That's probably due to the fact that they *DO HAVE* H265 licenses,
and they produce their own hardware (which is one of the few that keeps supporting H265 despite the current mayhem).
on the other hand, given the track record of Xiph (Vorbis, FLAC, Speex, OPUS, Theora, etc.), and of such IETF-aproved standards in general (Matroska),
there's going to be beer/free libraries supporting AV-1 under permissive license (short BSD and similar) that can be included easily even in commercial apps.
In other words :
- No, iPhone 8 & 9 will still not support AV-1 because Apple will still be trying to push H265/HEVC using their license into the GPU core of their "A" processors.
- But yes, Netflix and Youtube apps on iPhones, iPad and AppleTV will stream AOMedia AV-1 none the less, simply by embeding their own copy of "libavone.dynlib" or whatever. (Same for Google Chrome, Mozilla Firefox and other browser apps following IETF standards).
MP3 will probably be supported on pretty every device for the foreseeable. There's literally no reason not to.
There's a very clear reason not to :
Force users to rebuy the song they are used to listen to.
Armor piercing question: How is this a concern for the makers of MP3 PLAYERS and software that plays/encodes MP3s?
Hint: It isn't...
Yes it *DOES*. :
I'm quoting 1 sentence further into my post
If you happen to be in a vertically integrated market, where you both produce the playing device and content shop,
If the manufacturer of the music player and writer of software, happens to *ALSO* be the owner of the shop where the content is sold (and *USUALLY* also produce part of the content),
they have a strong incentive to phase out support for the older formats which are not featured in the shop.
If said manufacturer has a strong market share and brand loyalty (specially if users of their shop are already locked-in by their already-bought music collection, e.g. due to DRM), they can leverage it to try to pull the trick.
Currently Apple, could be in a position where they could realistically try to phase out MP3, and pressure their users to re-buy their old tracks as AAC from iTunes.
I'm not say that they *will*, just that they are in a position where *they could risk trying* even if they probably won't.
(Probably because they are already busy phasing out analog 3mm jacks because... huh... "reasons ?"... in order to pressure their users to re-buy new Apple approved earphones with lightning port or bluetooth wireless).
Historically, Microsoft has been in a slightly different position, a tried to pull one reminiscent of this :
Tried to build a music market - or at least a microsoft standard for music markets to use (remember the "Play for sure" era, before they tried it again with Zune ?), and indirectly tried to pressure hardware player manufacturer :
the certification for "Play for sure" specified an explicit exhaustive list of the format that a device should support.
They could not unlist MP3 - it was way too popular, would have been a commercial suicide.
But OGG/Vorbis was suspiciously missing (and the certification term worded in such an ambiguous way, that supporting OGG/Vorbis might exclude from being "Plays for sure" certified) - probably because it was a format competing favorably in terms of quality back then.
(On other hand, no name Asian manufacturer kept support for Vorbis, they just didn't advertise it, or sometime even advertised both "Plays for sure" and support for Vorbis).
Where the show was designed by the actor's race and sex instead of a plot and a casting call.
On the other hand, even the original serie, from the beginning has tried hard to be inclusive (the communication officier was a african american woman, the navigator comes from the other side of the iron curtain, etc.)
So trying to feature under-represented minory is absolutely nothing new in Star Trek.
The only key question is : are these characters otherwise well written, and are the actors portraying them good ? (or are their "under represented minority" the only noticeable thing about them). but that's hard to judge without watching 1-2 episodes of the serie.
(Which isn't available here around, at least not to me. So I can't judge)
AOMedia's AV-1 is attempting to be exactly that but for video.
A codec that is either patent free or whose patent are free.
Free code implementation.
supported by nearly anyone involved in video, including content providers (includes both Google and Netflix, so a sizeable portion of all video played), software makes (Mozilla, VLC, etc.) and hardware manufacturer (AMD, Intel, Broadcom, ARM, etc.)
And there are quite a few big developpers involved :
- Xiph (makers of Daala), Google (VP10) and Cisco (thor)
It's going to do exactly what OPUS did to audio.
probably within a year.
Until then it's either H264 / AVC if you can afford the patents or Google's VP-9.
Stay away from H265 / HEVC, it's a trap (that was the actual incentive to start AOMedia)
Vorbis, which seems to be better in (almost?) all ways except adoption.
Luckily, it's successor OPUS is better at this "adoption" part.
Until something better takes the world by storm, the patent expiration will only help the format become more widely available.
Let me intoduce your to this thing called OPUS.
(It's also by Xiph, the people behind Vorbis, but this time in collaboration with Skype).
It's patent-free, it's free.
it's accepted as a IETF web-standard, it's supported by web browser.
it's already used by lots of voice chat application (Skype - obviously - but also e.g.: WhatsApp)
your smartphone probably already supports it (if it's a recent enough version of android).
there are even informal standards to use it in Digital Radio Mondial (the digital cousin of AM radio).
And it has the best audio quality ever.
Beats nearly anything else in ABX tests.
Except maybe ultra small bandwidths ( 4 kbits ) which are beyond its intended usage anyway.
so it is taking the world by storm (chances are if an app on your smartphone deals with audio, it supports it)
it's just a very silent storm.
No, every piece of music is lossy because analog cannot be encoded into digital without an infinite amount of loss.
Usually, we talk in terms of music.
we don't try to record every possible vibration in existance in the universe, we try to record *sound*.
and the human body, due to limitation caused by laws of physics, has a very narrow set of vibration that it can hear and interpret as sound.
you can't hear ultrasound. there's no physiological way for you to hear them. thus there's no point in storing them.
There's a range of frequencies (tactile can feel up to dozens/hundrer of hertz, ears can feel up to somewhere between 10kHz and 20kHz).
There's a range of intensities (between impossible to hear, and causes pain/hearing loss).
By virtue of mathematics of information theory, every possible sound that you could ever hear can nicely fit within a 44kHz to 48kHz samplerate and 16bits to 24bits sample size.
Everything beyond that is just overkill, you're not physically equipped to percieve it. (That would be like trying to see UV, X-Ray, etc.)
---
BTW: A piano is also bound by the laws of physics and the amount of different vibrations it can produce isn't infinite either.
To get anything reasonable, I get a maximum 4:1 compression ratio comparing lossless to lossy, 11:1 would be like listening to a concert through a cellular connection.
Try using OPUS at 96kbps
(Also by Xiph, the same guys who made Vorbis - but in collaboration with Skype this time)
Resulting quality is incredibly close to lossless.
(It's also patent-free, consieded a IETF web standard and probably already supported by your current smartphone)
It's technically superior so why not?
Why not ?
Because the same developpers - Xiph - this time working in tandem with Skype - brought OPUS to you.
And that one busts nearly everyone else in ABX tests, is considered a IETF web standrad, and thus supported by most browsers and used by several on-line voice-chat apps (Skype, WhatsApp, etc.), there are even informal tools to support it inside Digital Radio Mondial (the digital cousin of AM radio), etc.
It's thus very likely to be supported by your smartphone (e.g.: recent versions of android do play OPUS) and will progressively see wider usage over time.
In other words, unless you're interested in very small bit rates ( 4 kbits , meaningless for the typical design of OPUS ) there is just no argument for not using it.
A thumb drive with .ogg files won't.
Surprisingly, more of the in-vehicle-infotainment computers support Ogg/Vorbis that you would have though.
(Software comes for free, requirement to add it as yet another supported format are minimal and it adds a nice additionnal checkbox on the feature list).
It's not widespread among standard DIN-format radios, but the big screens with Satnav support it.
(And stand-alone satnav too. It used to be the default audio format for Tomtom).
Also, most modern car IVI also feature an AUX-In jack and/or bluetooth audio link.
So basically cars play anything that your smartphone supports.
Most mp3 are drm free. Especially if you ripped them yourself.
Yup. Exactly.
And when they look at your collection, the RIAA feels deep disgust.
They would *definitely* prefer no to have to keep the support for your DRM-free MP3,
but have manufacturers only produce players that exclusively only play some modern DRMed music files (probably AAC based).
MP3 will probably be supported on pretty every device for the foreseeable. There's literally no reason not to.
There's a very clear reason not to :
Force users to rebuy the song they are used to listen to.
If you happen to be in a vertically integrated market, where you both produce the playing device and content shop, you a have a strong incentive to deprecate MP3 on some made-up reason (like low quality per bitrate in such an old format) (like all the modern bells and whistle that supposedly come with newer format), and produce devices that only play the modern format that is also used on your online shop.
Thus lots of users won't have any easy way to play all their old collection of song (they'll need to convert their MP3 into whatever new format is the current one. Of course, you'll try to make the converter specially NOT user-friendly).
The simplest for them would be to re-buy their song from your shop.
If you're only a small player in the field and you're the only one doing it, you're going to lose customers as they move to another manufacturer who won't deprecate their entire collections of old songs.
If you're a major player, holding a huge chunk of the market, the "running away" is going to be a tiny bit more complicated. You have your custommers locked in, let's proceed to make sure that they only play stuff that comes in your DRM'ed format from your store, not whatever old MP3 file they happened to still have laying around.
In other words : Apple could be in a great position to deprecate MP3 files support, forcing all iGadgets users to buy their songs from their store.
Plus no-one who is going to see this in the theater is going to not do so now that they can see the movie 20 minutes at a time on their laptop.
And also, people who are interested in getting pirated copies of the movies will get one at the thousands other leaks/bootleg/screen cap/DVD-screener/Web-DL/whatever that will pop up in the following months.
Even if Disney *did* decide to pay, and the blackmailer were "honest" and destroyed their copy, that wouldn't stop the countless others.
It's just basically about one more extra download link at your usual download/dtreaming spot.
It's just not worth for Disney to pay.
Really what happens is that a NAT router drops any packet that it can't figure out a destination for.
Nope.
They drop any packet, because that's the default rule in the iptables (sidenote: anyone with a modern modem that uses netfilter ?)
loaded into the linux kernel that runs on the MIPS (mostlikely) inside your modem/router.
The rest are exceptions.
On a NAT router the rules will be in the form 'if destination port is "6992", then replace destination ip with "192.168.2.13" and keep the packet'.
On a regular IPv6 router the rules will be in the fromo 'if destination IP is ":81a6:3d0f:5025:9243:5660" and destition port is "6692" then keep the packet'.
But the rest, on any sane modem implementation doesn't even leave the iptables rules.
It's kind of like the post office... they don't deliver mail for which they can't figure out the destination address.
Nope. :
To keep the metaphore
- security works because you have a post-office to begin with (filtering capabilities, thanks to iptables).
- the default for that post-office is to burn with a flamethrower anything that doesn't match known names and/or street numbers.
- the question of whether the destination address exists or not will never come, because most of the mail will have already been burnt before hand, on the grounds on not being on the list of allowed name or allowed street numbers. They never reach the postman's backpack / they're never scheduled for delivery.
It seems like a pedantic point, but it becomes important when you talk about IPv6. Computers behind NAT are protected because they don't actually exist on the Internet.
Nope.
They are protected because the metaphorical post office is trained to burn mail my default. Any sane router drops incoming traffic by default.
Even if they had public addresses, that could be reached from anywhere on the internet, they would be still protected because the default rules is to drop any packet that didn't get authorised by another rule.
The fact that the address needs to be rewritten is just icing on the cake.
They can only be reached via special tricks, and those tricks have to be implemented for the thing to work.
Those "special tricks" are just "yet another entry in the iptables".
In the very special sub-case of IPv4 behind a NAT, it happens that the rule also needs to rewrite the destination address. But that's about it.
Aside from that smal detail, everything is the same, including with IPv4 and public addresses. Including IPv6 and public addresses.
You could even imagine including IPv6 and private addresses, but it's not worth the hassle.
On the other hand, a device with an IPv6 address DOES exist on the internet.
Again : so what ?
The question of whether the IP is publicly addressable or not is completely perpendicular to the quesiton of whether the current ruleset will allow the packet through, or not.
(by default : it's not)
Unfortunately, protecting the IPv6 devices isn't as simple as just building routers that use the same filter as NAT routers do,
Yes, it is, as proven by nearly any modern sane router (random example : AVM's Fritz routers).
because NAT routers don't use a filter.
Wha.... ?
Here's an incredible surprise for you : nearly every modern modem/router runs a Linux kernel.
iptables are default feature of the linux kernel
iptables are necessarily present in modem/routers.
iptables ARE USED because you need it to put the default "masquarade out-going traffic rule".
iptables ARE USED because you need them to put additional port-forwarding rules.
You could build an IPv6 router that imitates what a NAT router does,
Manufacturer DO ALREADY build IPv6 routers that do what NAT router do and drop by default anything incoming, exce
It is now common practice to do a full memory reset just before any standardized examination.
I'm wondering :
Nowadays with extremely powerful (relatively to calculators) CPUs available in very small form factors -
has anyone attempted to built one (say a RaspPi Zero) inside a calculator sell,
programmed to mimick the "exam mode memory reset" in a believable way, but then unleash all its potential to the end user (full blown programmability, ability to use modern math language like R, Octave, Maxima, etc. Scripting with Python/Perl)?
NAT routers don't filter.* Any incoming traffic
* some also filter, but that's not really part of NAT
(Note: I was using "filter" in a very liberal way. Basically: they don't just pass blindly ethernet packets around as a hub/switch would.
Technically, yes, NAT router don't pay as much attention to the source IP as they pay to the destination port, so the applied rules are a bit unusual).
But most modem with NAT I've seen have their router set to drop most their inbound connection, unless addressed to a port that was white-listed : ...manually by the modem webinterface (forward port "6992" to the machine running bittorrent) ...ask by a machine over UPnP (skype running on a laptop asks router to have a port forwarded to the laptop) ...answering the port of an out-bound UDP request (so either a live video chat, or as part of a STUN firewall whole punching instead of classic UPnP) ...as part of an out-going TCP request (the answer of a HTTP request) ...as requested by a special protocol (as part of the in-bound TCP data channel in an FTP session, as specified in the port command)
-
-
-
-
-
Such modem systematically check any incoming packet for the destination port, and will block or forward it depending on the destination port.
(= so they look pretty close to the filtering work done by a classic firewall, except for the "not having to care the destination IP" part)
and in addition will remap the IP addresses (that's the extra part that NAT adds to the top of a regular firewall).
But again, the security brought by modern modems comes from the fact that they decide to drop or allow inbound traffic based on rules.
The IP remapping is additional mumbo-jumbo necessary to circumvent the limited amound of public IPs.
TL;DR: it's secure because the Modem's embed Linux' iptables. The fact that you can use a private IP range is just bonus.
Otherwise it responds, or doesn't, to traffic addressed to it, just like any other machine would.
I haven't seen a modern modem that responds to traffic it self by default.
All I've seen have: drop or forward rules (some forward rules built on the fly other manually set as mentioned above)
It usually takes special settings to actually have the router itself respond to external traffic. (usually to allow the ISP to administer it).
Or in other words : all the inbound traffic gets into the modem's embed Linux' iptables ruleset, from that point onward it's either sent to other machine or dropped, but under normal circumstances (i.e.: default settings) it never reaches the embed Linux' network socket layer.
The only exceptions are stuff like DHCPv4, RA for IPv6 and much more seldom telnet)