I agree.
If ATI cannot get their act together, somebody needs to put them out of business. As word spreads from towns who haven't even tested their ATIs, it will spread to others looking to replace sirens. It will only put people's lives at risks.
-
va_nuke_pe
- Registered User
- Posts: 111
- Joined: Thu May 18, 2006 3:47 am
- Location: Herndon, Virginia
How about knowing what you are talking about folks!
First of all, San Francisco wanted an outdoor VOICE warning system. So, they put in devices that could act as loudspeakers as well as make siren like sound. Second, programming an electronic loudspeaker to make a sound based on an mpeg file isn't exactly the hardest thing to do. The people there were used to hearing STL-10's, so they made them sound like STL-10's. It's always a good idea to give an attention-grabbing signal before giving a voice message. Third, because of difficulties in zoning and permits, the City of San Francisco decided that they would put the voice system components in the same locations as the STL-10's they were replacing. That is the root cause for all of the problems they are having getting intelligible voice. Let me explain so that maybe it will stick in some of your heads.
General rule - whatever makes for a great siren system tends to make for a bad voice system, and vice-versa. Loud, omni-directional devices used for siren system - GOOD! Loud, omni-directional devices used for voice system - BAD! Lower volume and fixed directional horns for voice system - GOOD! Lower volume and fixed directional horns for siren system - BAD!
Siren tone is single frequency (or two frequencies at a pre-set ration to each other). Sound travels through the air at ~1,110 feet per second.
The critical parameter for a siren tone is audibility - can it be heard over the back ground noise in the frequency band it is sounding in. If I am standing between two sirens and I get the exact same tone from each of them, then it reinforces the signal, even if I hear the closer siren first (in the real world, sound cancellation due to one wave being exactly 180 degrees out of phase with the other is nearly impossible to sustain although theoretically possible). Further, most sirens sound in the 500 Hz octave band - which is the octave band that tends to have the lowest ambient noise levels. Man-made sounds tend to be in the 250 Hz and lower octave bands. Natural sounds tend to be in the 1000 Hz and higher octave bands.
Voice is multi-frequency and complex - most voice is in the 500 Hz to 4000 Hz octave bands. The critical parameter for voice is intelligibility - assuming the listener understands the language, then can the listener understand what is being said. Signal to noise ratio (loudness above background noise for the specific octave band) is only one factor in intelligibility - there are several others that are at least as important. Making a voice device too loud actually decreases intelligibility. Further, if I am standing between two voice devices (closer to one than the other), then the the voice message from the second will run over the voice message from the first and start to blur it. Now put a listener in the middle of an array of voice devices at various distances. It's called "multiple arrival effect." Like listening to groups singing a round of "row, row, row your boat" at irregularly spaced intervals.
So, what helps for sirens hurts for voice.
A voice system is far more complex to design than a siren system. Further, it is affected more by the built environment (urban high rise, low to mid-rise, residential, open space) than a siren system. Weather affects voice more as well - wind and atmospheric turbulence especially. The simpleton 10 dB per distance doubled rule-of-thumb that dates back to slide rule days barely works for sirens - it isn't even close to describing what a voice system will perform like.
All you lovers of "hamburger stacks" - Federal Modulators and Whelen 2800/2900's are in for a rude surprise if you think they will work in a voice system array in a city environment. They make great point sources in open spaces where nothing else is nearby - but they are not the best choice for city scapes - fixed directional horns work best in those locations.
One more thing for all you electronics haters to consider - if you live in tornado alley and/or near a commercial nuclear plant - then a siren signal tells you what you need to know only because you have been reminded again and again what the siren signal means. These also tend to be where, let's face it folks, there do not tend to be a lot of visitors. Now let's go to a place that has a lot of visitors, say San Francisco. Visitors don't necessarily know what a siren signal is telling them to do - they may not even be able to read the language in the hotel brochure or understand the voice over the radio. Voice is the best way to alert and notify those folks what to do. And it is people who are outdoors who tend to be most at risk and have the least access to emergency alert methods. In those circumstances, voice warning in the most common languages, is the most effective way of reaching those people.
General rule - whatever makes for a great siren system tends to make for a bad voice system, and vice-versa. Loud, omni-directional devices used for siren system - GOOD! Loud, omni-directional devices used for voice system - BAD! Lower volume and fixed directional horns for voice system - GOOD! Lower volume and fixed directional horns for siren system - BAD!
Siren tone is single frequency (or two frequencies at a pre-set ration to each other). Sound travels through the air at ~1,110 feet per second.
The critical parameter for a siren tone is audibility - can it be heard over the back ground noise in the frequency band it is sounding in. If I am standing between two sirens and I get the exact same tone from each of them, then it reinforces the signal, even if I hear the closer siren first (in the real world, sound cancellation due to one wave being exactly 180 degrees out of phase with the other is nearly impossible to sustain although theoretically possible). Further, most sirens sound in the 500 Hz octave band - which is the octave band that tends to have the lowest ambient noise levels. Man-made sounds tend to be in the 250 Hz and lower octave bands. Natural sounds tend to be in the 1000 Hz and higher octave bands.
Voice is multi-frequency and complex - most voice is in the 500 Hz to 4000 Hz octave bands. The critical parameter for voice is intelligibility - assuming the listener understands the language, then can the listener understand what is being said. Signal to noise ratio (loudness above background noise for the specific octave band) is only one factor in intelligibility - there are several others that are at least as important. Making a voice device too loud actually decreases intelligibility. Further, if I am standing between two voice devices (closer to one than the other), then the the voice message from the second will run over the voice message from the first and start to blur it. Now put a listener in the middle of an array of voice devices at various distances. It's called "multiple arrival effect." Like listening to groups singing a round of "row, row, row your boat" at irregularly spaced intervals.
So, what helps for sirens hurts for voice.
A voice system is far more complex to design than a siren system. Further, it is affected more by the built environment (urban high rise, low to mid-rise, residential, open space) than a siren system. Weather affects voice more as well - wind and atmospheric turbulence especially. The simpleton 10 dB per distance doubled rule-of-thumb that dates back to slide rule days barely works for sirens - it isn't even close to describing what a voice system will perform like.
All you lovers of "hamburger stacks" - Federal Modulators and Whelen 2800/2900's are in for a rude surprise if you think they will work in a voice system array in a city environment. They make great point sources in open spaces where nothing else is nearby - but they are not the best choice for city scapes - fixed directional horns work best in those locations.
One more thing for all you electronics haters to consider - if you live in tornado alley and/or near a commercial nuclear plant - then a siren signal tells you what you need to know only because you have been reminded again and again what the siren signal means. These also tend to be where, let's face it folks, there do not tend to be a lot of visitors. Now let's go to a place that has a lot of visitors, say San Francisco. Visitors don't necessarily know what a siren signal is telling them to do - they may not even be able to read the language in the hotel brochure or understand the voice over the radio. Voice is the best way to alert and notify those folks what to do. And it is people who are outdoors who tend to be most at risk and have the least access to emergency alert methods. In those circumstances, voice warning in the most common languages, is the most effective way of reaching those people.
-
MattDean1003
- Registered User
- Posts: 1216
- Joined: Sun May 28, 2006 12:47 am
- Real Name: James Dean
- Location: Bogart, GA
Here's a Google video with the sirens going off in the background.
http://video.google.com/videoplay?docid ... iren&hl=en
-Matt Dean-
http://video.google.com/videoplay?docid ... iren&hl=en
-Matt Dean-
Return to “Main Outdoor Warning Sirens Board”
Who is online
Users browsing this forum: No registered users and 2 guests