How Do You Rate Wireless Mic Technology? [Archive]

Les Cargill[_2_]

September 19th 09, 02:30 AM

Soundhaspriority wrote:
> "Randy Yates" > wrote in message
> ...
>> What are folks' opinions on current wireless mic technology? Is
>> it "perfect"? If not, then where can improvement be made? Cost?
>> More link margin (e.g., to avoid dropouts)? Battery time?
>> --
>> Randy Yates % "Watching all the days go by...
>> Digital Signal Labs % Who are you and who am I?"
>> % 'Mission (A World Record)',
>> http://www.digitalsignallabs.com % *A New World Record*, ELO
>
> In replying, the following remarks are intended to apply to production sound
> wireless, not venue wireless.
>
> My particular frustration is the slowness with which computer related
> microwave technology has trickled in. Only a single company, Zaxcom, has an
> all digital system that is in wide use, with transparency approaching wired
> sound. Digital systems all seem to be mostly in-house engineered, lacking
> the economies of scale provided by computer networking. A Bluetooth
> transceiver chipset costs about a buck, but the wireless companies choose to
> reinvent the wheel at 1000X the cost.
>
> So I say to you, Randy, go engineer something based on cheap Bluetooth 2.1
> building blocks, make it sound great, and get it to us cheap. Use software
> to fix the delay time value, and code it into the audio stream, so it can be
> timeshifted in post.
>
> You'll make a bundle.
>

Once you get the drivers written... also, Bluetooth has a published
range of 10m - I have a wireless mouse that's useless past three
feet...

Once there is wireless USB for-real...

> Bob Morein
> (310) 237-6511
>
>

--
Les Cargill

Arkansan Raider

September 19th 09, 02:48 AM

Soundhaspriority wrote:
> "Les Cargill" > wrote in message
> ...
> [snip]
>> Once you get the drivers written... also, Bluetooth has a published
>> range of 10m - I have a wireless mouse that's useless past three
>> feet...
>>
> That's because your wireless mouse is either a Class 3 device, or a poorly
> implemented Class 2 device.
>
> 10m is the spec for a Class 2 Bluetooth device. Class 1 devices are specced
> for 100m, but in fact, there is no obstacle to longer range, even though
> there is no class for such range. Two extended range Bluetooth devices I am
> familiar with are the Trendnet TBW-104UB, and the AirCable adapter,
> http://www.merlinsystemscorp.co.uk/msc/air-cable-host-xr.html ,
> which is capable of miles with any antenna good for 802.11b/g, such as a
> cantenna.
>
> Bob Morein
> (310) 237-6511
>
>

Wow.

I'd never heard of that. Holy crap---they're claiming 10km as a class!

That's friggin' crazy. Wouldn't the FCC be after your butt for using that?

---Jeff

Randy Yates

September 19th 09, 02:52 AM

"Soundhaspriority" > writes:

> In replying, the following remarks are intended to apply to production sound
> wireless, not venue wireless.

Bob, what do you mean by the terms "production" and "venue"?
--
Randy Yates % "Maybe one day I'll feel her cold embrace,
Digital Signal Labs % and kiss her interface,
% til then, I'll leave her alone."
http://www.digitalsignallabs.com % 'Yours Truly, 2095', *Time*, ELO

Randy Yates

September 19th 09, 02:55 AM

"Soundhaspriority" > writes:

> "Mike Rivers" > wrote in message
> ...
>> Soundhaspriority wrote:
>>
>>> So I say to you, Randy, go engineer something based on cheap Bluetooth
>>> 2.1 building blocks, make it sound great, and get it to us cheap.
>>
>> Bluetooth - OK
>> Cheap - OK
>> Sound great - doubtful given the data reduction used in Bluetooth.
>
> Mike, there is no data reduction. Bluetooth data protocol is bit-for bit at
> 2.1 megabits/second.

I'm not sure what you mean by that, but there is indeed FEC, apparently
1/3 and 2/3 codes depending on conditions. Apparently the original BT
had a physical rate of 1 Mb/s, and the "EDA" allows 2 Mb/s and 3 Mb/s.
--
Randy Yates % "How's life on earth?
Digital Signal Labs % ... What is it worth?"
% 'Mission (A World Record)',
http://www.digitalsignallabs.com % *A New World Record*, ELO

Les Cargill[_2_]

September 19th 09, 03:28 AM

Randy Yates

September 19th 09, 03:43 AM

"Soundhaspriority" > writes:

> "Randy Yates" > wrote in message
> ...
>> "Soundhaspriority" > writes:
>>
>>> "Mike Rivers" > wrote in message
>>> ...
>>>> Soundhaspriority wrote:
>>>>
>>>>> So I say to you, Randy, go engineer something based on cheap Bluetooth
>>>>> 2.1 building blocks, make it sound great, and get it to us cheap.
>>>>
>>>> Bluetooth - OK
>>>> Cheap - OK
>>>> Sound great - doubtful given the data reduction used in Bluetooth.
>>>
>>> Mike, there is no data reduction. Bluetooth data protocol is bit-for bit
>>> at
>>> 2.1 megabits/second.
>>
>> I'm not sure what you mean by that, but there is indeed FEC, apparently
>> 1/3 and 2/3 codes depending on conditions. Apparently the original BT
>> had a physical rate of 1 Mb/s, and the "EDA" allows 2 Mb/s and 3 Mb/s.
>> --
> Well, sure, for the audio protocols that are part of the standard bluetooth
> stack. But you can start off with a clean slate. You don't need any of that.
> Data is data.

I'm really not connecting with you, Bob. Without some type of error
detection / correction scheme, you've got garbage, not audio. You can't
just use the raw, physical channel directly without EDC unless you have
an extreme amount of margin, and margin is usually very very expensive.
--
Randy Yates % "Remember the good old 1980's, when
Digital Signal Labs % things were so uncomplicated?"
% 'Ticket To The Moon'
http://www.digitalsignallabs.com % *Time*, Electric Light Orchestra

Arkansan Raider

September 19th 09, 06:29 AM

Richard Webb wrote:
> On Fri 2037-Sep-18 21:48, Arkansan Raider writes:
>> 10m is the spec for a Class 2 Bluetooth device. Class 1 devices are specced
>> for 100m, but in fact, there is no obstacle to longer range, even though
>> there is no class for such range. Two extended range Bluetooth devices I am
>> familiar with are the Trendnet TBW-104UB, and the AirCable adapter,
>> http://www.merlinsystemscorp.co.uk/msc/air-cable-host-xr.html ,
>> which is capable of miles with any antenna good for 802.11b/g, such as a
>> cantenna.
>> Wow.
>> I'd never heard of that. Holy crap---they're claiming 10km as a
>> class!
>
> ABout the only way you could do 10km at those frequencies
> would be a very efficient antenna, in the clear, for both
> transmit and receive. Both antennas would need quite a bit
> of gain.
>
> Still think with that much effective radiated power (thanks
> to antenna gain) the fcc would be all over their butts.
>
> Btw did you ever see my reply to you in another ng on
> adaptive vu metering JEff? I did reply to that comment of
> yours, but never saw it echoed back to me as is normal.
>
>
>
> Regards,
> Richard
> --
> | Remove .my.foot for email
> | via Waldo's Place USA Fidonet<->Internet Gateway Site
> | Standard disclaimer: The views of this user are strictly his own.

That's exactly what I was thinking about the FCC.

I don't remember seeing the adaptive VU metering reply--but then I may
be having a senior moment at 42 (it happens). You're talking about the
pro sound NG, right?

For some reason, once I plonked a couple of church/sound/FOH/idiot
threads, it seemed like a wasteland there--like tumbleweeds blowing
through. I don't see much in there anymore. Funny, I didn't even plonk
the posters, just the problem threads.

Try it again?

---Jeff

---Jeff

Arkansan Raider

September 19th 09, 06:33 AM

Richard Webb wrote:

> Btw did you ever see my reply to you in another ng on
> adaptive vu metering JEff? I did reply to that comment of
> yours, but never saw it echoed back to me as is normal.
>

Richard, if it means anything, I'm seeing your post as having been
posted at 2:06am, and it's 12:32am right now. Maybe some ISP glitches or
something out your way? Or maybe mine's acting up?

---Jeff

Richard Webb[_3_]

September 19th 09, 08:06 AM

On Fri 2037-Sep-18 21:48, Arkansan Raider writes:
> 10m is the spec for a Class 2 Bluetooth device. Class 1 devices are specced
> for 100m, but in fact, there is no obstacle to longer range, even though
> there is no class for such range. Two extended range Bluetooth devices I am
> familiar with are the Trendnet TBW-104UB, and the AirCable adapter,
> http://www.merlinsystemscorp.co.uk/msc/air-cable-host-xr.html ,
> which is capable of miles with any antenna good for 802.11b/g, such as a
> cantenna.
> Wow.
> I'd never heard of that. Holy crap---they're claiming 10km as a
> class!

ABout the only way you could do 10km at those frequencies
would be a very efficient antenna, in the clear, for both
transmit and receive. Both antennas would need quite a bit
of gain.

Still think with that much effective radiated power (thanks
to antenna gain) the fcc would be all over their butts.

Btw did you ever see my reply to you in another ng on
adaptive vu metering JEff? I did reply to that comment of
yours, but never saw it echoed back to me as is normal.

Regards,
Richard
--
| Remove .my.foot for email
| via Waldo's Place USA Fidonet<->Internet Gateway Site
| Standard disclaimer: The views of this user are strictly his own.

Randy Yates

September 19th 09, 11:51 AM

"Soundhaspriority" > writes:

>>>>> Mike, there is no data reduction. Bluetooth data protocol is bit-for
>>>>> bit
>>>>> at
>>>>> 2.1 megabits/second.

> [...]

> Randy, I'm not talking about the raw channel. Use the Bluetooth data
> protocol.

Bob,

Given that the maximum raw data rate is 3 Mb/s and, depending on channel
conditions, coding may be at a rate 1/3 code, I don't see how you could
get (reliably) anything more than 1 Mb/s.
--
Randy Yates % "My Shangri-la has gone away, fading like
Digital Signal Labs % the Beatles on 'Hey Jude'"
%
http://www.digitalsignallabs.com % 'Shangri-La', *A New World Record*, ELO

Scott Dorsey

September 19th 09, 12:53 PM

In article >, Randy Yates > wrote:
>"Soundhaspriority" > writes:
>> "Mike Rivers" > wrote in message
>> ...
>>> Soundhaspriority wrote:
>>>
>>>> So I say to you, Randy, go engineer something based on cheap Bluetooth
>>>> 2.1 building blocks, make it sound great, and get it to us cheap.
>>>
>>> Bluetooth - OK
>>> Cheap - OK
>>> Sound great - doubtful given the data reduction used in Bluetooth.
>>
>> Mike, there is no data reduction. Bluetooth data protocol is bit-for bit at
>> 2.1 megabits/second.
>
>I'm not sure what you mean by that, but there is indeed FEC, apparently
>1/3 and 2/3 codes depending on conditions. Apparently the original BT
>had a physical rate of 1 Mb/s, and the "EDA" allows 2 Mb/s and 3 Mb/s.

The number one issue with current wireless systems is frequency contention
and interference.

Moving to the most crowded band in the spectrum does not help that problem,
it just makes it a lot worse.

In general, I think the added latency involved in the digital transmission
systems is a severe problem... and the more serial error correction you do,
the worse the latency is. What you want is a nice wide bandwidth path
on a frequency low enough that your body doesn't block too much of it,
but high enough that your antenna is easily concealed, a frequency that
nobody else is using.
--scott

--
"C'est un Nagra. C'est suisse, et tres, tres precis."

Scott Dorsey

September 19th 09, 01:03 PM

Richard Webb > wrote:
>
>ABout the only way you could do 10km at those frequencies
>would be a very efficient antenna, in the clear, for both
>transmit and receive. Both antennas would need quite a bit
>of gain.
>
>Still think with that much effective radiated power (thanks
>to antenna gain) the fcc would be all over their butts.

The ISM bands were never intended for communications work, they were
intended as convenient dumping grounds for RF pollution. The fact
that ANY of this 2.4 Ghz stuff works given the interference conditions
is something of a miracle. Consequently, the FCC regulations for these
bands are a little different than for bands intended for communications
work.

In fact, the FCC cares only about field strength on that band, not
input power. Take, for example, your microwave oven. 1KW input
power, less than a microvolt per meter emission. That's okay.

On the other hand, stuff like the Pringles can antenna does indeed
violate the ISM band regulations. The FCC doesn't care all that much
though, because as mentioned above, it's intended as a dumping ground.

Most of the ISM bands are this way although there are some slightly
goofy combination field strength and input power requirements on the
13 Mhz one, which are specifically designed to accomodate the needs of
plywood manufacturers.
--scott

--
"C'est un Nagra. C'est suisse, et tres, tres precis."

Richard Crowley

September 19th 09, 03:25 PM

Soundhaspriority wrote:
> "Randy Yates" wrote ...
>> I'm really not connecting with you, Bob. Without some type of error
>> detection / correction scheme, you've got garbage, not audio. You
>> can't just use the raw, physical channel directly without EDC unless
>> you have an extreme amount of margin, and margin is usually very
>> very expensive. --

> Randy, I'm not talking about the raw channel. Use the Bluetooth data
> protocol.

The raw channel is not reliable enough for venue or production audio.
There is a great deal of error detection and re-transmission going on
underneath that you don't see in the end result data stream. Buy the
time you use the data detection and correction (including re-transmission
of data packets) it is no longer real-time enough to use for much of
any kind of live audio work. THAT is why Zaxcom, et. al. had to
do so much of their own development.

Arkansan Raider

September 19th 09, 03:43 PM

Soundhaspriority wrote:
> "Arkansan Raider" > wrote in message
> ...
>> Soundhaspriority wrote:
>>> "Les Cargill" > wrote in message
>>> ...
>>> [snip]
>>>> Once you get the drivers written... also, Bluetooth has a published
>>>> range of 10m - I have a wireless mouse that's useless past three
>>>> feet...
>>>>
>>> That's because your wireless mouse is either a Class 3 device, or a
>>> poorly implemented Class 2 device.
>>>
>>> 10m is the spec for a Class 2 Bluetooth device. Class 1 devices are
>>> specced for 100m, but in fact, there is no obstacle to longer range, even
>>> though there is no class for such range. Two extended range Bluetooth
>>> devices I am familiar with are the Trendnet TBW-104UB, and the AirCable
>>> adapter, http://www.merlinsystemscorp.co.uk/msc/air-cable-host-xr.html ,
>>> which is capable of miles with any antenna good for 802.11b/g, such as a
>>> cantenna.
>>>
>>> Bob Morein
>>> (310) 237-6511
>> Wow.
>>
>> I'd never heard of that. Holy crap---they're claiming 10km as a class!
>>
>> That's friggin' crazy. Wouldn't the FCC be after your butt for using that?
>>
>>
>> ---Jeff
>
> Jeff, I don't think so, because it's in an unlicensed band shared with
> 802.11b/g. Folks have been doing extreme range with 802.11 in that band for
> more than a decade with no complaint by the FCC.
>
> So now you might be asking, with all that unlicensed goodness, what would it
> do to a Bluetooth audio link? From the beginning, Bluetooth had the problem
> of coexisting with 802.11 networks without getting clobbered. Beginning with
> protocol version 1.2, they had it solved. In the most common Bluetooth
> mode, the link jumps randomly around 79 channels like a grasshopper. It does
> this regardless of whether there is a collision. The whole thing is written
> so that if a hop lands in an occupied channel, it simply resends the data.
>
> Inj other mode of Bluetooth, used for maximum throughput, Bluetooth can sit
> on a vacant frequency. In a typical wireless mike scenario, the educated
> user could choose, analogous to the way he does a careful channel survey
> before committing.
>
> The cost of channel hopping is latency, but, methinks, it is not bad for
> production sound, where the delay is compensated for by user setting in
> modern field production recorders. It would not be acceptable at all for
> venue reinforcement, but there, a dedicated channel could be reasonably
> used. Still, for venue wireless, it appears that spread spectrum has the
> advantage, with the extremely low latencies possible.
>
> Bob Mreoin
> (310) 237-6511
>
>

Thanks for that very informative post. I just learned a bunch.

Now, I don't know what I'm gonna' *do* with all of that info, but it's
most definitely filed away for future reference.

---Jeff

Randy Yates

September 19th 09, 03:45 PM

(Scott Dorsey) writes:

> In article >, Randy Yates > wrote:
>>"Soundhaspriority" > writes:
>>> "Mike Rivers" > wrote in message
>>> ...
>>>> Soundhaspriority wrote:
>>>>
>>>>> So I say to you, Randy, go engineer something based on cheap Bluetooth
>>>>> 2.1 building blocks, make it sound great, and get it to us cheap.
>>>>
>>>> Bluetooth - OK
>>>> Cheap - OK
>>>> Sound great - doubtful given the data reduction used in Bluetooth.
>>>
>>> Mike, there is no data reduction. Bluetooth data protocol is bit-for bit at
>>> 2.1 megabits/second.
>>
>>I'm not sure what you mean by that, but there is indeed FEC, apparently
>>1/3 and 2/3 codes depending on conditions. Apparently the original BT
>>had a physical rate of 1 Mb/s, and the "EDA" allows 2 Mb/s and 3 Mb/s.
>
> The number one issue with current wireless systems is frequency contention
> and interference.

That's good information, Scott. Can you provide a couple of data points
to give me an idea of where your pain threshold is? I.e., what
performance are you currently seeing that causes you to make this
statement?

> Moving to the most crowded band in the spectrum does not help that problem,
> it just makes it a lot worse.

Notice that my original query never asked for solutions, just areas
where you'd like to see improvement.

> In general, I think the added latency involved in the digital
> transmission systems is a severe problem...

Specify your maximum latency and let the designer determine how to meet
it.

> and the more serial error correction you do, the worse the latency is.

What is "serial error correction?" I can't recall seeing that term in
any of my studies in coding or information theory.

> What you want is a nice wide bandwidth path on a frequency low enough
> that your body doesn't block too much of it, but high enough that your
> antenna is easily concealed, a frequency that nobody else is using.

Again, it's best to avoid the tendency to attempt to design the
system. Instead, state your requirements.

On thing I'm not clear on: for venue work, what is the maximum range
requirement? Also, what is the maximum number of channels required?
--
Randy Yates % "And all that I can do
Digital Signal Labs % is say I'm sorry,
% that's the way it goes..."
http://www.digitalsignallabs.com % Getting To The Point', *Balance of Power*, ELO

Randy Yates

September 19th 09, 04:28 PM

(Scott Dorsey) writes:

> In fact, the FCC cares only about field strength on that band, not
> input power. Take, for example, your microwave oven. 1KW input
> power, less than a microvolt per meter emission. That's okay.

This is absolutely incorrect by my reckoning. The reason microwave ovens
can operate the way they do is because they are exempted by section
15.103 in [fccpart15].

@manual{fccpart15,
title = "\href{http://www.digitalsignallabs.com/part15-9-20-07.pdf}{Part 15---Radio Frequency Devices}",
organization = "Federal Communications Commission"}

In several other places there are conducted output power limitations.

> On the other hand, stuff like the Pringles can antenna does indeed
> violate the ISM band regulations.

Perhaps, but even if it does, not as much you might think. You have to
reduce the conducted power by 1 dB for every 3 dB that the antenna gain
exceeds 6 dBi (section 15.247.c.1.i). Given that a home-brew antenna
like this probably may not have the highest quality connectors and
cabling, that 1 or 2 dB might be lost there.
--
Randy Yates % "She has an IQ of 1001, she has a jumpsuit
Digital Signal Labs % on, and she's also a telephone."
%
http://www.digitalsignallabs.com % 'Yours Truly, 2095', *Time*, ELO

Arkansan Raider

September 19th 09, 05:21 PM

Richard Webb wrote:
> On Sat 2037-Sep-19 01:33, Arkansan Raider writes:
>> Richard, if it means anything, I'm seeing your post as having been
>> posted at 2:06am, and it's 12:32am right now. Maybe some ISP
>> glitches or something out your way? Or maybe mine's acting up?
>
> NOpe, it's my system, which is on universal time. sO I
> indeed posted it at 2:06 A.M. utc. WE're both in same time
> zone. THe bbs machine though is on utc.
>
> Regards,
> Richard
> --
> | Remove .my.foot for email
> | via Waldo's Place USA Fidonet<->Internet Gateway Site
> | Standard disclaimer: The views of this user are strictly his own.

Ah, roger that.

Wait. Today is Talk Like A Pirate Day.

AAAAAAAAARRRRRRRRRRRRRRRRRR!!!!

Rrrrrrrrrrrrrroger that, Jolly!

Buckle that swash!

---Jeff

Scott Dorsey

September 19th 09, 06:57 PM

In article >, Randy Yates > wrote:
(Scott Dorsey) writes:
>>
>> The number one issue with current wireless systems is frequency contention
>> and interference.
>
>That's good information, Scott. Can you provide a couple of data points
>to give me an idea of where your pain threshold is? I.e., what
>performance are you currently seeing that causes you to make this
>statement?

Well, a lot of it has to do with users. But here are some examples of
things I have seen when folks don't coordinate:

1. Williamsburg Scottish Festival... guy from some other state brought
in two Samson wireless mikes. One worked fine, the other one did not.
S-meter on the one that didn't work registered signal but wasn't
calibrated so nobody knows how much.

2. Unnamed celtic festival.... TV crew comes in with a UHF wireless boom
mike. When they turn it on, the lead singer's wireless suddenly drops
out.

4. People trying to use VHF wireless at an airport, a few hundred feet
from a Unicom 122 Mhz transmitter running 30 watts. You'd think
this would be fine. It wasn't.

5. Two churches across the street from one another. When the wireless
mike in one church is turned off, you can hear the pastor across the
street.

And these are just stories from this summer.

>> Moving to the most crowded band in the spectrum does not help that problem,
>> it just makes it a lot worse.
>
>Notice that my original query never asked for solutions, just areas
>where you'd like to see improvement.

Oh, I'd like a system that sounds better, and is more reliable, and
requires no frequency coordination but whose frequency can be locked
down in the event it's working at a place where frequency coordination is
required. More than 20 mS of latency is bad. Also it should have
longer range and longer battery life and cost less. And I want a pony.

>> In general, I think the added latency involved in the digital
>> transmission systems is a severe problem...
>
>Specify your maximum latency and let the designer determine how to meet
>it.

I think 20mS is maybe okay for stage applications. For production sound
work you MIGHT be able to live with twice that... but if you can hear
the sound outside and then a discrete echo in the cans, that's bad. Even
if you can't, consider comb filtering issues.

>> and the more serial error correction you do, the worse the latency is.
>
>What is "serial error correction?" I can't recall seeing that term in
>any of my studies in coding or information theory.

If you're using some kind of trellis encoding thing where the signal is
being split up among multiple carriers, you can think about error correction
on a single carrier vs. error correction on the composite parallel channel.

Optimizing error correction, while at the same time having optimal
latency, while at the same time dealing with narrowband noise within
the receiver passband by dropping individual carriers out as needed,
while at the same time dynamically adjusting carrier strength in order
to optimize power use, is left as an exercise...

>> What you want is a nice wide bandwidth path on a frequency low enough
>> that your body doesn't block too much of it, but high enough that your
>> antenna is easily concealed, a frequency that nobody else is using.
>
>Again, it's best to avoid the tendency to attempt to design the
>system. Instead, state your requirements.

I'm just the engineer. All I care is that it works and nobody has
to worry about it. Ask the performers what THEY want. They want something
small that is easily concealable and they want something that they can
move around with constantly without any worries. I just want something
that acts like a cable.

>on thing I'm not clear on: for venue work, what is the maximum range
>requirement? Also, what is the maximum number of channels required?

I have never actually seen a gig using more than 200 channels, but I
wouldn't be surprised if one exists. Figure if you have more than ten
or twenty, you probably have a frequency coordinator assigned to the
gig to deal with it. Figure that you're sharing the air with lots of
other users who are doing video links and wireless links to IFB and to
delay speaker stacks and to green room monitors, and you have intercoms
and lots of 460 Mhz walkie-talkies. And you have police and fire people
standing by on site, with their radios. And you have broadband noise
sources like the laptop computer being used at FOB to play spots back.

And you have some users who know they are under frequency coordination
but they don't talk to the coordinator because they are afraid that
the coordinator will say no to them.

Also, look at places like broadway shows, where they may only be using
fifty channels or so, but they are next door to another theatre which
is also using fifty channels.

Maximum range.... depends a lot on the gig, but the important part is
knowing what the range is. If it's only 50 feet, that's usually fine,
just as long as you know what it is. I know a lot of systems that
seem like they have a 50 foot range in rehearsal, but then once the
audience comes in, suddenly they don't any more.

God, I hate wireless.
--scott

--
"C'est un Nagra. C'est suisse, et tres, tres precis."

Richard Webb[_3_]

September 19th 09, 07:57 PM

On Sat 2037-Sep-19 01:33, Arkansan Raider writes:
> Richard, if it means anything, I'm seeing your post as having been
> posted at 2:06am, and it's 12:32am right now. Maybe some ISP
> glitches or something out your way? Or maybe mine's acting up?

NOpe, it's my system, which is on universal time. sO I
indeed posted it at 2:06 A.M. utc. WE're both in same time
zone. THe bbs machine though is on utc.

Regards,
Richard
--
| Remove .my.foot for email
| via Waldo's Place USA Fidonet<->Internet Gateway Site
| Standard disclaimer: The views of this user are strictly his own.

Richard Webb[_3_]

September 19th 09, 09:03 PM

On Sat 2037-Sep-19 07:53, Scott Dorsey writes:
> The number one issue with current wireless systems is frequency
> contention and interference.

Agreed.

> Moving to the most crowded band in the spectrum does not help that
> problem, it just makes it a lot worse.

OF course, and even trying to share it with all these
wireless networks etc. is a recipe for disaster imho.

> In general, I think the added latency involved in the digital
> transmission systems is a severe problem... and the more serial
> error correction you do, the worse the latency is. What you want is
> a nice wide bandwidth path on a frequency low enough that your body
> doesn't block too much of it, but high enough that your antenna is
> easily concealed, a frequency that nobody else is using.
WOuld agree, but that's the tough part, finding a chunk of
spectrum where the antenna can be small enough that isn't
occupied.

The latency issues would cause me to avoid anything such as
Bob suggests, but most of my acquaintance with wireless is
for live production where such latency from all that digital processing and error correction would be unacceptable under
any circumstance. ON a multi-track film shoot where you can slide stuff around, etc. it's no biggie I guess, but in the
world of live production such latency would be unacceptable.

The real problem here though, as you mention is finding
frequencies that aren't already overcrowded that meet the
other requirements, i.e. concealable antennas, etc.

YOur comments on the 2.4 ghz bands and others in another
post noted, and would agree with those as well.

AS I understand Randy's question, it's more about desirable
characteristics and specs than actual design.

To answer his question:

I'd say usable range of 500 feet should cover most venues
for live performance, that enables the production to place
receivers at foh position.

AS for number of channels, I"m of the opinion that too many
performers who really don't need it demand wireless anyway.
But channels is a hard one to nail down because many want
channels not just for their wireless mics but for in ear
monitoring, etc.

Regards,
Richard
--
| Remove .my.foot for email
| via Waldo's Place USA Fidonet<->Internet Gateway Site
| Standard disclaimer: The views of this user are strictly his own.

audiobooks

September 20th 09, 01:32 AM

> I'd say usable range of 500 feet should cover most venues
> for live performance, that enables the production to place
> receivers at foh position.
>
500 feet? Are you putting your foh in the parking lot? Distance would be
less than 150 feet more like 60 feet.

Richard Webb[_3_]

September 20th 09, 09:53 PM

On Sat 2037-Sep-19 20:32, audiobooks writes:

I wrote:
>> I'd say usable range of 500 feet should cover most venues
>> for live performance, that enables the production to place
>> receivers at foh position.

"> 500 feet? Are you putting your foh in the parking lot? Distance would
"> be less than 150 feet more like 60 feet.

True, but it's called over-engineering. IF I"ve got usable
rx at 500 feet I"m good even if I want my own receiver in
the remote truck *in* the parking lot. I also know I"m
going to have usable signal levels and snr at the usual
distances for foh. Nothing can be trusted like a signal
being fed me via a good piece of copper. The radio hop just ads one more variable I can't control to the equation <g>.

Regards,
Richard
--
| Remove .my.foot for email
| via Waldo's Place USA Fidonet<->Internet Gateway Site
| Standard disclaimer: The views of this user are strictly his own.

Scott Dorsey

September 21st 09, 04:10 PM

In article >, Randy Yates > wrote:
(Scott Dorsey) writes:
>
>> In fact, the FCC cares only about field strength on that band, not
>> input power. Take, for example, your microwave oven. 1KW input
>> power, less than a microvolt per meter emission. That's okay.
>
>This is absolutely incorrect by my reckoning. The reason microwave ovens
>can operate the way they do is because they are exempted by section
>15.103 in [fccpart15].

15.103 really only applies to digital devices. It uses as an example the
digital controller built into a microwave oven. It does not have anything
to do with the actual magnetron emission.

You need to look under Part 18 for all the fun ISM stuff. Most of these
unlicensed digital devices would not be considered acceptable under Part 15,
but are in fact allowed under Part 18.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

Mark

September 22nd 09, 10:33 PM

On Sep 21, 11:10*am, (Scott Dorsey) wrote:
> In article >, Randy Yates > wrote:
>
> (Scott Dorsey) writes:
>
> >> In fact, the FCC cares only about field strength on that band, not
> >> input power. *Take, for example, your microwave oven. *1KW input
> >> power, less than a microvolt per meter emission. *That's okay.
>
> >This is absolutely incorrect by my reckoning. The reason microwave ovens
> >can operate the way they do is because they are exempted by section
> >15.103 in [fccpart15].
>
> 15.103 really only applies to digital devices. *It uses as an example the
> digital controller built into a microwave oven. *It does not have anything
> to do with the actual magnetron emission.
>
> You need to look under Part 18 for all the fun ISM stuff. *Most of these
> unlicensed digital devices would not be considered acceptable under Part 15,
> but are in fact allowed under Part 18.
> --scott
> --
> "C'est un Nagra. *C'est suisse, et tres, tres precis."

Hi Randy,

I would say the best kick ass wireless mic would use a wideband analog
FM system to carry the audio and some heavy duty DSP at the receiver
with diversity Rx antennas (3 or more) and the DSP combines the
antennas not just picks the best one.. MISO I guess you would call
it. I prefer analog for the actual audio path simply becasue that
gives the best latency and battery life for the Tx. I think any
digitization of the audio at the quality these guys need is going to
add power consumption to the Tx and hurt battery life.

Battery life for the Tx is a major feature.
Small package for the Tx is a major feature.
Reliability of the link is a major feature.
High quality audio is a given for this class of user.
You have a lot of space and power avaialbe at the Rx.
Put the bells and whisltes there.

Mark

Randy Yates

September 23rd 09, 12:40 AM

Mark > writes:

> On Sep 21, 11:10Â*am, (Scott Dorsey) wrote:
>> In article >, Randy Yates > wrote:
>>
>> (Scott Dorsey) writes:
>>
>> >> In fact, the FCC cares only about field strength on that band, not
>> >> input power. Â*Take, for example, your microwave oven. Â*1KW input
>> >> power, less than a microvolt per meter emission. Â*That's okay.
>>
>> >This is absolutely incorrect by my reckoning. The reason microwave ovens
>> >can operate the way they do is because they are exempted by section
>> >15.103 in [fccpart15].
>>
>> 15.103 really only applies to digital devices. Â*It uses as an example the
>> digital controller built into a microwave oven. Â*It does not have anything
>> to do with the actual magnetron emission.
>>
>> You need to look under Part 18 for all the fun ISM stuff. Â*Most of these
>> unlicensed digital devices would not be considered acceptable under Part 15,
>> but are in fact allowed under Part 18.
>> --scott
>> --
>> "C'est un Nagra. Â*C'est suisse, et tres, tres precis."
>
> Hi Randy,
>
> I would say the best kick ass wireless mic would use a wideband analog
> FM system to carry the audio and some heavy duty DSP at the receiver
> with diversity Rx antennas (3 or more) and the DSP combines the
> antennas not just picks the best one.. MISO I guess you would call
> it. I prefer analog for the actual audio path simply becasue that
> gives the best latency and battery life for the Tx. I think any
> digitization of the audio at the quality these guys need is going to
> add power consumption to the Tx and hurt battery life.
>
> Battery life for the Tx is a major feature.
> Small package for the Tx is a major feature.
> Reliability of the link is a major feature.
> High quality audio is a given for this class of user.
> You have a lot of space and power avaialbe at the Rx.
> Put the bells and whisltes there.

Hi Mark,

There would be two serious problems using analog FM:

1. The signal quality just wouldn't be good enough. I would shoot for
24 bits at 48 kHz, and you're just not going to get anywhere near that
level of performance from an analog FM system, which I would roughly
guess could be pumped up to give maybe 20 kHz / 80 dB SNR (if you're
lucky). The actual performance depends heavily on the design of the
modulation (e.g., modulation index, minimum target CNR, etc.), which
unfortunately interacts very negatively with the link margin.

2. The required link resources. Analog FM is going to be orders of
magnitude less efficient in both power and bandwidth than a comparable
digital modulation with FEC.

I would also venture to guess that a major factor in the source device's
power consumption is the transmitter power, and since FM is less much
efficient, that drives your battery life the wrong direction.

But I do concede that the A/D is also a major factor in power
consumption using a digital modulation. If you consider the two power
requirements (FM transmitter power using the analog FM method, and A/D
converter power using the digital method) approximately equal, then I'd
say it's pretty clear digital is going to be the more optimal approach
overall due to the efficiency in channel use.
--
Randy Yates % "She tells me that she likes me very much,
Digital Signal Labs % but when I try to touch, she makes it
% all too clear."
http://www.digitalsignallabs.com % 'Yours Truly, 2095', *Time*, ELO

Mark

September 23rd 09, 02:13 PM

>
> * 2. The required link resources. Analog FM is going to be orders of
> * magnitude less efficient in both power and bandwidth than a comparable
> * digital modulation with FEC.

This is not obvious to me.
Analog FM can use a class C Tx output stage.

If you use QAM or some other complex digital modulation you need a
linear output stage.

QPSK might be OK.

Most analog FM wireles mic systems use companding and can achive the
required S/N and range with a 10mW to 50 mW Tx output power. And the
FM modulator / Tx is very small and simple and reliable.

I aggree digital is more power efficent for cellphones where the low
voice quality allows for low bit rate vocoder etc and you can turn off
the Tx during speech pauses etc, but that is not the case for a hi
quality wireless mic system.

Regards
Mark

Randy Yates

September 23rd 09, 03:37 PM

Mark > writes:

>>
>> Â* 2. The required link resources. Analog FM is going to be orders of
>> Â* magnitude less efficient in both power and bandwidth than a comparable
>> Â* digital modulation with FEC.
>
> This is not obvious to me.
> Analog FM can use a class C Tx output stage.

This is not what I was referring to by efficiency, Mark. What I'm saying
is the transmit SNR and/or bandwidth required for analog FM to achieve a
desired output SNR and/or bandwidth is going to be much higher than that
predicted by Shannon's capacity formula, and since it is possible to
approach the limit these days with digital systems, the digital system
will be more power efficient.

> Most analog FM wireles mic systems use companding and can achive the
> required S/N and range with a 10mW to 50 mW Tx output power. And the
> FM modulator / Tx is very small and simple and reliable.

I believe we need to differentiate between two classes of products.
Representative of one class is a church where the pastor wants to
be connected wirelessly. Sure, in this sort of venue, folks don't
care about 16 bit performance (or higher).

The other class would be studio work, where ambient noise levels are way
down and you're recording in high-resolution (digital) multi-track. Do
you really think people are going to want to run their $7000 microphones
through a companding stage?

It is this latter class of product I am considering targeting.

> I aggree digital is more power efficent for cellphones where the low
> voice quality allows for low bit rate vocoder etc and you can turn off
> the Tx during speech pauses etc, but that is not the case for a hi
> quality wireless mic system.

I disagree, based on the points covered above.
--
Randy Yates % "I met someone who looks alot like you,
Digital Signal Labs % she does the things you do,
% but she is an IBM."
http://www.digitalsignallabs.com % 'Yours Truly, 2095', *Time*, ELO

Randy Yates

September 23rd 09, 03:40 PM

Randy Yates > writes:
> [...]
> I believe we need to differentiate between two classes of products.
> Representative of one class is a church where the pastor wants to
> be connected wirelessly. Sure, in this sort of venue, folks don't
> care about 16 bit performance (or higher).
>
> The other class would be studio work, where ambient noise levels are way
> down and you're recording in high-resolution (digital) multi-track. Do
> you really think people are going to want to run their $7000 microphones
> through a companding stage?
>
> It is this latter class of product I am considering targeting.

I would think that there are also non-studio scenarios, for example, a
big name concert that's going to be recorded and released, where a
low-fi FM solution just wouldn't cut it.
--
Randy Yates % "Rollin' and riding and slippin' and
Digital Signal Labs % sliding, it's magic."
%
http://www.digitalsignallabs.com % 'Living' Thing', *A New World Record*, ELO

Scott Dorsey

September 23rd 09, 03:49 PM

In article >, Randy Yates > wrote:
>
>This is not what I was referring to by efficiency, Mark. What I'm saying
>is the transmit SNR and/or bandwidth required for analog FM to achieve a
>desired output SNR and/or bandwidth is going to be much higher than that
>predicted by Shannon's capacity formula, and since it is possible to
>approach the limit these days with digital systems, the digital system
>will be more power efficient.

The digital modulation isn't so efficient either. With trellis encoding
it's almost a wash with analogue, though.

>> Most analog FM wireles mic systems use companding and can achive the
>> required S/N and range with a 10mW to 50 mW Tx output power. And the
>> FM modulator / Tx is very small and simple and reliable.
>
>I believe we need to differentiate between two classes of products.
>Representative of one class is a church where the pastor wants to
>be connected wirelessly. Sure, in this sort of venue, folks don't
>care about 16 bit performance (or higher).
>
>The other class would be studio work, where ambient noise levels are way
>down and you're recording in high-resolution (digital) multi-track. Do
>you really think people are going to want to run their $7000 microphones
>through a companding stage?
>
>It is this latter class of product I am considering targeting.

Everybody cares more about reliability than sound quality. The sound quality
of current FM wireless systems isn't great, but it's good enough for most
everything people want to use wireless for. The reliability is not, and the
need for careful coordination seems to be a problem for a lot of people.

>> I aggree digital is more power efficent for cellphones where the low
>> voice quality allows for low bit rate vocoder etc and you can turn off
>> the Tx during speech pauses etc, but that is not the case for a hi
>> quality wireless mic system.
>
>I disagree, based on the points covered above.

For the most part, digital systems are _always_ going to require more
bandwidth for the same channel characteristics as analogue systems. The
beauty of the digital encoding, though, is that it allows you to readily
trade one characteristic off for another. If you want more bandwidth but
less dynamic range, or you want more dynamic range but less bandwidth, you
can trade them off readily.

If digital systems really _were_ as efficient at bandwidth utilization, we
would be able to put uncompressed digital voice into the same 25 KC channel
that we currently put FM comms into. We cannot.

But we _can_ do some very cool things.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

Randy Yates

September 23rd 09, 06:22 PM

(Scott Dorsey) writes:
> [...]
> For the most part, digital systems are _always_ going to require more
> bandwidth for the same channel characteristics as analogue systems.

That is utterly false. Here's a counter-example.

Consider an analog, wideband FM system with the following
characteristics:

SNR_o = 96 dB
B = 20 kHz (audio bandwidth)
F = 75 kHz (maximum frequency deviation)
beta = F / B (modulation index)
= 3.75

where SNR_o denotes the output (audio) SNR.

According to [schwartzcomm], SNR_o is related to the RF SNR, SNR_c, as

SNR_c = SNR_o / (3*beta^2).

If you derive the required RF, or channel, SNR, and then use that
and the channel bandwidth (150 kHz) to define the channel
characteristics (W and SNR), then you can use Shannon's capacity
formula,

C = W * log_2(1 + SNR),

to find that the theoretical capacity limit for that channel is about 4
Mb/s.

Now 16 bits is approximately 96 dB SNR, and so at 48 kHz a digital
system would require a channel capacity of 16 * 48000 = 0.768 Mb/s.

So analog FM in this scenario OVERUSES the channel by a factor of 5!

To put that in a little more accessible terms, see the chart in
Figure 1 here:

http://www.digitalsignallabs.com/shannonlimit.pdf

For binary modulation, Eb/No is the channel SNR. Thus at an SNR of just
20 dB, we a spectral efficiency of 10 b/s per Hz. So, theoretically, to
achieve 768 kbps, we need 76.8 kHz at 20 dB SNR. This kicks the ass off
of analog FM in both SNR and bandwidth required.

Now it's true that we can't reach the limit. But these days, with
advanced coding like LDPCs, we can get pretty close.

Now we could go into a myriad of other practical considerations, but
this should show pretty clearly that, at least from a channel use
point-of-view, digital beats the hell out of analog (at least wideband
analog FM).

--Randy

@book{schwartzcomm,
title = "Information Transmission, Modulation, and Noise",
author = "{Mischa~Schwartz}",
publisher = "McGraw-Hill",
edition = "fourth",
year = "1990"}

> The beauty of the digital encoding, though, is that it allows you to
> readily trade one characteristic off for another. If you want more
> bandwidth but less dynamic range, or you want more dynamic range but
> less bandwidth, you can trade them off readily.
>
> If digital systems really _were_ as efficient at bandwidth utilization, we
> would be able to put uncompressed digital voice into the same 25 KC channel
> that we currently put FM comms into. We cannot.
>
> But we _can_ do some very cool things.
> --scott

--
Randy Yates % "So now it's getting late,
Digital Signal Labs % and those who hesitate
% got no one..."
http://www.digitalsignallabs.com % 'Waterfall', *Face The Music*, ELO

Mark

September 23rd 09, 09:50 PM

On Sep 23, 1:22*pm, Randy Yates > wrote:
> (Scott Dorsey) writes:
> > [...]
> > For the most part, digital systems are _always_ going to require more
> > bandwidth for the same channel characteristics as analogue systems. *
>
> That is utterly false. Here's a counter-example.
>
> Consider an analog, wideband FM system with the following
> characteristics:
>
> * SNR_o = 96 dB
> * B * * = 20 kHz * * * *(audio bandwidth)
> * F * * = 75 kHz * * * *(maximum frequency deviation)
> * beta *= F / B * * * * (modulation index)
> * * * * = 3.75
>
> where SNR_o denotes the output (audio) SNR.
>
> According to [schwartzcomm], SNR_o is related to the RF SNR, SNR_c, as
>
> * SNR_c = SNR_o / (3*beta^2).
>
> If you derive the required RF, or channel, SNR, and then use that
> and the channel bandwidth (150 kHz) to define the channel
> characteristics (W and SNR), then you can use Shannon's capacity
> formula,
>
> * C = W * log_2(1 + SNR),
>
> to find that the theoretical capacity limit for that channel is about 4
> Mb/s.
>
> Now 16 bits is approximately 96 dB SNR, and so at 48 kHz a digital
> system would require a channel capacity of 16 * 48000 = 0.768 Mb/s.
>
> So analog FM in this scenario OVERUSES the channel by a factor of 5!
>
> To put that in a little more accessible terms, see the chart in
> Figure 1 here:
>
> *http://www.digitalsignallabs.com/shannonlimit.pdf
>
> For binary modulation, Eb/No is the channel SNR. Thus at an SNR of just
> 20 dB, we a spectral efficiency of 10 b/s per Hz. *So, theoretically, to
> achieve 768 kbps, we need 76.8 kHz at 20 dB SNR. This kicks the ass off
> of analog FM in both SNR and bandwidth required.

1) Using 76.8 kHz BW vs 150 kHz BW of radio spectrum is no real
advantage. I did not see an allowance for FEC overhead.

2) To achieve 10b/s per Hz implies a higher order QAM modulation which
requires a linear power amplifer in the Tx body pac and lower DC power
effiency.

3) 20 dB SNR is the THRESHOLD. In order for the link to be reliable
with fading, you will need a good 20 dB more RF power for the nominal
operating condition. DC power consumption wise considering the
linear vs class C amp an the A/D and all the digital stuff you end up
close or worse compared to the analog case.

I don't see a practical advantage for digital in terms of either BW or
power.

The big problem in wirelss mic for stage applications (music and
theatre) is propogation fading into a null. This problem is normally
addressed with (usually 2) diversity receiving antennas and as I said,
here is an opportunity to some cool things with DSP in the Rx. The
audio performance of analog companded FM systems is acceptable for
sound reinformcent applications where wirelss mics are normally
used.

I don't see the application for wireless in a studio setting where
"perfect" audio is needed. Studio guys with $7000 mics will never be
satified with the audio performance of any wireless device and they
have no compelling need for wireless. They LIKE to use their $500
cables and mic preamps attached to $7000 mics. :-)

Mark

Randy Yates

September 24th 09, 04:54 AM

Mark > writes:
> [...]
> 1) Using 76.8 kHz BW vs 150 kHz BW of radio spectrum is no real
> advantage.

70+ MHz isn't real? Tell that to the FCC.

> I did not see an allowance for FEC overhead.

That's because there isn't any such allowance. The capacity formula
already accounts for this overhead and provides you the (theoretical)
capacity given the bandwidth and SNR of the channel.

> 2) To achieve 10b/s per Hz implies a higher order QAM modulation which
> requires a linear power amplifer in the Tx body pac and lower DC power
> effiency.

That's true, and this is one of those "myriad" factors I was referring
to. We're not going to cover them all at this point since there are
way way way too many variables to account for.

However, note that the FM solution required 80 dB RF SNR (I didn't state
that number previously, but it does for the scenario I proposed). I
doubt that your Class C power amplifier is 60 dB more efficient than a
linear power amplifier.

> 3) 20 dB SNR is the THRESHOLD. In order for the link to be reliable
> with fading, you will need a good 20 dB more RF power for the nominal
> operating condition.

And you will also require margin for the FM solution as well. Again,
this is another of the "myriad" other considerations, and, again, I
believe that the extreme advantage shown in the ideal theoretical case
makes it very likely that the final digital system will still be orders
of magnitude more "efficient" (both power and bandwidth) than the analog
system when all these factors have been accounted for.

> DC power consumption wise considering the linear vs class C amp an the
> A/D and all the digital stuff you end up close or worse compared to
> the analog case.

Perhaps - hard to say without going through the details.

However, the issue being addressed in this subthread isn't battery power
but channel usage.

> I don't see a practical advantage for digital in terms of either BW or
> power.

If we're discussing channel usage, then I'd say you're blind.

> I don't see the application for wireless in a studio setting where
> "perfect" audio is needed. Studio guys with $7000 mics will never be
> satified with the audio performance of any wireless device and they
> have no compelling need for wireless. They LIKE to use their $500
> cables and mic preamps attached to $7000 mics. :-)

I do admit that I may have the use-case scenario wrong here. It would be
helpful if others would chime in and state if they would have any need for
a "high-end" wireless mic, capable of >20-bit/48 kHz performance.
--
Randy Yates % "How's life on earth?
Digital Signal Labs % ... What is it worth?"
% 'Mission (A World Record)',
http://www.digitalsignallabs.com % *A New World Record*, ELO

Mark

September 24th 09, 03:16 PM

On Sep 23, 11:54*pm, Randy Yates > wrote:
> Mark > writes:
> > [...]
> > 1) Using 76.8 kHz BW vs 150 kHz BW of radio spectrum is no real
> > advantage. *
>
> 70+ MHz isn't real? Tell that to the FCC.
>

the delta is 70 kHz not MHz..

Mark

Scott Dorsey

September 24th 09, 03:46 PM

In article >, Randy Yates > wrote:
>
>So analog FM in this scenario OVERUSES the channel by a factor of 5!

Yup, that seems about fair.

>To put that in a little more accessible terms, see the chart in
>Figure 1 here:
>
> http://www.digitalsignallabs.com/shannonlimit.pdf
>
>For binary modulation, Eb/No is the channel SNR. Thus at an SNR of just
>20 dB, we a spectral efficiency of 10 b/s per Hz. So, theoretically, to
>achieve 768 kbps, we need 76.8 kHz at 20 dB SNR. This kicks the ass off
>of analog FM in both SNR and bandwidth required.
>
>Now it's true that we can't reach the limit. But these days, with
>advanced coding like LDPCs, we can get pretty close.

We can't really get all that close, even with stuff like LDPCs. And some
of the things that help us get closer cause some latency issues.

I think the standard example is to take a single FSK carrier, and show
the math on that, then show what happens when you pack a huge number of
low-bandwidth FSK carriers into a single channel. Capacity goes up,
but it levels off as you increase the number of carriers.

Yes, there are other modulation methods that have higher channel utilization
than FSK, but the total numbers still aren't good.

>Now we could go into a myriad of other practical considerations, but
>this should show pretty clearly that, at least from a channel use
>point-of-view, digital beats the hell out of analog (at least wideband
>analog FM).

Dunno, you showed the math for FM properly, but I don't think the math for
the digital modulation is quite as good as you present it.

Even so, I don't think digital encoding is a bad thing at all for wireless,
_because_ it allows us to make bandwidth vs. noise floor tradeoffs so
effectively.
--scott

--
"C'est un Nagra. C'est suisse, et tres, tres precis."

Randy Yates

September 24th 09, 04:06 PM

(Scott Dorsey) writes:

> In article >, Randy Yates > wrote:
>>
>>So analog FM in this scenario OVERUSES the channel by a factor of 5!
>
> Yup, that seems about fair.
>
>>To put that in a little more accessible terms, see the chart in
>>Figure 1 here:
>>
>> http://www.digitalsignallabs.com/shannonlimit.pdf
>>
>>For binary modulation, Eb/No is the channel SNR. Thus at an SNR of just
>>20 dB, we a spectral efficiency of 10 b/s per Hz. So, theoretically, to
>>achieve 768 kbps, we need 76.8 kHz at 20 dB SNR. This kicks the ass off
>>of analog FM in both SNR and bandwidth required.
>>
>>Now it's true that we can't reach the limit. But these days, with
>>advanced coding like LDPCs, we can get pretty close.
>
> We can't really get all that close, even with stuff like LDPCs.

You're absolutely right, Scott. 0.0045 dB is still pretty far away from
the limit:

http://en.wikipedia.org/wiki/Noisy-channel_coding_theorem

(search for "LDPC").

Better to spend that 60 dB and fall back on analog FM.

> And some of the things that help us get closer cause some latency
> issues.

Sure, if you want to get within a gnat's ass of it. I don't think 5 dB
or so would require too much latency, and that would still give you a 55
dB advantage over analog FM.

So, I'm not quite getting you: tell me again how analog is better?
--
Randy Yates % "Though you ride on the wheels of tomorrow,
Digital Signal Labs % you still wander the fields of your
% sorrow."
http://www.digitalsignallabs.com % '21st Century Man', *Time*, ELO

Randy Yates

September 24th 09, 04:08 PM

Mark > writes:

> On Sep 23, 11:54Â*pm, Randy Yates > wrote:
>> Mark > writes:
>> > [...]
>> > 1) Using 76.8 kHz BW vs 150 kHz BW of radio spectrum is no real
>> > advantage. Â*
>>
>> 70+ MHz isn't real? Tell that to the FCC.
>>
>
> the delta is 70 kHz not MHz..

You're right, Mark. My mistake.

So, if an FM station is 70 kHz too wide, it doesn't matter too much to
the FCC, right? After all, it's not "real."
--
Randy Yates % "Ticket to the moon, flight leaves here today
Digital Signal Labs % from Satellite 2"
% 'Ticket To The Moon'
http://www.digitalsignallabs.com % *Time*, Electric Light Orchestra

Scott Dorsey

September 24th 09, 04:53 PM

In article >, Randy Yates > wrote:
(Scott Dorsey) writes:
>>>
>>>Now it's true that we can't reach the limit. But these days, with
>>>advanced coding like LDPCs, we can get pretty close.
>>
>> We can't really get all that close, even with stuff like LDPCs.
>
>You're absolutely right, Scott. 0.0045 dB is still pretty far away from
>the limit:
>
> http://en.wikipedia.org/wiki/Noisy-channel_coding_theorem
>
>(search for "LDPC").

I think you're talking about Gallagher codes. Sorry, when you implement that
with real-world modulation, you don't get even close to that. If you could,
the cellphone folks would be having a ball since their whole lives revolve
around optimizing channel capacity.

The encoding might be really efficient, but you still need to get it onto
an RF link. I think in the real world you're more in the QPSK or 16QAM range,
possibly with hundreds of individual carriers across the channel. The issue
is the modulation, not the encoding.
--scott

--
"C'est un Nagra. C'est suisse, et tres, tres precis."

Scott Dorsey

September 24th 09, 04:54 PM

In article >, Randy Yates > wrote:
>Mark > writes:
>
>> On Sep 23, 11:54Â*pm, Randy Yates > wrote:
>>> Mark > writes:
>>> > [...]
>>> > 1) Using 76.8 kHz BW vs 150 kHz BW of radio spectrum is no real
>>> > advantage. Â*
>>>
>>> 70+ MHz isn't real? Tell that to the FCC.
>>>
>>
>> the delta is 70 kHz not MHz..
>
>You're right, Mark. My mistake.
>
>So, if an FM station is 70 kHz too wide, it doesn't matter too much to
>the FCC, right? After all, it's not "real."
>--
>Randy Yates % "Ticket to the moon, flight leaves here today
>Digital Signal Labs % from Satellite 2"
% 'Ticket To The Moon'
>http://www.digitalsignallabs.com % *Time*, Electric Light Orchestra

--
"C'est un Nagra. C'est suisse, et tres, tres precis."

Scott Dorsey

September 24th 09, 04:56 PM

In article >, Randy Yates > wrote:
>Mark > writes:
>> On Sep 23, 11:54Â*pm, Randy Yates > wrote:
>>> Mark > writes:
>>> > [...]
>>> > 1) Using 76.8 kHz BW vs 150 kHz BW of radio spectrum is no real
>>> > advantage. Â*
>>>
>>> 70+ MHz isn't real? Tell that to the FCC.
>>>
>>
>> the delta is 70 kHz not MHz..
>
>You're right, Mark. My mistake.
>
>So, if an FM station is 70 kHz too wide, it doesn't matter too much to
>the FCC, right? After all, it's not "real."

In the RF world, you get whatever bandwidth you're willing to pay for. If
you want your channel to be 70 KC wider, no problem. You just pay more money
for it. The vast majority of RF encoding issues revolve around trying to
get more channel capacity for less money. All of the rest of them have to
do with compatibility with legacy hardware....
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

Mark

September 24th 09, 05:29 PM

On Sep 24, 11:08*am, Randy Yates > wrote:
> Mark > writes:
> > On Sep 23, 11:54*pm, Randy Yates > wrote:
> >> Mark > writes:
> >> > [...]
> >> > 1) Using 76.8 kHz BW vs 150 kHz BW of radio spectrum is no real
> >> > advantage. *
>
> >> 70+ MHz isn't real? Tell that to the FCC.
>
> > the delta is 70 kHz * not MHz..
>
> You're right, Mark. My mistake.
>
> So, if an FM station is 70 kHz too wide, it doesn't matter too much to
> the FCC, right? After all, it's not "real."
>

Randy,

per FCC CFR 47 Part 74.802, the BW of professional a wireless mic
( low power auxillary station) "shall not exceed 200 kHz" Any lower
BW "does not matter too much to the FCC" and is of no "real"
advantage.

Mark

Scott Dorsey

September 24th 09, 06:08 PM

Mark > wrote:
>
>per FCC CFR 47 Part 74.802, the BW of professional a wireless mic
>( low power auxillary station) "shall not exceed 200 kHz" Any lower
>BW "does not matter too much to the FCC" and is of no "real"
>advantage.

Note that part 74 ONLY applies to broadcasters. The only folks who have
a legal authority to use the auxiliary broadcast provisions are radio and
TV stations with existing broadcast licenses. So THOSE people can use
wireless with 200KC channels.

Everybody else... well... no one really knows. Maybe they're okay under
part 15? Maybe...
--scott

--
"C'est un Nagra. C'est suisse, et tres, tres precis."

Randy Yates

September 24th 09, 06:44 PM

(Scott Dorsey) writes:

> In article >, Randy Yates > wrote:
(Scott Dorsey) writes:
>>>>
>>>>Now it's true that we can't reach the limit. But these days, with
>>>>advanced coding like LDPCs, we can get pretty close.
>>>
>>> We can't really get all that close, even with stuff like LDPCs.
>>
>>You're absolutely right, Scott. 0.0045 dB is still pretty far away from
>>the limit:
>>
>> http://en.wikipedia.org/wiki/Noisy-channel_coding_theorem
>>
>>(search for "LDPC").
>
> I think you're talking about Gallagher codes.

Same damn thing, yes. Most folks these days refer to them as LDPC.

> Sorry, when you implement that with real-world modulation, you don't
> get even close to that.

Sorry, need to check with, umm, the DVB standard and several other
standards that have recently adopted them.

http://en.wikipedia.org/wiki/Low-density_parity-check_code

> If you could, the cellphone folks would be having a ball since their
> whole lives revolve around optimizing channel capacity.

They were forgotten until the latter part of the 90's - GSM had already
been defined by then.

> The encoding might be really efficient, but you still need to get it onto
> an RF link. I think in the real world you're more in the QPSK or 16QAM range,
> possibly with hundreds of individual carriers across the channel. The issue
> is the modulation, not the encoding.

I think DVB is fairly close to the real world, don't you?

Give it up, already, Scott. Digital comms beats the **** of any analog
modulation.
--
Randy Yates % "With time with what you've learned,
Digital Signal Labs % they'll kiss the ground you walk
% upon."
http://www.digitalsignallabs.com % '21st Century Man', *Time*, ELO

Randy Yates

September 24th 09, 06:57 PM

Mark > writes:

> On Sep 24, 11:08Â*am, Randy Yates > wrote:
>> Mark > writes:
>> > On Sep 23, 11:54Â*pm, Randy Yates > wrote:
>> >> Mark > writes:
>> >> > [...]
>> >> > 1) Using 76.8 kHz BW vs 150 kHz BW of radio spectrum is no real
>> >> > advantage. Â*
>>
>> >> 70+ MHz isn't real? Tell that to the FCC.
>>
>> > the delta is 70 kHz Â* not MHz..
>>
>> You're right, Mark. My mistake.
>>
>> So, if an FM station is 70 kHz too wide, it doesn't matter too much to
>> the FCC, right? After all, it's not "real."
>>
>
> Randy,
>
> per FCC CFR 47 Part 74.802, the BW of professional a wireless mic
> ( low power auxillary station) "shall not exceed 200 kHz" Any lower
> BW "does not matter too much to the FCC" and is of no "real"
> advantage.

You're confounding the point of this portion of this thread, Mark. My
response was directed to Scott's blanket claim that digital modulation
will always require more bandwidth than analog modulation. This is
clearly false, and the point of my example was to prove it.

Leap-frogging to the conclusion that we're going to be using 200 kHz
channels is also quite premature.
--
Randy Yates % "And all you had to say
Digital Signal Labs % was that you were
% gonna stay."
http://www.digitalsignallabs.com % Getting To The Point', *Balance of Power*, ELO