How an echo canceller distinguishes that an incoming stream of bits are of voice or data. So that it should not cancel data bits considering voice echo.
echo canceller
Started by ●January 24, 2004
Reply by ●January 24, 20042004-01-24
<pankajjd@rediffmail.com> wrote in message news:1e5efa11.0401240729.6b2a568c@posting.google.com...> How an echo canceller distinguishes that an incoming stream of bits > are of voice or data. So that it should not cancel data bits > considering voice echo.It's all data isn't it? The issue is reliably reconstructing the bits at the output. If there's echo / multipath, then that's still the objective. Or am I missing something here? Fred
Reply by ●January 24, 20042004-01-24
data call (e.g. fax,modem etc.) setup signals that the echo cancellers should be turned off. <pankajjd@rediffmail.com> wrote in message news:1e5efa11.0401240729.6b2a568c@posting.google.com...> How an echo canceller distinguishes that an incoming stream of bits > are of voice or data. So that it should not cancel data bits > considering voice echo.
Reply by ●January 24, 20042004-01-24
michael yarwood wrote:> data call (e.g. fax,modem etc.) setup signals that the echo cancellers > should be turned off. > > <pankajjd@rediffmail.com> wrote in message > news:1e5efa11.0401240729.6b2a568c@posting.google.com... > >>How an echo canceller distinguishes that an incoming stream of bits >>are of voice or data. So that it should not cancel data bits >>considering voice echo.What are turned off are not echo cancellers, but echo suppressers. Not at all the same thing. Once the suppressers are turned off, cancellation is done at the modem. Jerry -- Engineering is the art of making what you want from things you can get. �����������������������������������������������������������������������
Reply by ●January 24, 20042004-01-24
Jerry Avins <jya@ieee.org> writes:> michael yarwood wrote: > >> data call (e.g. fax,modem etc.) setup signals that the echo cancellers >> should be turned off. >> <pankajjd@rediffmail.com> wrote in message >> news:1e5efa11.0401240729.6b2a568c@posting.google.com... >> >>>How an echo canceller distinguishes that an incoming stream of bits >>>are of voice or data. So that it should not cancel data bits >>>considering voice echo. > > What are turned off are not echo cancellers, but echo suppressers. Not > at all the same thing.Why don't you fill us in, Jerry. -- % Randy Yates % "She's sweet on Wagner-I think she'd die for Beethoven. %% Fuquay-Varina, NC % She love the way Puccini lays down a tune, and %%% 919-577-9882 % Verdi's always creepin' from her room." %%%% <yates@ieee.org> % "Rockaria", *A New World Record*, ELO http://home.earthlink.net/~yatescr
Reply by ●January 25, 20042004-01-25
Randy Yates wrote:> Jerry Avins <jya@ieee.org> writes: > > >>michael yarwood wrote: >> >> >>>data call (e.g. fax,modem etc.) setup signals that the echo cancellers >>>should be turned off. >>><pankajjd@rediffmail.com> wrote in message >>>news:1e5efa11.0401240729.6b2a568c@posting.google.com... >>> >>> >>>>How an echo canceller distinguishes that an incoming stream of bits >>>>are of voice or data. So that it should not cancel data bits >>>>considering voice echo. >> >>What are turned off are not echo cancellers, but echo suppressers. Not >>at all the same thing. > > > Why don't you fill us in, Jerry.The wire pair from subscriber to local office carries a two-way signal. There, it is split into east- and westbound signals traveling separate paths. Those can be analog, but now are usually digital. There is an attempt to terminate the line to minimize reflection and the residual reflections are simply suppressed. Echoes are suppressed by turning off a channel when it carries no sound above a threshold volume (such as a pause between words or even between syllables), and turning it back on when talking resumes. This clips the beginnings of words, but not enough to be noticeable. It does, however, interfere with modem operation, even at 110 baud. The repeater amplifiers respond to a particular tone at the time a connection is established by disabling the suppresser circuits. It is then up to the modem to adapt to the echo and cancel it. The digital portions of the path may well be handled by cancellation rather than suppression, but that's not what the OP referred to. I've been told that true echo cancellation is used on satellite links, but I haven't checked that with what I believe to be a reliable source. Jerry -- Engineering is the art of making what you want from things you can get. �����������������������������������������������������������������������
Reply by ●January 27, 20042004-01-27
Jerry Avins wrote: (snip regarding echo suppressers)> The wire pair from subscriber to local office carries a two-way signal. > There, it is split into east- and westbound signals traveling separate > paths. Those can be analog, but now are usually digital. There is an > attempt to terminate the line to minimize reflection and the residual > reflections are simply suppressed. Echoes are suppressed by turning off > a channel when it carries no sound above a threshold volume (such as a > pause between words or even between syllables), and turning it back on > when talking resumes.I had thought that it just attenuated the signal, not turning it completely off. Enough that the echoes are not noticeable. I have had calls where they were off, and the echo is very distracting. > This clips the beginnings of words, but not enough> to be noticeable. It does, however, interfere with modem operation, even > at 110 baud. The repeater amplifiers respond to a particular tone at the > time a connection is established by disabling the suppresser circuits. > It is then up to the modem to adapt to the echo and cancel it. The > digital portions of the path may well be handled by cancellation rather > than suppression, but that's not what the OP referred to.> I've been told that true echo cancellation is used on satellite links, > but I haven't checked that with what I believe to be a reliable source.I have heard that the goal is to use satellite links for only one direction of a phone call. In both directions, the delay is enough to be distracting. Though for a satellite phone that obviously won't work. -- glen
Reply by ●January 27, 20042004-01-27
"Jerry Avins" <jya@ieee.org> wrote in message news:401343c9$0$7333$61fed72c@news.rcn.com...> Randy Yates wrote: > > > Jerry Avins <jya@ieee.org> writes: > > > > > >>michael yarwood wrote: > >> > >> > >>>data call (e.g. fax,modem etc.) setup signals that the echo cancellers > >>>should be turned off. > >>><pankajjd@rediffmail.com> wrote in message > >>>news:1e5efa11.0401240729.6b2a568c@posting.google.com... > >>> > >>> > >>>>How an echo canceller distinguishes that an incoming stream of bits > >>>>are of voice or data. So that it should not cancel data bits > >>>>considering voice echo. > >> > >>What are turned off are not echo cancellers, but echo suppressers. Not > >>at all the same thing. > > > > > > Why don't you fill us in, Jerry. > > The wire pair from subscriber to local office carries a two-way signal. > There, it is split into east- and westbound signals traveling separate > paths. Those can be analog, but now are usually digital. There is an > attempt to terminate the line to minimize reflection and the residual > reflections are simply suppressed. Echoes are suppressed by turning off > a channel when it carries no sound above a threshold volume (such as a > pause between words or even between syllables), and turning it back on > when talking resumes. This clips the beginnings of words, but not enough > to be noticeable. It does, however, interfere with modem operation, even > at 110 baud. The repeater amplifiers respond to a particular tone at the > time a connection is established by disabling the suppresser circuits. > It is then up to the modem to adapt to the echo and cancel it. The > digital portions of the path may well be handled by cancellation rather > than suppression, but that's not what the OP referred to. > > I've been told that true echo cancellation is used on satellite links, > but I haven't checked that with what I believe to be a reliable source.Sounds like by echo suppression you are referring to a non-linear process that acts like a noise gate (or expander if the level is merely reduced). This is in contrast to the linear process of the adaptive filter for echo cancellation. Makes sense to me. I've sometimes heard the first part of this called NLP (non-linear processing).
Reply by ●January 27, 20042004-01-27
Jon Harris wrote:> "Jerry Avins" <jya@ieee.org> wrote in message > news:401343c9$0$7333$61fed72c@news.rcn.com... > >>Randy Yates wrote: >> >> >>>Jerry Avins <jya@ieee.org> writes: >>> >>> >>> >>>>michael yarwood wrote: >>>> >>>> >>>> >>>>>data call (e.g. fax,modem etc.) setup signals that the echo cancellers >>>>>should be turned off. >>>>><pankajjd@rediffmail.com> wrote in message >>>>>news:1e5efa11.0401240729.6b2a568c@posting.google.com... >>>>> >>>>> >>>>> >>>>>>How an echo canceller distinguishes that an incoming stream of bits >>>>>>are of voice or data. So that it should not cancel data bits >>>>>>considering voice echo. >>>> >>>>What are turned off are not echo cancellers, but echo suppressers. Not >>>>at all the same thing. >>> >>> >>>Why don't you fill us in, Jerry. >> >>The wire pair from subscriber to local office carries a two-way signal. >>There, it is split into east- and westbound signals traveling separate >>paths. Those can be analog, but now are usually digital. There is an >>attempt to terminate the line to minimize reflection and the residual >>reflections are simply suppressed. Echoes are suppressed by turning off >>a channel when it carries no sound above a threshold volume (such as a >>pause between words or even between syllables), and turning it back on >>when talking resumes. This clips the beginnings of words, but not enough >>to be noticeable. It does, however, interfere with modem operation, even >>at 110 baud. The repeater amplifiers respond to a particular tone at the >>time a connection is established by disabling the suppresser circuits. >>It is then up to the modem to adapt to the echo and cancel it. The >>digital portions of the path may well be handled by cancellation rather >>than suppression, but that's not what the OP referred to. >> >>I've been told that true echo cancellation is used on satellite links, >>but I haven't checked that with what I believe to be a reliable source. > > > Sounds like by echo suppression you are referring to a non-linear process > that acts like a noise gate (or expander if the level is merely reduced). > This is in contrast to the linear process of the adaptive filter for echo > cancellation. Makes sense to me. I've sometimes heard the first part of > this called NLP (non-linear processing).What I wrote doesn't make a lot of sense. Echoes -- reflections from imperfect terminations and poor transitions in the run of a line -- happen only on the analog parts of a link. When I learned about telephony, it was all analog. By now, practically all local "centrals" are linked digitally. Only the echoes on analog analog lines in the four-wire part of the run are suppressed by squelching a channel when the level on it drops below some threshold. I used to be able to turn of the echo suppressors by whistling. It could be fun to see how bad the echo really was, and a good way to ensure that a conversation I knew would be unpleasant would also be short. Jerry -- Engineering is the art of making what you want from things you can get. �����������������������������������������������������������������������
Reply by ●January 27, 20042004-01-27
"Jerry Avins" <jya@ieee.org> wrote in message news:4016d364$0$11462$61fed72c@news.rcn.com...> > I used to be able to turn of the echo suppressors by whistling. It could > be fun to see how bad the echo really was, and a good way to ensure that > a conversation I knew would be unpleasant would also be short.That's way cool!
