Thread Subject: Subject: teitac-telecom Digest, Vol 2, Issue 9

Note

This archival content is maintained by WebAIM and NCDAE on behalf of TEITAC and the U.S. Access Board . Additional details on the updates to section 508 and section 255 can be found at the Access Board web site.

From: John Combs (jcombs)
Date: Fri, Nov 10 2006 5:05 PM

Hi everyone!

I just joined the teitac-telecom mailing list.

Here's my short self-introduction.

I started working in telephony testing/regulation in 1983 at GTE's
internal telecommunications testing laboratory. This lab was later sold
to an independent test lab chain called ITS (Intertek Test Systems) in
1994, where I continued doing telecom testing. In 1998 I moved to Cisco
Systems, where I am currently the manager of our
Accessibility/Safety/Telecom regulatory requirements/testing
department(s).

I have participated in the TIA TR-41 telecom subcommittees on and off
for the last fifteen years, representing my various employers. I have
also been involved in testing of accessibility features for most of my
career, including telephones for volume control & HAC, and TTY
compatibility with various phones & phone systems.

I would like to add some detailed background history to the discussions
I read in the "Subject: teitac-telecom Digest, Vol 2, Issue 9".

- John Combs

-----------------------------------------
1) TTY compatibility via a "digital to analog" adapter on a digital
phone system:

For those old enough to remember, the original "key systems" sold in the
US were the AT&T (Western Electric) 1A Key System, and a very similar
product from GTE (Automatic Electric) called a 10A2 key system.

"Key" is another term for "button," and it simply meant that these were
small multiline phone systems, generally with some/all phone lines in
the system available on all phones, with a button (key) for each line;
you pressed the key for the line you wanted, then made a call.

Here's a web page link with photos of several models of the key phones
for the 1A: http://mysite.verizon.net/dalderdi/phones/#key

These key systems were TTY compatible -- they had standard "G-style"
(round end) handsets on the analog phones, so a TTY could acoustically
couple with it. Since they were analog systems, and used analog trunks
from the local telco, RJ-11 jacks with analog lines could also be
provided for a TTY for direct connect.

Then FCC Part 68 was adopted, and the US telephone equipment market was
wide open for competition. Companies started selling newfangled
"electronic" key systems, one of the first being the TIE (Technicom
International), with their EK 6X16 system (six incoming analog trunks,
up to 16 proprietary digital telephones).
http://www.businesstelephone.com/tie_m-16ksu.asp

These low-end systems, such as the TIE Mod 16, generally had no
provision for an analog station (if a company had a fax machine, they
brought in a separate analog line for it). Some of the digital
proprietary phones could still acoustically-couple with a TTY, but as
more companies got into the telecom equipment business, it became trendy
to make stylish phones with new looks, to stand out from the
competition.

Unfortunately, quite a bit of that styling included flashy new handset
shapes. Many of these new handset shapes worked very poorly in the TTY
acoustic coupler cups, or not at all.

One handset which was totally incompatible with a TTY acoustic coupler
was the AT&T Merlin key system digital phone, with its infamous "tiny"
mouthpiece; unfortunately for TTY users, the Merlin was a huge seller
since AT&T Bell System still had nearly 85% of the phone lines in the
U.S.

Here's a link to photos of Merlin phones, scroll to the bottom of the
page for a closeup of the Merlin handset:
http://www.businesstelephone.com/att_merlin.asp

Frankly, most users found the Merlin handset uncomfortable, as you
couldn't hold it with your head against your shoulder (so you freed up
your hand) because of the small mouthpiece!

A TTY-using visitor to a company with an electronic key system often had
no way to use it, since there were no analog ports on the system, and
the phone handsets didn't fit the TTY acoustic couplers due to
mechanical shape.

This history is the reason for the first portion of Section 1194.23(a):
"Telecommunications products or systems which provide a function
allowing voice communication and which do not themselves provide a TTY
functionality shall provide a standard non-acoustic connection point for
TTYs."

And the wording was chosen very carefully, the phrase
"Telecommunications products or systems..." was intended to mean that
every phone did NOT have to be equipped with an analog port for a TTY
RJ-11 connection, but that the *system* had to support analog
connections for TTY, i.e., "...standard non-acoustic connection point
for TTYs." (One solution being an optional analog port module provided
alongside or inside a digital phone.)

Requiring every digital phone to have an analog port for TTY RJ-11 means
that only expensive models of digital phones can be sold, it adds
considerable cost to the design.

This clause was specifically targeted at the then-prevalent digital key
systems with no analog lines or digital-to-analog converters, and
proprietary digital telephones with handsets that were useless for TTY
acoustic coupling. And there were (and still are) lots of these key
systems out there.

Finally, like so many accessibilty features, 1194.23(a) has provided
useful additional functionality for non-accessibility purposes too.
Such as having an available analog port adapter for a fax machine, or
dial-up modem, so the company doesn't have to pay extra for dedicated
analog lines from the phone company.
--------------------------------------


2) 20 dB gain vs. 12 dB gain & FCC Part 68.317

There is a lot of confusion about acoustic gain for telephone handsets;
this is an area I hope the Access Board can clarify with their current
work. Even the FCC Part 68.317 section on Volume Control is not quite
as precisely worded as it should be, and I wish the TIA would provide
new language for FCC Part 68.317.

First, there is confusion about how to measure the acoustic gain.
Reputable telephone equipment manufacturers assume that the actual
meaning is "gain over the normal default receiver volume."

A few telephone equipment manufacturers have played games by claiming
the total volume range of the phone as "gain." So they set the volume
control to minimum, measure the loudness, set the volume control to
maximum, measure the loudness, and claim this entire range as their
"gain." Since most phones have 40-45 dB of volume range from minimum to
maximum, this allows them to claim a very high "gain" (impossibly
high).

But of course, the only usable gain for a hearing-imparired person would
be from the default volume setting of the phone, to the maximum volume
setting of the phone.

When we specify it this way, my testing over many years for many
manufacturers has shown that most business telephones have a bit more
than 12 dB of what I will call "true gain from default volume" -- just
enough to meet FCC Part 68.317.

Please note that FCC Part 68 only *mandates* a minimum 12 dB of acoustic
gain on all telephones sold in the U.S. And the gain should be
non-clipping (not distorted), but only up to 12 dB.

The 18 dB acoustic gain value in FCC Part 68 is actually the *maximum
permitted gain* a phone can have, *unless* the phone is equipped with an
automatic function to lower the gain back below 18 dB each time the call
is ended. (So if a different caller uses the phone next, and has normal
hearing, they won't get a superloud burst of audio that could damage
their hearing.)

I wasn't attending TIA at the time, but people who were told me there
was a lot of discussion between the TIA, FCC, and various hearing
imparied lobbyist groups, etc., with the lobbyists pushing for 20 dB (or
more) of true acoustic gain being required. The technical discussions
in the TIA showed that it was not possible to provide that level of gain
in a "line powered telephone", e.g., an analog loop-powered telephone on
a long loop -- there was not sufficient power available on the line.

It would have required an external AC wall adapter for extra power for
the phone (or batteries which had to be periodically replaced -- it is
also forbidden to use phone line power to keep batteries charged in a
phone).

It turns out that an IP phone powered solely by 48 VDC over Ethernet
(IEEE 802.3af) also lacks sufficient power to do really high acoustic
boost, while also trying to run phone displays, multiple
microprocessors, Ethernet switches, etc.

Buyers of phone systems greatly prefer phones which are line-powered --
they don't like to keep up with AC wall power packs, and the newer laws
about energy efficiency in products are beginning to mandate power
efficient phones.

So the reason the FCC has held their mandatory gain value at 12 dB for
all these years is because of the power limitations of the phone lines
(analog at the time, and power-over-Ethernet now).

If a higher minimum gain is imposed, it will mostly likely eliminate
low-cost line-powered phones from the market (or force all phones to use
AC wall power packs for additional power).

I did participate in a TIA discussion a couple years back, about the "12
dB FCC gain" vs. the "Section 508 20 dB gain" requirements, talking
about how (or if) there should be a harmonization effort. Several
manufacturers expressed the opinion that they were afraid of legal
liability for hearing damage to users with normal hearing, and would
never want to build a phone with built-in 20 dB or more of true gain.
Some speculated that the high volume levels would actually cause further
hearing damage to the user (who already had some hearing loss obviously,
since they wanted high volume boost).

Finally, it is worth mentioning that EVERY highly amplifed phone
amplifier/handset that I have ever measured does indeed distort the
audio. It is inevitable, for a couple of reasons.

Most of the energy in the telephone voice band (300-3300 Hz) is
contained in the lower frequencies of the audio range, say 300 Hz to
1000 Hz (83% of the total energy in telephony sound is in this part of
the band). But the lowest frequencies contain mostly vowel sounds (300
to 1100 Hz), and have the *least* contribution to intelligibility of the
conversation -- the consonants are in the range of 2000 to 3000 Hz and
are far more important for intelligibility.

Even though these add-on amplifiers have their own power source (AC wall
power pack or batteries), their power is not infinite, and I've found
that as the boost is increased to maximum, the designers emphasize the
amplification of the higher frequencies, and de-emphasize the lower
frequencies, so they roll off below 1000 Hz. This allows a higher
apparent amplification, as it takes less energy to boost the higher
frequencies, and that's more important for intelligibility.

So they are deliberately introducing a form of distortion of the
original phone signal (low frequency rolloff) because it lowers the
amplification power requirement, and most likely actually improves
intelligibility (at the cost of fidelity -- peoples' voices no longer
"sound like themselves").
-----------------------------------------


3) Network-based amplification

This would involve a caller sending some kind of signal back to the
phone system, saying to increase the volume so they can hear better.
Perhaps by pressing a DTMF key.

If this is internal on a private phone system (PBX or key system) this
is not illegal, but it could cause problems. If the system is analog,
by boosting the signal power to the phone, it would soon cause crosstalk
from that phone pair to other phone pairs, and the conversation would be
heard on adjacent phone lines (crosstalk).

It is not feasible at all on a digital phone system (PBX or key system
or PSTN) because the G.711 codec has a defined upper limit on how "loud"
a sound it can carry.

Also, if the call goes across the PSTN (perhaps a caller at a home
calling aross the PSTN to a PBX) it violates the US/Canadian telecom
regulations.

Specifically, TIA-968-A Section 4.5.1 if the sound being amplified is
being played back from stored voice, i.e., a voice mail system.
https://www.part68.org/SecureDocuments/TIA-968-A-Final.pdf

And it violates the TIA-968-A-1 Table 4.6 "Allowable Net Amplification
Between Ports" if the sound being amplified is live voice:
http://www.part68.org/SecureDocuments/TIA-968-A-1.pdf

Besides, the US/Canadian PSTN (and much of the rest of the world) is
fully digital now, using the G.711 codec. The highest signal power
which can be encoded in PCM G.711 is approximately +3 dBm, so that
wouldn't be useful for giving true gain boosts of 20-25 dB that people
are looking for!

So high gain amplification for hearing-impaired users has to be done at
the phone instrument itself.
-------------------------------

WebAIM is an initiative of:
Center for Persons with Disabilities (CPD) Utah State University