home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.dcom.isdn
- Path: sparky!uunet!zaphod.mps.ohio-state.edu!pacific.mps.ohio-state.edu!linac!att!cbnewse!cbnewsd!varney
- From: varney@cbnewsd.cb.att.com (Al Varney)
- Subject: Re: end-to-end digital on ISDN -- the real application
- Organization: AT&T Network Systems, Lisle, IL
- Distribution: na
- Date: Fri, 8 Jan 1993 18:04:18 GMT
- Message-ID: <1993Jan8.180418.20998@cbnewsd.cb.att.com>
- References: <PICONE.93Jan7174958@trdc001.trdc.ti.com>
- Sender: Al Varney <varney@ihlpl.ih.att.com>
- Lines: 80
-
- In article <PICONE.93Jan7174958@trdc001.trdc.ti.com> picone@trdc001.trdc.ti.com (Joe Picone) writes:
-
- >Perhaps I should explain our real application and let you ISDN wizards
- >out there solve the problems...
- >
- >I am interested in collecting large speech databases that represent
- >good samples of the variation of speakers, telephones, channels, and
- >the ambient conditions found in the "normal telephone network"
- >(whatever that means these days). Of course, we cannot sample all
- >dimensions of the problem (collect data in every city, from every
- >carrier , etc.).
-
- If that's the case ("...variation of ... ambient conditions"), why
- would you want to bias the results by collecting samples that attempt
- to bypass the D-A "reconversion" that occur for normal calls???
- I'd think you would want to collect samples from various callers AND
- from various types of line terminations. Are you sure that you aren't
- really interested in a "speech" database that DOESN'T include the effects
- of the telephone network?
-
- >Two scenarios we have used in the past are the following
- [details deleted]
-
- >Case 1 suffers from the fact that at the receiving site there is a fixed
- >analog line from the local office to the data collection machine.
- >This does impose an additional and unnecessary A/D conversion.
-
- But it represents the "real telephone network"...
-
- >Case 2 suffers from the fact that the data collection system is a PC
- >and T1s are unmanageable for us novices.
- >
- >So, now that Sun Sparcstation 10s have ISDN interfaces built in,
- >we can replace Case 2 with an ISDN interface. Sounds too good to
- >be true...
- >
- >The question was raised: how does the data collected in such a manner
- >compare to the T1 connection in Case 2. My instinct is that doing this
- >type of data collection with ISDN will be as good, certainly no worse,
- >than the T1 case, and definitely better than Case 1.
- >
- >Others have raised concerns that there might be, on the average,
- >more re-conversions performed.
- >
- >ISDN gurus please comment...
-
- I'm not one, but that's never stopped me before!! :-) ISDN on YOUR
- end would be very similar to Case 2 in preservation of the signal. It
- obviously would be more flexible than the PC system (does it ever drop
- samples?? What does it do with the A/B bits, if anything?).
- ISDN could also be cheaper than the T1, but obviously depends on usage
- and SW Bell's tariffs. You could run a side-by-side trial for a while;
- have the T1 system direct callers to call another 800 number and do the
- same thing into the ISDN system.
-
- ISDN on the CALLERS end would remove the big three sound mutilation
- sources in the customer-to-EO loop (cheap phones & other junk on the line
- [FAX,modem,other phones], loop noise/loss/echo, customer's wiring).
- But that would mean your samples would NOT represent the "real world"
- telephone network.
-
- ISDN vs. T1 could result in an added "re-conversion" (in the LEC),
- but if you are on an ISDN-capable EO, that's not likely in Dallas.
- The primary AT there is a 4ESS(tm) digital switch, so even calls that
- aren't direct will almost certainly stay digital.
-
- As I've mentioned before, I think you are unduly worried about
- "re-conversions" at an analog switch. In metro areas (and many others),
- the analog switch + analog facilities will generally be very short
- (<500 feet) connections between the 2 A-D converters. If the converters
- are good, they should add a very small error into the total call.
-
- You are going to get more distortion from "Toys-R-Us(sm)" cheap
- telephones than from a typical telephone network connection, IMHO.
- Maybe you should have the callers identify the brand/type of telephone
- they are using....
-
- Al Varney - just MY opinion
-
- Al Varney
-