December 14, 2017, 02:33:15 pm *
Welcome, Guest. Please login or register.
News:
 
   Forum Home   Search Login Register OBiTALK  
Pages: [1]
  Print  
Author Topic: HD VoIP solutions for a new home  (Read 11579 times)
SeanTek
Jr. Member
**
Posts: 32


« on: December 28, 2015, 12:48:45 pm »

I am looking at rewiring a new home so that it is only-IP (furthermore: IPv6 preferred!). That means no need for "legacy" communications technologies such as coaxial cable and RJ11 phone lines. If possible, I want everything to run over RJ45/Cat6 Ethernet (with PoE), and 802.11n/802.11ac Wi-Fi. (Television/video services can be distributed over IP, such as with HDHomeRun boxes.)

As a technical user but a relative newbie to the VoIP space, I am looking for solutions that use standards-based technologies where possible, are not terribly expensive or hacked together, offer some kind of encryption, and support high call quality across devices. Closed-loop systems like Ooma are attractive for ease-of-use and claim to offer "encryption", but it is unclear what Ooma means by "HD" in their marketing, or how they get "HD" to work outside of the Ooma ecosystem. HD voice can be used by other closed-loop systems such as Skype or Apple FaceTime: they control the software at both ends! I want a path to interoperability.

What VoIP solutions do people here recommend? I like the premise of using Obi1062 phones with Google Voice and other SIP providers (recommendations?), but it's unclear just how much better call quality will be. I would also like recommendations for ensuring HD voice (Opus codec or G.722) across different devices.
Logged
SteveInWA
Hero Member & Beta Tester
*****
Posts: 3967



« Reply #1 on: December 28, 2015, 04:39:41 pm »

Hi:

Wideband audio over VoIP ("HD Voice") is still emerging technology -- there is very little interoperability between carriers at this time.  Over time, as carriers make more progress, you'll eventually be able to call between, say, a Verizon or Sprint HD Voice-enabled handset and a SIP VoIP carrier, but it's not here yet.

Any of the OBi 10x2 IP phones would be fine for your application.  The all support HD Voice (OPUS and G.722 WB CODECs).  Calls placed between OBi IP phones over Obihai's own OBiTALK network use the OPUS WB CODEC, and they sound great -- the phones were specifically engineered with high-quality audio circuitry to support wideband audio.  I also tested calls between the two IP phones using two Callcentric SIP numbers (in other words, staying on the VoIP infrastructure, and not going over the PSTN), and they also used OPUS.  OPUS support will depend on your choice of service provider.  Calls that traverse the PSTN cannot support wideband audio, since the PSTN itself is limited to narrowband.

It's a similar story for encryption.  The OBi IP phones support it, but the service provider and other end would also have to support it.

OBi IP phones also support standards-based PoE, so use a PoE-equipped Ethernet switch, and you'll only need one cable to connect each phone!
Logged

SeanTek
Jr. Member
**
Posts: 32


« Reply #2 on: December 30, 2015, 01:17:56 am »

Thanks Steve for your informative response!

I guess that when two IP phones are using PSTN numbers on the same provider (both Callcentric, both Ooma, etc.), the provider does not have to transcode the audio traffic to a different format, so it can stay Opus throughout, is that right?

Also if one IP phone calls another IP phone using a SIP URI, no transcoding should be needed?

So the issue is traversing the PSTN, which seems to imply some decoding and re-encoding is necessary. I read up a bit (Googled it) on how the PSTN works, but can you or someone explain at what point in the PSTN narrowband is enforced, and how?

The whole PSTN thing seems very hand-wavy. If the PSTN is a vast network of junction points and communications links, and all those elements are digital, then I do not see why transcoding has to occur (and if so, transcoding to what encoding formats, exactly, if it's all digital). Even if the nodes do not support a particular codec, the nodes should just pass the digital data through.

If a link in the PSTN ciruit is an analog copper wire pair, then clearly some decoding on one end and re-encoding on the other end has to occur.
Logged
SteveInWA
Hero Member & Beta Tester
*****
Posts: 3967



« Reply #3 on: December 30, 2015, 02:08:01 am »

Hi:

"Transcoding" is not the issue.  Transcoding is when an already-digitized signal, which was digitized using one standard format, is decoded and re-encoded into another digitized signal.  For example, CD-Audio transcoded into MP3 files.

Short history lesson:  The telephone network, from the time of Alexander Graham Bell, was purely analog, consisting of a limited range of audio frequencies being carried over copper wire "loops".  For a local call, the loop was between your premises and the central office.  Between central offices were "long lines" (simply more pairs of wires).  At first, telephone operators (picture Lily Tomlin) plugged cords into jacks to connect the calls.  Later, electromechanical stepping relays switched the calls for direct-dialed calls.  Thus, the name Public Switched Telephone Network, PSTN.  Each call was "circuit-switched":  an end-to-end connection patched together.

Over time, this meant more and more bundles of wires, until the system couldn't handle it.  Bell Labs eventually developed digital telephony.  Each conversation was run through an analog-to-digital converter, changing it into a pulse-code-modulated stream of bits (similar to the bitstream on those Compact Discs).  Those bits were multiplexed (combined) with other conversations' bits and carried over the wires, and later, microwave towers, satellites and fiber optic cables.  To conserve expensive and limited bandwidth, and, importantly, to deal with the physics of signal loss and distortion, using the technology available at that time, scientists looked at the average frequency range of human speech, and decided how much of the bottom and top-end they could throw away, leaving an intelligible mid-range.  This became the standard used to this day over the PSTN, the PCM G.711 CODEC, which is a 64Kbps digitized bitstream, encoding frequencies at 8,000 samples per second,  between 300Hz and 3400Hz

All the equipment in the PSTN is designed only to handle 64Kbps narrowband -- every electronic component, and other piece of hardware and software is only engineered to carry that level of frequency range.  It simply isn't compatible with wideband audio.  Think of it as the least-common-denominator, similar to AM radio -- AM radio has a very narrow bandwidth, by design, and FM radio has a wider, but still-not HD bandwidth, until digital HD radio came along.

VoIP doesn't have that restriction, since it uses packet-switched (not circuit-switched) RTP protocol over UDP (one of the simpler protocols in TCP/IP), which doesn't care what bits are being sent (it's just buckets 'o' bits, as far as UDP is concerned).  So, any CODEC supported by the equipment at both ends can be used to transfer packetized data.  It doesn't care if it's Morse Code, or a HD YouTube video, or a HD Voice phone call, as long as you have adequate bandwidth on your internet connections so the connection doesn't become saturated (like you see when trying to watch HD Netflix video when everyone else is doing the same thing at once).

Thus:  IP phones, or any other VoIP hardware or software, can communicate using whichever CODEC they both support, as long as the call stays purely as TCP/IP data packets.   This means, using SIP, for example, one endpoint sends a "HELLO" to the other's URI, and then the endpoints negotiate and set up the call.  Once this is done, the bitstream just flies over the network.  

Commercial SIP VoIP service providers (Such as Callcentric) support calls either being sent purely as SIP, URI to URI, or via the PSTN.  Only the former method supports HD CODECs.  In the latter case, calls are limited by the bandwidth specification of the PSTN.  Services like OOMA, Skype and Google Hangouts similarly can either use direct TCP/IP or can step down to the PSTN as needed.

As a final note:  the large telcos that have to support the FCC-regulated PSTN aren't really making much if any money on it anymore.  Verizon and AT&T have dumped a lot of their landline business, and most of the legacy "Baby Bell" companies have merged to survive.  CenturyLink and Frontier are two good examples of companies that are scraping along, with piles of landlines, much of which they inherited from bankruptcies and dumping by other telcos.  The industry would like to get rid of the PSTN entirely, because it is so much more expensive to maintain than VoIP, but due to the hundreds of millions of people who depend on it, especially in rural areas, it's not going away anytime soon.
« Last Edit: December 30, 2015, 02:18:56 am by SteveInWA » Logged

SeanTek
Jr. Member
**
Posts: 32


« Reply #4 on: December 30, 2015, 02:37:17 am »

All makes a lot more sense now: namely, that even in the digital parts of the PSTN (which is to say, pretty much all of it these days), all the equipment is engineered to handle the PCM G.711 codec. Is it that the G.711 codec or its technical features (64Kbps, 300Hz-3400Hz, etc.) are essentially hardwired into the protocols, or rather, that there are codec selection mechanisms (e.g., in SS7) but G.711 is the only common codec that all parties can be expected to support?
Logged
drgeoff
Hero Member & Beta Tester
*****
Posts: 2721


« Reply #5 on: December 30, 2015, 05:17:37 am »

Even G.711 is not a true common denominator as there are two codecs, A-Law and mu-Law.  Each is used in large parts of the world.  Transcoding is necessary where they meet though it is a relatively simple operation.

For the record, before the introduction of PCM and Time Division Multiplexing there was very significant use of Frequency Division Multiplexing in which baseband analogue voice signals were single sideband modulated and stacked on carriers at 4 kHz intervals.  The complete signal would then be transported over coaxial cables or microwave links.
Logged
SeanTek
Jr. Member
**
Posts: 32


« Reply #6 on: December 30, 2015, 09:32:21 am »

For the record, before the introduction of PCM and Time Division Multiplexing there was very significant use of Frequency Division Multiplexing in which baseband analogue voice signals were single sideband modulated and stacked on carriers at 4 kHz intervals.  The complete signal would then be transported over coaxial cables or microwave links.

Got it. And FDM necessarily implies that the original analog voice signals are frequency-limited to a predetermined range, in this case, the range common to most of the human voice.

Prior to the introduction of these PCM and FDM techniques, there literally was one (effective) electromechanical circuit per phone call from one end to the other. I guess that would be up through the mid-1960s or so. Was audio quality markedly better, worse, more wide-band, or just different, back in those days? For those who remember, anyway. Smiley

It stands to reason that if it's just two copper wires, what really matters is the quality of the telephony equipment on either end of the line...no different than if you connected a mono ⅛" headphone jack over RJ11-style wires (modulo resistance and power issues, which would have to be matched). But for any telephony system operated by a common carrier, there would be power injectors, filter banks, surge arrestors, and all sorts of other equipment that would change the voice quality.
Logged
SteveInWA
Hero Member & Beta Tester
*****
Posts: 3967



« Reply #7 on: December 30, 2015, 10:31:38 am »

I don't know where you live, or how old you are, but in the good old days of copper loops and mostly analog equipment, call quality varied considerably, based on all the many factors that can impair any analog signal, just like the wide variation in sound quality of FM radio before it was available in digital format.  Even under the best circuit conditions, the telephones themselves weren't designed to reproduce "high fidelity" frequency bandwidth.  Remember, the microphones used to be made out of carbon  Roll Eyes.  The Bell system developed a standard to test and rate call quality, the Mean Opinion Score, or MOS.  It was an opinion score, because they had a panel of listeners rate the call quality, which, after all, was the goal:  reproduce "good enough" human voice.  The MOS for VoIP is now a calculated value that includes jitter and packet loss.  Read more about it here:

https://en.wikipedia.org/wiki/Mean_opinion_score

A typical "land line" phone call now sounds much like it did decades ago, in terms of bandwidth, but without the static, interference, crosstalk, etc.  Thinking back to the first digital PBX phone systems in office buildings, I remember how amazed we were over how quiet and clear the calls sounded, and we never thought "hey, this is narrowband".

G.711 is designed with a maximum MOS of 4.1.  I haven't seen this measurement applied to WB audio.  Let's not get into one of those audiophile debates over whether digitized audio sounds worse than analog audio, ok?  Suffice it to say that most people are now so accustomed to the highly variable, and often miserable sound quality of mobile phone calls, that most younger consumers don't think much about it...if they're even still talking on their phones instead of texting and playing games while they jaywalk in front of traffic.

This has gotten pretty far off-topic to your original goal, which was selecting telephone equipment for your new home.  Suffice it to say that any of the OBi phones would be fine choice, since they support both narrowband and wideband audio, from both a digital standpoint (selection of CODECs) and from an analog standpoint (audio circuitry engineered to reproduce wideband audio).  Being firmware-based, if some future WB audio CODEC came along, it wouldn't be very difficult to upgrade the devices to support it.
Logged

drgeoff
Hero Member & Beta Tester
*****
Posts: 2721


« Reply #8 on: December 30, 2015, 10:47:41 am »

You might find https://www.theitp.org/historic/Short_History interesting. Although the details and dates are specific to the UK, the general principles and techniques were fairly universal throughout the world.
Logged
SeanTek
Jr. Member
**
Posts: 32


« Reply #9 on: December 31, 2015, 08:30:04 pm »

You might find https://www.theitp.org/historic/Short_History interesting. Although the details and dates are specific to the UK, the general principles and techniques were fairly universal throughout the world.

Thank you! Yes, the read was interesting.
Logged
SeanTek
Jr. Member
**
Posts: 32


« Reply #10 on: December 31, 2015, 08:58:53 pm »

I don't know where you live, or how old you are, but in the good old days of copper loops and mostly analog equipment, call quality varied considerably, based on all the many factors that can impair any analog signal, just like the wide variation in sound quality of FM radio before it was available in digital format.  Even under the best circuit conditions, the telephones themselves weren't designed to reproduce "high fidelity" frequency bandwidth.  Remember, the microphones used to be made out of carbon  Roll Eyes.  [MOS...]

A typical "land line" phone call now sounds much like it did decades ago, in terms of bandwidth, but without the static, interference, crosstalk, etc.  Thinking back to the first digital PBX phone systems in office buildings, I remember how amazed we were over how quiet and clear the calls sounded, and we never thought "hey, this is narrowband".

G.711 is designed with a maximum MOS of 4.1.  I haven't seen this measurement applied to WB audio.  Let's not get into one of those audiophile debates over whether digitized audio sounds worse than analog audio, ok?  Suffice it to say that most people are now so accustomed to the highly variable, and often miserable sound quality of mobile phone calls, that most younger consumers don't think much about it...if they're even still talking on their phones instead of texting and playing games while they jaywalk in front of traffic.

This has gotten pretty far off-topic to your original goal, which was selecting telephone equipment for your new home.  Suffice it to say that any of the OBi phones would be fine choice, since they support both narrowband and wideband audio, from both a digital standpoint (selection of CODECs) and from an analog standpoint (audio circuitry engineered to reproduce wideband audio).  Being firmware-based, if some future WB audio CODEC came along, it wouldn't be very difficult to upgrade the devices to support it.

Well I am an older Millenial, which is to say, I grew up around Apple IIs, 5¼" floppy disks, NES, and MS-DOS (5.0). Not vacuum tubes or punch cards, except we made holiday wreaths out of the latter. Grin I also do Internet-related software engineering, so I have a good understanding of IP, but the PSTN has (up until now) been a very nebulous cloud-thing in diagrams.

Ultimately I see how these are engineering exercises: bandwidth is costly, as are pieces of equipment at the nodes and at the endpoints. Especially when multiplied over trillions of voice calls of varying lengths. Even nowadays as some wireless carriers (in the United States) are turning on HD voice services with VoLTE/Advanced Calling, I believe they are opting for much less bandwidth-intensive algorithms. I read that T-Mobile, for instance, uses AMR-WB (G.722.2; possibly at 24kbps) for its "HD Voice"—a far cry from Opus or SILK. I don't know what Verizon Wireless is using but I read that it is limited to 13kbps.

Fortunately when we are talking about wireline services, I am not concerned with conserving bandwidth because there are no data caps.

Suppose I get an Obi1062 in advance, for testing it out (the new home issue shouldn't matter, as long as it's connected to a quality Internet connection). Other than other Obi10x2 devices, what would be recommended pieces of hardware and services to test interoperability with? The Obi1xx and Obi2xx products don't use Opus or wideband or "HD" in any way, do they? I.e., do the analog phone lines on those products have higher resolution than the 300-3400Hz range common to G.711? Probably not, that would be over-engineering for the purpose...it's also unclear to me if the Obi1xx/2xx products can be contacted directly via SIP or direct IP. Would one expect to see (i.e., codec use) or hear (subjectively) a difference when calling between Obi1062 and Ooma products, which are more explicitly marketed as "HD", even though Ooma-to-non-Ooma is supposed to go through the PSTN?
Logged
SteveInWA
Hero Member & Beta Tester
*****
Posts: 3967



« Reply #11 on: December 31, 2015, 09:18:13 pm »

Sean, happy new year!

I am puzzled as to why you are so focused on the topic of wideband audio.  As I said earlier, it's an emerging technology.  Currently, there is very little interoperability between service providers, so the only realistic use is for SIP calls between SIP clients.  The OBi phones are ideal for that purpose.  There's no point to testing anything in terms of inter-carrier capability.  There is nothing to test with at this time, aside from calls that stay on the VoIP infrastructure between VoIP service providers that support wideband CODECs.

At some point in the future, when wideband audio calls between various VoIP or mobile carriers is supported, you'll be all set.  The PSTN will not be upgraded to support wideband.

Why don't you just buy a phone and get familiar with the features and call quality as it exists today?  The 1062 is overkill for residential use; a 1022 or 1032 would be fine.  The phones' VoIP capabilities and analog components are identical between the three models.  The 1032 and 1062 models just have more buttons you can program, and support business-class headsets with RJ-9 jacks, and the 1062 has built-in WiFi and Bluetooth support, whereas the cheaper models require optional dongles for those radios.  Because the phones are engineered to support wideband audio (not just the digital VoIP parts, but the phone's analog audio components), you will find that audio quality on conventional calls is superior to typical consumer-grade telephones.

See the spec sheet for more details.  http://www.obihai.com/docs/OBiPhoneDS.pdf
Logged

SeanTek
Jr. Member
**
Posts: 32


« Reply #12 on: January 01, 2016, 08:39:20 am »

Sean, happy new year!

I am puzzled as to why you are so focused on the topic of wideband audio.  As I said earlier, it's an emerging technology.  Currently, there is very little interoperability between service providers, so the only realistic use is for SIP calls between SIP clients.  The OBi phones are ideal for that purpose.  There's no point to testing anything in terms of inter-carrier capability.  There is nothing to test with at this time, aside from calls that stay on the VoIP infrastructure between VoIP service providers that support wideband CODECs.

At some point in the future, when wideband audio calls between various VoIP or mobile carriers is supported, you'll be all set.  The PSTN will not be upgraded to support wideband.

Why don't you just buy a phone and get familiar with the features and call quality as it exists today? [...]

Happy New Year!

Well you say it's an emerging technology, but Skype has had wideband audio since the mid-2000s, and other non-VoIP audio and video applications have had it long before that. Skype's traffic was almost 40% the size of the entire conventional international telecom market in 2013. This begs the question: why are we mucking around with annoying VoIP setup things, when we should all just use Skype like regular people?  Roll Eyes

I was a very early adopter of Skype (~late 2003) and used it regularly to keep in touch with folks, until Microsoft bought it and the quality of the network went considerably downhill. Also Skype has not really kept pace with the mobile revolution; I have not been able to get the performance and usability out of its smartphone apps like it was with their classic Windows (and Mac) clients. Then of course there are the closed-source/obfuscation issues.

Honestly, if I want free calling, I can just use my mobile phone for everything. Having a residential land line at all is increasingly anachronistic, but it still makes sense for the following use cases:
  • better audio quality (avoid dropped or garbled calls)
  • working at your desk, aka home office (in which case, it functions more like a business phone line, so business use cases apply)
  • (VoIP specifically) can unify many different phone numbers (essentially a "business" use case)
  • call "the house" and page whoever is in "the house"
  • "intercom" (from your cell phone to the house)
  • doorbell substitute
  • unlimited talk: don't use up minutes or data plan

The big minus remains "no SMS" (a separate thread, I know it's possible, but won't go there right now).

As you see, better audio quality is at the top of the list. That is why I am focused on the topic of wideband audio: we just entered 2016 and it's rather appalling that the two-line RJ11 pair persists when our mobile phones and computers are capable of so much more. But there is a lot to be said for the simplicity of two wires for home users...not to mention five 9 availability (last I checked, dumb phones don't crash).

I actually have acquired several VoIP desk phones over the last few years, not to mention a variety of softphones. The latest are the Grandstream GXV3275 (Android touchscreen phone) and the Grandstream GXP1450. (I also sold off a Grandstream GXV3240.) The GXV3275 is a bit of a dog (slow, doesn't install additional Android apps) and leaves me wondering why I didn't use the money to buy an iPod Touch, which would have been cheaper, a lot smaller, and would let my family members play games on it, not to mention FaceTime, which just works™. The GXP1450 is cheap but not particularly effective for HD. Firstly, it doesn't have a way to enter SIP URIs or direct IP addresses directly from the keypad, so VoIP-to-VoIP outbound calling is hindered. Secondly, it only supports G.722. Thirdly, it randomly disconnects from my current SIP provider after a day or two. Fourthly, it only supports two accounts so trying to test multiple SIP providers against it at once is very laborious.

None of these concerns would really exist if these phones were hooked up to a real Asterisk PBX, as the phones could be configured with XML or whatever management technologies are around, but I am not interested in running my own phone server for my residence.

On the choice of OBi1062: the price difference on Amazon is pretty negligible in the long run, and I don't want to deal with extra dongles for Wi-Fi and Bluetooth. Plus the OBi1062 has gigabit ethernet (including the extra port for other devices) and 6 line keys, which make it more convenient for switching between services. (As I hinted at, we have a Google Voice number whose sole purpose in life is for the "doorbell" function.)

I guess at this point, I am deciding which "science project" to tackle next: an Obi-based platform or a Wi-Fi only smart device, aka iPod Touch. As of this writing, a new pink/yellow/blue iPod Touch is cheaper (at $149, Walmart)! Plus it will do Skype/FaceTime/videoconferencing, and can even turn on my lights, control the thermostat, and manipulate my home theater systems, all with the appropriate apps!  Grin The biggest downside is that said iPod Touch will have a tendency to disappear at inappropriate times, while a "real" IP phone won't be disappearing on me since it is tethered to the wall.
Logged
SeanTek
Jr. Member
**
Posts: 32


« Reply #13 on: January 04, 2016, 05:59:49 pm »

To follow up with this, I am going to give Obi1062 a go. I realized that I can simulate the "iPod Touch in every room" with an iPhone, but not the other way around. Cheesy So Obi here we go!

Now there is the issue of service providers to choose from. I suppose that the Approved providers are all reasonable, but can anyone suggest at least the first one to start with? I would rather not open up three or five starter plans at once.

Ideally, I would like a provider that supports T.38 properly (or supports color faxing, don't ask, it's complicated...most e-fax providers perplexingly do not support color faxing) for occasional but reliable use, has voicemail, permits SIP URIs for direct SIP-to-SIP calling (see desire for Opus/HD audio codecs), and supports SMS somehow. SIP SIMPLE / XMPP would be great for SMS but from what I have read, this may be too much of an ask. I guess if I have to, I will settle for a web portal for the SMS bits.
Logged
SteveInWA
Hero Member & Beta Tester
*****
Posts: 3967



« Reply #14 on: January 04, 2016, 06:24:23 pm »

I use and recommend Callcentric, as do a lot of other forum members.

CC owns their own Competitive Local Exchange Carrier (CLEC), and they offer free inbound phone numbers (DIDs) in several NY State area codes.  Of course, you can also obtain a DID in your own locale if you wish, for a monthly fee.  Similar to other SIP ITSPs, they offer a la carte (separate) pricing for inbound or outbound calling.  You can start with pay-per-minute pricing to see if the service meets your needs, and if so, you can either stay per-minute, or select a monthly plan, based on usage needs.  Of course, none of this is on-contract, so you can change or cancel any time.  Voicemail is included, fax via G.711u works, and they also offer dedicated inbound DID fax mailbox service.  I've tested and confirmed that calls using OPUS work between CC numbers.

Voip.ms is another excellent choice.  I am not a customer, and so I can't comment on their support for WB audio CODECs.

The 10x2 phones support URI dialing, although doing so from the GUI would be difficult.  You can define speed dials on the phone to call URIs.

You can also get a Google Voice phone number and configure it on one of the phone's other SP slots.  GV generally works well for faxing, and GV numbers support text messaging.
Logged

SeanTek
Jr. Member
**
Posts: 32


« Reply #15 on: January 04, 2016, 11:24:18 pm »

Well, about Callcentric...
I am currently on their IP Freedom plan with a NY free phone number. Unfortunately the test went very poorly, both with VentaFax (and more recently, with Bria for iPhone). VentaFax is Windows software that sends and receives faxes, including color faxes--it's like one of the few fax apps on the market that actually does that, plus it includes direct SIP connectivity with T.38 and G.711 options. Strangely enough, when VentaFax picks up the SIP phone via Callcentric, the fax tone is the wrong tone. Instead of the reception tones (which sound like a high-pitched tone for a long second followed by shrill fluttering chirps), the Callcentric end emits the transmission tones (which sound like a short medium-pitched BEEP, repeating every second or two). My sending fax device therefore sits there emitting the BEEP, but it hears the same BEEP on the other end, so the endpoints basically continue BEEPing at each other and no fax transmission actually begins. Shocked For the record, I tried both T.38 and G.711 options.

I tried t38fax.com (a FoIP-specialized provider), and t38fax.com gets it exactly right. This should not be rocket science...anyway, that is my Callcentric experience. It's possible that the VentaFax FoIP software is messing it up, but I highly doubt it.

(In fairness, Callcentric's Fax Reception feature worked fine for receiving black & white faxes, if I want to dedicate a number solely to fax. Fax Reception for color faxes was a bust: it errored out without sending any pages.)
Logged
SeanTek
Jr. Member
**
Posts: 32


« Reply #16 on: January 10, 2016, 11:05:33 pm »

An update: I got as far as selecting Anveo as my provider: its feature list seemed to hit all the right points (SMS, fax, optional G.722). Unfortunately I have not been able to figure out how to set up additional SIP accounts to link additional (non-OBi) devices and softphones: in fact it appears to be hardwired to just one OBi phone. Unless there is some way around this, it looks like Anveo is out.

I guess one option is to switch to the open-ended "Starter" package that is not OBiTALK-specific, which shows that it allows up to 3 users/SIP accounts. But with the line fees and per-minute fees, the price is a bit steep.
Logged
drgeoff
Hero Member & Beta Tester
*****
Posts: 2721


« Reply #17 on: August 11, 2017, 10:18:31 am »

Prior to the introduction of these PCM and FDM techniques, there literally was one (effective) electromechanical circuit per phone call from one end to the other. I guess that would be up through the mid-1960s or so. Was audio quality markedly better, worse, more wide-band, or just different, back in those days? For those who remember, anyway. Smiley
There was low pass filtering in play long before FDM and PCM.  Longer circuits over pair cable had loading coils. https://en.wikipedia.org/wiki/Loading_coil
Logged
Pages: [1]
  Print  
 
Jump to:  

Powered by SMF 1.1.11 | SMF © 2006-2009, Simple Machines LLC