Wireless Notes -- Digital CellularVoice Latency/Voice-over-IP Latency and Voice Delay caused by Digitization Discussion and Cellular Carrier Comparison of Voice Latency between AT&T, T-Mobile, Sprint, and Verizon:

How CDMA (Sprint/Verizon), GSM (AT&T/TMobile) and other cellular protocols affect conversational latency and delay when calling from cellular phones to traditional landlines, other cellular/mobile phones, and to/from Voice-over-IP phones


[ Wirelessnotes Home | Interpage | Cellular Carrier Comparison | Digitized Latency Overview/General | Test your phone's Latency | Disclaimer | Verizon Dropped Call List | Verizon 3G/4G Data Drop List | Sprint Dropped Call List | T-Mobile Dropped Call List | AT&T Wireless Dropped Call List | Nextel Dropped Call List | Contact WirelessNotes ]


      Interpage OutVoice
Text-to-Speech and WAV voice broadcast and messaging service
promitional link. Click here for additional details on the
Interpage OutVoice service. Interpage DirectPage
service to replace or augment numeric pages by sending numeric
pages to cellphones/SMS, multiple pagers, email, voice and fax.
Click here for additional details. Interpage Voicemail service which
offers notification of hangup calls and all other messages received with
CallerID to cellphones, SMS, email and pagers, and will even send a
voice/WAV file to your phone or email and call you with your new
messages!. Click here for details.


General:

Overview: A Brief discussion on the latency and delay experienced in cellular, Voice-over-IP, and other digitized telephone end devices and how it affects and degrades normal voice conversation.

(Note: Before reading the discussion below, you may want to call in to hear and test the latency on your phone and/or of your current carrier; the dial-in voice latency test is available 24-hours a day, all year long, at: (802) 359-9100 (this is a regular, local number), and follow the recorded instructions. You may call with your number blocked if you wish for privacy purposes, but we in any case do not market or sell any information about you, nor will we contact you by any means to market to you.)

Have you ever placed a call on a cellphone/mobile and found the conversation with the called party to be "difficult" or strained, or just not the same as conversing in person or even on an older landline phone?

Although digital cellular telephony has allowed for smaller handsets with significantly longer battery time than analog mobile phones had previously offered, a significant trade-off has been the imposition of "latency" on every voice conversation which takes place via cellular service in the US and Canada (we are not aware as of October 2014 of any analog cellular systems still in operation on the North American mainland). (Additional details about US cellular carrier audio and call quality comparisons are available on the Cellular Carrier Comparison page, detailing, in part, audio differences of CDMA vs. GSM, as well as latency and audio delay issues of AT&T Wireless, Verizon, Sprint, and T-Mobile).

The same is true for Voice-over-IP (VoIP) residential (and to a lesser extent office) service, where the sometimes lower rates for monthly service can be offset by significant degradation in terms of voice latency (and other qualities which make it often inferior to traditional landline/"circuit switched" service).

This is not to say that all digital communication is "bad" per se, and if fact, most if not all long distance calls have been transmitted via digital trunk lines (lines which carry multiple calls for generally longer distances) for many years in the US and Canada, but these were designed to offer nearly no latency (similar to the end-to-end immediacy of the analog long-distance trunks of the Bell System and similar carriers), yet when implemented starting in the early 1980's, offered a significant improvement to the call quality the of the analog long-distance trunks which had previously been used for long-distance telephony in the US/Canada, and internationally as well.

Some readers may recall long set up times after dialing a long-distance number and a background "hiss" or "white noise" during a call; these were gradually eliminated with the implementation of digital trunks starting in force during the early/mid-1980's which made long-distance calls sound just like local calls. (Some carriers, such as AT&T, did a much better job of it than others, such as MCI, which was taken over by and whose network appears to be in part used by Verizon, and the difference in quality between a long distance call on AT&T as compared to the less-than-satisfactory call quality on Verizon (especially internationally) is often obvious.) The effect of digitized trunks (often, but not exclusively, using fiber optic cables) was even more noticeable on international calls, where the extreme latency of a satellite connection (often 2 seconds from the time the caller said "Hello!" to the time it was received at the distant end!) often rendered a call useless for ordinary conversation as any immediacy was vitiated by constant echoes and interjections of "Yes? Did you hear that?" due to the delay in receiving a response.)

Effectively, digital "switched circuit" telephony allowed for long distance and worldwide (IDDD) calling which sounded as good and as immediate (no latency) as a local call.

Cellular service underwent a similar transformation -- when cellular mobile telephone service was introduced in the US/Canada in the early to mid-1980s, all _voice_ connections were analog and thus offered effectively immediate voice conversation with no latency. The Advanced Mobile Phone Service (AMPS) used FM-radio signals for voice conversations, and a digital channel to control which frequencies a given mobile phone would access (as well as to perform ring, dialing, and other control and administrative functions). In effect, analog cellular, or AMPS, when implemented well, offered very similar call characteristics to a landline phone -- immediate conversation (no latency), no digital distortion (a metallic or "robotic" "twang" often evident in CDMA and to a lesser extent in GSM calls), and generally good voice quality. Unfortunately, like FM radio, AMPS's FM signaling meant that mobile calls would be subject to static and drop-outs similar to what a motorist would hear on their FM radio while driving. While somewhat inherent to analog communication, the sound quality issues were exacerbated by the (then) A and B carriers' hesitance to make significant investments both in call-quality and service area build-outs (eg, to cover the entirety of their service license areas or their "de-minimis" service contour requirements in FCC parlance) as they were at the time not convinced that AMPS would be a significant source of revenue, and the effective duopoly of the "wireline" (B) carriers (such as NYNEX, Bell Atlantic, Pac*Bell, etc.) and "non-wireline" (A) carriers (such as "Cellular One", "Dobbson Cellular", and "Metro Mobile":) and lack of significant competition outside of the duopoly allowed the A and B carriers to be lethargic and slow to add coverage to any area other than certain highway corridors, very large cities, and other areas where improved service use would immediately translate into higher revenue.

As service slowly expanded under the A/B carriers in the late 80's and early 1990's (hindered to no small part by a federal regulatory apparatus which did not keep up with the technology and treated AMPS as merely an improved version of the earlier and significantly more limited Mobile Telephone Service (MTS, 1940s) or Improved Mobile Telephone Service (IMTS, 1950's) of Bell and Independent service providers (especially with respect to roaming and system integration), as the subscriber base of cellular customers expanded, some carriers were approaching the limits of their network's ability to handle all calls in a given area, and blocking and failed calls (similar to IMTS' problems) became an increasingly prevalent issue and impediment to good, reliable mobile service.

In many cases, had a given carrier wanted to make appropriate investments in its analog network, by dividing cells into smaller areas, providing more cell sites, etc, many of these congestion and blocking issues could have been mitigated, but by the mid-1990's, customers were demanding smaller, handheld units which, due to their lower power, were even more prone at times to the vagaries of FM radio, as well as phones with longer battery life. (There were also fraud issues as well due to the less secure nature of ESN validation on analog networks). Carriers wanted a cheaper way to expand their networks without building out many, many more analog towers/sites, and thus the industry turned to digital cellular technologies to address these issues.

With the early-2000's introduction of digital cellular service in the US and Canada, and the subsequent phase-out of analog cellular commencing in 2007, a degree of immediacy and "real-time" communications was withdrawn from the North American voice/mobile landscape. When compounded with the decreasing number of traditional landlines and the increase in mobile-to-mobile (or mobile-to-VoIP) calls, a significant and noticeable delay or latency has become increasingly evident in (voice) telephony, and is in many ways a step backwards from the immediate, high-quality calls which were commonplace only a few years before.

In nearly all cases, calls placed to/from cellular/mobile phones use one of two digital protocols: GSM (which is also what most of the rest of the world predominantly uses), and CDMA (used by some carriers in the US/Canada, some carriers in China, and in a decreasingly small set of other countries; CDMA and GSM will likely eventually "merge" into one single protocol at some point in the distant future), with Voice-over-IP carriers (like Vonage) employing a variety of protocols (including, at times, GSM for the digitization of the audio component of VoIP calls) as part of a larger set of protocols collectively named "SIP". Generally speaking, in terms of cellular voice calls, the GSM digital protocol (used in the US by TMobile, AT&T Wireless, and some other smaller regional carriers, resellers, and MVNOs) offers sound quality and an overall call "experience" - in areas of strong coverage/signal strength - which is noticeably superior to the CDMA protocol (used by carriers such as Verizon and Sprint), and sounds much more like a traditional landline/local phone call.

While both protocols (GSM for T-Mobile and ATTWS, and CDMA for Verizon and Sprint) no longer suffer from analog static and bounces off of buildings (multi-path), when all other things (such as coverage and signal propagation) are equal, GSM clearly sounds better, more lifelike, and more like a traditional landline call, while CDMA often sounds "digitized", "metallic", and highly processed to the point that is is clearly obvious that the caller is using a CDMA cellphone. Music is also much more distorted over a CDMA phone and often "fades out" during long "music on hold" periods, giving the caller the impression that the called party who put the caller on hold has hung up.

As for VoIP calls, quality suffers to varying degrees based on the service provider, type of compression (GSM or others), and the customer's internet connection (speed, latency, packet loss, intergration and connection points to the internet backbone, etc, all of which do not currently affect cellular/voice calls), and thus call quality is very "hit or miss" depending on these factors as well as the overall quality of the given VoIP carrier. As the issues affecting a limiting the quality of VoIP calls are often beyond the control of only the carrier, and depend to a significant degree on the given customer's internet connection and that connection's carrier, most of the following discussion will treat VoIP as an aggregate service with some degree of latency over traditional digital long-haul telephony, the specifics of which will be relegated to another post. With respect to digital cellular voice telephony, one of the most significant conversational and voice-quality advantages offered by the GSM protocol is a somewhat lower "latency" than CDMA. That is, if a T-Mobile customer were to call a traditional landline, and say "Hello!", the called landline party may have a 1/10th of a second (100 millisecond) delay from the time "Hello" is uttered on the TMO phone to the time the called party hears the word "Hello". That is, it takes 1/10th of a second for the cellular/T-Mobile caller's voice/audio (the word "Hello") to reach the called landline party on their phone due to the inherent delay incurred in the digitization of the word "Hello" on the cellular phone, and the subsequent conversion back to analog on at the cellular carrier's "switch" for transmission over the landline network to the called party. (Note: these delay/latency times are not necessarily exact measurements, but are used to exemplify the differences in audio delay between landline, VoIP, TDMA and CDMA phones). CDMA phones offer a longer delay/greater latency (not by much, but noticeably, especially when a CDMA phone calls another CDMA phone), which, combined with the other "unnatural" voice transmission characteristics mentioned above, makes CDMA noticeably more difficult to have "landline-quality", natural conversations with.

As both GSM and CDMA are require digitization of voice on the cellphone/handset itself and conversion back to analog when the signal reaches the subscriber's carrier, latency is thus _inherent_ in both GSM and CDMA digital protocols (or any digital protocol), as it takes time for the analog (voice) sounds uttered into the cellphone to be converted to digital signaling (digitized) and the converted back to analog for transmission to the dialed number.

(Note: Most if not all calls in the US/Canada are digitized - cellular AND landline; however, the "codecs" (processors which digitize and "un-digitize") used in traditional landline telephone, such as long distance, are significantly faster and sample voice and considerably higher rates, which makes the digitization "effect" on long-distance calls un-noticeable.) While the latency from a single cell phone (GSM or CDMA) to a landline may not be noticeable by some, on cellphone to cellphone calls, or cellphone to Voice-over-IP phone calls (Voip, also digital, suffers some latency as well, which varies based on the given Voip carrier and the data network/internet connection used to carry the Voip call), each additional digital "step" adds more latency, making the delay more noticeable.

As an example, and to demonstrate and test the latency of a given cellular carrier (or Voip carrier, or any digital voice carrier), have "person A" and "person B" in the same room. "A" should then use a cellphone to place a call to "B", who will answer on a traditional landline (that is, local telephone service from an ex-Bell or independent telephone company like Verizon, ATT, US West, Frontier, etc. FiOS and Cable-Modem telephone service are OK as well since they offer (or come close offering) "telephone-company" "circuit switched" audio and latency characteristics for voice).

Cellular and Digital Telephony Latency chart 1, 
indicating no latency in direct conversation in a room

Although there is some minimal latency when two people speak to each other in a room due to the speed which sound waves take to travel through the air, the delay isn't noticeable, and we'll use that as the standard for "no latency" which is expected in conversation. (As depicted in Figure 1, above)

Now, to continue the "test", get two traditional landlines (no VoIP or FiOS VoIP; copper or Fiber-to-the-Curb replacements are fine) in the same room, and call from one to the other -- there will no appreciable delay, and "B" will hear "A" at the same time over the phone as by the sound waves transmitted via the air vibrations (talking) in the room. (Analog cellular was also like this, offering no appreciable delay in voice conversation.)

Cellular and Digital Telephony
Latency chart 2, indicating no latency difference between in-room
conversation and conversation which takes place over traditional local
landline (POTS) service

Figure 2, thus shows no effective difference in latency or conversational delay when two people are speaking in the same room or speaking to each other over a traditional, local landline ("POTS") call.

(NOTE: The test assumes that two people try it in a room/office which has two local landline "POTS" phones (NOT Voice-over-IP phones), and they are NOT extensions of the same line in different rooms. To accurately guage latency (or the lack thereof) over local landline service, there need to be two physical lines, one of which will call the other.)

Thus, if person "A" were to pick up one of the lines, dial to the other line, have "B" answer and say "Hello!", "A" would hear the word "Hello" simultaneously in the room/office as over the phone. There would be no noticeable delay or difference between the direct conversation in the room as compared to the one being conveyed via the analog/POTS telephone call. (In fact, if it were a large room and it took a while for the sound waves to reach from A <-> B, it may actually be faster over the phone than over the somewhat slower path via the air!)

Cellular and Digital Telephony Latency chart 
3 indicating the latency of conversation carried out over digital 
cellular phones as compared to more natural, latency-free conversation 
taking place between two people in a room
In Figure 3, as "A" speaks to "B" in same room, "B" will hear "A" first directly, and then over the phone with a slight delay, demonstrating the latency of digital cellular communications. If "person C" were to enter the room, and "B" were to place a three-way/conference call to "C"'s (digital) cellphone, then additional latency from "A" to "C" would be present, and "C" would hear "A" say "Hello!" in the room noticeably before hearing it on his cellphone. (NOTE: This depends on how "B"'s carrier conferences two calls; the additional delay to "C" may not be a linear function as the conference call to "C" may directly connect "C" to "B"'s connection at the cellular switching center, thus not introducing any additional significant delay. This is less likely to be true for VoIP and the delays will be much more noticeable and linearly compounded with additional conferees).

Cellular and Digital Telephony Latency chart 
4 indicating the compound effect of digitally-induced latency in 
three-way or multi-party conversations further increasing latency and 
adversely affecting normal conversation
By placing a series of tests like this, most cellular (and VoIp) subscribers may see for themselves the degree of latency introduced by their respective carriers. In many cases, the delay is quite noticeable, and thus quite distruptive to normal, two-way conversation.

Alternately, as noted above, you may call +1 (802) 359-9100, which will perform a "voice/echo test" which will immediately play back whatever you say as soon as our system "hears" your voice. This will allow you to try different phones and carriers to see which has the least latency when speaking. (The latency/echo test uses a digitized signal for part of the path to us, and will introduce a very slight delay, of 10 milliseconds or so, which is generally undetectable; however, it is technically not the functional equivalent of a local call via traditional landline service over copper, preferably run through an analog exchange (not that any seem to exist anymore...:( ).

In general, after performing these admittedly somewhat less-than-controlled-tests, the order of latency, from lowest to highest, would be:

  • 1. Traditional local phone-company landline ("local loop", "POTS") telephone service. Cable TV "digital" service comes close too, but the fastest (least latency) with the same conversation-like properties as talking to another person directly remains copper or similar "phone company" local service.

    (Note: The discussion below, pertaining to various "local loop" technologies, has grown beyond the scope of telephonic voice latency, and eventually will be made into its own article. For those not interested in the various nuances of local phone service and their effects on voice latency, or Verizon's manaical drive to FiOS or wireless and apparent hatred of copper, please skip to item 2.)

    As noted above, Voice-over-IP based carriers, such as Vonage, ViaTalk, or Google Voice, even though they may present themselves as "Phone Company replacements", all suffer from a good deal of latency as compared to traditional local service, which causes it to be difficult to carry on normal, "back and forth" conversations without the latency causing repeated interruptions and parties to the call cutting each other off and having to say (if they're polite! :) ) "Sorry, please go ahead...".

    Additionally, by "traditional" telephone or POTS service, we are not exclusively referring to copper, but to any other technology which approximates the sound quality and lack of latency as a direct copper pair of wires from a given POTS phone to the phone company's Central Office (CO). Thus, Fiber-To-The-Curb / FTTC services, such as AT&T's U-Verse, or some older Bell System techniques such as "Slick-96"/SLC-96 Subscriber Loop Carrier, eg, where the telephone company doesn't run a copper pair from the central office to each telephone, but instead has one or more T-1 voice circuits sent to a remote phone company box (sometimes with a little green or red light on it) and from that box runs copper wires to the subscriber over a much shorter distance, are all relatively latency-free and they were all designed to meet the strict Bell System or Bell Operating Company specifications for voice communication.

    Although the SLC-96 and other similar fiber based Fiber-To-The-Curb type technologies have had their problems, they generally perform well latency-wise and offer more or less the same degree of true sounding, delay-free conversation as a copper pair being run all the way from the Central Office.

    These services, when properly implemented, provide similar performance characteristics to POTS/copper lines all the way to the CO. As an example, New York Telephone invested heavily in a fiber-based SLC system in the early 1990s which was prone with problems, but eventually most were fixed and for voice service, offer nearly identical service characteristics as copper. (Other than generally not being able to support DSL; a good test to see if a line is copper through-and-through to the CO or if SLC or some fiber replacement is used is to inquire with the phone company if DSL is available - if it is, the given line is most likely copper all the way to the CO, if not, it is likely some SLC-ish service is in place. Some DSL variants can be offered via these SLC-ish type services, but they are generally slower and sparsley implemented.)

    Although there should be little difference in the service characteristics between well-implemented, well-maintained SLC-96 and newer fiber variants, one can discern between the two by looking at the local distribution boxes in a given neighborhood: The newer fiber variants can be seen as little green boxes with thick black cables (the fiber line back to the Central Office) on the back of houses and buldings, while the older SLC-96 boxes are generally on telephone poles with green/red lights clearly visible at night, are much larger, and as an older pre-fiber technology, use a few copper pairs from the CO to the box on the pole, which then distributes the lines to up to 96 subscribers in the given local area.

    As to Verizon's FiOS and similar all-fiber systems (eg, Fiber-To-The-Home, or FTTH), it depends which FiOS service is used. Verizon's FiOS is a trade name used to send cable-TV, phone, and internet/data over a fiber service which runs into the home, and from there the 'last mile' (or few feet) of service is from the Verizon Optical Network Unit (ONU) in the basement or utility closet, via household copper wire, to the various phones in the house. (This is also why FiOS isn't such a good idea in areas with power problems or frequent power outages - the ONU uses power provided from the customer (which also raises the customer's electric bill a bit), whereas the older SLC-96 and other Fiber-to-the-Curb technologies generally use power provided from the phone company central office, which lasts a lot longer. FiOS generally will last for 8 hours or less if the power goes out, which is not a very long time, especially at such times when landlines may be most needed.)

    FiOS is generally how Verizon brands the higher-priced bundled (TV, Phone, and Internet) service, and what they try to get people to sign up for (apparently often via less than scrupulous business practices -- a colleague of ours in New York City recently had his copper service go out, and -the very next day- somehow by -extreme- coincidence there was a FiOS desk with phone company representatives and lots of glossy FiOS literature set up in the lobby, with the oh-so-ever-friendly reps saying "Oh, we're getting rid of copper, so you need to move over to FiOS". Not only was this not true, and copper was not being abandoned, but the repair times quoted if one wished to keep the copper -a month!- were outrageous. Even more outrageous was that when one called Verizon/NYTel repair, instead of the usual disinterested rep at the "611" repair office, the (800) VERIZON IVR picked up that the calling customer was a "prime candidate" (victim?) for FiOS service and transfered the call to a perky lady from the Midwest who claimed to be repair, but was really just trying to covert the caller to FiOS, and said "Oh, I'm SO sorry you are having all these problems, but of course it's due to all that bad copper you still have there. But, wait! You're in so much luck! FiOS is available for you! I'm so happy for you!" When she was politely told that no one wanted FiOS and Verizon/NYTel just needed to fix the copper, she would not stop badgering about why she is "so concerned that you want to use such a terrible, old technology as copper..." (read: she was terribly concerned she wasted 5 minutes on the call and wouldn't be getting her commission!); the call got very tense after she refused to get repair, and the NY Public Service Commission needed to be called to remedy the issue... Amazingly, the estimate of a month before completion of the copper repair (that is, what they were called about to begin with!) became 2 days and the lines were back in service then. But of course this is all just a big misunderstanding and coincidence, and no one would ever think that Verizon is using tricks to get uninformed people to abandon copper and move over to FiOS...never...who would think such a thing? ;( ).

    In any case, there are essentially two FiOS voice services: FiOS Phone, which is the regulated Fiber-To-The-Home replacement for copper (or SLC-96 or FTTC-type services), and FiOS Digital Voice (or Phone), which is their unregulated (read: they can jack up your prices whenever they want and you have little if any recourse! - AVOID THIS!!!) VoIP-ish product which runs over the FiOS Internet connection.

    FiOS Phone should have more or less all the voice and latency characteristics of traditional copper/POTS voice phones (although we've had problems with modem calls, like faxing or home alarm systems even with this), although again, if the power goes out any FiOS service will not work after the limited batteries run out, and at all times the homeowner is paying the electric bill for FiOS whereas under copper/SLC-96/FTTC type services the phone company paid for the power. FiOS Phone is also what Verizon forces people on who don't want to opt for the higher-priced bundled/package deals with Cable-TV and Internet, and with the costs of the ONU box (which is what uses up the power and takes up space in the basement) being around $600, Verizon likely isn't too interested in having customers who are being forced off copper take this option, although in nearly all cases the customer, if they are told that Verizon is officially abandoning copper, has the right to obtain FiOS Phone service at the same rates and with the same features as their prior copper/POTS service.

    Additionally, in most states, new Verizon customers may also instruct Verizon to set up phone service with the regulated FiOS Phone product: Verizon, if they have abandoned copper or copper-ish products in an area or locality where new phone service is desired, may not refuse the regulated product over FiOS, although, with all the expenses they have to put in the FiOS ONU box, they often seem to be worse than used car salesmen in trying to steer new customers away from the regulated (generally cheaper) product, and instead try to "convince" them that the FiOS bundles (all unregulated) are better deals -- yes, indeed they are! -- for Verizon, which will likely never make the money back on a $600 box in the basement on a regulated service plan and has a very strong financial motivation to upsell people who only want/need good voice phone service on useless bundles of self-serving, unregulated services.

    Essentially, FiOS Phone is just a (less than adequate in many cases) replacement of copper with fiber, and the rates and services must remain the same as under copper, including regulated protection from the given state's Public Service Commission/Agency. While Verizon may claim it's more reliable than copper, if Verizon made the effort to maintain their copper network there'd be many fewer problems; it's just painfully obvious that they don't want to do this and want everyone off copper and on to FiOS or wireless. And, as noted above, FiOS Phone is regulated, and afforded the same protections as copper service/POTS lines by state regulatory commissions, which is another reason why Verizon likely would prefer customers to not migrate to FiOS Phone but move to one of the unregulated bundles instead. They try REALLY hard to convince you not to take the regulated product, so beware!

    FiOS Digital Voice/Phone is not the same as FiOS Phone, and is an unregulated product in most states (this may change over time as many states have realized how Verizon is pushing FiOS bundled over regular FiOS Phone and are becoming increasigly inclined to make sure that customers who are not aware of the distinction between FiOS Phone (regulated) and FiOS Digital Phone (generally unregulated) are afforded some protections). In any case, FiOS Digital Voice/Phone appears, from our experience, to be more of a VoIP-ish product (as it's name properly implies), and is not a "circuit switched" product (it doesn't meet all the requirements and specifications of traditional phone service), and appears to have increased latency. This is generally offset, though, by the admittedly very high speeds of data which FiOS can achieve, but it is still prone to more latency, or at least the potential for latency, as compared to traditional POTS service.

    Thus, in terms of local service:

    • POTS Service: Copper lines to the Central Office, SLICK-96/SLC-96 Subscriber Line Carrier, local neighborhood fiber distribution via SLC-like service, Fiber-To-The-Curb, etc, all should have no noticable difference in terms of latency. While the audio quality can vary from copper to SLC to fiber (Greenwich CT, which is served by NYTel/Verizon and not Southern New England Telephone/SNET or whatever they call themselves today, is a good example of poorly implemented SLC-96 service and has had problems for years), or between other technologies, the actual latency of calls should be the same regardless of which of the above technologies/local loop delivery mechanisms are used.
    • FiOS Phone: FiOS Phone is the regulated, replacement for copper in Verizon territory (not to be confused with FiOS Digital Voice, which is un-regulated and can offer lower standards than the regulated copper replacement product). It offers fiber "lines" (as a transport conduit instead of copper) from the Central Office all the way to the home. While there are concerns with respect to power outages and the ability to use fax and alarm monitoring systems (like ADT, et al), as well as those pertaining to no service during power outages as well as increased electric bills due to the ONU placed in the basement or utility room which the phone subscriber now has to assume, the FiOS Phone service is generally similar to traditional POTS service for voice calls. Thus, with all technologies (POTS-type, above, and FiOS Phone) being equally well-implemented and maintained (which, granted, is wishfull thinking!), the end customer should experience little appreciable difference in terms of latency and audio quality between the two.
    • FiOS Digital Voice/Phone: Fiber "lines"/conduit from the Central Office all the way to the home. Same issues as "FiOS Phone" above, with the additional problem of potentially increased latency due to the VoIP-ish service being offered. While in effect, due to the high speed of FiOS and it's cable company equivalents, VoIP should work relatively well over such circuits, they generally can take multiple, circuitous paths to reach the VoIP interconnection with the telephone network, and thus will likely still suffer a greater degree of latency.
  • 2. Traditional "circuit switched" long distance service (such as using AT&T or a Bell/LEC carrier's long-distance service); there should be no noticeable difference in latency between landline local calls and landline long-distance calls; if there are, the carrier may be using VoIP or some other poor-quality long-distance ("interexchange") carrier. Some long-distance carriers may use VoIP-ish or slower, more latent connections for calls outside of North America, which will cause delays in conversation and slow down (or make inoperable) services like faxing. AT&T Long Distance still seems to offer the best and most reliable international service with near local-sounding calls to many foreign destinations (but they have most of their customer service offshore and are, from our experience, difficult to deal with).
  • 3. Analog Cellular Service, which is no longer available in the US.
  • 4. GSM-based cellular service providers (the two primary carriers in the US being T-Mobile and AT&T Wireless).
  • 5. iDEN-based cellular service; the now defunct Nextel used to offer iDEN service in the US; it is still available via SouthernLinc and AirPeak in the United States and Telus and AirTel in Canada; slightly more latency than well-implemented GSM but not as much as CDMA.
  • 6. CDMA-based cellular service providers (the two primary carriers in the US being Verizon and Sprint). As noted above, CDMA also distorts and masks background noises and music much more so than other cellular transmission protocols.
  • 7. VoIP-based services, such as Vonage, which tend to add the most latency and make immediate conversations awkward and unnatural.
  • (Note: Some smaller cellular carriers may use VoIP for "backhaul" (connections between cell towers and their switching centers) and/or for long-distance, so latency may be exacerbated by backhaul and/or VoIP-ish interconnection/long-distance connections, but these are not a result of the given cellular ptotocol per se but instead likely originate in a given carrier's desire to save money and lack of concern over latency and call quality.)

    In the near future (2015 onwards), cellular carriers plan to introduce packet-based, IP-like voice service so that effectively voice calls will be carried over the same air interface/protocol/link to the cell tower as data. We haven't had a chance to try either a GSM or CDMA based phone which packetizes all calls (voice and data), and despite the promised improved sound quality, if steps are not taken to limit and control latency, may suffer from the same infirmities as the current GSM (to a lesser extent) or CDMA (to a greater extent) protocols.

    Overall, then, customers who enjoy the sound quality and immediacy of "landline"-type calls would be best to find a suitable GSM carrier instead of CDMA. Not only do GSM carriers appear to have less latency, but they also offer a more "lifelike"-sounding call experience as compared to the often overly-digitized sound quality of CDMA. (On the other hand, CDMA does mask background noise, which, although again detracting from its "lifelike", normal conversational qualities, is very useful in noisy places where the caller does not want any background noise to be conveyed to the person he is calling.


    Latency / Echo Test

    Wirelessnotes.org offers an automated latency/echo test number which can be called to determine the amount of latency which your current provider adds to a given call. The number is a local number in Vermont, and if you have a calling or long-distance plan with free domestic long-distance, you should not be charged anything to call it. It's available 24-hours a day, all year long.

    The Voice Latency Echo test number may be reached at: +1 (802) 359-9100

    Recorded instructions are provided upon calling, but generally, the Wirelessnotes.org Echo/Latency test number will play back whatever you say immediately, and you can call multiple times to see what differences there are between carriers. (Note that the Wirelessnotes latency test does introduce a slight degree of latency (about 8 to 10ms), but this shouldn't normally be detectable over properly configured traditional local or long-distance lines, and the delay should only become noticeable when called from a cell phone, Voice-over-IP phone, satellite connection, or other service which introduces a significantly greater amount of voice/audio latency.)

    For those concerned about the privacy of their number(s), please feel free to call with "PRIVACY" on (dial *67 first) if you don't want us to "see" your number; in any case we do not track or sell your information, nor will we contact you to sell to, or obtain additional information from, you. See our Privacy Policy for additional details.


    Contacting WirelessNotes.org

    We may be reached via:

  • E-Mail: Please mail to abcwireless-notes@interpage.net , removing the first three letters of the alphabet from the name/address.
  • Web-based Interface
  • Phone: +1 (802) 316-5000 during Business Hours.

  • Back to Main Wireless Notes Page
    Comparison of Cellular Carriers
    Interpage NSI Main Page

    All materials, including images INSI 2017. Last modified 04/10/2017