[Starlink] 300ms Telecommunication Latency and FTL Communication

[Starlink] 300ms Telecommunication Latency and FTL Communication

[Colin_Higbie]

Sebastian,

At 300ms RTT, that would mean the starting point for any communications are already at the threshold of unacceptability. I would think the strongest argument is that's at best a passable latency in absolutely perfect conditions, which never exist. "Pleasant" communication latency is sub-100ms, adding additional travel time to the actual servers involved and processing at each end, the ISP needs to do significantly better than that target to provide some margin for those other sources of latency, many controlled by fundamental physics sending the signal over distance.

By the way, on the original subject of quantum entanglement and FTL communication: the current theory and experimental data on quantum entanglement do not permit FTL communications. Yes, particles can exchange state information with their entangled particle FTL (tested and verified, possibly instantaneous or within a Planck-unit of time ~10^-44 seconds), but no information can be conveyed this way. Information transfer is still limited to the speed of light, as far as we know. This is because if particle A and B are entangled with respect to spin (i.e., if A has spin up, then B must have spin down and vice versa), and someone with particle A changes its spin, the only thing a person can do at particle B is query it once. The person at particle B will know the spin of A and B, but has no way of knowing if that's the spin before or after the person at A changed it. And because they can't check it multiple times (first check collapses the wave function), they have no way of knowing if it changed or when. The only way to check would be to use conventional communications, limited to the speed of light, which defeats any benefit to the FTL state change. 

While it's always possible we'll overcome what appear to be current limitations of physics, it's nothing that's likely in our near future. This is fairly fundamental quantum mechanical property and relates to the fact that you change the state and collapse the wave function when you observe the particle. Even attempts to use large numbers of particles in the hope of catching a statistical change across many has not been able to overcome this fundamental property of quantum entanglement. 

Best I've heard on this is the somewhat-famous-in-Physics-circles hypothesis that ER = EPR, referring to the paper by Einstein and Rosen on wormholes (the Einstein-Rosen bridge) = the paper by Einstein, Rosen, and Podolsky on entanglement, intended to mean that quantum entanglement is conveyed by General Relativity-style wormholes, so maybe we'll find a way to use the entanglement wormholes to send add'l information, but there's no evidence to suggest that's a possibility today. 

Cheers,
Colin


-----Original Message-----

Would you have any pointer for that study/those studies? Our local regulator thinks that 150 ms access network OWD (so 300msRTT) is acceptable and I am trying to find studies that can shed a light on what acceptable delay is for different kind of interactive tasks. (Spoiler alert, I am not convinced that 300ms RTT is a great idea, I forced my self to remote desktop with artificial 300ms delay and it was not fun, but not totaly unusable either, but then human can adapt and steer high inertia vehicles like loaded container ships...)

Sorry for the tangent...

Regards
        Sebastian

P.S.: Dave occasionally reminds us how 'slow' in comparison the speed of sound is ~343 m/second (depending on conditions) or 343/1000 = 0.343 m/millisecond that is even at a distance of 1 meter delay will be at a 3 ms... and when talking to folks 10m away it is not the delay that is annoying, but the fact that you have to raise your voice considerably...

Re: [Starlink] 300ms Telecommunication Latency and FTL Communication

[Sebastian Moeller]

Hi Colin,



> On 5. Jun 2024, at 19:58, Colin_Higbie <CHigbie1@Higbie.name> wrote:
> 
> Sebastian,
> 
> At 300ms RTT, that would mean the starting point for any communications are already at the threshold of unacceptability.

[SM] Not according to the ITU (114)
mouth-ear delay in ms (so OWDs)
0-200ms: users very satisfied
200-275ms: users satisfied
275-375ms: some users dissatisfied
375-600: many users dissatisfied
600-...: nearly all users dissatisfied

So even 150ms OWD still falls within the very satisfied range if the remaining delay is not to large... And even if er string two of these users together, we end up with worst case >300ms delay, but that sill only gets us into the "some users dissatisfied" which the regulator might find an acceptable trade-off in the context of guaranteed internet access parameters (where the idea is the 150ms OWD or 300ms RTT is not the target, but the threshold for being acceptable).

My gut feeling is these ranges are not actually measured in a way they are now interpreted (e.g. when testing transatlantic call delays users likely already had an expectancy of longer delay and simply judges these calls against a different yard stick). BUT unless I can demonstrate that the original studies resulting in these numbers are terminally flawed there is little chance that I can convince our regulator to take my word vor voice delays over the word of the ITU... so I need different, preferably newer data and focus on probably remote desktop usage as a relative novel use case without much encrusted ideas about acceptable latency...


> I would think the strongest argument is that's at best a passable latency in absolutely perfect conditions, which never exist. "Pleasant" communication latency is sub-100ms, adding additional travel time to the actual servers involved and processing at each end, the ISP needs to do significantly better than that target to provide some margin for those other sources of latency, many controlled by fundamental physics sending the signal over distance.

[SM] Personally I agree, yet I am not sure picking a fight over the VoIP numbers is going to be productive, as I have considerably less clout with the regulator than the ITU...

Regards
	Sebastian

> 
> By the way, on the original subject of quantum entanglement and FTL communication: the current theory and experimental data on quantum entanglement do not permit FTL communications. Yes, particles can exchange state information with their entangled particle FTL (tested and verified, possibly instantaneous or within a Planck-unit of time ~10^-44 seconds), but no information can be conveyed this way. Information transfer is still limited to the speed of light, as far as we know. This is because if particle A and B are entangled with respect to spin (i.e., if A has spin up, then B must have spin down and vice versa), and someone with particle A changes its spin, the only thing a person can do at particle B is query it once. The person at particle B will know the spin of A and B, but has no way of knowing if that's the spin before or after the person at A changed it. And because they can't check it multiple times (first check collapses the wave function), they have no way of knowing if it changed or when. The only way to check would be to use conventional communications, limited to the speed of light, which defeats any benefit to the FTL state change. 
> 
> While it's always possible we'll overcome what appear to be current limitations of physics, it's nothing that's likely in our near future. This is fairly fundamental quantum mechanical property and relates to the fact that you change the state and collapse the wave function when you observe the particle. Even attempts to use large numbers of particles in the hope of catching a statistical change across many has not been able to overcome this fundamental property of quantum entanglement. 
> 
> Best I've heard on this is the somewhat-famous-in-Physics-circles hypothesis that ER = EPR, referring to the paper by Einstein and Rosen on wormholes (the Einstein-Rosen bridge) = the paper by Einstein, Rosen, and Podolsky on entanglement, intended to mean that quantum entanglement is conveyed by General Relativity-style wormholes, so maybe we'll find a way to use the entanglement wormholes to send add'l information, but there's no evidence to suggest that's a possibility today. 
> 
> Cheers,
> Colin
> 
> 
> -----Original Message-----
> 
> Would you have any pointer for that study/those studies? Our local regulator thinks that 150 ms access network OWD (so 300msRTT) is acceptable and I am trying to find studies that can shed a light on what acceptable delay is for different kind of interactive tasks. (Spoiler alert, I am not convinced that 300ms RTT is a great idea, I forced my self to remote desktop with artificial 300ms delay and it was not fun, but not totaly unusable either, but then human can adapt and steer high inertia vehicles like loaded container ships...)
> 
> Sorry for the tangent...
> 
> Regards
>        Sebastian
> 
> P.S.: Dave occasionally reminds us how 'slow' in comparison the speed of sound is ~343 m/second (depending on conditions) or 343/1000 = 0.343 m/millisecond that is even at a distance of 1 meter delay will be at a 3 ms... and when talking to folks 10m away it is not the delay that is annoying, but the fact that you have to raise your voice considerably...
> 

Re: [Starlink] 300ms Telecommunication Latency and FTL Communication

[Colin_Higbie]

Sebastion, I was not providing any knowledge or data on acceptable latency for video calling. That is not my area of expertise (closest facet of my business merely involves web site responsiveness and start time for playing audio after buffering, both of which are much less sensitive to latency). I can state that, as a user, I would find 150ms measured ISP latency high, not intolerable, but noticeable - video conferencing is more sensitive to latency than pure voice (in my personal opinion, no study I've read on this specifically), because we watch people's faces for reactions to what we say as we're speaking. If there is a noticeable lag there, it disrupts the conversation. On the other hand, the same lag in a pure voice discussion, which is inherently less synchronous, would not be noticeable. 

In my prior post, I was using the 150/300 ms figure you provided and saying that IF that's the max acceptable figure for network latency, THEN that's already a problem to only hit that as the ISP because each participant also adds their distance and network delays. For those that are just as quick, that may be fine. However, assuming there's some form of bell curve distribution on latency, many of these will be longer, and some much longer than what your ISP provides to their customers. Therefore, to ensure a satisfactory experience with the majority of prospective video call participants on other networks, the ISP would need to provide a sufficiently low latency to accommodate these differences. Otherwise, a significant portion of the calls would be of poor quality. Obviously, they can't make up for a participant whose own latency exceeds 300ms, but they should not be the cause of poor communication with someone at 160ms latency. But that's just reasoning around your numbers, not data.

That said, here are some studies I found that may be helpful:

This one includes the 300ms round-trip time, but puts at the extreme outer range of acceptability:
"Defining 'seamlessly connected': user perceptions of operation latency in cross-device interaction"
https://www.sciencedirect.com/science/article/abs/pii/S1071581923000770

"What Are Good Latency & Ping Speeds?"
https://www.pingplotter.com/wisdom/article/is-my-connection-good/

A Cisco discussion that supports the 300ms round trip time:
"Acceptable Jitter, Latency and Packet Loss for Audio and Video on a WebEx Meeting"
https://community.cisco.com/t5/webex-meetings-and-webex-app/acceptable-jitter-latency-and-packet-loss-for-audio-and-video-on/m-p/4301454


These are behind pay walls or require academic credentials, so don't know if they are good or not, nor what conclusions they reach - they could even be the source of the 150/300ms figure, but I agree with you that seems high: 

"A Study of the Effects of Network Latency on Visual Task Performance in Video Conferencing"
https://dl.acm.org/doi/10.1145/3491101.3519678
https://www.academia.edu/98061737/A_Study_of_the_Effects_of_Network_Latency_on_Visual_Task_Performance_in_Video_Conferencing

"Effect of latency on social presence in traditional video conference and VR conference: a comparative study"
https://ieeexplore.ieee.org/document/10402741

"Determination of the latency effects on surgical performance and the acceptable latency levels in telesurgery using the dV-Trainer((r)) simulator"
https://pubmed.ncbi.nlm.nih.gov/24671353/


-----Original Message-----
From: Sebastian Moeller <moeller0@gmx.de> 
Sent: Thursday, June 6, 2024 3:22 AM
To: Colin_Higbie <CHigbie1@Higbie.name>
Cc: starlink@lists.bufferbloat.net
Subject: Re: 300ms Telecommunication Latency and FTL Communication

Hi Colin,



> On 5. Jun 2024, at 19:58, Colin_Higbie <CHigbie1@Higbie.name> wrote:
> 
> Sebastian,
> 
> At 300ms RTT, that would mean the starting point for any communications are already at the threshold of unacceptability.

[SM] Not according to the ITU (114)
mouth-ear delay in ms (so OWDs)
0-200ms: users very satisfied
200-275ms: users satisfied
275-375ms: some users dissatisfied
375-600: many users dissatisfied
600-...: nearly all users dissatisfied

So even 150ms OWD still falls within the very satisfied range if the remaining delay is not to large... And even if er string two of these users together, we end up with worst case >300ms delay, but that sill only gets us into the "some users dissatisfied" which the regulator might find an acceptable trade-off in the context of guaranteed internet access parameters (where the idea is the 150ms OWD or 300ms RTT is not the target, but the threshold for being acceptable).

My gut feeling is these ranges are not actually measured in a way they are now interpreted (e.g. when testing transatlantic call delays users likely already had an expectancy of longer delay and simply judges these calls against a different yard stick). BUT unless I can demonstrate that the original studies resulting in these numbers are terminally flawed there is little chance that I can convince our regulator to take my word vor voice delays over the word of the ITU... so I need different, preferably newer data and focus on probably remote desktop usage as a relative novel use case without much encrusted ideas about acceptable latency...


> I would think the strongest argument is that's at best a passable latency in absolutely perfect conditions, which never exist. "Pleasant" communication latency is sub-100ms, adding additional travel time to the actual servers involved and processing at each end, the ISP needs to do significantly better than that target to provide some margin for those other sources of latency, many controlled by fundamental physics sending the signal over distance.

[SM] Personally I agree, yet I am not sure picking a fight over the VoIP numbers is going to be productive, as I have considerably less clout with the regulator than the ITU...

Regards
	Sebastian

Re: [Starlink] 300ms Telecommunication Latency and FTL Communication

[Sebastian Moeller]

Hi Colin,

Thank you very much, also for the paper and analysis you sent after this.


> On 6. Jun 2024, at 15:43, Colin_Higbie <CHigbie1@Higbie.name> wrote:
> 
> Sebastion, I was not providing any knowledge or data on acceptable latency for video calling. That is not my area of expertise (closest facet of my business merely involves web site responsiveness and start time for playing audio after buffering, both of which are much less sensitive to latency). I can state that, as a user, I would find 150ms measured ISP latency high, not intolerable, but noticeable -

[SM] I agree, and there is the rub, unlike capacity where we often see hard limits like a VoIP call takes ~100Kbps, if link capacity is sufficiently smaller than 100 Kbps VoIP will not work at all (so there will be hard qualitative thresholds), with latency we often see slow degradation of quality... if e.g. VoIP works well at 100ms, it likely will still be virtually identical at 101ms and still reasonably similar at say 150ms... much harder to turn that into a convincing threshold...

> video conferencing is more sensitive to latency than pure voice (in my personal opinion, no study I've read on this specifically), because we watch people's faces for reactions to what we say as we're speaking.

[SM] I am happy to believe you on this, but ti turn this into something useful for my purpose I will need to find something published, preferably peer reviewed. But thanks to the pointer which should help in my search.


> If there is a noticeable lag there, it disrupts the conversation. On the other hand, the same lag in a pure voice discussion, which is inherently less synchronous, would not be noticeable.

[SM] Not sure I fully agree here, assuming video and audio arrive both with the same delay I would guess both suffer similarly from the delay... my gut feeling is as long as natural speech sequence stays intact, that is no unintended collisions due to both speaking at the same time, audio-only and audio-video should both be sort of OK...

> 
> In my prior post, I was using the 150/300 ms figure you provided and saying that IF that's the max acceptable figure for network latency, THEN that's already a problem to only hit that as the ISP because each participant also adds their distance and network delays. For those that are just as quick, that may be fine. However, assuming there's some form of bell curve distribution on latency, many of these will be longer, and some much longer than what your ISP provides to their customers. Therefore, to ensure a satisfactory experience with the majority of prospective video call participants on other networks, the ISP would need to provide a sufficiently low latency to accommodate these differences. Otherwise, a significant portion of the calls would be of poor quality. Obviously, they can't make up for a participant whose own latency exceeds 300ms, but they should not be the cause of poor communication with someone at 160ms latency. But that's just reasoning around your numbers, not data.

[SM] I am looking at the studies underlaying ITU-114 delay quality assessments, but I think that is a long shot in convincing the regulsator, these numbers are deeply entrenched by know, so I likely would need a more recebt study that conclusively shows that these numbers are way too high...


> 
> That said, here are some studies I found that may be helpful:
> 
> This one includes the 300ms round-trip time, but puts at the extreme outer range of acceptability:
> "Defining 'seamlessly connected': user perceptions of operation latency in cross-device interaction"
> https://www.sciencedirect.com/science/article/abs/pii/S1071581923000770
> 
> "What Are Good Latency & Ping Speeds?"
> https://www.pingplotter.com/wisdom/article/is-my-connection-good/
> 
> A Cisco discussion that supports the 300ms round trip time:
> "Acceptable Jitter, Latency and Packet Loss for Audio and Video on a WebEx Meeting"
> https://community.cisco.com/t5/webex-meetings-and-webex-app/acceptable-jitter-latency-and-packet-loss-for-audio-and-video-on/m-p/4301454
> 
> 
> These are behind pay walls or require academic credentials, so don't know if they are good or not, nor what conclusions they reach - they could even be the source of the 150/300ms figure, but I agree with you that seems high: 
> 
> "A Study of the Effects of Network Latency on Visual Task Performance in Video Conferencing"
> https://dl.acm.org/doi/10.1145/3491101.3519678
> https://www.academia.edu/98061737/A_Study_of_the_Effects_of_Network_Latency_on_Visual_Task_Performance_in_Video_Conferencing
> 
> "Effect of latency on social presence in traditional video conference and VR conference: a comparative study"
> https://ieeexplore.ieee.org/document/10402741
> 
> "Determination of the latency effects on surgical performance and the acceptable latency levels in telesurgery using the dV-Trainer((r)) simulator"
> https://pubmed.ncbi.nlm.nih.gov/24671353/

[SM] Thanks a lot, I will go look these up.

Regards
	Sebastian


> 
> 
> -----Original Message-----
> From: Sebastian Moeller <moeller0@gmx.de> 
> Sent: Thursday, June 6, 2024 3:22 AM
> To: Colin_Higbie <CHigbie1@Higbie.name>
> Cc: starlink@lists.bufferbloat.net
> Subject: Re: 300ms Telecommunication Latency and FTL Communication
> 
> Hi Colin,
> 
> 
> 
>> On 5. Jun 2024, at 19:58, Colin_Higbie <CHigbie1@Higbie.name> wrote:
>> 
>> Sebastian,
>> 
>> At 300ms RTT, that would mean the starting point for any communications are already at the threshold of unacceptability.
> 
> [SM] Not according to the ITU (114)
> mouth-ear delay in ms (so OWDs)
> 0-200ms: users very satisfied
> 200-275ms: users satisfied
> 275-375ms: some users dissatisfied
> 375-600: many users dissatisfied
> 600-...: nearly all users dissatisfied
> 
> So even 150ms OWD still falls within the very satisfied range if the remaining delay is not to large... And even if er string two of these users together, we end up with worst case >300ms delay, but that sill only gets us into the "some users dissatisfied" which the regulator might find an acceptable trade-off in the context of guaranteed internet access parameters (where the idea is the 150ms OWD or 300ms RTT is not the target, but the threshold for being acceptable).
> 
> My gut feeling is these ranges are not actually measured in a way they are now interpreted (e.g. when testing transatlantic call delays users likely already had an expectancy of longer delay and simply judges these calls against a different yard stick). BUT unless I can demonstrate that the original studies resulting in these numbers are terminally flawed there is little chance that I can convince our regulator to take my word vor voice delays over the word of the ITU... so I need different, preferably newer data and focus on probably remote desktop usage as a relative novel use case without much encrusted ideas about acceptable latency...
> 
> 
>> I would think the strongest argument is that's at best a passable latency in absolutely perfect conditions, which never exist. "Pleasant" communication latency is sub-100ms, adding additional travel time to the actual servers involved and processing at each end, the ISP needs to do significantly better than that target to provide some margin for those other sources of latency, many controlled by fundamental physics sending the signal over distance.
> 
> [SM] Personally I agree, yet I am not sure picking a fight over the VoIP numbers is going to be productive, as I have considerably less clout with the regulator than the ITU...
> 
> Regards
> Sebastian

Re: [Starlink] 300ms Telecommunication Latency and FTL Communication

[David Lang]

Sebastian Moeller wrote:

>> video conferencing is more sensitive to latency than pure voice (in my personal opinion, no study I've read on this specifically), because we watch people's faces for reactions to what we say as we're speaking.
>
> [SM] I am happy to believe you on this, but ti turn this into something useful for my purpose I will need to find something published, preferably peer reviewed. But thanks to the pointer which should help in my search.

one factor to point out, almost all video conferencing is you to server to other 
user, not direct you to other user. If you have two people in the same house on 
a call together, they suffer double the latency.

>> If there is a noticeable lag there, it disrupts the conversation. On the other hand, the same lag in a pure voice discussion, which is inherently less synchronous, would not be noticeable.
>
> [SM] Not sure I fully agree here, assuming video and audio arrive both with 
> the same delay I would guess both suffer similarly from the delay... my gut 
> feeling is as long as natural speech sequence stays intact, that is no 
> unintended collisions due to both speaking at the same time, audio-only and 
> audio-video should both be sort of OK...

the longer the latency, the more likely people are to talk over each other, 
because they don't see/hear the other person talking when they start. If the 
latency is low, they can stop quickly, but as the latency increases, they are 
talking longer before they hear the other person.

1. this means it's harder to figure out who started first and should continue
2. this means that there is a longer time period of multiple people talking

I agree that this is the same video vs audio. That's why I was thinking back to 
the early AT&T research I've heard from Internet lore (back when AT&T had a huge 
R&D section). It may be useful to look not only for long distance info 
(including microwave relays vs direct cables vs satellite relays) but also if 
they have any research on early conference calling.

David Lang

Re: [Starlink] 300ms Telecommunication Latency and FTL Communication

[Sebastian Moeller]

Hi David,

On 7 June 2024 16:55:54 CEST, David Lang <david@lang.hm> wrote:
>Sebastian Moeller wrote:
>
>>> video conferencing is more sensitive to latency than pure voice (in my personal opinion, no study I've read on this specifically), because we watch people's faces for reactions to what we say as we're speaking.
>> 
>> [SM] I am happy to believe you on this, but ti turn this into something useful for my purpose I will need to find something published, preferably peer reviewed. But thanks to the pointer which should help in my search.
>
>one factor to point out, almost all video conferencing is you to server to other user, not direct you to other user. If you have two people in the same house on a call together, they suffer double the latency.

>
>>> If there is a noticeable lag there, it disrupts the conversation. On the other hand, the same lag in a pure voice discussion, which is inherently less synchronous, would not be noticeable.
>> 
>> [SM] Not sure I fully agree here, assuming video and audio arrive both with the same delay I would guess both suffer similarly from the delay... my gut feeling is as long as natural speech sequence stays intact, that is no unintended collisions due to both speaking at the same time, audio-only and audio-video should both be sort of OK...
>
>the longer the latency, the more likely people are to talk over each other, because they don't see/hear the other person talking when they start. If the latency is low, they can stop quickly, but as the latency increases, they are talking longer before they hear the other person.

[SM] I fully agree, that is what I meant with unintended collisions... and as long as we are in the regime with little talking over each other I expect little differences between the modalities.


>
>1. this means it's harder to figure out who started first and should continue
>2. this means that there is a longer time period of multiple people talking
>
>I agree that this is the same video vs audio. That's why I was thinking back to the early AT&T research I've heard from Internet lore (back when AT&T had a huge R&D section). It may be useful to look not only for long distance info (including microwave relays vs direct cables vs satellite relays) but also if they have any research on early conference calling.

[SM] Thanks, will have a look at that as well.
>
>David Lang

-- 
Sent from my Android device with K-9 Mail. Please excuse my brevity.

Re: [Starlink] 300ms Telecommunication Latency and FTL Communication

[Colin_Higbie]

GREAT, GREAT point on the increased likelihood of unintentional interruptions and talking over each other the higher the latency. I had not thought of that before, but I think that's a compelling and easily measurable metric. It also gives a clear cost-based selling advantage to the ISP with lower latency, which should make it attractive to business: if an ISP can say to a business (or at-home worker): based on our 20ms lower latency, you will save 3 minutes per day in lost annoyances due to accidental interruptions, that has a value per person in those calls of $X. Therefore, you should switch your business to us.

I do still think, at least for me as a speaker, that video feedback that does not "feel" like it provides instant video feedback in the form of real-time facial expressions to show audience feedback for the speaker is more critical (I can stomach small delays on audio feedback after I ask a question, but not on video feedback, which I need while I'm talking), but I do think the interruption piece is also very real. Further, my point on video is almost binary – either I'm getting close enough to real-time feedback or I'm not and need to plod on without it, which diminishes its value as an argument for lower latency.

In contrast, the increase in accidental interruptions in audio (and video too presumably) would scale with latency (not binary means much better for an argument to the ISP). Every 1ms of additional latency would provide a small but measurable % increase in the amount of accidental interruptions. Conversely, the lower the latency, the fewer of these. This means that there's not merely a "good enough" level on latency (at least not above human reaction time), but rather the lower the better in a very tangible way. This should be a compelling and objective argument.

I'm not aware of any study, but it stands to reason that every interruption results in X seconds of lost productive talking time. Say X = 5s. While we don't know the slope of the function: #_accidental_interruptions = [some unknown slope] x (latency_in_ms - minimal_human_reaction_threshold_in_ms), where minimal_human_reaction_threshold_in_ms probably equals something like 20 - 50ms. As soon as latency > minimal_human_reaction_threshold_in_ms, #_accidental_interruptions becomes a positive number and lost productivity also increases at 5s/interruption (using above assumption). 

Even if there is no such study, this is clearly an easily studyable metric to establish the typical human reaction time (that's probably already known), the mean and median time lost per accidental interruption, and the key piece: the slope of the line or shape of the function that says how many additional accidental interruptions occur for every ms of added latency. Anyone here looking for an academic study with a simple metric to provide data to ISP on economic value in reducing latency that is simple enough that even a high-school kid could understand the meaning, this might be it.

Cheers,
Colin


-----Original Message-----
From: Sebastian Moeller <moeller0@gmx.de> 
Sent: Friday, June 7, 2024 11:33 AM
To: David Lang <david@lang.hm>
Cc: Colin_Higbie <CHigbie1@Higbie.name>; starlink@lists.bufferbloat.net
Subject: Re: [Starlink] 300ms Telecommunication Latency and FTL Communication

Hi David,

On 7 June 2024 16:55:54 CEST, David Lang <david@lang.hm> wrote:
>Sebastian Moeller wrote:
>
>>> video conferencing is more sensitive to latency than pure voice (in my personal opinion, no study I've read on this specifically), because we watch people's faces for reactions to what we say as we're speaking.
>> 
>> [SM] I am happy to believe you on this, but ti turn this into something useful for my purpose I will need to find something published, preferably peer reviewed. But thanks to the pointer which should help in my search.
>
>one factor to point out, almost all video conferencing is you to server to other user, not direct you to other user. If you have two people in the same house on a call together, they suffer double the latency.

>
>>> If there is a noticeable lag there, it disrupts the conversation. On the other hand, the same lag in a pure voice discussion, which is inherently less synchronous, would not be noticeable.
>> 
>> [SM] Not sure I fully agree here, assuming video and audio arrive both with the same delay I would guess both suffer similarly from the delay... my gut feeling is as long as natural speech sequence stays intact, that is no unintended collisions due to both speaking at the same time, audio-only and audio-video should both be sort of OK...
>
>the longer the latency, the more likely people are to talk over each other, because they don't see/hear the other person talking when they start. If the latency is low, they can stop quickly, but as the latency increases, they are talking longer before they hear the other person.

[SM] I fully agree, that is what I meant with unintended collisions... and as long as we are in the regime with little talking over each other I expect little differences between the modalities.


>
>1. this means it's harder to figure out who started first and should 
>continue 2. this means that there is a longer time period of multiple 
>people talking
>
>I agree that this is the same video vs audio. That's why I was thinking back to the early AT&T research I've heard from Internet lore (back when AT&T had a huge R&D section). It may be useful to look not only for long distance info (including microwave relays vs direct cables vs satellite relays) but also if they have any research on early conference calling.

[SM] Thanks, will have a look at that as well.
>
>David Lang

--
Sent from my Android device with K-9 Mail. Please excuse my brevity.

Re: [Starlink] 300ms Telecommunication Latency and FTL Communication

[David Lang]

[-- Attachment #1: Type: text/plain, Size: 7083 bytes --]

On Fri, 7 Jun 2024, Colin_Higbie wrote:

> GREAT, GREAT point on the increased likelihood of unintentional interruptions 
> and talking over each other the higher the latency. I had not thought of that 
> before, but I think that's a compelling and easily measurable metric. It also 
> gives a clear cost-based selling advantage to the ISP with lower latency, 
> which should make it attractive to business: if an ISP can say to a business 
> (or at-home worker): based on our 20ms lower latency, you will save 3 minutes 
> per day in lost annoyances due to accidental interruptions, that has a value 
> per person in those calls of $X. Therefore, you should switch your business to 
> us.
>
> I do still think, at least for me as a speaker, that video feedback that does 
> not "feel" like it provides instant video feedback in the form of real-time 
> facial expressions to show audience feedback for the speaker is more critical 
> (I can stomach small delays on audio feedback after I ask a question, but not 
> on video feedback, which I need while I'm talking), but I do think the 
> interruption piece is also very real. Further, my point on video is almost 
> binary – either I'm getting close enough to real-time feedback or I'm not and 
> need to plod on without it, which diminishes its value as an argument for 
> lower latency.

the problem is drawing the line at what latency is the problem. This is going to 
be different for different people and take quite a bit of experimentation to 
discover (and I would not be surprised if someone who is doing testing is 
learing how to cope with longer latencies, so you can't just increase/decrease 
the latency in a predictable manner, you would need to randomize it and ask 
after each call how good/bad it was)

> In contrast, the increase in accidental interruptions in audio (and video too 
> presumably) would scale with latency (not binary means much better for an 
> argument to the ISP). Every 1ms of additional latency would provide a small 
> but measurable % increase in the amount of accidental interruptions. 
> Conversely, the lower the latency, the fewer of these. This means that there's 
> not merely a "good enough" level on latency (at least not above human reaction 
> time), but rather the lower the better in a very tangible way. This should be 
> a compelling and objective argument.
>
> I'm not aware of any study, but it stands to reason that every interruption 
> results in X seconds of lost productive talking time. Say X = 5s. While we 
> don't know the slope of the function: #_accidental_interruptions = [some 
> unknown slope] x (latency_in_ms - minimal_human_reaction_threshold_in_ms), 
> where minimal_human_reaction_threshold_in_ms probably equals something like 20 
> - 50ms. As soon as latency > minimal_human_reaction_threshold_in_ms, 
> #_accidental_interruptions becomes a positive number and lost productivity 
> also increases at 5s/interruption (using above assumption).

I also think the interruptions will be longer with higher latency as it will be 
harder to sync up.

> Even if there is no such study, this is clearly an easily studyable metric to 
> establish the typical human reaction time (that's probably already known), the 
> mean and median time lost per accidental interruption, and the key piece: the 
> slope of the line or shape of the function that says how many additional 
> accidental interruptions occur for every ms of added latency. Anyone here 
> looking for an academic study with a simple metric to provide data to ISP on 
> economic value in reducing latency that is simple enough that even a 
> high-school kid could understand the meaning, this might be it.

or can anyone setup a simple router/pi image that can introduce arbitrary 
latency and then let's see if we can find some high school students looking for 
a science projct. That may not get the academia stamp of approval, but if we can 
get a bunch of people around the world to test things, we would get data back 
faster than waiting for a more academic process (and it could potentially feed 
into such a process)

it doesn't even need to be a full blown video conferencing program in use, but 
with more people on the call, there's more chance of interruptions.

David Lang


> Cheers,
> Colin
>
>
> -----Original Message-----
> From: Sebastian Moeller <moeller0@gmx.de>
> Sent: Friday, June 7, 2024 11:33 AM
> To: David Lang <david@lang.hm>
> Cc: Colin_Higbie <CHigbie1@Higbie.name>; starlink@lists.bufferbloat.net
> Subject: Re: [Starlink] 300ms Telecommunication Latency and FTL Communication
>
> Hi David,
>
> On 7 June 2024 16:55:54 CEST, David Lang <david@lang.hm> wrote:
>> Sebastian Moeller wrote:
>>
>>>> video conferencing is more sensitive to latency than pure voice (in my personal opinion, no study I've read on this specifically), because we watch people's faces for reactions to what we say as we're speaking.
>>>
>>> [SM] I am happy to believe you on this, but ti turn this into something useful for my purpose I will need to find something published, preferably peer reviewed. But thanks to the pointer which should help in my search.
>>
>> one factor to point out, almost all video conferencing is you to server to other user, not direct you to other user. If you have two people in the same house on a call together, they suffer double the latency.
>
>>
>>>> If there is a noticeable lag there, it disrupts the conversation. On the other hand, the same lag in a pure voice discussion, which is inherently less synchronous, would not be noticeable.
>>>
>>> [SM] Not sure I fully agree here, assuming video and audio arrive both with the same delay I would guess both suffer similarly from the delay... my gut feeling is as long as natural speech sequence stays intact, that is no unintended collisions due to both speaking at the same time, audio-only and audio-video should both be sort of OK...
>>
>> the longer the latency, the more likely people are to talk over each other, because they don't see/hear the other person talking when they start. If the latency is low, they can stop quickly, but as the latency increases, they are talking longer before they hear the other person.
>
> [SM] I fully agree, that is what I meant with unintended collisions... and as long as we are in the regime with little talking over each other I expect little differences between the modalities.
>
>
>>
>> 1. this means it's harder to figure out who started first and should
>> continue 2. this means that there is a longer time period of multiple
>> people talking
>>
>> I agree that this is the same video vs audio. That's why I was thinking back to the early AT&T research I've heard from Internet lore (back when AT&T had a huge R&D section). It may be useful to look not only for long distance info (including microwave relays vs direct cables vs satellite relays) but also if they have any research on early conference calling.
>
> [SM] Thanks, will have a look at that as well.
>>
>> David Lang
>
> --
> Sent from my Android device with K-9 Mail. Please excuse my brevity.
>

Re: [Starlink] 300ms Telecommunication Latency and FTL Communication

[Colin_Higbie]

Yes, of course it differs for different people. Statistics doesn't care. There will be a mean and median for any sample. Statistical analysis already expects no two people are identical. 

Good point on the interruptions being longer due to longer fix. That makes my original equation a bit more complex, but, to your point, it only intensifies the problem with each ms of added latency being worse than the one before. 

Cheers,
Colin



-----Original Message-----
From: David Lang <david@lang.hm> 
Sent: Friday, June 7, 2024 12:57 PM
To: Colin_Higbie <CHigbie1@Higbie.name>
Cc: Sebastian Moeller <moeller0@gmx.de>; David Lang <david@lang.hm>; starlink@lists.bufferbloat.net
Subject: RE: [Starlink] 300ms Telecommunication Latency and FTL Communication

On Fri, 7 Jun 2024, Colin_Higbie wrote:

> GREAT, GREAT point on the increased likelihood of unintentional 
> interruptions and talking over each other the higher the latency. I 
> had not thought of that before, but I think that's a compelling and 
> easily measurable metric. It also gives a clear cost-based selling 
> advantage to the ISP with lower latency, which should make it 
> attractive to business: if an ISP can say to a business (or at-home 
> worker): based on our 20ms lower latency, you will save 3 minutes per 
> day in lost annoyances due to accidental interruptions, that has a 
> value per person in those calls of $X. Therefore, you should switch your business to us.
>
> I do still think, at least for me as a speaker, that video feedback 
> that does not "feel" like it provides instant video feedback in the 
> form of real-time facial expressions to show audience feedback for the 
> speaker is more critical (I can stomach small delays on audio feedback 
> after I ask a question, but not on video feedback, which I need while 
> I'm talking), but I do think the interruption piece is also very real. 
> Further, my point on video is almost binary – either I'm getting close 
> enough to real-time feedback or I'm not and need to plod on without 
> it, which diminishes its value as an argument for lower latency.

the problem is drawing the line at what latency is the problem. This is going to be different for different people and take quite a bit of experimentation to discover (and I would not be surprised if someone who is doing testing is learing how to cope with longer latencies, so you can't just increase/decrease the latency in a predictable manner, you would need to randomize it and ask after each call how good/bad it was)

> In contrast, the increase in accidental interruptions in audio (and 
> video too
> presumably) would scale with latency (not binary means much better for 
> an argument to the ISP). Every 1ms of additional latency would provide 
> a small but measurable % increase in the amount of accidental interruptions.
> Conversely, the lower the latency, the fewer of these. This means that 
> there's not merely a "good enough" level on latency (at least not 
> above human reaction time), but rather the lower the better in a very 
> tangible way. This should be a compelling and objective argument.
>
> I'm not aware of any study, but it stands to reason that every 
> interruption results in X seconds of lost productive talking time. Say 
> X = 5s. While we don't know the slope of the function: 
> #_accidental_interruptions = [some unknown slope] x (latency_in_ms - 
> minimal_human_reaction_threshold_in_ms),
> where minimal_human_reaction_threshold_in_ms probably equals something 
> like 20
> - 50ms. As soon as latency > minimal_human_reaction_threshold_in_ms,
> #_accidental_interruptions becomes a positive number and lost 
> productivity also increases at 5s/interruption (using above assumption).

I also think the interruptions will be longer with higher latency as it will be harder to sync up.

> Even if there is no such study, this is clearly an easily studyable 
> metric to establish the typical human reaction time (that's probably 
> already known), the mean and median time lost per accidental 
> interruption, and the key piece: the slope of the line or shape of the 
> function that says how many additional accidental interruptions occur 
> for every ms of added latency. Anyone here looking for an academic 
> study with a simple metric to provide data to ISP on economic value in 
> reducing latency that is simple enough that even a high-school kid could understand the meaning, this might be it.

or can anyone setup a simple router/pi image that can introduce arbitrary latency and then let's see if we can find some high school students looking for a science projct. That may not get the academia stamp of approval, but if we can get a bunch of people around the world to test things, we would get data back faster than waiting for a more academic process (and it could potentially feed into such a process)

it doesn't even need to be a full blown video conferencing program in use, but with more people on the call, there's more chance of interruptions.

David Lang

Re: [Starlink] 300ms Telecommunication Latency and FTL Communication

[Michael Richardson]

[-- Attachment #1: Type: text/plain, Size: 1663 bytes --]


Sebastian Moeller via Starlink <starlink@lists.bufferbloat.net> wrote:
    > [SM] I fully agree, that is what I meant with unintended
    > collisions... and as long as we are in the regime with little talking
    > over each other I expect little differences between the modalities.

Back in 2002, the FreeS/WAN project moved from STU-III phones (I had two of
them with a two-line conference phone, so we could have three locations
connected) to H323 over IPsec [dog fooding].  Of course, each meeting started with 15-30
minutes to debugging: when IPsec didn't rekey right, we'd often wind up with
unidirectional communications. (It was a very good way to debug dog food issues)

A point that ||ugh made at the time was that he didn't think that we it was
actually useful to be in radio mode: that all of the subsequent digital
systems were designed to emulate the properties of the radio, which was that
the *general* as allowed to shout lounder than everyone else until they all
shut up and listened.

The question was our conferencing systems couldn't just accept audio from
each participant, put it in a queue, and play it out.  Some participants
might hear comments in different orders.  I don't know if anyone has tried
this.
It seems that such a system would be useful for in a Earth/Luna conference
where speed of light delays are 1-2s.  Beyond that, one gets into sending
video-grams as one sees in Expanse.

--
]               Never tell me the odds!                 | ipv6 mesh networks [
]   Michael Richardson, Sandelman Software Works        |    IoT architect   [
]     mcr@sandelman.ca  http://www.sandelman.ca/        |   ruby on rails    [




[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 511 bytes --]