Latency 1.4.....Improvement?

  • 5
  • Question
  • Updated 4 years ago
  • Answered
I just had a conversation with a "contesting" friend who said he avoids the Flex due to slow "turn-around" time particularly in SSB mode. It causes him to be one syllable behind the other radios in a contest thus loosing points. According to the QST review, the turnaround time is 138 ms for the 6300 and an even slower 140 ms for the 6700. Looking through reviews of other radios I see number like 38 etc. So I was wondering if ver 1.4 made any improvements in that area which we could pass along to others? I'm not a contester so I'm happy but apparently it is at least a perceived problem among contesters. 73  
Photo of Steve N4LQ

Steve N4LQ

  • 568 Posts
  • 92 Reply Likes

Posted 4 years ago

  • 5
Photo of George Molnar, KF2T

George Molnar, KF2T, Elmer

  • 1652 Posts
  • 600 Reply Likes
The 6000 uses relays to provide better signal quality, which may slow things down a bit. Perhaps up to 0.1 second per transmission. If a contest operator makes 1000 contacts, and transmits three times per contact, 3000 PTT times .1 second delay, or 300 seconds over the event. Five minutes. How is that costing contacts, really? Especially considering that the time from PTT to speech probably doesn't truncate anything from the exchange (what can you say in 0.1 second?)

There are several contest ops here, and I believe the consensus is that the abilities the rig provides more than outweigh an imaginary fault.
.
Photo of Ross - K9COX

Ross - K9COX

  • 345 Posts
  • 107 Reply Likes
The only relays that I know of that take that much time are usually hand operated with a very long throw.
Photo of W7NGA

W7NGA

  • 446 Posts
  • 189 Reply Likes

simply listen to your signal on an external receiver ... you might be surprised.

73's

dan  W7NGA

(Edited)
Photo of K2CM

K2CM

  • 265 Posts
  • 18 Reply Likes
An f 6.7 k, 2kw pep, a 10 element beam will certainly outweigh any small layency issue.
Photo of Steve N4LQ

Steve N4LQ

  • 568 Posts
  • 92 Reply Likes
According to the "contester"...The time lapse between when the other guy releases his PTT + time your Flex takes to reproduce his transmission in your ears + time to transmit your response is great enough that someone with a faster rig grabs the attention of the other guy first. It's a difference of over 100 milliseconds between the 6000 series and typical other rigs. I'm just quoting others here but I suspect it is a little more than just an "Imaginary fault". 
I thought maybe that maybe with ver 1.4, moving some of the cpu work to the 6000 might change the turn time?
Photo of Lee - N2LEE

Lee - N2LEE

  • 301 Posts
  • 154 Reply Likes
Steve what ever you do don't convince your contester friend to get a flex. Let him think the delay is a problem.

The last thing we need is another contester than can hear stations that the Flex users can hear. :)
Photo of W7NGA

W7NGA

  • 446 Posts
  • 189 Reply Likes

latency - it's enough that you cannot realistically monitor and assess your signal real-time on an external receiver, and I suspect it could be an issue for the consummate contester.

dan  W7NGA

(Edited)
Photo of Ken - NM9P

Ken - NM9P, Elmer

  • 4189 Posts
  • 1334 Reply Likes
The last time I looked, there is no interlock on the flex radio than prevents pressing the PTT before the other station's last syllable fades into the noise. A good contester knows his station and makes adjustments accordingly.

One can press the PTT a split second before the other station finishes his "QRZ" and be right there on top of everyone else. In fact, I often see stations begin transmitting several seconds before the DX ends his transmission, presumably on order to get the drop on the competition. (I often wonder if they know that the other station cannot hear them until the DX station stops transmitting....)

Yes, timing is often more important than raw power, and articulation is often better than compression, but much of that timing is in the footswitch or VOX, or even more..in the operator's mind.
Photo of Bill Roberts

Bill Roberts

  • 226 Posts
  • 46 Reply Likes
Is this contester named Archie Bunker?  Does he come configured with preconscrewed ideas?
Photo of Steve N4LQ

Steve N4LQ

  • 568 Posts
  • 92 Reply Likes
Bill that is so funny. I'l just tell those ole contesters they are Archie Bunkers...Yea that will fix'em. Man why didn't I think of that? 
Photo of Steve N4LQ

Steve N4LQ

  • 568 Posts
  • 92 Reply Likes
The Flex-5000 turnaround time = 29ms
The Flex-6700 turnaround time = 140ms
Per QST reviews
We went from one of the fastest to by far the slowest. 
Photo of Mike Hoing

Mike Hoing

  • 271 Posts
  • 43 Reply Likes
The change is ms is 111 The is little more than 1/10th of a second. I don't know about you but that's way down the list of things holding me back from working all entities, state, counties.

So far down the list that it never crossed my kind and now that I know I hope it never crosses my mind again. The ability to hear thing and weed them out is far more important then 1/10th of a second

Mike
N9DFD
Photo of caravanken

caravanken

  • 4 Posts
  • 0 Reply Likes
Here is some perspective: .111 is faster than Ashley Force Hoods average reaction time .115 in a Funny Car. It is .026 slower than her dad's at .085. 
(Edited)
Photo of Steve - N5AC

Steve - N5AC, VP Engineering / CTO

  • 1053 Posts
  • 1076 Reply Likes
Relays have nothing to do with these measurements.  The measurements are a result of signals passing through the signal chain in the receiver and the transmitter.  As one of my radio mentors was fond of saying "all the delay comes from the quality of the final filter."  This is a hard concept to wrap your head around if you are not familiar with it, but the net-net is that the better filter you provide at the final sampling rate, the more latency is introduced.  The FLEX-6000 is the first radio I am aware of to provide variable-depth filters.  We can significantly reduce the delay by reducing the filtering and we can do this dynamically.  Even the FLEX-5000 with PowerSDR did not have this capability.

Our read when we designed the signal chain for sideband was that a little latency was less important than good filtering.  So we "maxed out" the filtering capabilities for sideband.  The filter dynamically changes width in DIGx modes and in CW.  If you would like to try out a shorter latency with reduced filtering capabilities, try running in DIGU instead of USB and set your filter width to greater than 2kHz.  The latency will go down to what we consider to be a very minimal level.  

The ARRL and others are not used to measuring radios that have complete flexibility in how things like this are controlled and so they put the radio in a given position and took a measurement.  The measurement is accurate for that one setting, but is not a complete picture of what the radio can do (and in this case doesn't explain the tradeoff that we made).  I'm sure I would have done just what they did.  As you guys probably know by now, if you provide a compelling case for making a change in the radio there are very few things we cannot change.  You can try using DIGU and let us know if you prefer it like this.  There is actually a table in the owners manual with this information along with a description of how it works (see 29.6.4 on page 140).  
Photo of George Molnar, KF2T

George Molnar, KF2T, Elmer

  • 1651 Posts
  • 600 Reply Likes
Thanks for the info - learned something new today!
Photo of KY6LA - Howard

KY6LA - Howard, Elmer

  • 3731 Posts
  • 1583 Reply Likes
That explains why I found the latency so low when running DIGU and DAX Audio when remoting...
Photo of k0eoo

k0eoo

  • 618 Posts
  • 87 Reply Likes
Thanks Steve, FRS made the right choice.  Better filters serve us ALL, every day and every contact....
Photo of Steve N4LQ

Steve N4LQ

  • 568 Posts
  • 92 Reply Likes
Steve
Thanks for the informative reply and I now understand. So to answer our contesting friends concerns we need to explain to them that excellent filtering is more important than longer delay. 
I guess a "contest mode" would be a nice marketing tool. :*)
72
Photo of K6OZY

K6OZY, Elmer

  • 542 Posts
  • 212 Reply Likes
I think a "contest" mode for USB/LSB filters is a great idea.  We can test latency using DIGU/DIGL but the PROC is disabled so it's not a long term solution.
Photo of Steve - N5AC

Steve - N5AC, VP Engineering / CTO

  • 1053 Posts
  • 1076 Reply Likes
If you guys truly feel you are missing contacts due to the latency I want to hear about it.  If not, I would argue that the filtering is more important.  Agreed that DIGU is not a long term solution -- I just wanted to show you an easy way to test the difference.
Photo of Charles - K5UA

Charles - K5UA

  • 319 Posts
  • 89 Reply Likes
Woudn't a contest mode with poor filter skirts be counterproductive ? If there is one situation where you need good filter skirts ... It's in a contest. Surely no contester would give up these fantastic shape factors for a few msec less latency. How many seconds would be lost in a 48 hour contest asking for a repeat of the exchange because a poor filter skirt made copy impossible?
Photo of Tim - W4TME

Tim - W4TME, Customer Experience Manager

  • 9186 Posts
  • 3541 Reply Likes
"Wouldn't a contest mode with poor filter skirts be counterproductive ?"

Yes, but sharper filters result in greater signal propagation delays. 
Photo of Chris Tate  - N6WM

Chris Tate - N6WM, Elmer

  • 921 Posts
  • 260 Reply Likes
Steve I think the hard core contesters, (a group I belong too) would appreciate a balanced solution to this latency presented.  It WILL make a difference. Optimally a way to do this without much compromise would be key here.  and if there is no way to correct it without sacrifice, then it should be a option that can be invoked with a click... I am sure you wizards can make it happen ;-).  I know what the power contesters are looking for.. and the flex is on the radar scope of many of them.  a few tweaks in this regard could very well put it in the bulls eye.
Photo of KY6LA - Howard

KY6LA - Howard, Elmer

  • 3735 Posts
  • 1599 Reply Likes
You actually do not need to do that much to achieve a "Contest" Mode with the lowest latency in the market...

All you need to do is to route your TX and RX audio via a DAX Channel.. and then have your device (speaker and mike) listen to those channels..

I could provide your friend with screen shots as to how to do it as that is how I do it with iPad Remoting via DAX..before Remote was available in 1.4

BTW,,, at times I get hit with the contesting bug and I can be a serious contestor... I have won my ARRL section using SDR's and have placed 2nd in World and #1 NA in a JIDX SSB Contest... so I can speak with some authority about contesting..... 

It always helps to have a contesting station.. I have a 6700. SPE 2K-FA and a SteppIR MonstIR @85' located 600'ASL 1KM from the Pacific Ocean.. so by definition I will always do well in an Asia Pacific Contest without trying too hard.

So what is the difference between #2 in the world and #10 in the world when I was not trying too hard......

Basically its the operator's skill...

that 100ms latency is a mental factor.. if no one had pointed it out there is absolutely not doubt that  it would not be an issue..as a great operator can win with even a lowly ICOM IC-756 Pro3...
Photo of George - AB4FH

George - AB4FH

  • 63 Posts
  • 17 Reply Likes
Steve,
Could you give us some data that compares latency to filter skirts? While I'm not the contester that I used to be in my younger days, with the call WB4PDQ, I think this is worth further discussion. On CW a difference of 100 ms is very noticeable at 40 wpm contest speed.
Photo of Mike va3mw

Mike va3mw

  • 824 Posts
  • 198 Reply Likes
I am calling BS on this and he is just looking for an excuse.

I have done hundreds of SSB contests both with and without the Flex and there is no difference that would ever make a difference in my score. If he really knew contesting and it was a pile up, it isn't about being first in, but being heard when there are less people calling.

I also contest with a remote base with about 100ms latency (and I have been for almost 10 years) and I have never noticed a difference where it would make an impact on my score.

It is 100% operator skill and nothing to do with the change over time. va3mw
Photo of Barry N1EU

Barry N1EU

  • 495 Posts
  • 123 Reply Likes
What's clearly needed here is hard data, not words - what are the rx and tx latency measurements in milliseconds for the range of filter slope and audio connection settings?  I'd for one appreciate seeing that for the 6300 and 6500.

Barry N1EU
Photo of Steve - N5AC

Steve - N5AC, VP Engineering / CTO

  • 1053 Posts
  • 1076 Reply Likes
We'd have to go plot them, George, which we could do.  I don't have the plots ready today.  I can tell you that with the highest latency and number of taps, the filters are virtual brick walls.  If you compare the filters in CW mode to the finest crystal filters available in other radios, the shape factors we produce are significantly better.  Here's a post where several folks chimed in, did some plots and talked about the filters:

https://community.flexradio.com/flexradio/topics/cw_filter_shape_factors?topic-reply-list%5Bsettings...

The same filters used in CW are also used in sideband so the same numbers apply.
Photo of George - AB4FH

George - AB4FH

  • 63 Posts
  • 17 Reply Likes
Steve,
Excellent read, thanks. I didn't see any numbers for the associated latency of each bandwidth, however. Based on what Howard says below, I agree that latency will only matter to the average contester who does a lot of S&P. Hmmm...I think this means it is a spec that could be significant to a number of us average Joe's. I would propose, as someone else suggested, either having a contest latency switch, or make it prominently visible in the documentation how to choose low latency.
Photo of Michael - N5TGL

Michael - N5TGL

  • 308 Posts
  • 101 Reply Likes
I'm in the camp that the 140ms delay is just an excuse.  It truly is all about timing, understanding how the other operator is operating and jumping in at the right time.  Unless you have a big gun station, being the first to rattle off your call after "QRZ?" is likely to not get you an answer.
Photo of WA2SQQ

WA2SQQ

  • 475 Posts
  • 122 Reply Likes
In the digital photography world people who scrutinize image quality by viewing images at 400%, or quote numbers that just are not "visible" are called "pixel peepers". My take on this is they are complaining about this "spec" and using it as ammunition to support their decision to purchase the brand they have, over a Flex. Most of them probably wish they waited and bought a Flex! My Flex 6500 allows me to enjoy the hobby which is what it should be about.
Photo of Barry N1EU

Barry N1EU

  • 495 Posts
  • 123 Reply Likes
I would suggest that where this becomes a real issue in serious (i.e., high rate) ssb contesting is when the latency (sum of rx plus tx latency) is equivalent to a single spoken syllable.  In one likely scenario, what happens is that the dx will hear a quicker responder's signal and then your signal come in with that syllable (or character) delay and the dx gets your callsign wrong.  

Latency is a trade-off with filter slope steepness.  To throw out a number, perhaps a 2:1 minimum filter slope is needed for serious ssb contesting when the least amount of latency is desired (i.e., sacrificing filter slope steepness for less latency). 

Barry N1EU
(Edited)
Photo of Steve W6SDM

Steve W6SDM

  • 625 Posts
  • 283 Reply Likes
There are some really detailed high tech explanations here.  Here's my low tech view of the matter.  We're talking syllables and milliseconds here.  As a mediocre contester myself, I don't see how that can possibly impact your score unless you are one of the top three contenders and the winning difference comes down to that last single contact.  My guess is that this guy scores somewhere in the middle of the pack - otherwise he would be developing contest skills that worked around whatever characteristics the radio had.

This reminds me of the guys who speed the "5nn" up to 65 WPM in a CW response - how much time did they really save and did it make a difference?

Whatever latency may be characteristic of the Flex is much outweighed by the various advantages that a Flex offers to contesters.  I wouldn't be surprised if they didn't come out with a SDR-assisted category to level the playing field.
Photo of KY6LA - Howard

KY6LA - Howard, Elmer

  • 3735 Posts
  • 1599 Reply Likes
Winning Contests is about Rate. You don't run high rates by running Search and Pounce. You run rate by holding a frequency and letting everyone come to you. When holding a frequency the 100ms latency is not an issue PERIOD

I can see where 100 ms might slow u down in S&P But As I said you are not running high rates in S&P. Even then a BIg Station will easily dominate most everyone else who beat you to the 100MS punch.

So I can only surmise that the Latency issue is. Not an issue for Big contest winning stations but might be an issue for someone in the middle of the oack.
Photo of George - AB4FH

George - AB4FH

  • 63 Posts
  • 17 Reply Likes
Sorry, Howard, we "doubled". Lol
Photo of George - AB4FH

George - AB4FH

  • 63 Posts
  • 17 Reply Likes
BTW when Running a frequency, I often widen my receive slice bandwidth to make sure I don't miss any callers. Katashi Nose, the former KH6IJ, contester of fame, would often copy multiple stations in his head and answer them in rapid fire. I know, because I was one of the callers back in the '70s. Now that's a way to up your rate...but few of us mortal can achieve that.
Photo of George - AB4FH

George - AB4FH

  • 63 Posts
  • 17 Reply Likes
Howard, when using N1MM+ in CW as SOA, I've hit 100 contacts per hour rate in S&P for the first hour on a good band. The trick is to use CW Skimmer to load up the bandmap. Then to get to the next frequency, you just tap Ctrl < or > and it takes you to the next unworked station on the map. You only have to wait long enough for him to finish his call. I'm sure to win you need to Run some too.
Photo of Barry N1EU

Barry N1EU

  • 495 Posts
  • 123 Reply Likes
Latency is not an issue when you're running at a high rate?  If it's enough latency that you miss the first syllable/character of a fast responder, it WILL slow you down.
(Edited)
Photo of KY6LA - Howard

KY6LA - Howard, Elmer

  • 3735 Posts
  • 1599 Reply Likes
Because most of the time when you hold a frequency it is like a DX Pileup so you are trying to separate the callers in your head... Frankly the slightly later callers are usually heard better than the call stompers.

Americans speak at 5-7 syllables per second - so 100ms is perhaps 1/2 a Syllable lost...So if they are saying Kilo Yankee Six Lima Alpha.. the lost sound would be half of Ki syllable.. something one easily correctly integrates back in your head..

Non Native English speakers talk at much slower rates so you likely would not even miss that first 1/2 syllable.
Photo of Ken - NM9P

Ken - NM9P, Elmer

  • 4189 Posts
  • 1334 Reply Likes
Howard, I agree with you on phone.  On CW it might be a little bit different.  You might totally miss the first letter of someone's call.

BTW... There are many other factors to reduced rate.  For example, running to FAST on CW requires more repeats because you are sending over the heads of the majority of contesters.  I have seen many articles that recommend speeds between 28-32 WPM, and no more than 35 WPM in a contest, especially if your call or report has lots of "dits" in it, (because they are the ones most prone to copying errors).  I cannot fathom why some people see the need to send at 40, 45 or 50 WPM in a contest.  They have gone well past the point of diminishing returns in the CW exchange game.  

When rag chewing with friends or handling traffic with experienced operators, I can understand the advantage of very high speeds.  But in contest style exchanges with operators of various skill levels, I think excessive speed is counter-productive.  If nothing else, it wastes other people's time as they try to copy your call and decide if you are fair-game or a dupe.  Frankly, I think that is kind of rude.

Ken - NM9P
Photo of George - AB4FH

George - AB4FH

  • 63 Posts
  • 17 Reply Likes
For clarification, I think the situation that latency becomes important in a contest is when you are doing Search and Pounce, i.e. moving frequency looking for the next station calling "test". In this mode, the other station will respond to the first, readable station that answers his call. I'm sure that being one tenth of a second behind, given equal readability, will cause you to not get the contact. So, now, you either have to find another station or wait to call the same station again. As Howard has said before, everything in a contest is about rate or contacts per hour. Each missed contact slows the rate. Imagine missing 10 such contacts per hour at a 100 contact per hour rate.
Photo of Ken - NM9P

Ken - NM9P, Elmer

  • 4189 Posts
  • 1334 Reply Likes
If you are S&P, and know your station lags a 100ms on beginning of your transmission, just develop the habit of being a little quicker on the mike.  End of story.  

I run S&P all the time because I don't have enough antenna or power to RUN a frequency.
If that is the case, my best chance is often not to be the first syllable anyway, because I will likely get stomped by the stronger stations.  Sometimes my best chance is to be the first of the second wave calls.  "Timing" doesn't always mean being first, but often it means being the "smartest."
Photo of George - AB4FH

George - AB4FH

  • 63 Posts
  • 17 Reply Likes
Ken (corrected), I think there are different cases. At a leisurely 30 Q's per hour I could afford to wait. But when at 100 Q's per hour S&P with N1MM+ and CW Skimmer loading the bandmaps, I often skip to the next station rather than wait for the exchange to finish. I work up the band like this and catch him on my way back down. My experience shows this works better for me. In fact if a station is spending too much time repeating CQ TEST over and over, I move and try to catch him one of the times he's near the end . My particular enjoyment is to work the contest for a few hours, enough to get 300+ Q's. This leaves more time for my family and other activities. In other words, I enjoy the exhilaration of speed and intensity, not the butt-in-the chair time of winning.

FWIW, you are absolutely 100% spot on when working a DX pileup while not in a contest.
BTW, I am a CW contester. ;-)
(Edited)
Photo of George - AB4FH

George - AB4FH

  • 63 Posts
  • 17 Reply Likes
To elaborate further, it's good to run SO2V. Then, while waiting, switch to the other slice/VFO and make a Q...switch back to VFO A and make the Q you were first waiting on.
Photo of Bill -VA3WTB

Bill -VA3WTB

  • 3792 Posts
  • 914 Reply Likes
Perhaps his mind has to much latency. What do you do about that? Latency problems, I just can't buy it....
Photo of Rich McCabe

Rich McCabe

  • 1120 Posts
  • 239 Reply Likes
Interesting. I guess this explains why I cant seem to listen to my 6500 audio on another receiver.  I thought my brain was not working correctly.  This is the first radio I have tried to EQ via monitor on another device and although all by radio buddies do it, I could not make it work for me.  Maybe my drain in not bamaged.
Photo of Tim - W4TME

Tim - W4TME, Customer Experience Manager

  • 9186 Posts
  • 3541 Reply Likes
You should be setting the EQ with all processing turned off, such as PROC os it will not "color" your audio and will also decrease the latency a little when listening to your transmitted signal in a second receiver.
(Edited)
Photo of Joe, KQ1Q

Joe, KQ1Q

  • 79 Posts
  • 21 Reply Likes
What is the proven basis that a 100 mS latency differential results in loosing a "quick draw" contest pounce? That is 1/10th of a second. According to this research paper, the average English syllable length is about 200 mS: http://www.researchgate.net/profile/Steven_Greenberg5/publication/222530556_Temporal_properties_of_s...

This means on average, the latency might result the other party getting a 1/2 syllable head start, given equal human response times.

It would be interesting to set up a double-blind test with variable Rx latency (radios not required) and see if unprepared subjects can even detect 100 mS, or if others can truly leverage that to their advantage. Until that is investigated I don't see pulling limited resources off higher priority tasks to implement something with unproven benefit, based on an allegation with multiple possible causes besides latency.
Photo of Barry N1EU

Barry N1EU

  • 495 Posts
  • 123 Reply Likes
Joe, that sounds about right to me - I think 200 msec is the critical delay where it's going to definitely impact you at times.

Barry N1EU
(Edited)
Photo of Joe, KQ1Q

Joe, KQ1Q

  • 79 Posts
  • 21 Reply Likes
Yes, that was my point. If (just supposing) 200 mS is the critical delay threshold, and the Flex 6000 is 1/2 of that, how would it cause a major problem? Actually I'm not sure anyone authoritatively knows what the latency threshold is where it provably and measurably causes a disadvantage in this one narrow area of contest performance.

This is especially so since changing to lower latency would involve losing other filtering advantages. Any hypothetical advantage from reducing the 100 mS latency might be offset by worse filter performance. If poorer filtering caused just 1 in 20 QSOs to require a single 3 sec retransmit due to unintelligibility, overall contest performance would ironically be worse by using faster latency.
Photo of Peter

Peter

  • 21 Posts
  • 4 Reply Likes
I have done tests on a couple radios regarding RX latency. This because a while ago I had a 6300 tuned to a Colorado time server frequency and a Kenwood 590 also. Noticed a perceivable delay between the 2 audio streams which made me wonder. Using a Rigol scope I was able to compare RX latency between several radios. Not surprisingly, the most basic, simple radios (like the MFJ Cub) had the lowest latency. Reason: No DSP and other complex circuitry. The Flex had the absolute highest delay around 150ms if I remember correctly. A K3 has about 15ms. TX latency (CW) I have not looked into, if anybody has done tests in that respect I would like to hear from them.

This conversation is no longer open for comments or replies.