Hello.
This is Jack. I have been a member for several months. I have read lots of info and have tried several set up variations. I’m still not getting decent latency. First of all, I am using wired ethernet. My only available connection is not fiber. I’m using an external audio device. I’m using wired headphones. I have purchased and am using a new audio interface. ( Focusrite Scarlett Solo 4th generation). and I have tried assorted settings. (Perhaps you could enlighten me on what settings you feel are optimal). My desired reason for using JackTrip would be to play in a jam situation or teaching a class. Other than the lack of fiber internet, I should have it all. Is it doable without fiber internet?
Thanks,
Jack
1 Like
Hi Jack and welcome back. I did not read anything in your post about the distance between you and others to the studio server where you are connecting. Latency has more to do with distance than any of the other details you have mentioned above, so a distance of more than 300 miles to the server by anyone connected will cause more latency.
And I want to add that even with fiber, we face the speed of light limitation.
Fiber Internet offers latency that is 4 to 10 times better than any other type of connection. It’s nearly impossible to get below 25 milliseconds (the typical threshold often referenced for live music performance) without it. Depending on what you are trying to do, it may be required.
Hi Synthia,
I appreciate you getting back to me. I am a newbie here. I really want to get this to work. As far as your question: So far, I have just been testing my latency by opening up my studio, following the basic guidelines, connecting a guitar or microphone, connecting to San Francisco (I am in Oakland)and observing the latency. I’m not connecting with anyone else. My latency will show up as high as 300 ms. I know this is a pretty basic, uninformed question. But, do I have to be connected with someone, regardless of their distance, to get an accurate latency reading? Much thanks. Jack
Thanks, Mike. I will be doing live jams and one on one music lessons, with a desire to play over a track at the same time. Sounds like I might have to wait for fiber connection.
Jack
Hi Jack,
Thank you for your question. You should be able to get an idea of the
latency with just your connection to the Virtual Studio in San Francisco
from your location in Oakland, probably less than 20 miles away.
The high latency you are reporting makes me wonder: by any chance is
your internet service connected to a mesh network?
I looked up the definition of mesh network. I don’t believe I have that. I do have 2 neighbors in my apartment building doing wi-fi off of my network.
Additional users on your network, whether WIFI or ethernet, can impact
the availability of bandwidth and amount of latency.
Regarding mesh networks: very popular in the Bay Area, the providers
advertise the low cost and wide service area of a mesh network. They can
give you ethernet inside the building but outside is connected to a WIFI
mesh in order to serve a larger number of users and lower cost to each.
Also Jack, the latency fluctuations that you have reported in previous
posts is typical of WIFI’s dips and spikes of data flow. If you and two
additional people are using your WIFI, it is not unusual to have high
and erratic latency readings.
Which begs the question: are you using wired Ethernet cable from your computer to the router? You can have an Ethernet connection without it being connected to a fiber optic network.
Thanks so much for getting in touch again. To answer your last question: I am using an Ethernet cable from my computer to the router.
I will do a little experiment and have my neighbors turn off their computers as I test Jack trip.
Great idea Jack. Please let us know your findings.
One last tip is to disable Wifi on the router and the computer as well as Bluetooth on the computer. wireless devices tend to compete for computer resources and add varying amounts of instability and latency in the process.
Standing by…
Who is your Internet provider, and what type of connection do you have (Cable, DSL, etc)? 300ms is insane. Is that the number you see reported by JackTrip’s latency indicators? There are plenty of providers in the bay area that can consistently get you under 35ms. You can search for Fiber Internet availability via this link, but you can search on that website for other providers as well.
Checking in with you Jack, we’re so glad you’re here. I have been thinking about how you want to use JackTrip and that reducing latency is a critical need for you.
Please let us know if you have had a chance to test JackTrip again without your neighbors’ connections, or other ways to reduce the latency you are experiencing.
Hi Synthia!
Thank you so much for getting back to me. I did indeed check with turning off my two neighbors Wi-Fi. Sorry to say, nothing changed. Still coming in at around 300 MS. I think I might just let it go for right now. I will get back in touch with you after I get a fiber connection. Much appreciation.
Jack
1 Like
ok, understood. The latency reported here indicates that your internet
service could be going to WIFI outside the building. And there might be
non-fiber provider options that are not using WIFI outside the building
and offer reduced latency without the big jump to fiber.
Good luck with everything Jack, and thanks for sharing - we’re here if
you have any other questions or concerns.
Hi there - within a metropolitan area, distance does not matter much. Beyond that, I don’t think any remote playing tool works well to make music. ATT fiber provides lower latency than Comcast over coax, and even there’s a difference between the latency that ATT provides for 1 Gb/s vs. 500 Mb/s customers. Having said that, I play regularly over the net in a duo, in which one of us has 1 Gb/s fiber, and the other one 500 Mb/s Comcast coax. We tried Jacktrip, and it did not work, but Jambridge did (it’s more bare-bones, more tolerant of network deficiencies, and the sound is very good)
Actually, I evaluated multiple remote, interactive music playing tools (Jamulus, Jamkazam, Jacktrip and Jambridge), and there are only two that I think are usable (sound not degraded, consistent absence of clicks, cracks, etc, and reasonable latency): Jacktrip (JT) and Jambridge (JB). Next I measured the true latency of JT and JB. Both use uncompressed audio and (can) have superb sound quality. Jacktrip is more feature rich, but Jambridge has lower latency (22 ms vs. 34.5 ms), is easier to use, and free. I started using JB to play classical music and tangos during the pandemia, and nowadays I use it regularly.
Most of the remote playing software packages advertise latencies for only portions of their system. Instead of those under-estimates, I measured the complete, end-to-end time, including the remote playing software, mics, headphones, adapters, network, etc., which is what’s relevant to musicians. If you want to see the full setup, waveforms, and how I measured the latencies accurately, see the PDF I put in my Google Drive: Performance of Jacktrip vs. Jambridge.pdf - Google Drive
–Dan
Hi Dan, your latencies seem rather high. Are you sure your setup is optimal? By optimal I mean nothing wireless, fast audio interfaces.
Under optimal conditions in a metropolitan area I would expect RTT latencies way under 10ms.
With optimal setups I regularly play with people more than 800km away from JackTrip servers and it works.
@dchapiro These forums are for JackTrip users, so please refrain from using them to promote your own projects, especially doing so with misinformation, and reposting the same thing across several different inactive threads (our software flagged these duplicate promo posts as spam. I allowed this one but removed the others).
There are a number of obvious things that are factually incorrect in your PDF. For example, JackTrip is free open source software (MIT license). JackTrip can also be used as a peer-to-peer client. It started out only supporting peer-to-peer, but we later added more options because there are rarely any advantages, including with regards to latency. Only in extreme cases (or extreme settings configurations) does peer-to-peer provide latency advantages.
My bigger concern is that your measurements indicate you clearly have something setup incorrectly. Our published results demonstrate 25 ms from San Francisco to Portland using default settings, and it’s very easy to reproduce these. Having been doing this for years, anything over 25 ms on Fiber in the SF bay area is not remotely believable to me.
With a few strategic settings changes in JackTrip, such as bumping up the buffering settings or using a location on the other side of the world, it’s pretty easy to generate latencies over 100 ms latency as well…