ND5 XS 2 (see my profile and my previous posts).
ND5 XS 2 (see my profile and my previous posts).
That’s what I thought I read. My understanding was the drop outs were effecting the 1st gen streamers. At least someone is getting to enjoy Tidal with their Naim.
Well Naim are aware on the beta forum, but they are looking for input. So if anybody with the know how to sniff using wireshark as per my post with a legacy streamer who is being affected could help Naim by sharing their findings.
Frustratingly I can’t assist this time, as I have now a latest generation streamer, and their updated architectures are not affected… come on there must be some other IT engineers out there with a legacy streamer who is affected…
I’ve had a couple of days width limited drop out…today every track is dropping out, arhhhh!
NDX in NZ
Yep, I’ve abandoned using Tidal direct to my UC2 and Uniti2 and now have 3rd generation Apple TVs connected and using Airplay from the Tidal iOS App and Mac desktop App.
I’d like to help, but no idea what all of that actually means… I mean, it all worked fine until last week. Not very good.
i have problem with TIDAL on my ND5 XS. I can’t play without drop outs. my internet is OK. radio and TIDAL on my phone works OK in the same network.
i see one solution change TIDAL on Spotify… it is sad.
I watch this topic, and I hope we find some solution for this problem. TIDAL is (was) my general source of music
Mariusz from Poland
Thanks, but as I said it was for any IT or network engineers out there, it’s just basic network monitoring
I don’t mind giving it a go if it’s tinkering on my Mac, I might learn something.
Okay… it will mean me re-signing up for tidal but I can try this for you. I have to change my switch so I’ll get a ‘dumb’ hub from work today and see what happens tonight. I’m not sure if my problem is 100% the same as those reported thnough. What I was seeing is some albums simply not playing at all which I think is a different issue… something about incompatible files.
You will need a managed switch that can support port mirroring or SPANing as well. A Cisco 2960 for example can be configured to provide a SPAN port.
In my case I have an old ‘dumb’ hub… we keep this at work for exactly this sort of thing
You really need to port mirror or SPAN. A network hub can be used instead of a switch but obviously as the hub is half duplex it affects the TCP latency characteristics, and almost certainly this is the area that is causing the issues, so best use a switch and full duplex monitoring to support the regular mode of streamer full duplex operation.
I could get a managed switch with port mirroring. So the thoughts here are this is a latency issue? In that case would we need to also factor in what people are using for a backbone and even their broadband service so neigh on impossible to test/factor for?
Yes… in the early days this was quite a struggle… I helped Naim in this area with various ‘real world ‘ scenario traces. There is limited TCP window space on the first gen streamers and so are susceptible to TCP flow issues with higher latencies as seen on the internet with respect to the throughputput required for lossless. On home networks this is not an issue as TCP latency is typically minimal unless on a very poor Wifi .
The fixes in the end was to vary the TCP flow start strategy, and optimise the TCP stack on the streamer to reduce processing time … ie streamer induced latency… to the minimum.
Rememember it’s the dynamic you are looking for in higher TCP latency occasions… this will happen from time to time on an otherwise suitably functioning access…
I have a NetGear GS105 switch and a NZ Telecom router/modem.
Mike, thanks, however that equipment can’t support the analysis capabilities required.
If latency is what I call buffering, then yes, I think so? But, why it’s suddenly arisen is another question, has Tidal changed how it’s been delivered (i.e. higher resolution = less available latency = quicker drop outs?).
Mike, TCP latency is the time it takes for the transport protocol to send and confirm packets from host to host. A series of packets can be sent and then acknowledged on mass, individually, or if packets are lost due to over run or rare network loss they can be requested to be re sent. This transport flow and exchange determines the overall speed of confirmed or assured data transfer.
There is a relationship between throughput and TCP latency for a given over all data latency. The latter is determined by the TCP stack memory on the hosts. This overall relationship is determined by both hosts either end of the TCP communication. If Tidal change their TCP flow characteristics for whatever reason this could off balance the Naim streamer which has been finely tuned in this area due to the limited resources available in the first generation streamers. This is also why a proxy can help… as it creates a new TCP flow with the streamer with home network characteristics.
The internet itself is like an elastic pipe in terms of the time taken for data to traverse it… the time is constantly changing.
btw… and this might be relavent, what ‘quality’ tidal package are people using that are having issues? I’m only using Tidal as an aid for music purchases (trial listening) so I’m on the lower ‘high’ quality rating so towards the lowest end of traffic demands.