Well,
- it's no complete call. Thus no final BYE, thus no phone's statistics available (assuming you configured "Stats In Bye" to Yes)
- you didn't configured "RTP Packet Size" to 0.020
- for a some reason, every packet is captured twice (you should verify the packets are not send twice over air but it's caused by capture itself)
To compensate [3] I removed duplicates before further analysis. I removed packets unrelated to particular call as well. The resulting file is attached so you may verify my conclusions.
There's no problem with the second stream (0x1ebdedff) - no packet lost, max delta as well as jitter is good. The first stream (0xc53e3fda) is different beast. 6.8% packet lost is unacceptable. Burst of 124 packets missing (#10571-10693) mean about 2.5s of silence. Not surprisingly, the sound quality is low.
OK. I filtered out duplicates as well as unrelated packets. Analysis of resulting file (wireless variant) follow:
5% packet lost is still so high. And even 142ms of delay. Note that for phone call the delay of 100ms is considered maximum for standard quality calls. I mean end-device-to-end-device(!) delay. You are exceeding overall limit in first hop.
I need not to listen the stream to assume the call quality will be low.
It is clear.
What's not as clear is the cause of packet los and delay measured and it will be hard to debug.
The packets may be lost because
- they has not been transmitted from phone at all because of
- firmware bug
- wireless environment condition
- they has been transmitted, but lost later on
- air (so they has not been received by AP)
- wire (so they has been lost in wired part of transmit)
You need to capture packets from air as well as monitor wireless environment to distinguish those cases. Unfortunately, it require special equipment to analyze thinks like it.
Some smart AP with detailed statistics may help a lot, but not those cheap one.
Well, there's something simple and cheap you can try to decide the SPA525G2 is guilty or not.
Use a wireless-to-wired bridge (please not the cheapest crap you found around) on phone's side. Connect phone using wire to it.
In such scenario phone doesn't affect wireless operations in any way. It's the bridge who is sending/receiving packets to/from air.
Repeat the test. If the results become acceptable then issue is related to phone. If there will be still packet lost and so high delay, then phone is not causing the issue.
