Sorry if this post is redundant but I could not find anything specific to this issue…
My problem is that I am having Audio Quality issues within my local network. If I call straight from X-lite to Asterisk (let’s say to record VM msg or listen to it) the quality is very bad. Lots of Crackling, hissing, etc. That result is both with 711u or a. If I change my Codecs to GSM, it is much better but not good enough…for my standard
From what I read, the g711 should be of highest quality because less compression. Why is it that in my case, it produces the worst? My calls externally are just as bad.
My current Asterisk Setup.
Asterisk@home 1.5 (also tried every other releases)
2400AMD Athlon
512MB RAM
80GB
1Gigabite NIC
X-lite phone software
low network traffic
I just don’t understand why I get all this sound quality problems. I’ve also tried with a different System but get the same results. Is there something I am missing in my setup? Perhaps its something simple. Most forum search results come back with audio issues but mostly all related with the provider. In this instance, I haven’t reached the provider yet.
Simply using a SOHO device. USR8054 to be precise.
It’s a 10/100 switch with Router. So far, I’m just staying within the local network. I’ve made sure nothing is downloading hogging all the local switch bandwidth.
[quote=“yanicg”]
My problem is that I am having Audio Quality issues within my local network. If I call straight from X-lite to Asterisk (let’s say to record VM msg or listen to it) the quality is very bad. Lots of Crackling, hissing, etc. That result is both with 711u or a. If I change my Codecs to GSM, it is much better but not good enough…for my standard
From what I read, the g711 should be of highest quality because less compression. Why is it that in my case, it produces the worst? My calls externally are just as bad.
I just don’t understand why I get all this sound quality problems. I’ve also tried with a different System but get the same results. Is there something I am missing in my setup? Perhaps its something simple. Most forum search results come back with audio issues but mostly all related with the provider. In this instance, I haven’t reached the provider yet.
Any help would greatly be apreciated!!!
Regards,
Yanic[/quote]
Yanic:
So you are one of the many people that believed the “no hardware needed” tale? You can have crystal clear audio between two human users, simply because each one of them has a SIP phone with DSPs inside. However, if you try to have the Linux software to play a prerecorded greeting, there is NO WAY you will achieve anywhere near the level of audibility that business people have come to expect from a PBX.
Don’t waste your time with codecs or looking at the configuration files.
If there was a way to get some decent “software-only” audio, the Asterisk developers would have made it the default. Those terrible sounds are a very bad advertising for Asterisk, and its developers should come clean and have a disclaimer saying that the “no hardware needed” is not feasible in a realistic business setting.
I assembled several IVRs and made comparisons, keeping microphones, etc. constant among systems. My recordings are always made in a professional audio studio here in Boston.
There’s no magic here, DSPs are still required.
Do yourself a favor and get a hardware board if you want a better sound quality.
My experience of bad sound quality is very similar to yours, Yanicq. I can record my voice on my computer perfectly, but when i try making calls through asterisk using a softphone on that computer, it sounds bloody appalling.
I’m not sure exactly what’s going on here, but i think it’s an interaction between two or more different things - hardware, software and, maybe ethernet stuff.
However, it’s very easily fixed. Forget about using softphones like xten and invest in a cheap hardphone, like the Grandstream BudgeTone 101. That’s what i did and the audio quality went from virtually unusable to as good as a normal landline phone.
Softphones are crap, basically, although they do seem to work ok in some circumstances.
Thanks a million for both your replies. Very apreciated and will prove to be helpful in the future. I’m just starting off with Asterisk so we’ll see.
Telephony: Not that I believe “hardware not required” Just all new to this so don’t know what to expect between different asterisk tools. By Hardware board, what do you mean exactly? hardphone or other device like some sort of sound card?
WillKemp: I will take your advice and try a hardphone. Maybe this is what I’m looking for.
[quote=“yanicg”]Thanks a million for both your replies. Very apreciated and will prove to be helpful in the future. I’m just starting off with Asterisk so we’ll see.
Telephony: Not that I believe “hardware not required” Just all new to this so don’t know what to expect between different asterisk tools. By Hardware board, what do you mean exactly? hardphone or other device like some sort of sound card?
Thanks again![/quote]
You need to make sure that everywhere there is a voice, you must have some sort of DSP around to encode it.
For example: one person talking to another person.
If one of them is using a softphone, perhaps the software cannot take advantage of the DSP in the sound card which were placed there to record voices, not necessarily to support human conversation.
In the above scenario, all you need is 2 hardphones, and you’ll be fine. No voice hardware needed in the Asterisk server, just the Ethernet card.
Now let take a look at another scenario: a person calls to your Asterisk box (it doesn’t matter whether from POTS or SIP) and hears a recording. You better have some sort of hardware that will assist the Linux kernel to properly “pronounce” the recording, otherwise the software-only quality will be crappy, I don’t care what codec you use.
I think what you’re saying is fundamentally true, but i don’t think it’s necessarily quite as black and white as you make it sound. In my (limited) experience of hardware/software combinations running asterisk, i’ve noticed some very big differences in the audio quality that different machine and o/s version combinations produce.
I’m not sure what’s going on exactly, but it doesn’t seem to be a straightforward matter of processor power/speed and RAM. I think the specific architecture of the machine plays a major part in it, but i’m fairly sure there’s an interaction with the operating system influencing it too.
For example, the laptop i’m writing this on (an IBM Thinkpad A20m) seemed to perform much better as a softphone when i was running the 2.4 kernel than it does now i’m running 2.6. I must point out, however, that the 2.4 kernel i was using included patches for low-latency etc - not that this necessarily had an effect on the way the softphones worked, as they weren’t able to take advantage of that stuff.
Also, going by the experiences that other people have written about on this forum, there seems no doubt that very good results can be obtained using standard PC hardware. However, some very bad results can be obtained this way too!
This issue needs considerably more investigation, i feel.
