Sound quality

[Ross Clarkson 1041620]

Firstly, i am Ross Clarkson, not Ross Carlson, hee hee.

 

Ok, so i think i have managed to get myself sorted and connected to VATSIM thanks to all your help and the PRC. I have just spent sometime at EGLL listening to TWR.

 

Just a quick question: The sound quality from TWR was terrible. Is this something i can fix or will this be my headphones? They are not the best!!

 

Thanks in advance and i can not wait for my first flight.

 

Mr Clarkson.

[Norman Blackburn]

Ross,

 

In each of VATSIM's clients there are tick boxes to simulate radio distortion - make sure to turn these off.

Try changing the output from wav to directsound (or vice versa).

You mention headphones - are these USB or analogue ones? USB, in my experience tend to be a little better quality.

[Mike Phyrio 988306]

You gotta do some more testing man. Just one connection doesn't let you know a thing. I have people in different places that sound like they're inside a barrell chewing on rocks and others who you'd swear were sitting right next to you. So, do some more tests before you freak out.

[Ross Clarkson 1041620]

Cool, thanks for your replies. I was going to invest in USB ones soon anyway so that is next on my list. Will also try switching to Directsound as currently on Wave.

 

Cheers guys,

[Ross Clarkson 1041620]

Ok, i have spent quite a lot of time now at various airports. Controller sound quality is terrible to the point where i can not understand my instructions so i have to ask for text!! I can hear other pilots much better.

 

Will USB headphones really make any differnce or is there some settings within VATSIM somewhere?

 

Help would be greatly appreciated as i have figured everything else out!!

 

Regards,

 

Ross.

[Anthony Atkielski 985811]

Controller sound quality is often poor, with high-frequency attenuation, lots of white noise, and distortion. Sometimes turning off the VHF simulation helps a little, although the result doesn't sound quite as realistic.

[Brad Littlejohn]

Controller sound quality is often poor, with high-frequency attenuation, lots of white noise, and distortion. Sometimes turning off the VHF simulation helps a little, although the result doesn't sound quite as realistic.

 

I still believe this is wrong, as pilots hear me fine as a controller, and controllers hear me just fine as a pilot.

 

To show that, in another recently posted thread, Mike Evans, who helped with the RW protocol that is used by all of our clients, said the following:

 

the RW protocol uses 300 BYTES per second of audio.. * number of people in the channel.. it's really trivial bandwidth.

 

You can not say that controller quality is poor and pilot quality is not, if they both use the same protocol and consume the same amount of bandwidth per channel, per server.

 

In short, all of it is poor, or none of it is poor. I see that you haven't controlled on here, and have indicated in the past that you have no intention to control, so it is really hard for me to see how you can justify or quantify your statement.

 

BL.

[Ross Clarkson 1041620]

The thing is, the controller seems to be able to hear me fine but i can not even understand what he/she is saying at all! I dont really want to spend £60 on some USB headphones if they are not going to make a difference.

 

How do i turn off VHF?

 

How do others understand or is it just something that i will get used too??

 

Regards,

 

Ross.

[Brad Littlejohn]

The thing is, the controller seems to be able to hear me fine but i can not even understand what he/she is saying at all! I dont really want to spend £60 on some USB headphones if they are not going to make a difference.

 

How do i turn off VHF?

 

How do others understand or is it just something that i will get used too??

 

Regards,

 

Ross.

 

USB Headphones should be fine. I use them myself as both pilot and controller, and haven't had any issues whatsoever.

 

The VHF button is togglable in the options section in SB3 (Not sure where in FSInn), but how that helps in 'realism' without knowing how it truly sounds from left or right seat in the pilot's/FO's headset is beyond me. I can guarantee they sound don't sound distorted or scratchy or anything of the sort (and this being from the left seat of a B757 and a NorCal Approach controller's headset). In fact, it's one of the clearest broadcasts I've ever heard.

 

But definitely uncheck the VHF box. It does sound clearer and much closer to how it really sounds.

 

BL.

[Bill Casey]

Ross (Clarkson!) - since you're a UK pilot then drop me a line (PM or e-mail, check my profile on here for e-mail), we'll meet up at an airfield somewhere with a controller and we'll go through the setups and work out what your issue is. That's the sort of thing us Vat-Uk new pilot mentors do. If you could first pop onto the front page of the Vatsim-UK website and look for Teamspeak on the left hand side menu, get set up on that and then we can talk to each other "off air" and be able to clarify/fix everything for you quite quickly.

 

For clarification (no pun intended), in general controller comms are not dreadful at all, they're quite often clearer than being on the phone but if the phraseology is not familiar and they say it quickly then it may seem dreadful to a new user. We'll get you sorted out one way or the other.

[Anthony Atkielski 985811]

I still believe this is wrong, as pilots hear me fine as a controller, and controllers hear me just fine as a pilot.

 

It's very specific to each controller most of the time. Some controllers are consistently very difficult to hear, others are routinely crystal clear.

 

It has occurred to me that location might have something to do with it, as I've had experiences in the past where audio would sound better from one location than from another (via different VATSIM servers). That's a bit mysterious, though, since the sound doesn't go through VATSIM servers (?), from what I understand.

 

Since there's no way to test one's own audio over the network I don't know how I sound. However, the occasional long-delayed echoes I get on the network sound clear (but they are abnormal so I don't know if I can trust how they sound).

 

You can not say that controller quality is poor and pilot quality is not, if they both use the same protocol and consume the same amount of bandwidth per channel, per server.

 

I didn't say anything about pilot quality.

 

In short, all of it is poor, or none of it is poor.

 

Nope. Some controllers are nearly impossible to understand, others are very clear. It's something that correlates with the controller, which implies that it has to do with the controller's audio configuration or geographic location, but it's specific to controllers. I can switch to a different frequency and a different controller will sound completely different from the first.

[Julien Pham 1042198]

Today in France I was listening to LFLB (Chambery) approach, and the controller sounds crystal clear, but a pilot did not, the pilot's sound was too low, but I doubt this is the same problem. Perhaps the pilot did not speak loud enough, or too far from the microphone...

[Anthony Atkielski 985811]

Today in France I was listening to LFLB (Chambery) approach, and the controller sounds crystal clear, but a pilot did not, the pilot's sound was too low, but I doubt this is the same problem. Perhaps the pilot did not speak loud enough, or too far from the microphone...

 

Pilots are affected in the same way that controllers are. As a pilot, it makes little difference to me (except for the occasional ear-splitting "beep" or blast of white noise when some pilots transmit), but no doubt controllers have a difficult time with pilots who have audio problems, and I hear them tell the pilots so in the worst cases. Actually, in some cases I cannot hear and/or understand the pilot but the controller can, so either he is receiving different audio from what I hear, or he has incredibly good hearing.

[Julien Pham 1042198]

Perhaps the controller has put the sound volume to the max

As a pilot I have engine sound and external sounds as well... the ATC has not so perhaps it can help...

[Garry Morris 920567]

I am guessing it has more to do with a combination of the particular headset/microphone the garbled person is using, and/or their microphone sensitivity settings. If the mic is set too sensitively, the controller will max out the input and cause distortion. Otherwise, all headsets are definitely *not* created equal. While you don't have to spend $100's, getting the $4.99 special at WalMart may not be the best idea either.

[Jonathan Thurston 917551]

I have noticed many times while flying that I can not understand what a pilot is saying, yet the controller can. When I am controlling, I have very few issues with not being able to understand pilots surprisingly. I guess it is something different in the client?

[Brad Littlejohn]

I still believe this is wrong, as pilots hear me fine as a controller, and controllers hear me just fine as a pilot.

 

It's very specific to each controller most of the time. Some controllers are consistently very difficult to hear, others are routinely crystal clear.

 

Then it is very hard to conclude that controller sound is often poor when it varies from controller to controller. That simply is not the case, and all controller sound quality should not be compounded into a single entity.

 

It has occurred to me that location might have something to do with it, as I've had experiences in the past where audio would sound better from one location than from another (via different VATSIM servers). That's a bit mysterious, though, since the sound doesn't go through VATSIM servers (?), from what I understand.

 

Correct here, Sound does not go through the VATSIM servers, however, at the aforementioned 300bytes, network congestion is hardly a problem.

Since there's no way to test one's own audio over the network I don't know how I sound. However, the occasional long-delayed echoes I get on the network sound clear (but they are abnormal so I don't know if I can trust how they sound).

 

You can not say that controller quality is poor and pilot quality is not, if they both use the same protocol and consume the same amount of bandwidth per channel, per server.

 

I didn't say anything about pilot quality.

 

In short, all of it is poor, or none of it is poor.

 

Nope. Some controllers are nearly impossible to understand, others are very clear. It's something that correlates with the controller, which implies that it has to do with the controller's audio configuration or geographic location, but it's specific to controllers. I can switch to a different frequency and a different controller will sound completely different from the first.

 

But what you're not understanding is that both pilot and controller sound uses the same protocol, and has been pretty much standard since this network's inception, and SATCO's inception at that. So it would be everyone's sound quality here, as it comprises of the same data that both pilot and controller clients use. So it would be the protocol having the issue, making both pilot and controller clients have the same audio problems, not just one side.

 

You can test how you sound through your client's audio calibration. That gives a good indication of how you sound over the network, as it basically routes the 300 bytes of data back to you via loopback. Squawkbox, FSInn, ASRC, Euroscope, and VRC all have this functionality.

 

BL.

[Anthony Atkielski 985811]

I am guessing it has more to do with a combination of the particular headset/microphone the garbled person is using, and/or their microphone sensitivity settings. If the mic is set too sensitively, the controller will max out the input and cause distortion. Otherwise, all headsets are definitely *not* created equal. While you don't have to spend $100's, getting the $4.99 special at WalMart may not be the best idea either.

 

Well, acoustically, one problem I hear is excessive white noise, which implies over modulation, digital systems produce white noise when they are overdriven. So they might just be too close to the microphone or have it turned up too high. The other common problem is absence of high frequencies, which at worst turns the audio into a burbling rumble. Without the high frequencies, a lot of phonemes sound the same when spoken. I don't know what's causing that. It is probably something at the analog end because just digitizing the audio wouldn't selectively drop high frequencies. However, if the audio is going through a codec (?), things get more complicated.

 

Another problem is extremely variable volume, for both pilots and controllers. I often have to tweak my voice level setting either because the controller is too hard to hear or (more rarely) because he is too loud. In this particular case, pilots are even more of a problem, since some of them come blasting in with an ear-splitting volume, and you can't really turn it down because then you can't hear the controller, who has a normal volume setting. The variations in volume are in fact realistic, but hearing pilots who are ten times louder than controllers (even though they have tinier transmitters) is perhaps somewhat less so.

[Garry Morris 920567]

I've seen some of that as well Anthony to be sure.

[Norman Blackburn]

I was listening to Oakland and Los Angeles Centres last night, and, by way of coincidence I noticed Anthony was one of the many pilots online.

 

Aside from the excellent controlling I observed, the quality of the voice communications was superb. Was this your finding also Anthony?

[Anthony Atkielski 985811]

I was listening to Oakland and Los Angeles Centres last night, and, by way of coincidence I noticed Anthony was one of the many pilots online.

 

Aside from the excellent controlling I observed, the quality of the voice communications was superb. Was this your finding also Anthony?

 

I flew a couple of places yesterday evening and I don't remember if there was a controller change or not. I recall that the controllers were pretty clear (minimal noise or high frequency loss) but the volume was a bit low for both centers. Usually I compare the volume to engine noise. I have settings for the different aircraft I fly that I use with SB to make sure the voice overrides cockpit noise. The settings are pretty constant after zillions of flights, so if I have to crank them up I know that the controller is a bit softer than usual, and if I have to crank them down I know that he's a bit too loud. I usually have the SB voice volume set to 42 in the small Bonanza and Baron, and 100 in the big iron.

 

I know there are occasions when controllers tell a pilot that they cannot hear anything at all of what he said, whereas I hear the pilot perfectly. Conversely, sometimes controllers are manifestly hearing a pilot with no problem, while all I hear is a blast of white noise. That has to be a network issue.

 

Sometimes I'm flying along and every single transmission is preceded by a half-second blast of white noise. If I disconnect and reconnect, usually it goes away, although sometimes it seems to require a sim restart.

 

On very rare occasions (I can only remember one specific instance) a controller will tell me that I'm completely incomprehensible. If I disconnect and reconnect to a different server, it goes away. I can't hear any change at my end, but the controller tells me that I've gone from totally incomprehensible to loud and clear.

 

So there are many unresolved mysteries here. A bandwidth of 300 bytes per second is only 2400 Hz, which isn't enough for straight sampling of human voice, so there must be codecs in the chain somewhere. Perhaps they are a source of some of the problems. Simple sampling at low bandwidth can produce white noise but not intermittent frequency attenuation, and in any case just sampling at 2400 Hz would make people very hard to understand.

 

I note that sometimes there's a big difference between checking the "simulated VHF" box in SB or leaving it empty. The simulation does sound realistic, but it also seems to introduce a huge amount of unrealistic noise into the audio at times. Turning it off doesn't sound like radio at all, but sometimes it reveals distortions that don't hurt the audio much but might mess up the VHF simluation.

[Brad Littlejohn]

I know there are occasions when controllers tell a pilot that they cannot hear anything at all of what he said, whereas I hear the pilot perfectly. Conversely, sometimes controllers are manifestly hearing a pilot with no problem, while all I hear is a blast of white noise. That has to be a network issue.

 

This might be, but also might be something happening between you and the voice server. Is there a pattern to this, or does it happen at random?

 

On very rare occasions (I can only remember one specific instance) a controller will tell me that I'm completely incomprehensible. If I disconnect and reconnect to a different server, it goes away. I can't hear any change at my end, but the controller tells me that I've gone from totally incomprehensible to loud and clear.

 

So there are many unresolved mysteries here. A bandwidth of 300 bytes per second is only 2400 Hz, which isn't enough for straight sampling of human voice, so there must be codecs in the chain somewhere. Perhaps they are a source of some of the problems. Simple sampling at low bandwidth can produce white noise but not intermittent frequency attenuation, and in any case just sampling at 2400 Hz would make people very hard to understand.

 

Answers my question above. It could definitely have to do with connectivity between your machine/network and the server's end. There's not much that VATSIM can do about routes between the network, as if they are saturated, all we could do is notify those who own that link in the route, and tell them they have a problem. But I still don't see it as a network issue on our server's end, let alone the controller's side.

 

As for the codecs used for voice, I think they would be less of a problem, as this is the first that I've heard of such an issue with the quality of voice, and we've been using the same protocol and codecs for it for the past 8 years. But there is this that might come into play.

 

Could there be the chance that someone was listening to the frequency/channel you were on, using Roger Wilco, and an improperly configured Roger Wilco at that? That would definitely cause some of the heartburn you have been witnessing.

 

BL.

[Anthony Atkielski 985811]

I know there are occasions when controllers tell a pilot that they cannot hear anything at all of what he said, whereas I hear the pilot perfectly. Conversely, sometimes controllers are manifestly hearing a pilot with no problem, while all I hear is a blast of white noise. That has to be a network issue.

 

This might be, but also might be something happening between you and the voice server. Is there a pattern to this, or does it happen at random?

 

It is frequent but seemingly random. However, specific controllers/pilots often have the same audio issues each time I hear them.

 

On very rare occasions (I can only remember one specific instance) a controller will tell me that I'm completely incomprehensible. If I disconnect and reconnect to a different server, it goes away. I can't hear any change at my end, but the controller tells me that I've gone from totally incomprehensible to loud and clear.

 

So there are many unresolved mysteries here. A bandwidth of 300 bytes per second is only 2400 Hz, which isn't enough for straight sampling of human voice, so there must be codecs in the chain somewhere. Perhaps they are a source of some of the problems. Simple sampling at low bandwidth can produce white noise but not intermittent frequency attenuation, and in any case just sampling at 2400 Hz would make people very hard to understand.

 

Answers my question above. It could definitely have to do with connectivity between your machine/network and the server's end. There's not much that VATSIM can do about routes between the network, as if they are saturated, all we could do is notify those who own that link in the route, and tell them they have a problem. But I still don't see it as a network issue on our server's end, let alone the controller's side.

 

As for the codecs used for voice, I think they would be less of a problem, as this is the first that I've heard of such an issue with the quality of voice, and we've been using the same protocol and codecs for it for the past 8 years. But there is this that might come into play.

 

Could there be the chance that someone was listening to the frequency/channel you were on, using Roger Wilco, and an improperly configured Roger Wilco at that? That would definitely cause some of the heartburn you have been witnessing.

 

BL.

 

I don't know anything about Roger Wilco (which was before my time on VATSIM), so I can't really say. Did Roger Wilco have potential interference issues if it wasn't configured correctly? Wasn't it just another client, or did it work in a different way?

[Brad Littlejohn]

I don't know anything about Roger Wilco (which was before my time on VATSIM), so I can't really say. Did Roger Wilco have potential interference issues if it wasn't configured correctly? Wasn't it just another client, or did it work in a different way?

 

Absolutely. Roger Wilco had/has the potential for the following, and more times than not creates the following issues:

 

1. Poor reception of voice. Roger Wilco must be set to use 'slow modem', regardless of what type of internet connection you have. If it is not set like that, half conversations are received and sent between controllers and pilots. The only way to alleviate that is to disconnect from the voice channel, reconnect, and hope that it goes away. Users who use it cause tons of havoc over the voice channels. Unfortunately, it is needed for use with Squawkbox 2.3 or older, all of which are still available (AVC was the replacement for this, but not many people use it).

 

2. General havoc created on the channels. RW doesn't require authentication to VATSIM, so basically anybody can connect to any channel with it. All they need is to make up a callsign, and they're on. Step on people, create noise, ramble off expletives, the whole lot; you name it, they can use RW to cause the trouble.

 

3. spyware. Newer versions of RW (greater than Mk.II) contains spyware, which was the main reason AVC was created, so it could alleviate a lot of RW's problems. But it had to be voluntarily downloaded, and not an absolute requirement. If people still wanted to use RW, they could, but still had to deal with the problems above.

 

All of these contribute to network and poor sound quality. So there are a lot of things to consider before totally pinning it on the network.

 

BL.