85 replies to this topic

[Sybreed]

Niko seems to think the two problems are the same... I have my doubts personnally, which I told him in a PM I sent yesterday (linking zaitsev's rapidshare link). But, who knows, maybe it's the same issue?

[Acid Phreak]

i think this is the same issue but are more anoying for 7.1 setups.

and for the lolz MWO has no 7.1 feature i thought.

[Sybreed]

GM Reppu, on 22 December 2012 - 11:22 AM, said:

See this is what Open BETA is for.
Anecdotal:
My computer at work does not experience any sound corruption whatsoever.
My computer at home does experience sound corruption like you describe. (Betty sometimes trips on her own tongue, and Mech wrecks seem to buzz) ***note I use a 7.1 headset. I could drop it to stereo for testing but I'm lazy

It is literally impossible to test every possible configuration and scenario in a controlled testing environment, and we greatly appreciate the help our Open BETA testers are providing.

Eh, we're doing our best to find the cause of the issue ourselves, although it's hard to find what particular settings each of us are using, especially since soundcards differ from one computer to the next and info about audio systems aren't as easy to detail as, say, graphic cards. Acid Phreak has an interesting theory about a possible cause. I got jack in my case, but I'm no computer genius.

I hope we'll find it soon, I'm waiting for this fix to activate my premium timer...

[Zaitsev1984]

I have my doubts too, I don't think it's the same, the 7.1 thread says that the corruption is only present in some channels and it's present all times, here we have stuttering in every channel and mostly under heavy load.

But who knows, may be it's the same, or may be while trying to solve the 7.1 issue they'll solve our problem too.

About the 11923 tag, maybe it's meant to describe sound related threads, let's hope it's that and not a tag about hypochondria.

Let me know about any other information or test I could perform that could be useful. I've tried some other things with no success.

Edited by Zaitsev1984, 22 December 2012 - 08:52 PM.

[Stonefalcon]

Was it 7.1 causing that sh.t? OMFG my ears bled trying to listen for mechs while that crap was going on.

[Acid Phreak]

i found more proof,

to be more clarify for an possible fix,

the soundbanks from betty as example are original compressed in ADPCM, my rebuilded betty soundbank are compressed in PCM.

so i just wanted to know what if happens if i compressed my rebuilded soundbank in ADPCM, Surprise the soundbank samples got corrupted and ingame i got stutter from betty.

so now I know what is wrong with all the corrupted sounds ingame. if u(PGI) have the source files and working with fmod this is a half hour(max) work to fix that. (just switch the compressing in the soundbanks)

maybe this fixed also the 7.1 issue but i have no 7.1 headset. maybe i can have permission to share the "fixed" soundbank for getting proof? or just you (PGI) should look into the compressed soundbank on your own


http://en.wikipedia....code_modulation
http://en.wikipedia....code_modulation

now i got all the relevant data, im sure thats is it.

now its in PGI´s hand to fix that issue

have a nice Christmas

greetz
AP

[Zaitsev1984]

OMG, I'm not sure if devs are really reading this thread, I haven't had enough time to study the details about Acid Phreak's research but sounds very promising, ¿is there any other way we could get in touch with dev team?

[Zaitsev1984]

I got a little improvement by:

- Removing a software called "Asio4all" that came with FL Studio,

- Launching the game from "Razer Game Booster", it's free software. I didn't expect this one to work because none, CPU, GPU or Memory seems to be too much stressed during gaming, but it helps a little.

None of this is a final solution, the stuttering is still present, but in my particular case makes it a little more tolerable.

Edited by Zaitsev1984, 25 December 2012 - 08:57 PM.

[Acid Phreak]

hi,

after a little reading and head scratches, i think the buffer size is to low for the ADPCM compression that why this ingame begins to stutter (not this corrupted sample thing). because the ADPCM is a cpu ressource eating compression.

Zaitsev1984, on 25 December 2012 - 08:46 PM, said:

I got a little improvement by:

- Removing a software called "Asio4all" that came with FL Studio,

- Launching the game from "Razer Game Booster", it's free software. I didn't expect this one to work because none, CPU, GPU or Memory seems to be too much stressed during gaming, but it helps a little.

None of this is a final solution, the stuttering is still present, but in my particular case makes it a little more tolerable.

this is a indication for this buffer size problem!


from the fmod header (fmod.h)

Setting the 'decodebuffersize' is for cpu intensive codecs that may be causing stuttering, not file intensive codecs (ie those from CD or netstreams) which are normally altered with System::setStreamBufferSize. As an example of cpu intensive codecs, an mp3 file will take more cpu to decode than a PCM wav file. If you have a stuttering effect, then it is using more cpu than the decode buffer playback rate can keep up with. Increasing the decode buffersize will most likely solve this problem.

http://gaodayihao.gi..._8h_source.html

Edited by Acid Phreak, 26 December 2012 - 04:19 AM.

[Zaitsev1984]

It looks very promising, we don't have any means to tweak this value, do we?

[Acid Phreak]

maybe, not sure

look here http://freesdk.cryde...TREAMBUFFERSIZE

not tested yet, should be paste in the user.cfg. as example (PC default value)

s_StreamBufferSize = 131072

begin with the default value!
use 196608 or higher for testing!
look in the omicron log for errors after a game!
note that is only for the stutter!
I'm really not sure if this is the correct variable

i am just a experienced modder

Edited by Acid Phreak, 26 December 2012 - 09:03 AM.

[Zaitsev1984]

I'll try it this night and post results, let's hope it works, I'm crossing my fingers!!

Edited by Zaitsev1984, 26 December 2012 - 10:26 AM.

[Zaitsev1984]

I've been a little busy but I tried the Acid Phreak's workaround for 2 matches ¡with no stuttering at all!, I invite other players who're having the same issues to give this solution a try, it only takes 1 minutes to do and it's easily reversible.

Acid Phreak if you want to take a look on my omnicron log file let me know and I'll upload it.

[Ramses2020]

Acid Phreak, on 26 December 2012 - 09:01 AM, said:

maybe, not sure

look here http://freesdk.cryde...TREAMBUFFERSIZE

not tested yet, should be paste in the user.cfg. as example (PC default value)

s_StreamBufferSize = 131072

begin with the default value!
use 196608 or higher for testing!
look in the omicron log for errors after a game!
note that is only for the stutter!
I'm really not sure if this is the correct variable

i am just a experienced modder

I just tried using a value of 262144 (256KB), which should be more than enough buffering. I got no difference. Nice try though :\

[Sybreed]

Ramses2020, on 26 December 2012 - 10:04 PM, said:

I just tried using a value of 262144 (256KB), which should be more than enough buffering. I got no difference. Nice try though :\

same results

[Acid Phreak]

sry if that not works,

is possible that this variable not working on your rig or overrode from the code or its false!

i have just response to a request! and did not test it myself!

interesting that seems working for Zaitsev1984 @

yes i have interest but just to see if there are specific fmod message with this variable during a battle. edit out your ip - pc name and other stuff u dont want to share before post it^^

and note to all guys and girls who test it, keep in mind that this can cause side effects! if u recieve other bugs or somethings like this delete the variable and look for your bug/error Before u call the support!

Edited by Acid Phreak, 27 December 2012 - 06:02 PM.

[Sybreed]

just fyi, I installed MWO on my laptop and I have the same issue there aswell..

[Ramses2020]

After doing some digging, I found three variables that have significant impact to audio capability, and have doubled the default values. The amount of active sounds are noticeable, and Betty's stuttering has been significantly reduced. I put this is my user.cfg file.

s_StreamBufferSize = 262144
s_ADPCMDecoders = 64
s_MaxChannels=128

The s_StreamBufferSize may or may not do anything in the context of MWO (it's for fmod streams), but it was brought up before so I decided to keep it. The s_ADPCMDecoders variable definitely has an effect on Betty's fluidity. The s_MaxChannels variable helps to keep more sounds playing.

Making this change has improved immersion by a noticeable magnitude. It's still not perfect, though. Based on the fact that I just doubled the defaults makes me think there is some serious inefficiency going on in the client. A single laser might take its own channel, as does any footstep sound, launcher sound, etc and in a game like this, that can add up fast.

Try it out, it might work for you too.

[Dirty Old Man]

Noticed that the crackling happens most of the time when I am panning from one direction to another while there are mechs around me... if its just action coming from the front of me, then the sounds are alright.

[sycocys]

Acid Phreak, on 23 December 2012 - 10:04 AM, said:

i found more proof,

to be more clarify for an possible fix,

the soundbanks from betty as example are original compressed in ADPCM, my rebuilded betty soundbank are compressed in PCM.

so i just wanted to know what if happens if i compressed my rebuilded soundbank in ADPCM, Surprise the soundbank samples got corrupted and ingame i got stutter from betty.

so now I know what is wrong with all the corrupted sounds ingame. if u(PGI) have the source files and working with fmod this is a half hour(max) work to fix that. (just switch the compressing in the soundbanks)

maybe this fixed also the 7.1 issue but i have no 7.1 headset. maybe i can have permission to share the "fixed" soundbank for getting proof? or just you (PGI) should look into the compressed soundbank on your own


http://en.wikipedia....code_modulation
http://en.wikipedia....code_modulation

now i got all the relevant data, im sure thats is it.

now its in PGI´s hand to fix that issue

have a nice Christmas

greetz
AP

The compression isn't the major issue at play - its the oversaturation of frequency ranges that is causing most of the issues. There are several ways to deal with this, by far the easiest is running some narrow band equalization filters on the audio files so they don't bunch up.

Alternatively they could re-write the sound engine so that these filters only happened based on a priority filter chain, but something like that would probably add a minimum of a year of a dedicated programmers time and juice up the games spec requirements.