Does your phone have the HTC-Soundbug?

[Guest Drewtech]

I think this was why I had a problem hearing it when I tried the test file for the first time. However, the test file is the only place that I hear any tsss on a regular basis. I spent about 30 hours over the last week listening to a vieriety of music (all wma files) with no real problem. I heard extraneous noise (which is not the regular tss sound) a few times but only on about 3 songs and always in the same place on those songs. I haven't had a chance yet to delete them and recopy them to see if it is in the copy process.

I use WMP10 to synch my files to the C500. Is that how the rest of you are doing it and, if not, does this make any difference.

[Guest Pondrew]

I'm not a sound engineer so can someone help me understand why a .wma file encoded at 96kbps wouldn't have the sound while a .mp3 at 340kbps would? What does that mean for the sound bug? Is it a hardware bug and if so why does different encoding produce different results?

[Guest pharma79]

Hi bazzo , can you check it if WMA@96 does any good ? If it would (that i doubt ) i would convert all my songs to *.billgates :lol:

[Guest bazzo]

Hi, i can tell you why WMA@96kbps didn't develope the Tss-noise in this case. when encoding audio to a lossy format, they do a cutoff at a specific high frequnency. even when encoding to mp3 with lame --alt-preset standard (the ultimate way to encoder mp3 IMHO) there is a cutoff at 19kHz. wenn encoding to lower bitrates, the cutoff-frequency will be decreased.

I dont know at which frequ. the WMencoder will cutoff when encoding@96kbps, but i assume that this frequency is below the test-tones frequ. so you produce a WMA-file with nothing but silence.

due to the tsss-sounds dependence to the level of the inputsignal, the Tsss will not be produced. simply try again with a higher bitrate.

But when encoding real music to WMA, the tsss-sound will be back.

to soften the tsss-noise in real music, encoding at lower bitrates or a manual cutoff will do the job, but you will lose the higher frequencies, which will make the music sound dump and boring. even if you can't specifically hear them.

regards bazzo

[edit]

WMA@96 cut-off at 15kHz

WMA@128 cut-off at 17,5kHz

[/edit]

Edited February 27, 2005 by bazzo

[Guest bazzo]

Advice: if you have the Bug with the testfile, but you think that it doesn't appeare with real-music, rip the favourite song from your girlfrinds/wifes cd, encode it (or not) and let her hear it via your smartphone.

I've made the experience, that women have a much more sensitivity to audiophile issues, even if the most of them are not interested in overkill-HiFi.

e.g.

my girlfriend mentioned the tss-noise after listening to Eric Prydz "call on me" after a few seconds in the first audition. I dont like this song at all, no idea if its known in UK, but its really hard to hear the noise in this song, i've listenened to it for a couple of times an have not yet recognised the Tss-sound in this track. you should know that my girlfrind isn't interested in HiFi at all, she is even lucky when listening to Robbie W. on a clockradio.

regards bazzo

Edited February 27, 2005 by bazzo

[Guest moosery]

Advice: if you have the Bug with the testfile, but you think that it doesn't appeare with real-music, rip the favourite song from your girlfrinds/wifes cd, encode it (or not) and let her hear it via your smartphone.

I've made the experience, that women have a much more sensitivity to audiophile issues, even if the most of them are not interested in overkill-HiFi.

e.g.

my girlfriend mentioned the tss-noise after listening to Eric Prydz "call on me" after a few seconds in the first audition. I dont like this song at all, no idea if its known in UK, but its really hard to hear the noise in this song, i've listenened to it for a couple of times an have not yet mentioned the Tss-sound in this track. you should know that my girlfrind isn't interested in HiFi at all, she is even lucky when listening to Robbie W. on a clockradio.

regards bazzo

<{POST_SNAPBACK}>

oh yes, it's known!!!

[Guest pharma79]

Hi, i can tell you why WMA@96kbps didn't develope the Tss-noise in this case. when encoding audio to a lossy format, they do a cutoff at a specific high frequnency. even when encoding to mp3 with lame --alt-preset standard (the ultimate way to encoder mp3 IMHO) there is a cutoff at 19kHz. wenn encoding to lower bitrates, the cutoff-frequency will be decreased.

I dont know at which frequ. the WMencoder will cutoff when encoding@96kbps, but i assume that this frequency is below the test-tones frequ. so you produce a WMA-file with nothing but silence.

due to the tsss-sounds dependence to the level of the inputsignal, the Tsss will not be produced. simply try again with a higher bitrate.

But when encoding real music to WMA, the tsss-sound will be back.

to soften the tsss-noise in real music, encoding at lower bitrates or a manual cutoff will do the job, but you will lose the higher frequencies, which will make the music sound dump and boring. even if you can't specifically hear them.

regards bazzo

[edit]

WMA@96 cut-off at 15kHz

WMA@128 cut-off at 17,5kHz

[/edit]

<{POST_SNAPBACK}>

Bazoo are you sure about cut-off values ?

Because I`m sure a file encoded at WMA@96 is way better in terms quality than a MP3 @160 kbps or maybe even 196kbps , so if you confirm that this removes the problem on ALL files, I think this bitrate covers the hardware audio quality limits of a TYPHOON very good and we dont need better bitrate .

Please if you can confirm that this does the job for any music file , then i have to go shopping for a 512 mb MINI-SD ASAP :lol:

[Guest bazzo]

hi,

i saved a white noise in Adobe-Audition as wma in different bitrates. Then i had a look at the spectrum of the files and mesured the frequencies where the cutoff is visible. so these values should be correct. But i dont know if there is a generally cutoff in wma and mp3. This could be different on different encoders.

for my mp3s i use lame with --alt-preset standard. in this profile the cutoff at 19kHz is set. Perhaps you can change the cutoff if not using APS. No idea if there is a WMA-encoder that is as configurable as lame.

As test 3 shows, the level of the TSSS depends on the Inputfreqency, the higher the inputfreq. (with same level), the higher the level of the TSSS. So if you cutoff the high frequ. in the inputfile, there are only the frequencies below the cut-off-freq. which can develope the TSSS, but the level in the TSSS corresponding to these frequencies is logarithmicaly decreasing.

Perhaps that is the reason why some of our users didn't recognise the TSSS in real music, even if they heard it with the testfile. They simply have "bad"-encoded-files. In this case "bad"-can be good. :lol:

The lower the cut-off is established, the less audible the TSSS will be. But it will NOT disappeare.

I know that WMA is better at lower bitrates, but i can't belive that a "good" mp3@160 is more worse than a WMA@96. Even mp3@160 is hard to differ from cd, if lame-encoded. Perhaps if you compare the WMA@96 to the orig. CD, it even sounds better subjectiv. But for me its more important, that a compressed Audio-file sounds more similar to the source, than sounding "better". What you call better sound could perhaps be done with an EQ to the mp3. But I think for WMA, MP3 and OGG issues, www.hydrogenaudio.org is the better (best) forum.

reagards Bazzo

[Guest Drewtech]

This is going to be my loast comment on this topic simply because I am not a sound expert and I have done everything I can to identify a repeatable and consistent problem on my C500.

My final test has just been to rip an audio CD track to MP3 at 320 kbps (the highest that Nero Media Player can manage) and WMA at 128 kbps (also using Nero). Both files were copied to my SD card using Windows Explorer. I have listened to both extremely carefully and there is no tsss sound in either file, either occasionally or on a repeating regular frequency. I have listened to both files via my PC and they sound the same as on my phone. I have listened to the CD track again and the difference in quality between that and the ripped tracks is noticable in one or two places but is negligible.

My conclusion, therefore, is that for the music I listen to on my handset and the quality of file I normally use (wma 128 kbps) I do not have a problem. I am going to stick with that and stop looking for the problem (in case I find it).

For anyone interested the track I did all the final comparison on was Jan Garberek: Rites (Track 1 from CD1 of the album Rites).

[Guest bazzo]

hi,

this is absolutly ok! this thread was never meant to make people looking for a bug they didn't notice. If you dont hear it, and it dont disturbes your music pleasure, then i really envy you. Hearing the sound in my testfile, and hearing it in real music is a huge difference. And it mainly depends on the kind of music listened to. If you're content with your phone, stay it, and definitly stop looking for the bug!

I did my tests cause of i noticed it in real music and it really drives my crazy. perhaps it was a huge mistake to expect my "SDAmusic" to be an proper replacement to an mp3-player. even if the sound between the noises is excelent!!

the frequncy-response and the pulse-response are HiFi, and to my ears it sounds like HiFi, thats what makes me so sad about this issue. :lol:

regards Bazzo

Edited February 27, 2005 by bazzo

[Guest alexis_winters17]

Erm, not entirely off topic but it seems a few of us have been having a different sound issue...the phone's gone very quiet...and a hard reset hasn't helped!

[Guest pharma79]

Bad news ,

Its still there there as bad as always :lol: BAZOO are you sure you didnt get any tiss with wma96 ? (i doubt it can solve problem at all when you get the bug on any file played on the phone , even midi files ) but the TSS sound is subjective to type of music played .

Im hoping less & less for any cure. Maybe just let it go and wait for another time & phone ?

Edited March 1, 2005 by pharma79

[Guest bazzo]

Bad news ,

Its still there there :D BAZOO are you sure you didnt get any tiss with wma96 ? (i doubt it can solve problem at all when you get the bug on any file played on the phone , even midi files ) but the TSS sound is subjective to type of music played .

Im hoping less & less for any cure. Maybe just let it go and whait for another time & phone ?

<{POST_SNAPBACK}>

i didn't get ist with the 17640Hz-Testtone @wma96 becouse of the 15kHz-cuttoff. In real music even wma@96 developes the tsss.

regards bazzooooo

[Guest slickdesign]

Yeah the Tss sound happens in any audio format. I think its the soundcard driver for the phone that's screwed up.

Edited February 28, 2005 by slickdesign

[Guest celica1.8st]

Are we any further in deciding if it is hardware or software?

If it is hardware - then I guess it could be a capicitor or someting which charges then discharges?? Maybe some kind of noise reduction filter gone wrong?

If it is software, then would it be a buffer or something?

Does the fact that you don't hear the sound when on a phonecall have any significance? Does the phone use the soundcard driver, or is it seperate software/ or even different hardware wired up to the same output? Is the phonecall reduced quality? If it is reduced quality maybe it's just that the worst are cut off with the freq cut off, and the rest are not noticeable.

My current fix for this is to encode everything at WMA@64 it sounds acceptable to me through the headphones, and I can only notice a tsst in one place on my phone. (can't notice it at all still on my wifes phone). This must cut off all the worst tssts, and the others mustn't be too noticeable.

I am still going to repeat bazzos tests, just got to find a 2.5mm to 3.5mm adapter and make sure it works ok for headphones etc

[Guest bazzo]

"If it is hardware - then I guess it could be a capicitor or someting which charges then discharges?? Maybe some kind of noise reduction filter gone wrong?"

I dont know much about e-technics, but i thing with small SMD-caps you need a huge resitor to get these long periods of 16,68s, i doubt that an RC-part with a tiny cap and a huge resitor is abled to reproduce these 16,68s that accurate, but i may be wrong. although, nice idea...

I thought bout buffering-issues too, i searched some the registry for sounddriver related keys, i only found some interesting settings for the GSM-sounddriver, (bitrate, low-pass, etc.) maybe thats the indicator, that there is a different soundmodule for GSM. I didn't found buffersetting keys so far, but i search for them again today.

by the way:

I found a way to extract the tsss-noise from real music, simply downmix an audiofile to mono. Upload it to your phone. Than play it an record the output back to your PC. The most Audio-Lab-Softwares, are abled to filter all signal which is similar on both channels. All that should stay in the waveform is a slight hiss, and the TSSS at original level.

regards Bazzo

[Guest bazzo]

hi mates,

i got the following eMail today, from T-Mobile Service. I sent theme a link to my german analysis.

E-Mail von T-Mobile

Sehr geehrter Herr  xxxxxxx,

vielen Dank für Ihre E-Mail an T-Mobile.

Die von ihnen angesprochene Beeinträchtigung ist bekannt, und dem Hersteller mitgeteilt worden. Auch andere Auffälligkeiten, die bei uns auftauchen werden an den hersteller gegeben. Leider wird ein Update etwas dauern, aber wenn es verfügbar ist, finden sie es auf unserere Homepage unter www.t-mobile.de.

Bitte haben sie Verständnis dafür, das wir nicht mit Daten aus privat erstellten Foren arbeiten können, da diese Angaben für uns sehr zeitaufwendig wären und teilweise auch nicht immer auf einen Endgerätefehler deuten, sondern auf eine Fehlbedienung. Eine Trennung ist da schwierig, da man keinen direkten Zugriff auf den Ersteller und das jeweilige Gerät hat. Anmerkungen unserer Kunden geben wir selbstverständlich umgehend weiter.

Mit freundlichen Grüßen

xxxxxxx

T-Mobile Kundenservice

Technisches ServiceCenter

TRANSLATION:

Dear Mr. xxxxxxx

Many thanks for your eMail to T-mobile.

The bug you mentioned is known and forwarded to the manufacturer. Although other noticeable problems were forwarded to the manufacturer. Im sorry that the update will take a while, when its available, you can download it at our homepage (www.t-mobile.de). Please have appreciation, that we can't work with data from private forums.....

Best regards

X.XXXXXX

T-Mobile customer service

Technical ServiceCenter

regards bazzo

[Guest ocker]

Yeah, my HTC Feeler (XPhoneII) does it.

I recall that my MP3 CD player (iRiver SlimX iMP-350) had a similar problem where it would *pause* for a fraction of a second every 8 minutes on really long MP3s.

Turns out it was happening because the buffer could only hold 8 minutes of music. So every 8 minutes, there's this section in the music that hadn't been stitched up properly.

Good news is that iRiver were able to resolve their problem, so there's hopefully no reason why this one cantt be resolved, either.

[Guest mkent_barbados]

Well, I have a US ver. SP3i with Rockies mainboard and WMP10, and I thought I didn't have the sound bug, but I hear it on your test--so thanks alot! Not regular however, ran it 4 times. Times heard as follows: 17 & 25 secs, 12 & 29 secs, 17 & 33 secs, and 15 & 32 secs (using the time elapsed while playing). Not on auto-repeat. Well, can't be glad about this because I have never heard it during use and didn't think I had it, but now its bugging me!!

[Guest moosery]

Well, I have a US ver. SP3i with Rockies mainboard and WMP10, and I thought I didn't have the sound bug, but I hear it on your test--so thanks alot!  Not regular however, ran it 4 times.  Times heard as follows: 17 & 25 secs, 12 & 29 secs, 17 & 33 secs, and 15 & 32 secs (using the time elapsed while playing).  Not on auto-repeat.  Well, can't be glad about this because I have never heard it during use and didn't think I had it, but now its bugging me!!

<{POST_SNAPBACK}>

nothing's perfect..just ignore it ;o)

mine's just the same, and if I forget about it then it doesn't bug me anymore!

[Guest spark9990]

Well its bugging me. i hope Orange or HTC pull their fingers out and release some sort of fix.