Sound crackling/distortion with certain headphones?
- Thread starter trza2k
- Start date
Hi there,
Also very curious, if anyone have had luck with Westone 4r pair with i9505?
Many thanks,
Also very curious, if anyone have had luck with Westone 4r pair with i9505?
Many thanks,
Ok Maybe its better to start a new thread. Jensigner is probably the best user as his knowledge seems pretty good. Better to have all the tests he has done on a new thread and it will be more professional when we start directing tech blogs etc to it.
What do you think Jensigner. ?
Sent from my GT-I9505 using xda premium
This 40 page thread sounds more impressive than a 1 page, plus it will seem like we are trying to find workarounds and are willing to buy new earphones for Samsungs screwup - which I'm not prepared to do.
Jensigners info is here and in many other places if someone wants to find it - why split up our cause?
I just tried mine (they had been back to westone for repair and came back yesterday).
I get distortion on tracks with heavy bass. Gutted.
Not sure what folk are doing but will try to read the thread later. Is this across all units or just some? Has samsung got a fix?
Sent from my GT-I9505 using Tapatalk 4 Beta
Bad, bad, bad!! Anyway thank you!
Please note guys, that Westone 4r are rated 31ohm @1khz:
http://www.westone.com/music/index.php/products/personal-listening/westone-4r.html
Has any software update fixed the problem with your JVC headphones or have you found a workaround. .?
The impedance at 1KHz isn't very relevant here, its the impedance in the lower frequencies that are important.
From measured results at Goldenears the impedance less than 1KHz is much lower, although still higher than the lower bound threshold for crackling as we have reckoned it so far at about 18ohm. There may be some variability in production though, since their measured nominal impedance at 1KHz is around 24ohm rather than 31ohm.
I think we should just keep the info in this one thread in this forum. I'm still waiting to get official feedback from Samsung Mobile support. If I don't hear anything by this Friday, I'll call them to try to force a response.
Regarding removed posts from GSMArena in this thread:
Samsung i9500 Galaxy S4 vs Samsung i9505 Galaxy S4
I really can't understand this. I thought my post there as well as DocRambone's were objective and didn't seem to counter any of the posting guidelines. Over the years I have posted in MANY different types of professional technical forums what I consider to be objective technical comments and I can't remember EVER having a post removed without advising me of the reason. At a minimum I'd expect a PM about why my post was removed. So I think we can only conclude from this behaviour that GSMArena, for whatever reason, has some vested interested in not reporting negative performance aspects of Samsung's flagship phones. You be the judge.
Regarding removed posts from GSMArena in this thread:
Samsung i9500 Galaxy S4 vs Samsung i9505 Galaxy S4
I really can't understand this. I thought my post there as well as DocRambone's were objective and didn't seem to counter any of the posting guidelines. Over the years I have posted in MANY different types of professional technical forums what I consider to be objective technical comments and I can't remember EVER having a post removed without advising me of the reason. At a minimum I'd expect a PM about why my post was removed. So I think we can only conclude from this behaviour that GSMArena, for whatever reason, has some vested interested in not reporting negative performance aspects of Samsung's flagship phones. You be the judge.
lol, and they ALSO removed your single-line comment about GSMArena being sensitive to that topic.
Here's the graph: http://rinchoi.blogspot.com/2012/09/brainwavz-b2-updated.html
Peaks at around 90 ohms in the highs. Difference between minimum and maximum is only 50 so it shouldn't be a problem.
Take a look at the impedance sweep of the XBA-3:
Ranges from below 10 to 90 - that there would be a problem for weaker amps.
Last edited:
The HTC has plenty of other issues dude. I had one but needed more storage and removable battery as work outside away from a plug socket.
The sound is good if you have the right headphones. This is obviously totally unacceptable but that's the way it is.
Personally if I had the choice again I would wait till HTC release the next one with a better battery and 64gb (not here in uk) they may even improve the camera to an acceptable megapixel rating...
Off topic sorry. Anyway I am really. Pleased with my akg451s they are amazingshame my £150 headphones are now packed away
Sent from my GT-I9505 using xda premium
If rumours (via GSMArena) be true that the SG Note 3 will have the Qualcomm Snapdragon 800 (and presumably that means Qualcomm audio), here's hoping that Samsung understands the problems being discussed here and gets the headphone audio output right. Maybe they'll send me a pre-production sampler for independent verification of superb audio down to say 5 ohm headphone load :fingers-crossed:
Some people mentioned that only low impedance headphones has issues.
Well...
I did tested the following:
Shure SE215, SE315 & SE515.... all of them, on a hungry track they clip above 11 volume, sometimes even after 9.
They seem to have an 27Ohm impedance... but they still clip the sound, when on i9300 they work perfectly with the same songs, same headsets.
Well...
I did tested the following:
Shure SE215, SE315 & SE515.... all of them, on a hungry track they clip above 11 volume, sometimes even after 9.
They seem to have an 27Ohm impedance... but they still clip the sound, when on i9300 they work perfectly with the same songs, same headsets.
My thoughts
I currently have Galaxy S4 and have encountered this "bug". GS4 having such a powerful chipset should be able to handle any task without breaking a sweat -- so I'm leaning on it being a software issue.
I encountered this after I applied my bass-heavy eq while playing JT's Mirrors and I was on my Urbanears Plattan. I noticed the cracks when the screen turned off. Then when I woke it up, it was gone. Just to add, I wasn't running anything load-heavy in the background, I think. I'm on EDGE connection at that time, auto-sync, power-saving mode, air-view, air-gesture, multi-window and blocking mode turned on. I don't think this should put too much stress on the device. What I did is close all background stuff and clear the memory, and reboot (maybe 2-3 times) then it was gone. Tried playing the same file and never encountered the cracking again even I have tons of apps running in background.
Saying this, I think Samsung needs to optimize their software. Replacing headphones shouldn't be a proper solution. After writing the above, I just realized that this could never be a hardware issue as I haven't found any news about the HTC One with the same chipset, just slightly underclocked, having the same issue.
I currently have Galaxy S4 and have encountered this "bug". GS4 having such a powerful chipset should be able to handle any task without breaking a sweat -- so I'm leaning on it being a software issue.
I encountered this after I applied my bass-heavy eq while playing JT's Mirrors and I was on my Urbanears Plattan. I noticed the cracks when the screen turned off. Then when I woke it up, it was gone. Just to add, I wasn't running anything load-heavy in the background, I think. I'm on EDGE connection at that time, auto-sync, power-saving mode, air-view, air-gesture, multi-window and blocking mode turned on. I don't think this should put too much stress on the device. What I did is close all background stuff and clear the memory, and reboot (maybe 2-3 times) then it was gone. Tried playing the same file and never encountered the cracking again even I have tons of apps running in background.
Saying this, I think Samsung needs to optimize their software. Replacing headphones shouldn't be a proper solution. After writing the above, I just realized that this could never be a hardware issue as I haven't found any news about the HTC One with the same chipset, just slightly underclocked, having the same issue.
If this was a hardware fault wouldn't the HTC ONE and Xperia Z also have the same problem?
Sent from my GT-I9505 using XDA Premium HD app
Sent from my GT-I9505 using XDA Premium HD app
Please people, if you are going to post information please try and grasp the issue properly first.
I just looked up measurements for the Shure SE215 and in the bass frequencies its impedance is most certainly not 27ohm, looks about 17ohm by Inner Fidelity AND Goldenear measurements.
The SE315 I could only find measurements on via Goldenears and that gets a bit more interesting, since its impedance less than 1KHz is about 21ohm and we would not have expected that to crackle necessarily based on our other information to hand.
Couldn't find SE515 measurements at all.
Regarding the Urbanears Plattan mentioned above - I'm wondering if you have another issue there. Their impedance is more like 40ohm throughout most of the entire frequency spectrum, and the crackling the rest of us are mentioning here happens regardless of whether the screen is on or not. If you have PowerAMP try setting audio thread priority to highest and buffer to massive and see if that fixes it with screen off.
Last edited:
Similar threads
Top Liked Posts
-
There are no posts matching your filters.
-
6FYI
Hi People...
I just wanted to let you know about my research... I'm a GS3 owner (Qualcomm DAC)... this problem is largely related to the Qualcomm DACs more then the phone...
The reports that the GS3 (Qualcomm DAC) are unaffected are true but only because the GS3 ships with the 3.0 kernels not 3.4... when I rooted my GS3 and installed a ROM with the newer 3.4 kernel I have major crackling/distortion using my sony XBA-4 at 60% volume or more. Going back to the 3.0 kernel... 100% crystal clear sound.
There were major changes to the codecs and drivers in the kernel... I'm tying to make a 3.4 kernel but backport the 3.0 drivers for the GS3 so I can run android 4.2+
Someone with a GS4(Qualcomm DAC) and some kernel knowledge should try and do the same... I have proof that it's the kernel that made this problem much worse then the 3.0 kernels.5
I'm working on it. I've contacted a guy on eBay who says that he's received all of the parts, and he's making and testing right now. If they work, I'll let all of you know.5S4 Crackle/Instability is Fixed!
I have completed my tests on the new update for the SGH-I337 version:
I can confirm that the crackle instability is completely absent now. Tests were done at 250 Hz and 1kHz with 4.7, 10 and 33 ohm headphone loads at all volume levels and digital levels from zero to 15/15 volume using default Samsung Music player with no EQ.
The maximum output level for full digital volume is identical to that before the update: 0.88 Vpeak ( 0.62Vrms). The frequency response is also identical to before-update so absolutely nothing was tweaked in the frequency response. The background noise level (during playback) is ~ 3uVrms in 20 kHz BW so is really dead quiet, even for very high-sensitivity phones and identical to before-update value providing an amazing S/N of ~ 106 dB. Distortion characteristics are also unchanged. In a nutshell, all headphone audio characteristics I've measured with the new update are identical to those before the update as discussed here except that the buzz/crackle instability is gone! Note that, just as before (and as measured and discussed in the page above) the distortion characteristics WILL increase as the headphone impedance is lowered but this is normal. How noticeable is this distortion? Listening with my Senn HD598 (50ohm) with PARALLEL resistors of 10 ohm (or 4.7ohm) to deliberately increase harmonic and intermodulation distortions produced no detectable sound difference to my ears playing jazz and pop! In this case, simulating very low impedance iems, the distortions would be above 0.1%. No crackle, no buzz, no noticeable deteroriation in audio that I could perceive.
How much power will the S4 provide to drive headphones?
100 ohm phones: 4mW max
33 ohm phones: 11 mW max
10 ohm phones: 32 mW max
The most severe test was driving a 100% digital amplitude sine wave at 15/15 vol into 4.7 ohm loads. This just starts to show fizzie artifacts at the very bottom of the sine-wave and is definitely showing the current limit at a pretty amazing ~ 150mA peak.
So all thumbs and toes up to Samsung for fixing this problem! Would be great to know what was actually done to fix the crackle but I suspect we'll never know.3
I'm no expert on this, but here is my take:
- my SGH-I337 (similar to i9505) has an internal "output impedance" of ~ 2.5 ohm (measured)
- high-quality headphones, particularly balanced armatures ones, from what I've seen have quite a variable impedance as a function of frequency which shapes the sound by the headphone designer
- the best scenario is that the output impedance driving these headphones is very LOW so that the internal audio sound of the S4 is exactly what the headphones are getting. This will happen if the output impedance is very low (typically < 1/10) compared to the minimum headphone impedance
- if you ADD series resistance using a resistive adapter, you modify the frequency-dependent output signal (an unintended EQ effect) across these types of headphones and therefore color the sound. This is not so much a problem with dynamic headphones where the impedance vs frequency tends to be much flatter so added series resistance just drops the level at all frequencies uniformly without significant tonal coloration
This is why it is always better to add an external headphone amp ... they "buffer" (don't load down) the phone's headphone output, and they are designed to typically have an output impedance of < 1 ohm (and the lower the better). Designing this type of amp, with very low output impedance capable of driving headphone (say ~ 8 ohm) with very low distortion is difficult and requires very good audio design experience.
Since the S4 output inpedance is ~ 2.5 ohm, that is a VERY good match for stock earbuds at ~ 35 ohm or an extremely good match for my Senn HD598 (at nominal 50 ohm). For headphone impedances that vary with frequency and with Z(nominal) < 20 ohm, there will be tone shaping, and even more with the added series resistance of an impedance adapter.
If you get a resistive adapter, the amount of tone coloration you get will depend on what your headphone impedance looks like vs frequency. You should therefore pick the MINIMUM possible resistance value that eliminates the buzz. However the subtle "tone coloration" of a resistive adapter will probably be completely acceptable to most listeners compared to that garbage oscillation problem.
---------- Post added at 09:50 AM ---------- Previous post was at 09:34 AM ----------
Just a reminder: here is the link to several S4 articles I've written, include one on the audio problem:
Samsung Galaxy S4
I will be adding an article shortly on the different audio test .wav files I've mentioned here.
Since most of the articles are reasonably complementary to Samsung's S4 features, I can't imagine that Samsung would consider me to be a proverbial sh!t disturber3Coming soon: FiiO E17 vs S4 vs QuadBuffer headphone audio
I just received a FiiO E17 DAC/headphone amp. I'm going to do some RMAA and audio clipping measurements on the FiiO headphone output using the digital SPDIF input (for say 16 ohm loads) and see how it compares with the direct headphone out of the S4 and with the S4 into the E17 analog-input and then into 16 ohm. Oh yes, I'll also include the QuadBuffer amp (which is just a high-performance amp, not a DAC) I designed. Stay tuned!
First Impressions:
I tested the E17 for an output level of ~ 0.9Vpeak, so comparable level to max. output of the S4 phone.
After testing the analog chain using the Line-In (bottom), and the DAC/analog chain using the Optical SPDIF in and USB playback, I find that:
- the best audio performance is obtained using the Line-In
- next best is the SPDIF input (which as expected has the lowest crosstalk)
- surprisingly the USB playback has the poorer audio performance. The THD is still low but the other type of distortion (IMD) is a lot higher.
Line-In testing results were almost identical with open headphone-out load and with 15 ohm loads (the E17 is only supported down to 16 ohm headphones). With open load (actually into 3 kohm), the performance is extremely good and so this functions perfectly well as a standard "line out" .... no need to use the doc line-out adapter really.
More results to come in combination with the S4
Chart below shows RMAA audio performance results (from Left to Right) with:
S4 connected directly to 22 ohm headphone loads
S4 connected to Line-In of FiiO E17 with E17 output loaded with 22 ohm loads
S4 connected to Line-In of QuadBuffer Amp with QuadBuff loaded with 22 ohm loads
Results are for a signal level roughly equal to the maximum possible output level of the S4 (~ 0.8Vpeak).
The big improvement in THD (total harmonic distortion) is evident with a separate headphone amp. The QuadBuffer is marginally better than the E17 in distortion and noise but much better in crosstalk. However the E17 can deliver up to ~ 3.3Vpeak output level and can therefore drive higher headphone impedances well (up to ~ 300 ohm according to specs):
