Skip to main content
Topic: Sound "drop-outs" every ten seconds or so (Read 4127 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

Re: Sound "drop-outs" every ten seconds or so

Reply #25
Juha :
I am not able to use this kind of technical tool. I just got this view of the CPU actitvity (I think! the bottom line says % of processor time). In blue, the time window during wich I had two "drop-outs" at least. The CPU % shows nothing clear. The peak at the beginning is, I think, the start of Foobar.



Case : You are probably right... but I still find the instructions somewhat ambiguous. Since an issue during the flash of the BIOS would have serious consequences, I'd rather find another option (or wait until I get a replacement battery !).

Re: Sound "drop-outs" every ten seconds or so

Reply #26
Juha :
I am not able to use this kind of technical tool. I just got this view of the CPU actitvity (I think! the bottom line says % of processor time). In blue, the time window during wich I had two "drop-outs" at least. The CPU % shows nothing clear.
...

It's quite tricky tool I can say. You see the green + icon there above the graph window ... that allows you to set counters. I'm on Ubuntu again so I can't tell much ATM ... maybe you should look counters for CPU, GPU, DPC, device drivers activities (you can add quite many counters to follow and each counter has it's own line color in graph)?

Re: Sound "drop-outs" every ten seconds or so

Reply #27
I have tried to identify the parameters that show "high (or preak) levels".

Here is for example a picture of the curves showing "C1, C2, C3 Transitions/s" (I have no idea of what it is). But only those indicated "instance _Total", "Instance 0,_Total" "Instance 0,1" and "Instance 0,_ " seem to be frequently very high close to or at 100 %...C3 Instance 01 even seem to be almost continuously at 100 %.
The others (Instance 0,2 and 0,3 show much lower levels). Once again I have do idea of what it means.




Re: Sound "drop-outs" every ten seconds or so

Reply #29
Thank you for suggesting this link ; I learned some interesting information. However, to really understand what all these data mean, and how to spot the cause of the drop-outs, I would need to be really more competent.

I remain unable to understand why it worked perfectly well until suddenly I began to have this problem without any major change in my system.
Since the sound is clean when played through the PC loudspeaker, with no interruptions (suggesting that the CPU is working well in that case), but is interrupted every 10 sec when played through the USB, does it mean something about the possible cause ?

Re: Sound "drop-outs" every ten seconds or so

Reply #30
Thank you for suggesting this link ; I learned some interesting information. However, to really understand what all these data mean, and how to spot the cause of the drop-outs, I would need to be really more competent.

I remain unable to understand why it worked perfectly well until suddenly I began to have this problem without any major change in my system.

Begging the question: were their any minor changes?

Quote
Since the sound is clean when played through the PC loudspeaker, with no interruptions (suggesting that the CPU is working well in that case), but is interrupted every 10 sec when played through the USB, does it mean something about the possible cause ?

There is a huge difference in complexity between the internal  audio interface and one that is out in USB-land.

The internal interface is clean and simple - channel interface, converter and out the analog port. There are only 3 layers.

The USB interface has at least one more level of indirection or separation.

USB data flow is channel to USB interface to USB device to SP/DIF or optical to converter and out the analog port.

I count 5 or 6 which is obviously more than 3.  Any one layer can be a point of failure.

Re: Sound "drop-outs" every ten seconds or so

Reply #31
Thank you for your clarification, I understand better. I candidly tought that the digital music signal was sent almost unmodified from the HD to the USB, the conversion Digital>Analog being processed essentially by/in the DAC...Far from the reality.

The problem began a couple of weeks/months ago. I actually cannot remember of any change at all at that time. In case of a recent modification to my system (e.g. new software installed, new antivirus or change in power supply parameters), I think I would have connected the dots. But of course I may have changed something minor that I forgot...Unless it has something to do with MS or FireFox updates - but in that case the issue would have been more broadly noticed.

I hope I will find a solution. All ideas welcome.

Re: Sound "drop-outs" every ten seconds or so

Reply #32
By the graph you hosted few posts earlier it looks like those periodical spikes goes to "danger zone" (beyond 100%) ... does the interval of spikes match with drop-outs? Answer given on site behind the link I posted above suggests something as a solution for this type of an issue ... have you tried those already?

Re: Sound "drop-outs" every ten seconds or so

Reply #33
.. does the interval of spikes match with drop-outs?

No, the drop-outs are approx. every 12-15 seconds, whilst the spikes seen on the graph are two or three times more frequent : one every three-four seconds.

The page you quoted suggest to "tweak" the Bios, however I have not seen this kind of parameters in my Bios.

Re: Sound "drop-outs" every ten seconds or so

Reply #34
Thank you for your clarification, I understand better. I candidly tought that the digital music signal was sent almost unmodified from the HD to the USB, the conversion Digital>Analog being processed essentially by/in the DAC...Far from the reality.
No, it is correct. If you don't do any DSP, data is send unaltered over the USB to the DAC.
There the DA takes place.

As your system is high on DPC latency, the problem is most likely upstream.
The USB hub receives its data to late to maintain a permanent stream to the DAC.
A likewise problem: https://hydrogenaud.io/index.php/topic,92998.0.html#entry783934
TheWellTemperedComputer.com

Re: Sound "drop-outs" every ten seconds or so

Reply #35
Quote
data is send unaltered over the USB to the DAC.

But I thought that Foobar (or other players) did somehting to the signal ? If they don't, then what is the use of Wasapi, Asio, etc... ?

And since the music played through the computer loudspeaker is uninterrupted, does this mean that the internal audio circuitry (Nvidia, or else) compensate or fills-in the drop-outs ?

Reading the page that you suggested (thank you), most of the suggestions are around Bios power management adjustments that do not exist in my Bios user interface, or with CPU poser management that are beyond  my computer knowledge.
However, one idea was easy to test : I tried to plug my DAC into my (powered) USB Hub : surprise, the drop-outs happen approximately at the same frequency, but are much shorter, like 2-3/10th of a second instead of 5-6/10th (a rough approximtation, but the difference is audible). This may suggest that the USB power supply can have something to do with the problem ?



Re: Sound "drop-outs" every ten seconds or so

Reply #36
A media player reads the data (the audio file)
As DACs don’t understand a bloody thing about audio formats (FLAC, MP3, etc) the file is decoded=converted to PCM
If you don’t use DSP e.g. volume control, EQ, volume leveling, etc, the media player sends the audio unaltered to the audio driver.

In case of win this is by default DS (Direct Sound).
As it is a mixer by design each stream will be converted to float, mixed, dithered en converted back to integer.
If the sample rate differs from the sample rate as set in the win audio panel, sample rate conversion will be applied.
Unfortunately, the win sample rate converter is not the best there is.
http://archimago.blogspot.nl/2015/11/measurements-windows-10-audio-stack.html

You can bypass DS by using ASIO or WASAPI as these protocols talk straight to the audio device hence  the processing as inherent to the win mixer will be avoided
http://www.thewelltemperedcomputer.com/KB/WASAPI.htm


TheWellTemperedComputer.com

Re: Sound "drop-outs" every ten seconds or so

Reply #37
This is off-topic but I have to object to those Archimago's Windows resampling tests. RMAA test I ran for DirectSound output getting resampled from 44.1 kHz to 48 kHz in 2014 produced excellent results: link. For reference same sound card without resampling gave these results: link. This test was done in Windows 8.1 as it predates Windows 10.

And I ran just now archimago's test with a simple Realtek soundchip using Windows 10. Windows mixer was set to 96 kHz 24 bits mode for output and line-in and I played 44.1 kHz sample files using foobar2000 DS output. Recording was done with Audition in 96 kHz mode. 19 + 20 kHz sines playing with Windows resampling:

And same file recording with SoX resampling to 96 kHz with highest quality:


Frequency responses for sine waves, Windows resampler:

SoX resampler:


I don't think there's anything to complain about Windows' resampler.


Re: Sound "drop-outs" every ten seconds or so

Reply #39
I candidly tought that the digital music signal was sent almost unmodified from the HD to the USB, the conversion Digital>Analog being processed essentially by/in the DAC...Far from the reality.
The most "straightforward" type of audio device should be the PCI ones, anything based on external connection (USB/FW/TB) have an additional layer of complexity.

Of course it is not a good explanation to your drop out issue since a lot of people using USB/FW/TB have no problem at all.

About the power management features, if you cannot find them in BIOS, it probably due to the fact it is a laptop and Dell doesn't want you to shorten your battery's life so such options are not exposed. You may try the Presonus' method which don't need to tocuh the BIOS:

http://support.presonus.com/hc/en-us/articles/210046213-Windows-Vista-7-and-8-Advanced-Power-Options-Settings

If it works, that's fine, but remember doing so may shorten your battery's life and make your CPU hotter. If it doesn't work, then the problem is probably not related to power settings.

If you want to troubleshoot some potential driver issues by disabling or uninstalling some devices, I recommend making a full disk backup image using tools like Macrium Reflect Free before doing so, just in case you or Windows did something wrong.

BIOS flashing should be the last resort due the potential risk of bricking your laptop, no matter how small the chance it is.



Re: Sound "drop-outs" every ten seconds or so

Reply #40
Roseval : thank you for your additional explanations, really interesting although some of it is beyond my level.

Bennetng : I did some modifications following the presonus recommendations, with no improvement. Thank you for the advice about the back-up, I'll make sure to have a safety net when doing significant actions.

I've found a cheaper battery provider, we'll see when I receive it, if the zero charge battery was the root cause, after all... Whent I think of it, it is possible that the issue began when the battery went completely out of service. This 100 percent empty status was unnoticed at first, since I always work on mains, and so I would have had no clue that the cause could be this.

 

Re: Sound "drop-outs" every ten seconds or so

Reply #41
Finally, some good news ! 
(not sure double posts are allowed, but I tought it may be easier to read this way..?).

Facts :
- I replaced my dead battery by a new one, not a Dell at an extortionate 180 $, but an equivalent at 35$ !
- The multiple drop-outs, 5-6 per minute, have now disappeared. (*)
- Also, I discovered that my charger didn't charge the new battery ; it does power the computer, but does not charge.
- I replaced it by a similar used Dell charger which does charges the new battery (but not the old dead one, though).

My comments :
- the drop-outs came from the computer power management system badly handling the dead battery / faulty charger. Can't tell exactly how.
- I also discovered a Dell Battery Gauge option in the W7 Audio & hardware / Power Management screen. It show the status of the battery, and also offers a "Disable battery charge" option. If checked, it says that "the charge will stop until the next start of the computer". See here below. I do'nt know the use of this, and why an aircraft icon...?



(*) Since I actually had one drop-out in fifteen/twenty minutes listening, but it may be a file coding bug, I'll double check. At least that will mean 99 per cent of the issue solved. If it happens again, I'll try with the "disable battery charge" checked or not, just in case.

Again, thank you to all of those who spent time to read my post and try to help. It has been very useful anyways since the root cause was, as suggested, actually linked to the power management, and in fact at the very source of the power (dead battery or "dead battery + faulty charger" management).

Re: Sound "drop-outs" every ten seconds or so

Reply #42
Finally, some good news ! 
(not sure double posts are allowed, but I tought it may be easier to read this way..?).

Facts :
- I replaced my dead battery by a new one, not a Dell at an extortionate 180 $, but an equivalent at 35$ !
- The multiple drop-outs, 5-6 per minute, have now disappeared. (*)
- Also, I discovered that my charger didn't charge the new battery ; it does power the computer, but does not charge.
- I replaced it by a similar used Dell charger which does charges the new battery (but not the old dead one, though).

My comments :
- the drop-outs came from the computer power management system badly handling the dead battery / faulty charger. Can't tell exactly how.
- I also discovered a Dell Battery Gauge option in the W7 Audio & hardware / Power Management screen. It show the status of the battery, and also offers a "Disable battery charge" option. If checked, it says that "the charge will stop until the next start of the computer". See here below. I do'nt know the use of this, and why an aircraft icon...?



(*) Since I actually had one drop-out in fifteen/twenty minutes listening, but it may be a file coding bug, I'll double check. At least that will mean 99 per cent of the issue solved. If it happens again, I'll try with the "disable battery charge" checked or not, just in case.

Again, thank you to all of those who spent time to read my post and try to help. It has been very useful anyways since the root cause was, as suggested, actually linked to the power management, and in fact at the very source of the power (dead battery or "dead battery + faulty charger" management).

Interesting. I have used a goodly number of generic replacement laptop batteries, and generally but not universally have had good success with them. I did this in the context of where I've been supplying refurbished laptops to students of the High School  Diploma Equivalency program that my wife volunteers at.

I picked a common model of name-brand laptop, acquired sick and dying ones via various means, and put them back into service by replacing failed or substandard parts and software. Of course, most of them showed up with dead or dying batteries, no power supplies at all, dead hard drives, bad optical drives, weak wifi cards, corrupted OS's, a surprising that were good but had an unknown password, and occasional failed system boards and broken power jacks. Used but good parts are plentiful and inexpensive. The worst were sacrificed to fix others. All received SSDs which gave them a second life as a pretty fast machine. I also put in some upgraded battery packs which made a slightly heavier machine that ran up to half again as long on one standard charge which is a nice convenience. Inexpensive but smart looking padded pouches and a new mouse completed the package. I let the school's administration pick the recipients.

It is common for there to be two capacities of external power supply for a given make and model of laptop. One will charge it, or run it, but not both at the same time. The other larger capacity one will run the laptop and charge it at the same time. The low prices you saw for the batteries are duplicated for the power supplies, so there is no reason not to get the larger model.  Acommon mode of failure is for the power supply to work, but lose its ability to provide full power. This often preceeds total failure. The usual batter failure is to simply be unable to run the machine for the usual amount of time which also preceeds total failure in many cases. 

I managed to buy a few bogus batteries and power supplies, so Caveat Emptor.  The bogus batteries were either down on capacity or failed after a few charging cycles. The bogus power supplies, ditto. The bad ones were no cheaper than some that proved out to be just fine.

 
SimplePortal 1.0.0 RC1 © 2008-2019