PDA

View Full Version : Squeezebox3 - Complete FREAKOUT & stereo damage warning



Free Lunch
2005-12-05, 12:04
So I've had a pair of new sb3's since Friday. The idea is to replace
my very early slimp3. The hope is that 6.X will work well with the new
hardware (it has never worked well with my slimp3, so I have stayed on
5.4.1).

The sb3 has been playing almost constantly since Friday. Today, I was
eating lunch with the sb3 playing at a low volume.. Suddenly, massive
ear splitting static from my hifi. The sort that sends your lunch to
the floor while you make a frantic effort to mute the pre-amp. The
noise was traced to the analog rca outputs of the sb3.

The static occurred with every track I played.. I began a network
traffic capture and I placed an urgent call to slim support with the
hope of isolating this 'show stopper'.

Basically, support reported that this problem has been observed twice
before. The Xilinix chip 'somehow' gets corrupted during a firmware
update, causing sample rate problems on the DAC. I was asked to
re-program the chip (very simple, just a power cycle of the sb3 while
holding '1' on the remote). That seems to have cleared the problem
for now. Not sure why the problem did not occur sooner. The firmware
was upgraded when I first powered on the unit Friday. Never had
anything like this happen during years of using a slimp3. Perhaps
there should be more self-integrity checking?

I have been shopping for some new amps and plan on putting the other
sb3 into a new listening room. I have been considering directly
attaching the sb3 to the amps. This experience reinforces what my gut
was saying all along, basically "Attaching a squeezebox directly to an
amp? Are you f'n nutz??!!". Yeah, imagine a pair of 250 watt
monoblock tube amps sending max volume static and distortion to your
speakers..

Things were going quite well up until this point. Unfortunately, I
cannot say that this was the smooth intro to the 'latest & greatest'
product versions that I was hoping for. But, hey, the new case looks
really.. uh.. big.. and new.. Still haven't decided whether to send
them back during the eval period.



FL

kefa
2005-12-05, 13:20
if this was a dac problem - does this mean that it couldn't happen to the digital output...this is just pass-through right?

seanadams
2005-12-05, 16:41
FL,

Sorry about the trouble. FWIW we have always cautioned against connecting directly to an amp unless levels are appropriately limited on the analog side.

Sean

enduser
2005-12-06, 03:47
FL,

Sorry about the trouble. FWIW we have always cautioned against connecting directly to an amp unless levels are appropriately limited on the analog side.

SeanIt looks like your (in)famous speaker-killer bug is still there.

BTW, this issue cannot be solved with your 'caution' advice. A preamp doesn't help vs. just using an amp. The volume control is not effective either.

The only way it can be solved is to add a pro-audio signal limiter to your audio output chain. This protects your speaker system (and amps) with a slight degradation in sound quality.

Read up on the feature of DBX limiters and speaker control processors called PeakStopPlus:

For overall speaker protection, our new PeakStopPlus™ does all previous circuits one better. With a new design, the PeakStopPlus represses those unwanted transients from blowing your drivers while minimizing the distortion common to many other "hard" limiters.

You can get one of these or something like it to understand what I am talking about:

http://www.dbxpro.com/1066.htm

It sounds like a feature you should add to the SB. Maybe you can license the code from them :-)

Cleve
2005-12-06, 05:17
To be fair - do most sources (ie, cd players, tuners, tape decks, phono cartridges) have any type of "level protection" built in? I don't think so. Regardless, situations like this make me glad I have McIntosh amplification with Power Guard and Sentry Monitor. My loudspeaker investment is kept safe and protected from a component going "berserk".

enduser
2005-12-06, 07:43
To be fair - do most sources (ie, cd players, tuners, tape decks, phono cartridges) have any type of "level protection" built in? I don't think so. Regardless, situations like this make me glad I have McIntosh amplification with Power Guard and Sentry Monitor. My loudspeaker investment is kept safe and protected from a component going "berserk".It doesn't matter what sort of input source; there is almost always something that can go wrong. Maybe you think that grooves on a record are only so wide, so there is some maximum voltage. But that doesn't mean that sub-harmonics on a record cannot blow out your speakers given enough juice.

Hence in the analog world you have high-pass filters, low-pass filters, compressors, limiters, etc. All sorts of tools to make sound work and not damage your expensive equipment.

In the digital world, it takes a stream processing architecture that validates the stream and filters it with a digital peak limiter. You could also do compression and high/low pass filters all in the digital domain. This approach would rely on level-setting your analog gain stages appropriately.

In my experience with SB1/SB3, I don't think the SB does anything to provide safe signals. Sean could of course correct me on this.

T
2005-12-06, 07:55
> In my experience with SB1/SB3, I don't think the SB does anything to
> provide safe signals.

In my experiance, almost no consumer or professional equipment does, except
as an option at extra cost.

Tom

Khuli
2005-12-06, 08:04
In the digital world, it takes a stream processing architecture that validates the stream and filters it with a digital peak limiter...

I maybe missing something here, but aren't digital signals limited anyway to the maximum number of bits in the signal?

I was under the impression that (digital) peak limiters were used to prevent excessive clipping when amplifying signals or recording from an analogue source, not for output protection...

enduser
2005-12-06, 08:27
I maybe missing something here, but aren't digital signals limited anyway to the maximum number of bits in the signal?

I was under the impression that (digital) peak limiters were used to prevent excessive clipping when amplifying signals or recording from an analogue source, not for output protection...An analog peak limiter is usually used before the A/D converter so that you don't get garbage for a digital signal. Because as you may guess, if you exceed the signal level that digital can represent, that data is gone. There is no way to peak limit data that did not get captured.

Similarly, for D/A, unless you have calibrated the analog gain stages, running a digital peak limiter is of marginal utility. Hence most peak limiting is done in the analog domain.

What I was referring to was for digital-to-digital. If you are outputting a digital signal and you see something that is "bad data", you should filter it, not send it. In the old SB there was no way to fix these bugs because they were part of a non-programmable chip. I was told in the new SB models that these bugs are fixable. As the SB has two digital outs, the stream(s) that go to these outs should be filtered for bad data.

Aaron Zinck
2005-12-06, 09:08
>
> BTW, this issue cannot be solved with your 'caution' advice. A preamp
> doesn't help vs. just using an amp. The volume control is not effective
> either.
>

To the contrary--a preamp surely does help in this situation. The "garbage"
sound that's been described is likely output at 100% of the SB3's potential
volume output, and is certainly output at the same level irrespective of the
SB3's volume setting. Bearing that in mind, if you're not using a preamp
you're relying on the SB3 itself to attenuate its output to the appropriate
level for listening (maybe you have it set to 30% volume) but if the SB3
malfunctions and suddenly sends a loud signal then you're looking at 100%
SB3 garbage output straight into your amp with no attenuation. However, if
you're using a preamp then you're likely using your SB3 at full volume
output all the time and relying on the preamp for attenuation. This results
in a greatly reduced volume differential between the 100% volume music
signal and the 100% volume static/garbage signal. Another way to say it is
that if your preamp's set to, say 30% volume level then anything coming in
will be attenuated to 30% its original level before reaching the amp.

enduser
2005-12-06, 09:35
>
> BTW, this issue cannot be solved with your 'caution' advice. A preamp
> doesn't help vs. just using an amp. The volume control is not effective
> either.
>

To the contrary--a preamp surely does help in this situation. The "garbage"
sound that's been described is likely output at 100% of the SB3's potential
volume output, and is certainly output at the same level irrespective of the
SB3's volume setting. Bearing that in mind, if you're not using a preamp
you're relying on the SB3 itself to attenuate its output to the appropriate
level for listening (maybe you have it set to 30% volume) but if the SB3
malfunctions and suddenly sends a loud signal then you're looking at 100%
SB3 garbage output straight into your amp with no attenuation. However, if
you're using a preamp then you're likely using your SB3 at full volume
output all the time and relying on the preamp for attenuation. This results
in a greatly reduced volume differential between the 100% volume music
signal and the 100% volume static/garbage signal. Another way to say it is
that if your preamp's set to, say 30% volume level then anything coming in
will be attenuated to 30% its original level before reaching the amp.Yes, you can use the preamp as a brute-force and simplistic gain control. I was not saying that you cannot do that. But it mostly not a good solution for listening to music.

While the preamp does allow you to set a maximum gain, it has no knowledge of the volume of the incoming music. So you will have some music that is too quiet and some that is too loud.

While you can protect your speakers by setting the preamp gain (volume) to very low, you will not be able to listen to much of your music. This is because for most preamps, you would have to set the volume VERY VERY low to prevent any possible signal from having a harmful effect on your amps/speakers.

This is why a peak limiter not a volume control is needed.

There is no nice solution other than building in a quality peak limiter or running a scanner/filter on the digital data stream before it gets sent out the digital port for digital-to-digital situations.

With a digital stream processor you could then decide to set the digital outs on the SB to fixed and then control the gain with your preamp. This also has issues of working/not working well with all your music. And makes the customer work with two remotes vs. one. But it would prevent bad digital data from going out of the SB and into your main system.

All I know is that the SB puts out bad data far more often than any other digital source that I've tried. It is something to be aware of and take appropriate measures, however compromised those measures may be.

Dan Goodinson
2005-12-06, 09:48
enduser Wrote:
>>
There is no nice solution other than building in a quality peak limiter
or running a scanner/filter on the digital data stream before it gets
sent out the digital port for digital-to-digital situations.
>>

Would that not be the same as replay gain? My (admittedly limited)
understanding of reply gain is that at encode (or after encode) you can
build-in a sort of limiter to make the volume distribution more even
across the track. Would this, in combination with a pre-amp, achieve
the desired result?

pfarrell
2005-12-06, 10:03
On Tue, 2005-12-06 at 17:48 +0100, Dan Goodinson wrote:
> enduser Wrote:
> >>
> >There is no nice solution other than building in a quality peak limiter
> >or running a scanner/filter on the digital data stream before it gets
> >sent out the digital port for digital-to-digital situations.
>
> Would that not be the same as replay gain?


Depends on the cause, if there is any real problem.

If the SqueezeBox got totally confused, and did something really
weird like swapping the internal representation of the audio bytes,
than the signal would be very loud (the lowest bit usually varies the
most and would be the highest) and random sounding. Nothing
done on the server could help.

Even professional studio gear is not aimed at chopping 30 dB out
of a signal, limiters and compressors usually just wack off the top
few dB of signal. Taking more than 6 dB out of a signal usually removes
all the music. Only pop stations do that.


--
Pat
http://www.pfarrell.com/music/slimserver/slimsoftware.html

enduser
2005-12-06, 10:06
enduser Wrote:
>>
There is no nice solution other than building in a quality peak limiter
or running a scanner/filter on the digital data stream before it gets
sent out the digital port for digital-to-digital situations.
>>

Would that not be the same as replay gain? My (admittedly limited)
understanding of reply gain is that at encode (or after encode) you can
build-in a sort of limiter to make the volume distribution more even
across the track. Would this, in combination with a pre-amp, achieve
the desired result?
Replay gain is a means of taking disparate pieces of music, mastered at different levels, and making them come out of your system at the same volume level. The goal is to alleviate or eliminate your need to turn the volume up and down on a per song ( or other frequent ) basis.

Replay gain allows you to get your volume at a comfortable level and enjoy your music. Unfortunately this comfortable level, when faced with digital noise at max volume, is still enough to blow speakers.

As digital data is present on the server (for the most part) before being sent to the SB, what would be useful is a scanner that looks for digital noise in the input stream. This would help the overall signal chain quite a bit.

However, what is really needed is what radio stations use which is essentially a real-time mastering engine. This includes compression, equalization, gating, peak limiting, etc. This is the only way to get quality sound from a wide variety of music. It requires a lot of grunt to run such a mastering engine. If the SB system were a bit more open, the sound streams could be bussed into existing PC-based digital mastering tools and then out to the SB for on-location D/A. Maybe that will happen one day ;-)

Alex Twisleton-Wykeham-Fiennes
2005-12-06, 10:20
On Tue 6 December 2005 17:06, enduser wrote:
> Dan Goodinson Wrote:
> > enduser Wrote:
> >
> > There is no nice solution other than building in a quality peak
> > limiter
> > or running a scanner/filter on the digital data stream before it gets
> > sent out the digital port for digital-to-digital situations.
> >
> >
> > Would that not be the same as replay gain? My (admittedly limited)
> > understanding of reply gain is that at encode (or after encode) you
> > can
> > build-in a sort of limiter to make the volume distribution more even
> > across the track. Would this, in combination with a pre-amp, achieve
> > the desired result?
>
> Replay gain is a means of taking disparate pieces of music, mastered at
> different levels, and making them come out of your system at the same
> volume level. The goal is to alleviate or eliminate your need to turn
> the volume up and down on a per song ( or other frequent ) basis.
>
> Replay gain allows you to get your volume at a comfortable level and
> enjoy your music. Unfortunately this comfortable level, when faced with
> digital noise at max volume, is still enough to blow speakers.
>
> As digital data is present on the server (for the most part) before
> being sent to the SB, what would be useful is a scanner that looks for
> digital noise in the input stream.

I would be very interested in an algorithm that can distinguish noise from
music. Please post any online references to this.

> This would help the overall signal
> chain quite a bit.

Are you suggesting that the fault that you have experienced was introduced at
the server and then streamed down to the Squeezebox whereupon the Squeezebox
then output the supposedly valid yet somehow corrupt information out of the
SP-DIF outputs which your DAC then output at a level that was higher than
your normal music?

I'd really like to refresh my memory of what this thread is *really* about.
Can someone post a link to the original post, or any further post that
actually defines what the problem that is being discussed is? (I'm reading
this via email and regularly wipe the contents of my slim folder so I don't
have the original post).

> However, what is really needed is what radio stations use which is
> essentially a real-time mastering engine. This includes compression,
> equalization, gating, peak limiting, etc. This is the only way to get
> quality sound from a wide variety of music. It requires a lot of grunt
> to run such a mastering engine. If the SB system were a bit more open,
> the sound streams could be bussed into existing PC-based digital
> mastering tools and then out to the SB for on-location D/A. Maybe that
> will happen one day ;-)

Personally I can safely say that the day that Slim Devices add in further
limiting to the already stupid levels applied as a matter of course by the
mastering engineers of the world that I'll roll-back to my last version of
the firmware that doesn't include this "feature" and stick with that. Music
is already *way* too compressed and adding another level of compression would
generally just make it worse.

However, fortunately for yourself, the idea of plugging in compression into
the system on the server (which I believe is what you are suggesting) is
already a) completely open and b) under a license that lets you do exactly
that so you should have no problem with it.

Alex

T
2005-12-06, 10:56
> what is really needed is what radio stations use which is
> essentially a real-time mastering engine. This includes compression,
> equalization, gating, peak limiting, etc. This is the only way to get
> quality sound from a wide variety of music.

The purpose of "real-time mastering engine", as you call them, is NOT to
get 'quality' sound, but rather get maximum loudness (which is why they are
normall called "loundess processors"). Audiophiles, and all radio engineers
that I know, and I know at lot after 20+ years as a designer/manufacturer of
professional audio equipment, hate them. Only the marketing guys like them.

Tom

enduser
2005-12-06, 11:06
> what is really needed is what radio stations use which is
> essentially a real-time mastering engine. This includes compression,
> equalization, gating, peak limiting, etc. This is the only way to get
> quality sound from a wide variety of music.

The purpose of "real-time mastering engine", as you call them, is NOT to
get 'quality' sound, but rather get maximum loudness (which is why they are
normall called "loundess processors"). Audiophiles, and all radio engineers
that I know, and I know at lot after 20+ years as a designer/manufacturer of
professional audio equipment, hate them. Only the marketing guys like them.

TomI agree. I am not saying the system is good for audiophiles! JGH I am not. But they are used in almost every commercial radio station, are they not? ;-) Because they deliver *consistent* loudness. And maybe you didn't notice, but radio stations have a lot of pro equipment sold to them by pro manufacturers. It is simply part of how radio works. If it didn't have consistent loudness, it would be very difficult to listen to the radio given any ambient noise. And much of how this is done would be applicable to "replay gain".

Aaron Zinck
2005-12-06, 13:11
> Yes, you can use the preamp as a brute-force and simplistic gain
> control. I was not saying that you cannot do that. But it mostly not a
> good solution for listening to music.

This is a solution that all of the music-listening world uses. You need a
gain control of some sort. Period. If you want to do some sort of dynamic
leveling then there are many solutions for that from live dynamic
compression to solutions like replaygain. I, for one, don't want a dynamic
compressor in line with my music listening.

>
> While the preamp does allow you to set a maximum gain, it has no
> knowledge of the volume of the incoming music. So you will have some
> music that is too quiet and some that is too loud.

I will have some music that is louder than others, that is true. But your
assessment that that is not what I want is presumptious. Besides--this is
far removed from the issue that started this thread so I don't really see
your point here.

I'm not going to bother addressing all of your other statements one by one.
Sean's recommendation of running your SB3 into a preamp before an amp stands
as making sense. I agree that the only absolute way to avoid damage from
spikes like this is with a limiter. However, that doesn't mean a preamp
won't *help* the situation (and help it tremendously). You seem to be in
the my-way-or-the-highway camp of arguing. There's no middle ground with
you, no shades of grey. You seem to say "if a solution leaves even a remote
possibility of damage being done then that solution has no merit at
all--you're all wrong and only my solution is right". A preamp can be a
natural limiter of sorts as parts of its gain stage are likely to clip
before allowing the full brunt of a shrieking SB3 through to the amps (by
the way, for the benefit of anyone reading this post out of the context of
this thread it should be noted that this is apparently an extremely unusual
problem). This is not to say that there's no potential for damage, just
that a preamp would help mitigate possible damage. And, by the way, any
piece of electronics can malfunction at any point in time. A limiter can go
bad and start feeding crap into your amps. There's no guarantee against
that. I'm not saying that that doesn't make a limiter a good solution--I
just wanted to remind you that it's not fail-safe.


>
> All I know is that the SB puts out bad data far more often than any
> other digital source that I've tried. It is something to be aware of
> and take appropriate measures, however compromised those measures may
> be.
>

I've been involved in this mailing list (and forums) as either a "lurker" or
active participant for over 2 years now. This is generally a very helpful
and friendly community of folks with similar interests. I've watched your
posts over the past couple of days and you seem to have come on here with an
axe to grind. The tone of your posts seems designed to irritate and
agitate. I obviously can't stop you from participating and don't desire to
censor your comments, but perhaps some self-censoring is in order on your
part. If this is a community that you desire to be active in then please
treat people with respect. Your comment about the SB3's digital output
seems pretty baseless to me and that's why it's disrespectful to all of the
people on here who have tried to make this device great. Have you had the
problem described? What problems have you had with the SB3's digital outs?
What use does your comment serve if you're not going to point out what's
specifically wrong that needs to be fixed? I think we can all acknowledge
that this is a device that's on the bleeding edge. All devices have some
problems and I trust SlimDevices to fix the problems with their devices more
than any other company. There are occasional problems with it, but most of
us are here because we love the product and we love being geeks and figuring
stuff out and improving the product. I'm willing to do this because I've
seen SlimDevices show incredible receptiveness to constructive criticism and
fresh ideas. Their products are improved at breakneck speed. In this
particular circumstance it would appear that the problem isn't any kind of
Slim engineering error, but rather an error by one of their suppliers. I'm
sure they'll do whatever's in their power to isolate the problem and work on
getting a fix or making it right with those who are seeing the issue.
Repeatedly offering disparaging comments, a big ego, and whining aren't a
way to endear yourself to the community. If you don't like the SB3 then
return the thing and leave the forums--that's your perogative. No one's
forcing you to live with an SB3. It has a full, 30-day money back guarantee
and if you don't like it you can send it back and go tell all your friends
how crappy the SB3 is. If, however, you decide to keep the device then
don't act like you've been duped or swindled. You had 30 days to assess the
device. You know its (in my mind relatively few) shortcomings and if you
have constructive criticism and a respectful attitude towards the members of
this community then come on in and we'll work together to make the thing
even cooler.

Remember: I'm not saying don't criticize the device--just do it the right
way.

enduser
2005-12-06, 13:42
> Yes, you can use the preamp as a brute-force and simplistic gain
> control. I was not saying that you cannot do that. But it mostly not a
> good solution for listening to music.

This is a solution that all of the music-listening world uses. You need a
gain control of some sort. Period. If you want to do some sort of dynamic
leveling then there are many solutions for that from live dynamic
compression to solutions like replaygain. I, for one, don't want a dynamic
compressor in line with my music listening.

>
> While the preamp does allow you to set a maximum gain, it has no
> knowledge of the volume of the incoming music. So you will have some
> music that is too quiet and some that is too loud.

I will have some music that is louder than others, that is true. But your
assessment that that is not what I want is presumptious. Besides--this is
far removed from the issue that started this thread so I don't really see
your point here.

I'm not going to bother addressing all of your other statements one by one.
Sean's recommendation of running your SB3 into a preamp before an amp stands
as making sense. I agree that the only absolute way to avoid damage from
spikes like this is with a limiter. However, that doesn't mean a preamp
won't *help* the situation (and help it tremendously). You seem to be in
the my-way-or-the-highway camp of arguing. There's no middle ground with
you, no shades of grey. You seem to say "if a solution leaves even a remote
possibility of damage being done then that solution has no merit at
all--you're all wrong and only my solution is right". A preamp can be a
natural limiter of sorts as parts of its gain stage are likely to clip
before allowing the full brunt of a shrieking SB3 through to the amps (by
the way, for the benefit of anyone reading this post out of the context of
this thread it should be noted that this is apparently an extremely unusual
problem). This is not to say that there's no potential for damage, just
that a preamp would help mitigate possible damage. And, by the way, any
piece of electronics can malfunction at any point in time. A limiter can go
bad and start feeding crap into your amps. There's no guarantee against
that. I'm not saying that that doesn't make a limiter a good solution--I
just wanted to remind you that it's not fail-safe.


>
> All I know is that the SB puts out bad data far more often than any
> other digital source that I've tried. It is something to be aware of
> and take appropriate measures, however compromised those measures may
> be.
>

I've been involved in this mailing list (and forums) as either a "lurker" or
active participant for over 2 years now. This is generally a very helpful
and friendly community of folks with similar interests. I've watched your
posts over the past couple of days and you seem to have come on here with an
axe to grind. The tone of your posts seems designed to irritate and
agitate. I obviously can't stop you from participating and don't desire to
censor your comments, but perhaps some self-censoring is in order on your
part. If this is a community that you desire to be active in then please
treat people with respect. Your comment about the SB3's digital output
seems pretty baseless to me and that's why it's disrespectful to all of the
people on here who have tried to make this device great. Have you had the
problem described? What problems have you had with the SB3's digital outs?
What use does your comment serve if you're not going to point out what's
specifically wrong that needs to be fixed? I think we can all acknowledge
that this is a device that's on the bleeding edge. All devices have some
problems and I trust SlimDevices to fix the problems with their devices more
than any other company. There are occasional problems with it, but most of
us are here because we love the product and we love being geeks and figuring
stuff out and improving the product. I'm willing to do this because I've
seen SlimDevices show incredible receptiveness to constructive criticism and
fresh ideas. Their products are improved at breakneck speed. In this
particular circumstance it would appear that the problem isn't any kind of
Slim engineering error, but rather an error by one of their suppliers. I'm
sure they'll do whatever's in their power to isolate the problem and work on
getting a fix or making it right with those who are seeing the issue.
Repeatedly offering disparaging comments, a big ego, and whining aren't a
way to endear yourself to the community. If you don't like the SB3 then
return the thing and leave the forums--that's your perogative. No one's
forcing you to live with an SB3. It has a full, 30-day money back guarantee
and if you don't like it you can send it back and go tell all your friends
how crappy the SB3 is. If, however, you decide to keep the device then
don't act like you've been duped or swindled. You had 30 days to assess the
device. You know its (in my mind relatively few) shortcomings and if you
have constructive criticism and a respectful attitude towards the members of
this community then come on in and we'll work together to make the thing
even cooler.

Remember: I'm not saying don't criticize the device--just do it the right
way.I don't have a lot of time to spend figuring out the very best way to talk about something for a company that ripped me off for $200+. Having almost blown some very expensive speakers with the SB1, I really don't have any patience left for the raft of digital noise bugs that apparently still lives in the SB3 ( the original thread topic ).

I do have bad experiences with the company and it will take time for good experiences to outweigh the bad.

You may not like the way I am saying something and you will just have to live with that.

On the technical side of things, I think there are countless algorithms for identifying noise and unwanted sounds and many shipping products. While you might have been using a SB for two years, you don't have 25 years of audio background. Do your research before claiming things like there's no way to do this. I am not your lab assistant or tutor. If you are interested, do the work.

Aaron Zinck
2005-12-06, 14:05
> I don't have a lot of time to spend figuring out the very best way to
> talk about something for a company that ripped me off for $200+. Having
> almost blown some very expensive speakers with the SB1, I really don't
> have any patience left for the raft of digital noise bugs that
> apparently still lives in the SB3 ( the original thread topic ).
>
> I do have bad experiences with the company and it will take time for
> good experiences to outweigh the bad.
>
> You may not like the way I am saying something and you will just have
> to live with that.
>
> On the technical side of things, I think there are countless algorithms
> for identifying noise and unwanted sounds and many shipping products.
> While you might have been using a SB for two years, you don't have 25
> years of audio background. Do your research before claiming things like
> there's no way to do this. I am not your lab assistant or tutor. If you
> are interested, do the work.

Despite having no idea what my background is, you have proceeded to
patronize and insult me. I am a computer engineer and am a semi-pro sound
and recording engineer with 10 years of experience. I didn't claim that
there's no way to do what you're suggesting. I simply offered a defense for
Sean's preamp point because your argument against it was factually lacking.

I might point out that you still haven't offered any specifics about your
problem with the SB1. You seem much more interested in complaining and
insulting than anything else.

bossanova808
2005-12-06, 14:16
I had an LG DVD player blow the tweeters out of my $1800 speakers.

These things can happen. None of these things are perfect.

LG didn't even bother to respond to our contact about the issue. At least Slim Devices have been responsive. Hell, you didn't actually lose anything anyway - only 'almost' - count your blessings!

T
2005-12-06, 14:20
> On the technical side of things, I think there are countless algorithms
> for identifying noise and unwanted sounds and many shipping products.

You're thinking is wrong.

> While you might have been using a SB for two years, you don't have 25
> years of audio background.

I do. And 23+ of it is professional digital audio design.

Again, your thinking is wrong.

Tom

Aaron Zinck
2005-12-06, 14:39
> As digital data is present on the server (for the most part) before
> being sent to the SB, what would be useful is a scanner that looks for
> digital noise in the input stream. This would help the overall signal
> chain quite a bit.

As per the OP, this isn't where the problem originated. Instead,
(apparently) the DAC was in a funky state that resulted in noise from the
analog outs. This is entirely independent of the stream being sent to the
SB3.

> However, what is really needed is what radio stations use which is
> essentially a real-time mastering engine. This includes compression,
> equalization, gating, peak limiting, etc. This is the only way to get
> quality sound from a wide variety of music. It requires a lot of grunt
> to run such a mastering engine. If the SB system were a bit more open,
> the sound streams could be bussed into existing PC-based digital
> mastering tools and then out to the SB for on-location D/A. Maybe that
> will happen one day ;-)


This is a ludicrous charge. I challenge you to show an example of a more
open solution than the SB3. In fact, your suggestion of sending the streams
to external processors then onto the squeezebox is already possible. This
is how lame is used to compress audio for streaming to the SB, for instance.
Other folks have discussed putting FIR filters inline, and certainly your
suggestion could be implemented. It's important to note, howerver, that
this would not fix the problem as reported by the original poster.

freelunch
2005-12-07, 09:09
>[color=blue]
However, if you're using a preamp then you're likely using your SB3 at full volume output all the time and relying on the preamp for attenuation. This results in a greatly reduced volume differential between the 100% volume music
signal and the 100% volume static/garbage signal.

Not all pre-amps have remotes so many folks will run the pre-amp volume very high and use the sb remote. My pre does have a remote but I don't like juggling. My pre is/was set to 80+% volume when the freakout occurred. I also send analog to the pre via a UA5 USB device. It has a real volume knob and has never freaked out ;-) The volume knob is invaluable for mastering type work where I want to quickly change from low to high vol, etc.

FWIW, in many years of use I have never had a 'freak out' with my slimp3. Other than a skipping CD, I've never had a FO happen with any of my other audio gear. Of course it could happen but it hasn't hit anyone I know.

I often leave my slimp3 running while I am out of the house (to give the impression someone is home). I'm glad this didn't happen while I was gone for the weekend (new SB3 on friday, left on saturday, returned on sunday). Aside from the equipment damage, I can't imagine what the neighbors would have thought. No doubt they would have eventually called the cops...

I've read the posts here and I think those suggesting filters or limiters as a fix have missed the problem. As it was described, the Xilinix chip set the incorrect sample rate on the DAC (is there a bug ID for that problem?). I'll bet some work-arounds could be put in that would periodically check the sample rate, etc, in an effort to work around this problem.. But that would be a fix for only this problem. No doubt some other bug will come along and cause a similar situation.

I already have a heavy duty dedicated music box. At this point, I wonder if my use of slimp3/sb is best suited for casual listening and as a user inferface to a more reliable music delivery mechanism. I am one of those people who Really Needs 24/96 support. That doesn't to exist today and we haven't heard when it might. So I'm wondering if I should be focusing on using a 24/96 USB audio device. The slimp3/SB would just be the interface. I'm thinking the transcoding hooks would invoke a program which does the actual music delivery to the USB audio device (at 24/96 or whatever).... Hmm...

jonheal
2005-12-07, 11:52
Since I haven't received my SB3 yet, I'm a little ignorant, but from what I've read, the first thing you do before connecting your SB3 directly to an amp is set the SB3's preamp volume to zero, and then work your way up from there until you have an appropriate output voltage to drive your amp.

Assuming that "freelunch" set the preamp level within the SB3 to something less than 100%, how could the SB3 have emitted a 100% signal unless the problem is in the analog stage AFTER the DAC?

qwerty
2005-12-07, 12:14
Another p#ss!ing contest in the Slim forums.

Doesn't this belong in the Audiophile section? You know, the place where they care more about gear and counting the bits than actually listening to music? ;-)

Kurt
2005-12-07, 12:25
"Doesn't this belong in the Audiophile section?"

Tis the season!

jonheal
2005-12-07, 12:33
Another p#ss!ing contest in the Slim forums.

Doesn't this belong in the Audiophile section? You know, the place where they care more about gear and counting the bits than actually listening to music? ;-)

Well, if there is a bug that could potentialy shred your speakers, whether they cost $30 or $3,000, that seems like a problem of general interest to everyone. (Butterfly wings a pretty delicate, you know.)

Kurt
2005-12-07, 12:36
Point taken. Agreed!

ianjohnson_nz
2005-12-07, 17:07
Well, if there is a bug that could potentialy shred your speakers, whether they cost $30 or $3,000, that seems like a problem of general interest to everyone. (Butterfly wings a pretty delicate, you know.)

Indeed. But it's a "bug" with, to date, no concrete analysis and no similar reported behaviour from any other users. It's fair to assume that if SB had a habit of blowing speakers, there'd be more than one person moaning about it.

freelunch
2006-03-01, 07:25
Had another Squeezebox3 freak out this morning.

The power to my house was interrupted for a few seconds and the sb3
lost power. When it came back up, it was sending a buzzing signal to
the right channel of my amp at a moderate volume. There was no
slimserver running at the time.

After I restarted the slimserver, I turned the sb3 volume all the way
down and unpaused.. I was greeted by the wonderful sound of full
blast music despite the null volume. Good morning indeed.

I re-programmed the xilinix chip and it seems fine now.

It would be nice if the hardware/firmware had better corruption
checking and if the outputs were muted by default.. The temporary
buzzing is obviously less of a concern than the full output blasts.

Any progress on understanding/detecting xilinix corruption or
mitigating through firmware changes?


FL

Ben Sandee
2006-03-01, 09:08
On 3/1/06, Free Lunch <freelunch (AT) gmail (DOT) com> wrote:
>
> Any progress on understanding/detecting xilinix corruption or
> mitigating through firmware changes?


I thought they had fixed the Xilinx corruption issues w/the 33 or 35
firmware release on 6.2.2 branch/6.5 trunk. Have you been keeping up to
date? You didn't say what firmware you are running.

Ben

freelunch
2006-03-02, 19:52
Ah yes, the version. I was running the bug fix nightly from Feb 14:

SlimServer_6_2_x_v2006-02-14.tar.gz

And it looks like that version installs firmware version 33.


On 3/1/06, Ben Sandee <tbsandee (AT) gmail (DOT) com> wrote:
> On 3/1/06, Free Lunch <freelunch (AT) gmail (DOT) com> wrote:
>
> > Any progress on understanding/detecting xilinix corruption or
> > mitigating through firmware changes?
>
>
> I thought they had fixed the Xilinx corruption issues w/the 33 or 35
> firmware release on 6.2.2 branch/6.5 trunk. Have you been keeping up to
> date? You didn't say what firmware you are running.
>
> Ben
>
>
>
>