milesp
milesp's picture
Offline
Last seen: Never ago
Joined: Mar 24 2010 - 10:44pm
How bad for your speakers is source clipping?
Lamont Sanford
Lamont Sanford's picture
Offline
Last seen: 1 year 1 month ago
Joined: Mar 31 2006 - 8:32pm

It could damage your speakers. But you probably already know that. One the others will come on here and give you the mechanics on how it will damage your speakers.

Demondog
Demondog's picture
Offline
Last seen: 6 years 11 months ago
Joined: Feb 22 2009 - 5:01pm

Good question. My own guess would be that clipped source material would consist of much less energy than typical amplifier clipping. I can't back that up though. I think that the degree of source clipping would make all the difference, since it seems that many recordings are clipped these days.

dbowker
dbowker's picture
Offline
Last seen: 2 weeks 6 days ago
Joined: May 8 2007 - 6:37am

I think a clipped "source" is more about dynamic compression- where maybe it's all loud- or not I guess, depending.

In any case, I think it's not related at all to amp clipping, which I believe is when it can't handle the signal anymore and you get massive distortion running into your speakers- usually heard as pops or drop outs and is very bad for all concerned. Just don't turn your amp up to "11" and you'll be OK.

cyclebrain
cyclebrain's picture
Offline
Last seen: 2 months 1 week ago
Joined: Jun 16 2006 - 11:40pm

Not really sure what is meant by source clipping but a clipped signal has a lot of very high frequency energy in it. Energy above the original audio bandwidth. If you have clipping from a digital source it will not pass the high frequency effects of the clipping because of the limitations of bandwidth being limited to half the sample rate. Also any bandwidth limitations of the preamp/amp will remove the high frequency content of clipping. Amplifier clipping will pass clipping energy directly to the speaker.

Orb
Orb's picture
Offline
Last seen: Never ago
Joined: May 28 2009 - 12:51am

Yeah amp clipping and digital clipping are two different things IMO.
If a digital source (music) is clipping it can still be played with low level power from an amp, what you hear is distortion that can sound crappy (bear in mind some sounds are deliberately distorted such as those from guitars and some synth sounds that sound interesting when applied in the right musical context).
So while the music is clipping (should be very rare due to studio using compressors-limiter), the amp can be playing this at any level within its limits and is not the same as the amp clipping as the distortion is low energy (watts) as far as the speakers are concerned.
However one thing comes to mind and I am not sure if digital source clipping is actually bad for digital amps as the volume/etc are all also digital, so these may require some sort of designed protection (this may or may not also apply to PWM Class D amps but no idea).
Anyone able to expand on it from the perspective of digital and PWM Class D amps?

Anyway, the clipping that IMO is dangerous is the amp clipping, which is caused by the amp reaching the maximum limits of its voltage rail.
This is bad because you get distortion, but more importantly the power output is much higher than that of a sinewave.
The reason being a square wave is at the maximum for the whole cycle, while a sinewave only has a brief peak at its maximum.
So I think its the sustained power boost caused by the square wave that is the problem, combined with distortion to create the worst case scenario of sustained high energy and distortion.

Oh now what is interesting to consider, how do digital equipment (DACs/filters) handle 0dB FS and beyond sinewaves.
Not all digital equipment are equal in how they even handle the maximum possible loudness (full scale) sinewave.

This becomes interestingly because the poorer modern recordings associated to the Loudness Wars may be hitting the full scale often by using hardware and software that enables recordings to be made much louder/hotter by the use of limiters to reduce clipping dynamics to 0dBFS.

Two good articles to show how dBFS considerations apply to the poor modern recordings, and real life example on the behaviour difference on digital sources resolving this.
Bear in mind 0dBFS is not technically clipping but just on the verge.

Please read both as they link well together.
http://mastering-media.blogspot.com/2008/10/diy-mastering-part-5-how-loud-is-too.html
And
http://www.audioholics.com/education/aud...layback-systems
Oh the following from wiki gives a brief comment on limiters/dynamic range compressors that can be used to create poorly recorded/mastered albums so they suffer the "Loudness war" without technically clipping.
http://en.wikipedia.org/wiki/Dynamic_range_compression
For a visual the wiki link I provided also shows an example:
http://en.wikipedia.org/wiki/File:Clipping_compared_to_limiting.svg
Cheers
Orb

lionelag
lionelag's picture
Offline
Last seen: 4 years 2 months ago
Joined: Apr 18 2007 - 9:40am

Digital clipping at the source is not nearly as bad as overloading a preamp or amp and then putting that into the speakers. You're playing back a reproduction of a distortion, not abusing your equipment to create a new one.

I'm curious to know what recordings you're talking about, though. (The worst example of digital clipping in a commercial release that I know about is the remaster of Iggy and the Stooges album "Raw Power," where they stupidly let Iggy Pop loose at the mixing board during "Search and Destroy." The digital distortion is painful, and so bad that ten years on, they're reissuing the original, non clipped mix. Yes, even people who like punk rock have ears. ) Some of the recent Hendrix releases (esp. Electric Ladyland) come close, but I can't think of anything I've heard, in any genre, other than the aforementioned Stooges recording, which actually crosses the 0dbs line in digital.

bertdw
bertdw's picture
Offline
Last seen: 2 months 2 weeks ago
Joined: Sep 18 2007 - 5:41am

Metallica's "Death Magnetic" is a recent example.

http://mastering-media.blogspot.com/2008/09/metallica-death-magnetic-clipping.html

Orb
Orb's picture
Offline
Last seen: Never ago
Joined: May 28 2009 - 12:51am

Yeah agree.

Take a look at the 1st link I provided, those dBFs figures are averages, with peaks hitting 0dBFS and also clipping in some cases.
Ian also gives a summary next to those records, of course there are plenty more examples out there.
Especially if you also consider recordings where limiting is used and they hit 0dBFs often, which as I mentioned is handled differently between various CD players/DACs.

Edit:
Nice analysis of the Death Magnetic album:
http://mastering-media.blogspot.com/2008/09/metallica-death-magnetic-clipping.html
Cheers
Orb

Elk
Elk's picture
Offline
Last seen: 3 years 7 months ago
Joined: Dec 26 2006 - 6:32am

If the top of the waveform is truly chopped off it will be akin to a square wave with lots of random high frequency energy not present in the original. Played loudly enough I imagine this could damage teeters.

Orb is absolutely correct that some DACs/players do not handle 0dBFS well and generate intersample distortion exceeding 0dBFS. While this doesn't sound good, it isn't going to hurt anything however.

The level of compression and limiting used on the Metallica album is mind boggling. All compression is distortion of a kind, but can be a useful tool. At this level everyone can hear what happened to the sound.

The only question is whether one finds this desirable. An amazing number like compression on rock and pop recordings.

Orb
Orb's picture
Offline
Last seen: Never ago
Joined: May 28 2009 - 12:51am


Quote:
If the top of the waveform is truly chopped off it will be akin to a square wave with lots of random high frequency energy not present in the original. Played loudly enough I imagine this could damage teeters....Snip

Yeah there is something in that and I was keeping it a bit too simple, the key as you say is energy combined with the square wave/distortion.
I tend to agree that even without the amp voltage rail clipping, if the music is played loud enough (so enough energy-watts at the speaker) a clipped song due to recording could possibly damage speakers.

But then we also need to consider that an amp voltage rail clipping is more prolonged due to it presenting a square wave even for possibly good recordings along with the bad ones.
So again I feel it may be pretty rare for a recording to do damage (unless the speaker is really underspecc'd for the watts and room size and the amp is more than able to drive to loud levels), say compared to an amp clipping.

Heh just thought of it, many mastering/recording studios are close to 80db (either a bit above or below) in average listening ignoring peaks.
A real life consideration then is; how often do the studios need to replace their speakers due to source clipping for those crap album recordings they do?
But lets not forget I bet most utilise some sort of limiter/compressor to stop excessive clipping.
Would be interesting to know, anyone aware of cases where it is not used?

Cheers
Orb

Elk
Elk's picture
Offline
Last seen: 3 years 7 months ago
Joined: Dec 26 2006 - 6:32am

Orb, I think you are completely correct. A clipping amp has got to be more damaging. The OP's question is great as it is making us think.


Quote:
A real life consideration then is; how often do the studios need to replace their speakers due to source clipping for those crap album recordings they do?

Good one.

I know of some that replace tweeters on occasion but my impression this is primarily on the "impress the client" speakers which are played LOUD.

The signal chain is often like we are accustomed to, such as Pass amps into B&Ws.

cyclebrain
cyclebrain's picture
Offline
Last seen: 2 months 1 week ago
Joined: Jun 16 2006 - 11:40pm

Since the original post refered to digital sources, I will also post regarding digital sources. Yes, a clipped signal creates multiple high frequency harmonics. But any digital signal is limited in its high frequency content by the sample rate of the data. One half of the sample rate. So even if there is much high frequency energy created by clipping of the source data it will be removed by the limitations created by the sample rate.

Lamont Sanford
Lamont Sanford's picture
Offline
Last seen: 1 year 1 month ago
Joined: Mar 31 2006 - 8:32pm

I don't think it matters what the frequencies, sample rates, and so forth might be for the original poster. The question referred to what sort of damage will it cause to components. Specifically, speakers. Let's just say at a listening sound level of 90dB and a typical rated 100 Watt amplifier and speakers a pair of 3 way box rated at 100 Watts. A hypothetical set up for reference.

struts
struts's picture
Offline
Last seen: 2 years 10 months ago
Joined: Feb 1 2007 - 12:02pm

An interesting question. I think he bottom line is that playing clipped programme is most unlikely to damage your speakers, unless you play at crazy high volume levels in which case you risk driving the amp into clipping anyway.

As to whether tyring to reproduce clipped waveforms stresses the drive units in some minute way that might result in cumulative damage over time I can't say for certain, but somehow I doubt it.

Jim Tavegia
Jim Tavegia's picture
Offline
Last seen: Never ago
Joined: Sep 1 2005 - 4:27pm

I tend to agree with DBowker on this one. Limiting is used on just about all of the discs we own in some form or another. It is mostly your amp and how much headroom you have at the levels you like to listen to. We have no way of knowing how many "overs" we have on the discs we all own.

This does raise an intersting point about digital "overs" and what effect they can have on digital source components and outboard DACs as they try and output these distorted wave forms.

JA could chime in here and let us know of the sonic ramifications of audio past 0db. We can see how well just about all of the currnet crop of digital sources can handle square waves, as most are near perfect in JA tests. it might be nice to see what happens to that same square wave at +3 or +6 DB once you've run out of bits to cover it.

Could it be possible to damage the inputs of your amp or preamp? Probably not, but you do have me curious.

Freako
Freako's picture
Offline
Last seen: 2 years 11 months ago
Joined: Jan 17 2010 - 8:29am

About amp clipping: This is rarely dangerous for the speakers. Let's imagine a setup with speakers spec'ed for 150 watts and a transistor amp with 50 watts per channel. This could easily be a dangerous scenario for the speakers if playing very loud. But if the amp was a 250 watts model, the speakers could be driven much closer to their limit, and sometimes even beyond it. The dangerous clipping mostly occurs with underspec'ed amps.

About digital clipping: I have no clue...

cyclebrain
cyclebrain's picture
Offline
Last seen: 2 months 1 week ago
Joined: Jun 16 2006 - 11:40pm

Why doesn't JA respond? Maybe because despite all of your well intentioned posts this thread is making no sense. I must back out now and allow you to continue with your thoughts.

Glotz
Glotz's picture
Offline
Last seen: 13 hours 16 min ago
Joined: Nov 20 2008 - 9:30am

I think the point here was to say, yes, both forms of clipping at the source OR the amp would be deadly to your speakers, in certain situations. An amp clipping is exponentially worse, but source clipping and re-amplfying it very loudly is not good practice for your speakers OR your ears, especially if you are pushing either amp or speaker to the limits.

Simply put, if you know that a power amp clipping is producing a great deal of audible distortion, it could wreck a driver at a high enough volume level.

The same applies to source clipping to a lesser degree if driving distortion at a high enough level. I mean, if you're driving distortion that is being re-amplified again, and even if the extent of the distortion isn't causing the amp to clip, it is asking the amp to to 'unnatural' things, and it can be a recipe for disaster if you play it too loud. Knowing your equipment's limits (esp. speakers) is critical to long life.

And honestly, I believe the same applies to the source MUSIC to even a lesser (but still valid) extent as well, if you are driving metal or heavy rock, as you are pushing your equipment to the level in trying to recreate the glory. The combination of distortion created at the source only gets much worse as you amplify and re-amplify said crap. Any significant combination of high-level distortion and equipment that can't take the punishment equals blown drivers.

I've seen JA's recent, very astute responses in the magazine regarding his hatred of compression, and I think there is a video recently I've seen of his discussion. It's sad affair, but time will pass when the tenets of good hi-fi and sound engineering practices will reinvent themselves, as many artists decide to work and distribute with greater control of what they are producing.

Moreover, there are still a lot of labels that are doing smart work, or at least artists that have greater control over their sonic destiny than before. Perhaps the label is willing to bend as they are smaller or newer and trying to earn new clients. One thing is for sure, there are myriad more record labels than ever before, of all different ilks.

Log in or register to post comments
-->
  • X