Can "Digital" damage my gear?


Faulty (or malicious) CD/CD-R or digital download can contain large levels of high frequencies outside of hearing range - for instance 20kHz. Such frequency at full power will overheat tweeter and most likely damage output Zobel network in my class D amplifier. Stereophile Test CD contains such tracks and warns against playing them at full loudness. My DAC will mute anything other than valid S/PDIF but will play any frequency at any level.

What do you think? Can bad or malicious recording from CD/CD-R or server damage my speakers or amp. Is there a way to prevent it?
kijanki

Showing 9 responses by kijanki

Mlsstl - I don't expect noise/EMI pickup of full amplitude but circuit malfunction is always possible. I don't have control over that. Digital "malfunction" appears to be in addition.

Al - I wouldn't know how to start. Inspecting huge files with Sound Forge seems to be time consuming and would have to be done for every digital file. It would be great if somebody can write piece of software that automatically detects malicious footprint (like antivirus).
Eldartford - I have only one 0.5m XLR (locking) IC. My speaker cables have Cardas bar type (common knob) very secure connector on amp side while on speaker side I have very tight spade connection (I tend to overtight). No loose connection here.

Fast fuse (straight wire) wouldn't affect audio but would protect my tweeters. Amp is different story.
Mlsstl -

"I've got over 40,000 songs on my digital server. I have never come across a maliciously recorded digital song."

That is very good to hear - thank you.

"I'd be fascinated to hear of one blown by a malicious digital recording."

- that's exactly what people said about computer viruses at the beginning. Computer virus won't physically damage your hard disk. In worst case you have to reformat and reinstall software (there was a virus overheating Pentium processors). Music virus can damage very expensive gear.

I know it is not likely but many people move to servers and in my opinion its only matter of time. It doesn't have to be pirated CDR or downloaded music - it's enough to play internet soundbite or video using main speakers.

Software engineer, I work with, says that Fourier analysis of such size files will take forever. He participates in SETI program and does analysis of small files on his computer. Ideal would be small crude scanner program that looks for high amplitudes above 2kHz.

I hope I'm just paranoid since I'm planning eventually to move to server (already use DAC) and get much better speakers.
Al - that sounds doable but now I started worrying about viruses. Main household computers will become servers for music. It is already happening. I'm afraid that eventually somebody will get an idea to write virus that will destroy tweeters or amps (as they wrote virus overheating Pentium processor). Separate computer not connected to internet might be better.
Mstll - I wouldn't ask the question if I could give an example. In class D amp it will destroy amp - just look at any Icepower datasheet.

Separate computer is good idea if you can afford it. Computer infected by virus might play destructive file without you knowing it. They did it to overheat Pentium and I suspect it will be done for audio. Chances are small but repair cost is high.

As for viruses being serious issue from the beginning (being treated seriously) - you must be kidding? People didn't want to use antivirus programs (initially free) and many don't use it still. Firewall was unheard of concept (until 1988) and not present in Windows 3.1/95/98. First IBMs went on sale around 1981 but first Norton anivirus, as well as firewall appeared around 1992 because of damages that viruses did and not before (remember Lehigh or Jerusalem viruses affecting .com files or Brain virus affecting boot sector?). Networks might be different story but it took big losses to make people believe. Whole networks went down because of viruses before people got serious about it. I think that people got serious in 1999 when Melissa virus created "epidemic".

Sidssp - 25kHz at full blast will take tweeter fast (minutes) as well as my amp and I won't even hear that - sometimes it is the case when amp's output stage goes to high frequency oscillation (instability). That was main reason, other than clipping, for tweeter fuses that Eldartford mentioned.

Al - Remember what is at stake: my amp and the tweeters.

I do Windows at work (I refuse to put-up with Windows stupidity and clumsiness for free). Windows still have a lot of dangerous services enabled by default like Telenet or Remote Registry Access.

I enjoy my 1 year old MacMini but still remember years of being confused and scorned by Windows. Never again (for free).

Antivirus might be useless with new unknown virus. Mac does it much better than Windows executing applications in "sandboxes" etc. The only viruses I know for Mac are Windows applications' viruses.
"I have had some BIG pops on an LP, that I would worry more about then anything digital ?"

Can they damage amplifier?
Mlsstl - I asked because I checked data sheet for Icepower 200ASC used in my Rowland 102 and it shows that full power blast of 20kHz will damage amp (Zobel network) in 20ms and 20W oscillation will do the same in about 1s. This would imply I wouldn't even have time to react (and is inaudible).
Mlsstl - I'm probably paranoid but digital offers new way of destroying speakers and even amp in addition to good old ones. It just seems strange to me that one can have amp and speakers destroyed by just playing defective song.

Class D amps have output filter (Zobel network) designed to withstand full power at 20kHz for only 20ms. Normal music carry very little energy at high frequencies but modified digital file can. I suspect it might be only problem of class D amps since inductor, often in series with the output, in traditional class A or AB is more robust.

Avoiding internet downloads from questionable sides is definitely advisable. It is a little bit like virus except virus doesn't destroy expensive hardware (and virus can detected).