Does power(watts) affect listening ?

[thevortex]

Hi thevortex,

NOTHING WILL HAPPEN just play the music at the SPL u are comfortable with. U will 'hear' trouble approaching before anything happens to your speakers whether u keep that soft clipping engaged or off.
Rgds

Thanks dinyaar!

 

[Asit]

Thanks Flanker. The thing is I am getting a NAD C372 soon and these are going to be paired with EPOS M12.2's. The NAD is pretty powerful at 150W where the EPOS's are not rated for that power. Their maximum power handling is 120W. For that reason I thought that soft clipping might be useful in my setup. Any thoughts?

I am not a person who pushes the system very loud normally. So, I think it ought to be OK. But still would like to double check to make sure that the speakers are not damaged due to being paired with a powerful amp.

I do not see how soft clipping is going to help you, because that is going to take place at the limits of your amp and that is higher than the rated power handling of your speakers.

Long before that you will do some serious damage to your ears. But I do not think you should be worried about it at all, if you keep your amp at any reasonable volume level. My quad 99pre/909power audition was done with the Epos M12.2, the same speakers as yours. The Quad 909 is capable of delivering 150wpc@8 Ohm. The general volume level of that audition was very very loud, and at times the person at the store increased the volume even higher, I guess, to show me the muscles of the amp. I could not take it in that well-damped auditioning room of SKS Traders Kolkata and had to come out of the room. So I think it would be okay. The only danger is when it is operated by a kid (for example) in your absence, and if he just playfully increased the volume to insane level for some time, the speaker will eventually give up. This is just my impression. Need confirmation from people.

This style of pairing of a speaker and an amp where the safety rating of the speaker is significantly below the continuously deliverable rms power of the amp is a relatively modern phenomenon, I think. When I bought the core of my current system in 1989, people actually suggested me otherwise. For example, my salesman in Germany actually recommended the kind of pairing I ultimately ended up buying: HK 6300 amp at 60 watts of rms power and the speaker safety limits were 100/150 watts (nominal/max). At that time most people told me that was the ideal kind of pairing. Does that make sense, fellows?

 

[dinyaar]

Hi Guys,
Am sorry i did not include the kids factor here. Kids can cause a speaker failure and so do some crazy teens.

Generally the speaker does not just blow and as mentioned by quite a few on various threads u will hear the speaker really straining before it blows. U will immediately turn down the volume as soon as u hear the first signs of this. I dont really know how much this soft clipping can really do in any case. Maybe someone (i mean cranky) can explain in detail.
Personally i have used a 100 wpc/200 WPC/300 wpc/600 wpc on my 805s (rated at 120W unclipped max) and there were no issues. But I did warn my wife when the 4BSST/7BST was around to be careful when she listens in my absence.
Asit the salesperson must have recommended the combo keeping normal listening SPL /the speakers sensitivity and the amps drive in mind. Secondly this was the conservative norm till the 90s and have seen/heard of many similarly mated combos.
Rgds

 

[thevortex]

Sounds logical, Asit and Dinyaar. I am not going to use soft clipping when I get my NAD.

The other reason I am anxious is that I have not been a great fan of the NAD sound signature. Got this NAD C372 because there was a deal going on and I wanted a stereo amp. Not the best way to procure audio equipment I know. But this amp comes very highly recommended - reviews are very good as far as I have read. So, lets see. Crossing fingers here.

What I have thought of and also have received recommendations for is to even use the NAD as a poweramp if I happen to absolutely abhor it for music. Hopefully that will not be the case.

 

[Asit]

Cranky, you have posted some of that stuff in my amp thread.

I raised the possible dangers on the speakers by unattended use by kids (teens included), but you have brought forth the more important issue of damage to the ears of the kids (small ones, most teenagers are expected to have the presence of mind to pull back the volume when it starts hurting their ears). Thanks.

You have also nicely explained the two different design philosophies: one with low linearity, highly sensitive speakers (to be used with relatively lower powered amps) and the other with lineraity, but low sensitivity.

I suppose, in the above, by linearity, you mean a flat impedance vs frequency curve. If it's something else, please point it out.

It's true a majority of mainstream speakers we see in the market today has sensitivity of 85 -89 db. There are a few in the low nineties, for example the Def Techs.

There are only a few brand-named home speakers that have about 100db. For example, the Zu Druid speaker and Emarald Physics. There may be a few others I cannot remember at the moment. I do not know if this is the right thread to discuss this, but I want to know if these speakers compromise seriously on the above-mentioned linearity to produce the high sensitivity. Now, if these are supposed to be driven by SET amps or similarly low powered amps, and I understood what linearity is, then what really is the advantage, because the amp will have a hard time coping with large impedance slopes across a small change of frequency. Cannot there be a compromise somewhere around 88 - 92 db region (doesn't need to be there, just a guess) where one gets the best of both worlds?

I have seen the Harbeth designer and owner proudly stating that for the last 30 years his speakers have had a sensitivity of around 86 db, but they are easy to drive, and actually that is true from user reports. Similar things I hear about the Usher tiny dancers, you have them.

One more question: Speakers in public address systems or to be used in larger places like big auditoriums have very high sensitivity of above 105db or higher. Now does that mean they compromise seriously on lineraity. That would be really scary then. And they would really need some monster amps to drive them despite a high sensitivity. Am I confusing things unnecessarily or making extensions of your statements to a domain where they are not applicable?

 

[ajinkya]

Linearity = absence of distortion.

There are essentially two types of distortion in a speaker system. One is in the frequency domain, the other in the time domain. It is arguable which one is more offensive, as the ear is less sensitive to frequency variations than time (localisation of sound source, for example, is a function of timing discrimination).

Anyway the fact is that high linearity (at least in the frequency domain) requires higher self-damping,

Cranky,
Sorry but I could not understand the meaning of either linearity or distortion from the above post. Is it possible to explain with some simple examples, since I don't see the relations between all the terms you speak of.

Thanks,
Jinx.

 

[Asit]

Okay, now I have another question.

With SS amps, we know the THD is given by something like 0.0x%, for tube amps it is considerably higher (not necessarily bad), and I understand it from the different patterns of distortion in the two cases.

What sort of number for distortion is good for speakers? What ballpark region? Unfortunately, this is NOT a spec that is usually quoted in technical specs for speakers. It may be a difficult number to quote, firstly there are different kinds of distortion, and secondly I am sure it will depend on a few things like the power delivered from amp and frequencies, so to quote meaningfully a single number for distortion may be difficult. I am just thinking aloud and of course speculating.

However, in case of my very old Canton speakers, the user manual (see my first post in my amp thread) actually quotes a number for the distortion, and it claims to be less than 0.7%.

My first question is: Does this spec, the way it is quoted (for my Cantons), without any further reference to anything else (like power, db etc), make sense?

And the obvious next one is: If it makes some sense, how is that number "less than o.7%"? Is that good, mediocre or bad?

 

[ajinkya]

for electrical/electronics and audio engineering we have adopted 'distortion' as a generic term that indicates how much a component deviates from its intended function, which is supposed to be linear.

In a speaker, there can either be a delay between the application of the input and the acoustic output due to mechanical reasons (this is time domain distortion), or there may be irregularities between the applied electrical signal and the acoustic output. This is frequency domain distortion.

Does that simplify the explanation?

Cranky,

Thanks, it's a little more understandable now
So by 'linear' you mean the typical Linear Time-Invariant (LTI) systems in textbooks. That I understand and it's a good property to have in a system.

However, I still don't understand the 'time distortion' definition. If there is a delay between input and output which is constant, we still get an LTI system (because of the time-invariance property). Are you talking about non-constant delays between input signal and cone movement?

Regarding 'freq distortion', is that just distortion due to the non-linear behaviour of the system? I remember triodes (and valves in general) have a very small linear region, which can go into non-linearity over time because of thermal changes. When that happens, higher-order terms enter the input-output relation and cause nonlinear behaviour. Is this what you meant by 'freq distortion'?

Thanks in advance for the explanation,
-Ajinkya.

 

[Asit]

However, I still don't understand the 'time distortion' definition. If there is a delay between input and output which is constant, we still get an LTI system (because of the time-invariance property). Are you talking about non-constant delays between input signal and cone movement?

It has to be non-constant time delay of some sort, I guess. If it is constant, then the listener would not detect it, because the listener would get the whole sound time-delayed. It is just like: if in an infinite space dimension, every physical object is translated by a constant amount, nothing would change, nobody would know (this is called space-translation invariance). However, I am wondering what this time-delay may be changing with. Again, my guess would be with frequency. That means different frequencies would get delayed by a different amount and at the listener's end, the sound will seem a bit incoherent. Does that make sense, guys?

 

[gobble]

Sounds plausible. Its quite possible a sine wave will change as it is delayed in traversing a circuit - a phase change by a few degrees?

Regards

 

[brejeshvarma]

Dear Cranky,

I cannot help writing this.The way in which you put things together is so beautiful.You make complex things much more clear.You must be a good teacher by profession !
Thanks a lot.

 

[marsilians]

I second Brajesh for the clarity in Cranky's prose. Having said that, the following is a really interesting article a friend of mine shared with me. though it talks about distortion wrt amplifiers, the concepts are fundamental to any components. definitely worth a read IMHO.

http://www.passlabs.com/pdf/articles/distortion_and_feedback.pdf

PS. Nelson Pass is a respected name in audio, for those not familiar with that name.

For the inquisitive, worthwhile to check out the following as well:

1. Parts distortion - due to components
2. Intermodulation distortion
3. Effects (speed) distortion

Each of this affects the signal along the path and its amazing despite all this we still love to listen to music (and people). Our ears (and brain) are very adjusting against many odds.

 

[Asit]

Cranky, thanks for the efforts. It's one thing to know, it's another to put some effort in to make that knowledge available to others, that too for free. :clapping:

But, you still have not answered some of my queries listed in post #34. In your free time, please do. It was a few posts back, may be it escaped your notice.

 

[Asit]

Cranky, you were worried about having a discussion too far outside the main focus of the thread. Well, towards the end of this post, I shall try to bring back the amp and raise a point that would be directly connected to this main theme, i.e., power affecting listening.

Thanks a ton. Because of you and encouragement from others in these queries, I think I am finally getting some grip on the subject of speaker distortions, others are also in the same boat, I would guess.

To summarize, the frequency domain distortion is very similar to that in an amplifier and therefore also measured using similar parameters, i.e., relative presence of harmonics, especially the second and the third one. Although you have not explicitly mentioned in your post above, the graph presented above is an example of frequency-domain distortion of the mini-dancers (not your tiny dancers).

Now coming to the time-delay related speaker distortions, is there any connection between your statement that this type of distortion is roughly related to soundstaging and imaging, and another statement in my amp thread that at the power-limits of an amp, there is collapsing of soundstage?
It could very well be that your two statements are independent causes of collapsing of soundstage and not related at all.

 

[R Narayanan]

Thanx venkatcr. Your explanation on speakers (sensitivity, power and impedance) is excellent and easy to understand for "aam aadmi's" like me.

 

[Nikhil]

Wow thanks for digging up this thread! Lots of very useful info on here ...