How important is S/N Ratio....?


Over the years I have owned many amps....rated at different S/N ratios. As a example 80,90,100 or 120 db down....but some sound more transparent than others regardless.Also the higher the rating theoreticly is should sound better....right? Does your amp sound better than the advertised rating? If it does.....tell me 3 things that stand out about your amp.

wavetrader
If you listen to loud-all-the-time rock music S/N is of no concern. For quiet classical music 90 dB is OK. 100 dB or more is for neurotic audiophiles who put their ear to the speaker when no music is playing. (By the way, most of my stuff is spec'd at 100 dB or more. It's no big deal).
Be careful as there is no defined way to measure this - it is the noise with respect to a "reference level". If you stick to pro audio gear than the S/N ratio on balanced outputs is usually referenced to 1.23 Volts.

If you use consumer RCA then S/N is often referenced to 0.316 volts or whatever the manufacturer "decides" (4 or more times less stringent)!

So the S/N specification on pro gear is significantly better than the same spec on consumer gear and on consumer gear it is misleading to compare specifications anyway!

Oh and by the way - the reason gear may sound different is that you can actually hear about 20 - to 30 db BELOW a noise floor (yes our hearing is quite good). So what matters more is the type of noise...if it is just thermal background white noise that is totally unrelated to the audio signal then you have a great piece of gear. If you have equivalently low levels of noise but when music is playing you have within it some correlated noise that follows with the music (harmonically or IMD related) then this gear will sound much worse (even if they both share the same "on paper" S/N ratio)

This is why gear sounds different and some gear has a blacker background where subtleties stand out a wee bit more. Remember our ears are like harmonic analyzers - they pick up vibrations that are cyclical in nature... so white noise is much more benign than correlated noise which has a specific frequency content and which changes in response to the music signal.
if it is just thermal background white noise that is totally unrelated to the audio signal then you have a great piece of gear

I thought I would add that this is why analog tape hiss or Vinyl road noise (apart from wow and flutter) is not a big deal - we hear the music through this provided it is random noise.
" If you have equivalently low levels of noise but when music is playing you have within it some correlated noise that follows with the music (harmonically or IMD related) then this gear will sound much worse (even if they both share the same "on paper" S/N ratio)"

So what your saying is distortion effects the S/N or not...when measured.

"if it is just thermal background white noise that is totally unrelated to the audio signal then you have a great piece of gear"

Ok so more inner detail...subtleties....Clarity......Focus

But let's ponder this.....if distortion effects what we hear......wouldn't the S/N....true measurement...measure the distortion present in the amplifier. So the higher the S/N the lower level of distortion present in the device..amplifier.

That would in theory effect....soundstage,imaging percieved and also bandwith of the amplifier.

Also the componet with the poorest S/N would be the weakest link in the chain...hmmmm.
Specs are near-meaningless. At best, advisory.
Don't worry about them. Would you junk out an amp you liked with a 90db snr and replace it with an amp you Didn't like with a 110db snr? Didn't think so.

from wavetrader:
If you have equivalently low levels of noise but when music is playing you have within it some correlated noise that follows with the music (harmonically or IMD related) then this gear will sound much worse (even if they both share the same "on paper" S/N ratio)"

Just what does above quote mean in standard English?
But let's ponder this.....if distortion effects what we hear......wouldn't the S/N....true measurement...measure the distortion present in the amplifier.

What I am saying is that S/N on its own is pretty meaningless(unless it is really bad and then it is a warning flag). It is just the background noise level present. This is why many specs are given as THD + N...
Just what does above quote mean in standard English?

I tried to say that noise at specific frequencies and that may be modulated by the music signal is far worse then just background wide spectrum random hiss...

This is why tape and vinyl works well even though it has a high level of surface or background noise.

Sorry for the bad english - I type fast and rarely check.
So the consensus so far is......

Signal to Noise ratio.....is meaningless...inconsequetial to the sound of your amplifier...

Sorry I don't buy it...true there might be ten ways to measure it....so confusion abounds....but when a lower S/N level can be quantified.....it has to have a relation to actual levels of distortion present....

Wether the cart goes before the horse or viceversa.....I hear a difference especially in the Amplifier componet.
hmmmm... Well building an amplifier with exceptional Signal to Noise ratio will mean super stiff power supplies and better components. So I expect the sound will be much better in such a component.
PS. I can't hear any noise in my system when I press my ears up to the speakers between track change. So I guess this is good news. But I use active ATC speakers with Bel Canto DAC3 with everything connected to a PS audio PPP so that might be expected.
but when a lower S/N level can be quantified.....it has to have a relation to actual levels of distortion present....

Agreed but I think several replies here are simply trying to say that the link is not as strong as you suggest. Otherwise people with analog gear (vinyl or tape) would be very unhappy - and they are not! Distortion and Signal to Noise can be separate issues and don't have to be related. You can have one bad and the other good and vice versa. Of course in excellent gear both will be good.
Shadorne.....

"Agreed but I think several replies here are simply trying to say that the link is not as strong as you suggest. Otherwise people with analog gear (vinyl or tape) would be very unhappy - and they are not! Distortion and Signal to Noise can be separate issues and don't have to be related. You can have one bad and the other good and vice versa. Of course in excellent gear both will be good."

Interesting as tape has a very large bandwith.....I really can't say about vinyl as I moved to digital source long ago and have never thought about going back......

Of course I am talking about the high upper end here....as the build quality of the componets and circuit design...seperate the levels of sound performance.

Well let me ask you then......what are 3 qualities of a amplifier that combines both excellant thd and S/N levels...
Well noise in coming from Vinyl is a part of the playback which I think is totally different from the noise generated from an amplifier. Noise coming from the playback of analog sources is inherent to the technology itself, where physical contact is required. However, the noise from an amplifier is more like a flawed power supply and interference issue.
I know I am intruding to a 2 person discussion here but theories do suggest that an amplifier with low noise means a good power supply and a product of a good circuit designer. Low distortion generally means an amplifier that gives less fatigue to the listener. Of course different distortions contribute to a different sound. Since these topics have long been covered and tested before why not let's move on to an issue on the effect of amplifier damping on a loudspeaker because it seems that there is a trend of going for a very high damping in the amplifier industry of late, and with good results.

Long time audiophile magazine reader here.
Best
Alexander
Damping factor has been discussed many times on these forums and it is also misleading to pursue this number. I maintain that above a damping factor of 50 then you gain next to nothing. In fact, extremely high damping factors may be indicative of very high amounts of feedback and just like an amplifier with a specification from 10 Hz to 400 Khz I am wary of these kind of amps (Why do I need to amplify signals out to 400 Khz - surely it is better to get rid of this unwanted stuff). To me - there is danger in chasing excessive specifications which do not add anything to the sound quality but run the risk of making the amplifier less stable under difficult operating conditions....just two cents.

Oh and three things to look for in an amplifier:

1) More Power

2) More Power

3) More Power

...forgive me a cruel chuckle and a wink to Mini-Me!
Alexander....join the party....this seems to be not such a hot topic.....but let me put this in terms of my real experience..

After years of spending thousands on cables and other tweaks....to improve the noise floor....and it did remove grunge....I started replacing capacitors in my componets.

The preamp was first....major difference....then the coupling caps in my dac....again increased clarity,transparency....I was on to something here because these 2 caps were low noise and frankly the change was indeed stunning...Next was the amplifier...

Recapping and changing the 2 power supply caps was so much better...not incramentel but I would say a quantum leap...now people talk about live....well in theory a microphone should pick up everything within it's range. Sound,reverb,decay,air movement and changes in pressure,ect. Well in the best digital recordings for once I feel like I am hearing everything that a mike can pick up.

Now just changing capacitors...why that much difference..The tech sheet says reduces distortion by a 1000 pct and has a noise level of -135 db down or more....these are super low noise....and they are because I can hear it.

I won't go into the what equipment or circuits....you may or may not get this kinda of sound quality.....I just got lucky and upgraded some equipment that I had for years.

But the most improvement I heard was in the amplifier....in every aspect....you name it....everything.

Like I said I dont know if THD is the cart or the noise is the horse.....but the capacitors are "super low noise".
Now we are talking. If you talk high quality components then yes there is a difference. This is much easier to appreciate than "specifications" in a brochure.

Eliminating transformers from the signal path (often non linear). Installing better quality capacitors - absolutely important if poor quality ones were used. (also bear in mind that some types of capacitors age and can dry out - so in this case it can make a HUGE/NIGHT AND DAY difference on a piece of gear that is ten+ years old) High quality AIR core inductors - absolutely. Resistors are usually less of a problem but wirewounds are recommended in high power applications.

Also just as a general rule - eliminating capacitors/resistors/inductors from high power paths (such as in crossovers) also helps a lot.

I have no problem agreeing with that - 100%
"I maintain that above a damping factor of 50 then you gain next to nothing. In fact, extremely high damping factors may be indicative of very high amounts of feedback and just like an amplifier with a specification from 10 Hz to 400 Khz I am wary of these kind of amps"

Shadorne - I agree 100%.

Sane designer would try to design an amp with not more than few percent THD before feedback and then improve it using feedback to only about 0.1%. Finaly he would limit bandwith to what amplifier had before feedback was applied to prevent TIM. 100kHz bandwith as a result might be nice to prevent phase shifts but 400kHz is unnecessary and might be harmful.

Damping Factor is often limited to about 100 by the choke in series with the woofer. Also DF measured in static condition doesn't mean much. Transient response of power supply (due to usage of good capacitors like slit foil etc) might be more important.

As for "More Power" I would suggest "Less Power". On average 200W amp doesn't sound as good as 100W amp that cost the same and is only slightly louder (it needs 10x power for 2x loudness).
It's pretty easy to agree....on this once I clarify...now we can get our hands around this...

I must say just recapping wasn't actually my point....It was the quantum leap when using these ultra low noise devices...
135db and I can hear it!!!!!
Human hearing dynamic range is somewhat lower than that...
A sound at 135db would burst your ear drums or render you incapable of appreciating hi-fi.
Therefore, a sound '135db' down is either....So far below hearing limits as to be nearly, if not completely unmeasureable OR referenced to a sound so loud that it would Hospitalize you.
I am not even going to go to the 'Weakest Link' end of town where who cares if your caps are 135db, when your output devices and the REST of the circuitry are no better than 100db?
I repeat:: Specs are nearly meaningless. If you purchase something based on specs rather than quality, you may be surprised. It could go either way!
Magfan - large dynamic range of 120dB or more is absolutely necessary to reproduce real sound of the large jet planes taking off when I see it on TV.
As for "More Power" I would suggest "Less Power".

Au contraire baby, More Power is More Mojo
-135 db down is meaningless only when....there is no sonic difference discernable....but when I add two low noise PSU capacitors and 10 ultra low noise caps to a SS amplifier and the sound is unmeasurably better...I guess the poor guys that spend 30K for a amplifier can't tell the difference either...BTW the upgrade cost me 750 bucks....including labor.
Wavetrader,
I will love to upgrade the components n my components too, but have you ever wondered whether it might be detrimental to the flow of current when you use different solders to what was originally used on the circuit boards? Like the interactions between different alloys and the also the matter of whether the components will be mounted as stably as the ones originally supplied. Nowadays it's all surface mount. These are the main concerns that are preventing me to do such risky modifications. But if the amplifier is made by hand when purchased, I won't have any problems with that.
Wonderful.....I am no tech...but this amp has pull out CB's so the recapping was not that involved...the thing that was a challenge was fitting the low esr power caps. The solder issue you mention is better left to those that make their living...doing such things.

I will have a matching pre-amp with these ultra low noise caps vs a unit with some Nichicon gold installed.....if I get a quantum leap with the ULN type.....well it will be interesting.
Wavetrader - if you do this yourself be sure you know what you're doing. Installing low ESR caps with certain voltage regulators might cause oscillation.
Kijanki.....thanks for the warning.....I am lucky as I have one of the best techs in the country 10 minutes away.
Kijanki.....Let me clear that up....my tech Fred does all this work for me.....We have been friends since 1975. Fred has been doing this professionally for 40 years.
I found it interesting at RMAF that Roger Sanders insisted that more power was better. He was biamping his electrostatic speakers with 600 (top) and 1000 (bottom) watt amps. His reasoning had to do with headroom that would improve dynamics.

Not sure I agree fully, but thought I'd mention it in light of the comments here.
http://www.cs.cf.ac.uk/Dave/Multimedia/node271.html

This article claims 96db range....from dead quiet to loudest.
other articles claim up to 120db. 'Pain Threshold' is usually quoted at about 135db. I don't know what OSHA says, but I'll go out on a limb here and say that an accurately reproduced Jet takeoff at 120db will, if repeated monthly, take the edge off your HF hearing within a year or less.
At that PEAK level, the sound of a car door slamming or a gunshot would be enough to hear next door.
At this point, 'ya gotta wonder if your speakers are up to it! A 120db slam is pretty ambitious and even if your speakers CAN do that (What kinds CAN do that?) I rather suspect you would need your own substation to power the multiple kilowatts of amp necessary to keep up.
I haven't done the math, but you'd need speaker of fairly high sensitivity and power handling capacity with huge dynamic limits.
I tend to agree with those saying damping factor is not that important after a certain minimal value, but that bein said, nobody has mentioned the other half of the equation, that being speaker 'Q'. A critically damped speaker (Q=.707) coupled with a reasonable damping factor may actually produce thin bass.....Is this the opposite of what the fans of 'bloom' hint at? I'm not clear on that point.

Anyway, good luck, all you fans of accurate jet take offs and I'll be sure to buy stock in a hearing aid company. soon.
There is no "standard" dB level for a jet plane. Some business jets are so quiet that we don't know they have arrived at our little local airport until they taxi up to the terminal. On the other hand, I have had occasion to work with an F15 fighter in a flight testing project. Let me tell you, a full afterburner takeoff from a couple of hundred feet beside the runway is an awesome experience. We were required to wear ear plugs, but even so it was loud. I believe that even one unprotected event would do damage.
Wavetrader - that's good!

Clio09 - more headroom is always better if you can afford it. If you have let say $1k to buy an amp would you be better with mediocre 1kW amp or great 100W amp. 1kW amp should be twice louder but not always. Amps are rated for nominal power and 100W amp might have very large peak power being designed by good conservative company. 1kW amp that cost the same as 100W one will have most likely worse parts since power supply, case and heatsinks have to be 10x larger and something has to make-up for their cost.

Much worse if you choose SS amp over tube amp that you like only because of 200W over 100W (and 200W tube is too expensive). Another example might be selling 50W class A amp to get 100W class AB amp etc. There is a very little difference in sound level between 100W and 200W.
You guys are funny....jet plane takeoffs...
Keep in mind I didn't start the Jet takeoff thing.
I lived near El Toro MCAS and heard F-4 takeoffs which even with them being as easy as possible...no afterburners or aggressive climbouts, you could hear them for several minutes echoing thru the canyons. I was at an airshow and witnessed a flyby of a Concord! Man, that thing was LOUD.

This whole thing needs to be thought thru more thoroughly.
How quiet is your room? 40db at 0-dark-30? Let's say you can play it pretty much as loud as you'd like...not likely but we can dream! Subtract 40 from 120 and you've got 80db range. This is more than Vinyl is capable of and stretches the practical (NOTE: practical, not measureable) CD limits.

Sorry, this is a non-starter and a fine example of wishful thinking and specsmanship taking over!

There are great numbers of Credible Systems which use hi sensitivity speakers coupled to low power amps. These systems en toto draw less than my bedside reading lamp.
Eldaford - Jet plane was a joke but I think you mixing landing and taking off. During landing they are very quiet.

I read once that 120dB was equivalent to standing next to jet engine during take-off (should I say running very fast). To visualize what 120dB level is try to imagine large GE Jet engine at full power in front of your door (neighbors might not care too much for it).
Just for those interested I found some info.....

"Ordinarily, when the electrons flow through common conductors such as an electrolyte, a manganese dioxide, an organic semiconductor and so on, non-linear distortions generate corresponding to each conductor (average -100db) as shown in Figure 1 which indicates distortion characteristics. However, in the case of Black Gate, the distortion showed such a surprising low value (-174db) as if no electrons flew. The distortion noise was almost zero. "

excerpt from this link

http://www.acoustic-dimension.com/blackgate/techEcap.htm
Wavetrader - as far as I know electrons don't flow. Electric current is an electromagnetic wave (while electrons move only about 1/2" per second) which is also difficult to explain using example od stacked balls since space between electrons is in order of 100000 time their size. One theory says that electrons emit photons but this doesn't add up if we assume that photons can move in any direction. It is probably better explained here:

http://www.geocities.com/natureoflight/pg4.html
Electron flow = current.....that's what I thought...
I have had occasion to work with an F15 fighter in a flight testing project. Let me tell you, a full afterburner takeoff from a couple of hundred feet beside the runway is an awesome experience.

I didn't know there were some aeronuts here on A'gon. This is the World's best aerobatic pilot of all time - Jim Leroy.

BTW - Topgun is pretty awesome on HT with the volume cranked. However flying these things and cranking those g's is not all that pleasant - believe me I have first hand experience being up there...
You guys make me laugh though. F15 taking off. 135 db spl. Too loud. That is nothing. You guys are just wussies. Here I am at the airfield with my 153 db SPL system - just kidding ;-)