Do Tech-Specs Have Any Relation to Sound quality ?

[istari knight]

Okay, Ive been hunting around different manufacturers sites Enjoying the pics & reading the specs.......as you do

And wondered ifyou can get any idea of the products quality from the specs ?

I'm talking SNR here - i dont fully understand what it is but would like to find out....

Ive checked Copland , Primare ,Denon & a few others and all of their amplifiers have worse SNR to that of my Technics.....

A Denon pma-sa1 has a SNR of 105dB

A Copland cta-405 has an SNR of 90dB

& a Primare I30 has an SNR of 100dB

My Technics has a SNR of 124dB

(all SNR figures are Line level)

So Please can anyone explain what effect SNR has on sound quality ?

I'd just like to understand it fully so i know what I'm reading means

Many thanks , James.

 

[Davewhityetagain]

istari knight wrote:

Okay, Ive been hunting around different manufacturers sites Enjoying the pics & reading the specs.......as you do

And wondered ifyou can get any idea of the products quality from the specs ?

I'm talking SNR here - i dont fully understand what it is but would like to find out....

Ive checked Copland , Primare ,Denon & a few others and all of their amplifiers have worse SNR to that of my Technics.....

A Denon pma-sa1 has a SNR of 105dB

A Copland cta-405 has an SNR of 90dB

& a Primare I30 has an SNR of 100dB

My Technics has a SNR of 124dB

(all SNR figures are Line level)

So Please can anyone explain what effect SNR has on sound quality ?

I'd just like to understand it fully so i know what I'm reading means

Many thanks , James.

The problems with specs from company web sites is, you do not know how they measured it and what they did to get the result they want.

This is one area most of the time, group tests in mags is of use you find they print what the maker says and what they get in the lab

A good example is the Naim 5i cd player reported as 50 watts into 8 ohms by the company

more than one test report has been wrote up as 65 watts 8 ohms

going the other way I can think of an amplifer being sold as 200 watts yet being nearer 170 watts

tech specs are a starting point to be taken with a pinch of salt

 

[Sastusbulbas]

My Technics SU-MA10 is silent when on, at the tweeter there is no noise or hiss, it is very similar to the power amp you have specification wise.

My Classe is a little more noisy, and you can hear slight hiss at the tweeter, its S/N is around -100db- I think

The more sensitive the speaker the more noticable this is, when I was using the Nu-Vista 300 power amp I found the hiss too intrusive, its S/N ratio is around -88db with -113db claimed.

My Krell pre is around -100db and slightly better than my previous AR valve pre. (I was hoping to get a pre amp which was silent)

My Krell power is around -118db yet quieter than the Technics I have.

I guess the lower the S/N ratio the better, as less noise is added to the signal, and dynamics are increased.

The above are with 90db sensitivity 4ohm speakers.

 

[istari knight]

Okay thanks for the input - I understand it a bit better now

I suppose the only way to tell whats better is to hear them.....Oh the hassle !

Might just be content & stick with what ive got

James.

 

[dudywoxer]

Whatever you read the final arbiter has to be your ears, the specs for the amps you have are good,as are the tech specs for Naim, for instance, yet they have a totally different presentation.

Read as much you can, and learn from the specs as to waht you can expect in terms of power, noise levels. ablity to drive difficult loads etc, but you have to listen to decide what really floats your boat. Any half way decent dealer will give you all the time you need to listen and decide.

 

[Killahertz]

So Please can anyone explain what effect SNR has on sound quality ?

In reality Davewhityetagainis quite right, we rarely know how and what was measured. However, assuming all else is equal, then the difference in SQ between the units you listed (based solely on their varying S/N ratios) would be absolutely nothing.

Yes, you may be able to hear hiss close to the loudspeakers as you turn the volume up (with no input), but this will have no effect on SQ when a signal is present - the ratio of the signal to that noise is just so great. Further, don't take the audibility (or not)of hiss (with no input signal) as an idicator of quality - no hiss at all could mean that the amplifeir in question mutes it's output when a signal is not present. Again, hiss or not could be indicative of susceptibility to interference, not strictly as a result of good/bad S/N performance.

As an indicator of relative worth (or lack thereof), consider that vinyl has an S/N ratio of about 60dB - what use, then, is an amplifier with 100dB?. And, despite the audibility of inter-track hiss, is vinyl rendered unlistenable by such a 'low' S/N?.

 

[meninblack]

Signal to noise ratio is just another way of expressing the dynamic range of an amplifier.

The dynamic range is the ratio between the noise floor of the device (output amplitude with no signal) and the output amplitude of the maximum undistorted signal the device is capable of.

SNR is the ratio between the noise floor and some arbitrary reference output level - without knowing this reference level, the measurement cannot be used to compare amps.

Hiss from the speakers with no signal is determined by the amplitude of the noise floor, not by SNR.

As stated, vinyl has a dynamic range of about 60 dB, CD about 100 dB. An amp with a dynamic range of less than 100 dB won't reproduce CD accurately, although it might be OK with vinyl.

 

[SteveB]

Things that do make a big difference are:-

Frequency response. Most amps are ruler flat though. Not so speakers. 0.5dB variation is easy to hear in my experience if it's over an octave or two.

Polar response of speakers, and how the speakers react with the room. Wide dispersion usually causes more problems than it solves in typical rooms, others may disagree.

Distortion. Usually very low and you would have thought insignificant, but there is some evidence though to show that the only sub 0.1% distortion that just makes music sound wrong is 7th harmonic stuff and vanishingly small amounts are audible evidently. Who puts that in their specs? Electronics do sound different though, so there must be a reason whatever it is, we're probably just not measuring the right things.

Can't say SNR was ever a big thing in overall perceived sound quality, in fact if you google stochastic resonance you'll see that a little white noise helps your hearing's sensitivity to fine detail.

 

[JamPal]

SteveB wrote:

Things that do make a big difference are:-Frequency response. Most amps are ruler flat though. Not so speakers. 0.5dB variation is easy to hear in my experience if it's over an octave or two.

Polar response of speakers, and how the speakers react with the room. Wide dispersion usually causes more problems than it solves in typical rooms, others may disagree.

Distortion. Usually very low and you would have thought insignificant, but there is some evidence though to show that the only sub 0.1% distortion that just makes music sound wrong is 7th harmonic stuff and vanishingly small amounts are audible evidently. Who puts that in their specs? Electronics do sound different though, so there must be a reason whatever it is, we're probably just not measuring the right things.

Can't say SNR was ever a big thing in overall perceived sound quality, in fact if you google stochastic resonance you'll see that a little white noise helps your hearing's sensitivity to fine detail.

It is odd that two devices with the same technical spec and roughly the same measured outputs can sound different, but they really do. Or do they?

I read about a taste test recently undertaken, where two identical chocolate puddings were labelled differntly. 1 was Chocolate pudding the other was Luxury Belgian double Chocolate desert. Guess which one people prefered. They all said the Luxury desert was much better than the Chocolate pudding. :Not Sure:

 

[mosfet]

That would be the power of pudding expectation.

 

[Sastusbulbas]

In my view the lower the S/N ratio the better, as I said increased dynamics and a lower noise floor is what I find with power amps that have a good S/N ratio.

Out of all the amps I have had the ones with the lower S/N ratio were the quietest, from silent to full power without a music signal present, and there was no signal muting.

These are the amplifiers that played well late at night and at low levels also. And it is particulary noticable with sensitive speakers.

So I assume sound quality IS affected by poor signal to noise ratio, though it may not be important if you listen in a noisy enviroment to compressed music with no dynamics, or your preferences for quality are elsewhere.

When I used the Nu-Vista, (and BAT) the hiss and noise was quite noticable, it far outweighed the noise from my valve pre's , valve phono's and vinyl system, and detracted from my enjoyment of the music particulary on quiet days and with late night listening.

And if you are hearing hiss from a pre or power when no signal is present it has to be considered capable of masking fine low level treble details. If I am sitting 3m away from the speakers and I am hearing hiss, I only want to be hearing that hiss if it is present on the recording.

I also think that no matter the source, be it vinyl or CD the better the S/N the better the dynamics, therefor the better the music.

I myself think that there should be no hiss or hum produced by the amp if possible, it should only be the music that is amplified, any hiss or noise at the pre stages in particular are usually exagerated , and I feel mask low level treble detail and ambience.

Audioholics had a thread regarding LP versus CD, which I think was pointing out that the dynamic range of CD is not as high as it is made out to be, and that vinyl is quite capable in this area, I would have to double check though.

 

[Killahertz]

Sastusbulbas wrote:

And if you are hearing hiss from a pre or power when no signal is present it has to be considered capable of masking fine low level treble details. If I am sitting 3m away from the speakers and I am hearing hiss, I only want to be hearing that hiss if it is present on the recording.

In orderto induce audible levels of hiss, and certainly at levels audible at the listening position, how far are you into the volume dial?. Surely, (and in my experience) you need to be 80% plus for levels to become just audible. Yet for maximum system playback level to be reached considerably sooner - and hence negate the affect of noise on the playback signal?.

PS: TheLP vs CDdynamic range test ishere (although i'd err on the cautious side - the process is flawed and the conclusions likewise).

http://www.audioholics.com/techtips/specsformats/index.php

 

[Sastusbulbas]

Killahertz wrote:

Sastusbulbas wrote:

And if you are hearing hiss from a pre or power when no signal is present it has to be considered capable of masking fine low level treble details. If I am sitting 3m away from the speakers and I am hearing hiss, I only want to be hearing that hiss if it is present on the recording.

In orderto induce audible levels of hiss, and certainly at levels audible at the listening position, how far are you into the volume dial?. Surely, (and in my experience) you need to be 80% plus for levels to become just audible. Yet for maximum system playback level to be reached considerably sooner - and hence negate the affect of noise on the playback signal?.

PS: TheLP vs CDdynamic range test ishere (although i'd err on the cautious side - the process is flawed and the conclusions likewise).

http://www.audioholics.com/techtips/specsformats/index.php

Hi KHZ,

The above quotes on hiss and noise levels were in most instances with the volume turned down, and no music playing, using 90db sensitivity 4ohm speakers mainly, and a few others.

In some instances it was power only, power and pre, valve pre, phono stages, DAC, integrated etc etc. (I went through a stage where I wanted utmost silence from the system so tried various components in isolation before deciding to sell my valve pre's and buy a Krell pre)

MF Nu-Vista 300 power for instance , very noticable at around 2m to 3m and in some cases during the day with average ambient noise levels, even with pre switched off, and it completely swamped the little amount of hiss which was present in an Audio Research Valve pre.

This was the power amp regardless of volume, and it made listening to ambient or classical a pain, and listening at low level unbearable, as much as I loved the endearing qualities of the amp I couldnt live with it.

If I turn my Technics SU-MA10 amp up full, there is less hiss noticable at the tweeter than I get from an AR LS7 in standby.

When my Krell power is on it produces no discernable hiss at the tweeter whatsoever, though my Classe power produces some hiss and hum which is noticable with my head next to the tweeter.(though with less sensitive speakers or balanced operation this is reduced)

My Yamaha E800 produces less hiss at the speaker than most high end pre's I have had. (though its S/N ratio is great it is not of the greatest quality)

The noisiest pre I have used was a Bat 3Ki , which was producing noticable hiss at 3m, and when the remote was used a buzzing could be heard, it was though enjoyable otherwise, just too noisy for me.

Of course with low sensitivity 84db speakers at 8ohm the above levels are not as noticable, but in my system with my 90db 4ohm speakers they were quite noticable, I dont know if its all due to sensitive speakers , or noisy mains , or listening enviroment though. (Though Nu-Vista was also noticable at a dealers with less sensitive speakers and 90db 8ohm speakers, as was the BAT)

The quiter equipment does seem to be the stuff with the better S/N ratio though ?

For instance the Technics integrated and Krell power have similar S/N ratios of around -118db and are quiet.

My Audio research and Krell pre's are around -100db and are just noticable at the tweeter and around 1m in quiet conditions.

And yet I still prefer valve pre amps like Audio Research and EAR.

PS, I still read somtimes the early magazines where they used to comlain about CD and its poor dynamic range ?

The link was interesting reading though.

 

[boseman]

Sastusbulbas wrote:

I guess the lower the S/N ratio the better, as less noise is added to the signal, and dynamics are increased.

Wasn't the concept of chest beating with techincal quotes originally an American hi-end thing? Mark Levinson et alintroduced this method to 'prove' their products were better... forgetting that a goodtechnical performancedoes not always mean good sound.

Please correct me, if I am wrong butI thought that a higher value for SNR is technically better.

For example, a SNR value of 105 dB is technically better than a value of 95 dB. Yes?

My monoblocks are around 95 dB and I can hear some slight hissing from the tweeters when my head is right next to the speakers. However, I cannot hear the hissing from more than 5 cm away so this hissing does not intrude when listening to music.

My monoblocks quote a different figure for dynamic range (I can't remember the figure just now), and I am sure it is more than adequate, to say the least.

 

[Killahertz]

Sastusbulbas wrote:

Hi KHZ,The above quotes on hiss and noise levels were in most instances with the volume turned down, and no music playing, using 90db sensitivity 4ohm speakers mainly, and a few others.

Of course with low sensitivity 84db speakers at 8ohm the above levels are not as noticable, but in my system with my 90db 4ohm speakers they were quite noticable, I dont know if its all due to sensitive speakers , or noisy mains , or listening enviroment though. (Though Nu-Vista was also noticable at a dealers with less sensitive speakers and 90db 8ohm speakers, as was the BAT)

PS, I still read somtimes the early magazines where they used to comlain about CD and its poor dynamic range ?

The link was interesting reading though.

Hi S,

I have to admit that of all the amps i've tried, i've never had one that had noise at no or low volume travel. Indeed, I checked the three amps that I have at the moment, and all were noise-free (audibly speaking) until around 75% of volume travel (with no signal present). Although once noise did begin it rose from just audible to (in one case) very audible at >85% of volume travel. This would be affected by sensitive loudspeakers - but i've only ever used modesly sensitive 8 Ohm nominals. That said, that alone shouldn't have such a major effect on the audibility of hiss. As you say, perhaps there was an added susceptibility to mains or other noise.

PS: The test is interesting as it attempts to bring objectivity to theory. But, there are just too many variables in the method and conclusions for me. That said, i'd like to see a body likeAudioholics re-run the test under more controlled conditions. I wouldn't mind a go myself, but have no analogue front end to compare - although I may have a crack at the CD element.

 

[Umberto]

My Bow Warlock prewas quite noisy. I confirmed it was the pre that was making the hiss and slight mains hum (which was inaudible when the music was playing) by switching off the pre. Tec specs were 100dB s/n ratio. My valve setup still has a tiny amount of mains hum at no volume and the pre has a s/n ratio of 95dB but there is no hiss audible at no volume.

So the moral is get a dem of the equipment first and don't go by tec specs as a rule.

 

[rockmeister]

it was brought home to me that specs dont count, when I read that both the speakers I then owned, (Snell J11's,) and the amps (Audio innovationations First audio monoblock valves), "measured badly" according to most hifi mags. (In the 80's any mag worth it's salt produced charts and analysis to show what the specs were). Amusingly, all had to admit that the combo for some reason, actually sounded bleeding fantastic ans often better than amps that measured up with 'good' figures. Sound output can be tuned in any hifi product and often this involves introducing some measure of distortion somewhere.

Best not to ask why, but just to listen and smile IMO

 

[Sastusbulbas]

boseman wrote:

Sastusbulbas wrote:

I guess the lower the S/N ratio the better, as less noise is added to the signal, and dynamics are increased.

Wasn't the concept of chest beating with techincal quotes originally an American hi-end thing? Mark Levinson et alintroduced this method to 'prove' their products were better... forgetting that a goodtechnical performancedoes not always mean good sound.

Please correct me, if I am wrong butI thought that a higher value for SNR is technically better.

For example, a SNR value of 105 dB is technically better than a value of 95 dB. Yes?

My monoblocks are around 95 dB and I can hear some slight hissing from the tweeters when my head is right next to the speakers. However, I cannot hear the hissing from more than 5 cm away so this hissing does not intrude when listening to music.

My monoblocks quote a different figure for dynamic range (I can't remember the figure just now), and I am sure it is more than adequate, to say the least.

Sorry but I took the minus before the number IE -95db to to indicate a lower value.

Hence -118db is quieter than -88db ?

 

[meninblack]

If signal/noise is coming out negative, then they are not calculating it in the standard way. Should always be positive as it is the ratio of two quantities which are both (i think) positive - sound pressure levels, essentially.

For a negative SNR you would need +ve signal and -ve noise, or vice versa. That would be weird.