A SINAD of 80 or SINAD of 100 Can You Really Tell The Difference?

[jomo]

That is why @amirm's chase to highest SINAD makes sense, doesn't it? If a SINAD is at levels around 120dB, you will neither have distortion nor noise. Aim to the top and be safe.

No. The value of SINAD is not a useful parameter other than providing a gauge of combined distortion and noise. The value is pretty useless in practice because it mixes values that are individually important in different situations.

If you want a amp with no hiss, you don't use SINAD. The "quiet" target should use SNR, power of amp, sensitivity of speakers, and noise floor of room. The target varies depending on these parameters. For a very quiet room, powerful amp, and sensitive speakers, you may want an SNR greater than 120. What SINAD? Don't know.....likely greater than 95 dB. Unfortunately, I made the mistake of mixing SINAD and SNR too.

Yes, SINAD and SNR commonly correlate, but that is not the point. The value for SINAD is being used as if it is SNR. Because of this, this is driving target noise and distortion down to levels that are way beyond necessary for silent, transparent listening

 

[Blumlein 88]

No. The value of SINAD is not a useful parameter other than providing a gauge of combined distortion and noise. The value is pretty useless in practice because it mixes values that are individually important in different situations.

If you want a amp with no hiss, you don't use SINAD. The "quiet" target should use SNR, power of amp, sensitivity of speakers, and noise floor of room. The target varies depending on these parameters. For a very quiet room, powerful amp, and sensitive speakers, you may want an SNR greater than 120. What SINAD? Don't know.....likely greater than 95 dB. Unfortunately, I made the mistake of mixing SINAD and SNR too.

Yes, SINAD and SNR commonly correlate, but that is not the point. The value for SINAD is being used as if it is SNR. Because of this, this is driving target noise and distortion down to levels that are way beyond necessary for silent, transparent listening

I generally agree with you on this. However, describe a situation where something with a SINAD of -100 db can cause a problem. THD would be at least that as would noise. Likely both would individually be somewhat lower.
I do understand and have complained some gear with SINAD of only -70 or -80 db can be just fine if THD is in that range with noise much lower. So not all gear lower on the SINAD list is a problem though it probably is a bit more often than not.

 

[jomo]

I generally agree with you on this. However, describe a situation where something with a SINAD of -100 db can cause a problem. THD would be at least that as would noise. Likely both would individually be somewhat lower.
I do understand and have complained some gear with SINAD of only -70 or -80 db can be just fine if THD is in that range with noise much lower. So not all gear lower on the SINAD list is a problem though it probably is a bit more often than not.

The fact that you have to refer to THD and SNR separately to confirm if the SINAD is ok proves how useless this parameter is. Certainly when SINAD is high, all "problems" go away. All problems except reliability, apparently... Add to that, SINAD is being used as the "new" SNR and also used where only distortion is typically reported. We should report the core data (THD, IMD, multitone, SNR) such that we can use the data directly in the appropriate field situation. One could care less what the SNR is when music is playing. One could care less what the distortion is when the music is off.

 

[Blumlein 88]

The fact that you have to refer to THD and SNR separately to confirm if the SINAD is ok proves how useless this parameter is. Certainly when SINAD is high, all "problems" go away. All problems except reliability, apparently... Add to that, SINAD is being used as the "new" SNR and also used where only distortion is typically reported. We should report the core data (THD, IMD, multitone, SNR) such that we can use the data directly in the appropriate field situation. One could care less what the SNR is when music is playing. One could care less what the distortion is when the music is off.

I do believe all of those parameters are already reported in Amir's reviews. SINAD is used for the chart. Nice way to sidestep answering my question btw.

 

[AnalogSteph]

However, describe a situation where something with a SINAD of -100 db can cause a problem.

Some 130 dB/V IEMs on an amplifier that can do 6-9 Vrms, if said SINAD is dominated by noise.

SINAD considerations implicitly require a (somewhat) normalized playback volume. Under these conditions, SINAD >100 dB is a non-issue for most anyone, of course.

 

[jomo]

I do believe all of those parameters are already reported in Amir's reviews. SINAD is used for the chart. Nice way to sidestep answering my question btw.

They are. But the easy route for many is to go to the SINAD......then use it incorrectly.

I did not sidestep your question. An amp at -100dB SINAD will likely have a noise level suitable for any situation, but to confirm this you go to the SNR. This says to me that a much more useful gauge than SINAD is SNR. Distortion.....pfff.

 

[sarumbear]

An amp at -100dB SINAD will likely have a noise level suitable for any situation, but to confirm this you go to the SNR.

Why is there a need for confirmation? Can a device that has a SINAD value of 100dB has a noise level higher than that?

 

[ampguy]

I don’t think I can. My monoblock A300s have a published SINAD of 84 or so at 4ohms/5W, higher at 8, but can’t tell sound difference with a much higher SINAD amp. If prices same, or affordable, I’m happy with 90 or higher.

 

[jomo]

Why is there a need for confirmation? Can a device that has a SINAD value of 100dB has a noise level higher than that?

No it can't. However, 100 dB SINAD does not provide precise noise information to make a proper decision/calculation. SNR does, which is provided by Amir. A typical 100dB SINAD amp has an SNR of 115 to 125ish. If you have a virtually anechoic room, super efficient speakers, want to hit 120 dB peaks at your ear AND have the amplifier silent readying for this onslaught, you need to use the SNR to determine if the amp will be silent, not SINAD.

 

[sarumbear]

No it can't.

I take that you are agreeing high level SINAD is enough to tell us that it’s a transparent device with low noise and distortion, which was my comment.

 

[theREALdotnet]

Why is there a need for confirmation? Can a device that has a SINAD value of 100dB has a noise level higher than that?

It cannot – at 4V input signal.

I think the issue here is that distortion is worst at full tilt and therefore measured at high input level. Noise, on the other hand, is most bothersome at low (or no) input. It will be inaudible at high input, no matter what. More importantly, the SNR will be low. With noise, the absolute level is more interesting.

Shoehorning the noise performance of an amp into SINAD is akin to producing a compound performance figure that measures both noise and distortion with no (or very low) input signal.

 

[Hayabusa]

The fact that you have to refer to THD and SNR separately to confirm if the SINAD is ok proves how useless this parameter is. Certainly when SINAD is high, all "problems" go away.

Also a problem that does not go away is that these measurements are done with long FFT's which causes impulse errors to be spread out to a lower level.
Lets say there is glitch in the DSP code that gives out a 0dBFS single sample spike every few seconds, the FFT would show this at a much lower level.
Audibility of this spike is much higher than you would expect from the numbers.

 

[tp1]

Differences in Sinad are most definitely audible but not in the manner that some may think. The actual residual noise may not be audible on an amp with poor Sinad but with one with superior Sinad you will notice advantages . Firstly in a home cinema setting the lower Sinad will reveal low level detail such as ambient sounds which will be reduced on the lower Sinad amp. The background will be “blacker” and the soundstage will appear bigger by virtue of the new sounds emanating from all speakers. Dynamics are better defined too.

I was very disappointed in the Marantz AV8805 I bought after having used a Denon AVP- A1HD for years. Don’t get me wrong, the Marantz is indeed a great processor but the difference in low level detail , ambience, soundstage and dynamics compared to the legendary Denon AVP was there to be heard every day. I ended up selling it and buying a Lyngdorf MP-50 which restored the sound I was used to. It wasn’t until later when I saw the test results for the Marantz 8805 on the ASR site did I have data to explain why the differences were there. the Denon, Marantz and Lyngdorf processors were all used using the same source gear , speakers and power amps. ( I have since upgraded power amps too)

 

[Trell]

Differences in Sinad are most definitely audible but not in the manner that some may think. The actual residual noise may not be audible on an amp with poor Sinad but with one with superior Sinad you will notice advantages .

You’re aware that noise is part of SINAD by definition?

 

[tp1]

You’re aware that noise is part of SINAD by definition?

Of course

 

[Trell]

Of course

But that is not apparent in your post, thus my question. Your post still does not make sense, though.

 

[tp1]

But that is not apparent in your post, thus my question. Your post still does not make sense, though.

The title if the thread asked If you could tell the difference between a Sinad of 80 or Sinad of 100. My point is that you will not necessarily hear any residual noise from equipment with a poor Sinad score compared to one with significantly better Sinad. But you will hear differences between the pieces of equipment because of Sinad. The residual noise and distortion of the lower scoring gear will not be able to produce sounds recorded at a much lower level than the main audio , compared to the better scoring kit.

The example I gave of sounds recorded at a low level was the ambient sounds for a scene in a multichannel AV soundtrack. These are barely noticeable ordinarily as we usually focus on the main dialog or action on screen, but when they become less apparent on a different processor or amp, you do notice. In my case I did hear the differences between the Marantz, and the Denon & Lyngdorf processors, differences which were Sinad related. Marantz 8805 has Sinad of 92 while Lyngdorf MP-50 has Sinad well over 100. Not sure what Denon AVP measured at due to its age but it was built to a standard that no other Japanese company has since attempted.

Apart from the ambient sounds which create a more immersive experience, the lower noise level adds a sense of greater space to the soundstage. I put this down to a better reproduction of the more delicate harmonics . With music only playing , lower level detail is better retrieved, and for the lack of a better word, there is a “blacker” background.

I am trying not to overstate the differences because the 8805 is still a very good processor and unless you live with something better for a while I doubt you will have any complaints at all.

I hope this helps

​

 

[Astrofly]

Where are the ABX tests for proving such a distinction? If someone posts something about the audibility of amplifiers, cables, DACs, etc, then the quick response on ASR is to recommend ABX tests as the proven way to discern a difference. Ok, great, so then isn't that fair to expect in this instance as well? There seems to be a lot of speculation on the audibility of SINAD measurements. Post some tests, and then we can actually know if chasing the highest possible SINAD is worthwhile.

 

[tp1]

Where are the ABX tests for proving such a distinction? If someone posts something about the audibility of amplifiers, cables, DACs, etc, then the quick response on ASR is to recommend ABX tests as the proven way to discern a difference. Ok, great, so then isn't that fair to expect in this instance as well? There seems to be a lot of speculation on the audibility of SINAD measurements. Post some tests, and then we can actually know if chasing the highest possible SINAD is worthwhile.

I understand scepticism when it comes to audio claims however you have already been provided the measurements that determine differences in between the products. I am not trying to push any particular point of view, only relate my own experience . I am one of the people who who look for 3rd party verified measurements to distinguish between products, which I believe is one of the founding purposes of the Audio Science Forum. In this particular case the measurements came after my listening experience which helped explain the differences and reinforced what I had heard.

The problem with your request to do ABX testing is that apart from the difficulty in arranging it, I haven't seen an ABX test yet where all parties were happy with its parameters or conclusions. And on that subject, it is not for me to prove that item X sounds better than item Y due to Sinad differences . The proof is there in black and white that there are sonic differences. Have you done ABX testing that proves differences in SINAD is inaudible?

 

[CleanSound]

I am certain I cannot tell the difference, and I'm willing to bet anyone $100 bux that they can't neither.

BUT, if the cost is not astronomically more for more SINAD, I would still want more SINAD; headroom for other distortions in the signal chain, and pride of ownership of excellent engineering are my driving factors. But I will not sacrifice other important factors for SINAD, QC, the aesthetic looks, function, etc.