PS Audio M700 Monoblock Amplifier Review

[amirm]

I would maintain that very few of us run a system that offers up 96dB dynamic range on the music we listen to. We would lose our hearing fairly quickly if we did, and our neighbours would regularly call the noise police.

You maintain wrong. 96 dB momentary peaks for music would mean 76 dB average. No way you would go deaf or even have your neighbor complain. There are cars out there where at speed have higher average noise level and we don't see people going deaf in them.

This whole argument comes from misreading of OSHA safety standards which are for noise and continuous basis, not momentary peaks in music. Here are the OSHA standard:

As indicated, you can have 95 dB SPL for whopping 4 hours per day and still be safe!

You can go up to 110 dB and still be OK for half hour.

So no, that is a myth. You don't go deaf at 96 dB SPL peak. Average will be at 76 which is not even in the above table.

 

[amirm]

My first beef is with lumping noise in with THD.

I didn't mix the two. The complaint came from me measuring dynamic range:

 

[Racheski]

Well Thomas agreed to move it if someone gathered up a list of posts. I'll say I was too lazy to do it myself.

If you have a noise floor of 40 db spl, then 96 db on top of that would suggest 136 db SPL capable playback. I think that is too simplified, and Amir has pointed to his article based upon Dolby testing and other results as to why that isn't the correct way to view it. I agree on that being overly simplified. But if one took instrumented measures and didn't elaborate that is what you'd think. And you could say the way Amir does his testing would leave someone with that impression. One could go into many nuances of the question. My first beef is with lumping noise in with THD. But I'll stop there. I understand why Amir tests like he does, and I'm okay with it.

The power question can be almost divorced from dynamic range or linked somewhat to it. Depends upon context and purposes.

And if I'm going to be completely fair to the poster, there are no formal ASR community guidelines for posting. So who am I to say he shouldn't post whatever he wants, whenever he wants, wherever he wants.

 

[Vasr]

Now I am confused. If I understand it correctly, the review measurement looks for 96 db peaks above DUT noise (typically far below floor noise). Makes sense to me.

To hear that full dynamic range in a "noisy room" (typical room), we would need the peaks to be 96db above the floor noise level by cranking the amp up so I can hear the lowest parts above the floor noise, right? So it would have to be at some 30-40db above that whose peaks could be literally ear-splitting levels. That also makes sense to me.

What is the disconnect? What have I missed or misunderstood? Not an expert here with an opinion one way or the other.

PS: A neighbor's ability to be annoyed at the music loudness is usually far below the level at which it would destroy the listener's ear drums.

 

[amirm]

My sensitivity to that question is because of the argument that is used to beat people over the head that any level of fidelity is good enough. "Your room had 40 dB noise, you listen at 70 dB so you only need 30 dB of dynamic range" It is a wrong argument and was the motivation behind me writing that article for WSR magazine. So to come here and throw it at me as if this is something I have not heard was rude. It seems to still be continuing with the same tone and stance.

 

[amirm]

To hear that full dynamic range in a "noisy room" (typical room), we would need the peaks to be 96db above the floor noise level right by cranking the amp up so I can hear the lowest parts above the floor noise? So it would have to be at some 30-40db above that whose peaks could be literally ear-splitting levels. That also makes sense to me.

What is the disconnect? What have I missed or misunderstood. Not an expert here with an opinion one way or the other.

You have to read the article I wrote. Here it is again: https://www.audiosciencereview.com/forum/index.php?threads/dynamic-range-how-quiet-is-quiet.14/

 

[Colonel7]

Because you can't have recorded signal beyond peak digital, it would just be hard clipped, and the engineer won't have put it in the recording.

Let me respond in kind. This is not true. Sound engineers do this. Prove me wrong. Because I've seen it. But the proof is up to you.
-------
See how non-confrontational that is?

 

[GaryMnz]

You maintain wrong. 96 dB momentary peaks for music would mean 76 dB average. No way you would go deaf or even have your neighbor complain. There are cars out there where at speed have higher average noise level and we don't see people going deaf in them.

This whole argument comes from misreading of OSHA safety standards which are for noise and continuous basis, not momentary peaks in music. Here are the OSHA standard:

View attachment 77342

As indicated, you can have 95 dB SPL for whopping 4 hours per day and still be safe!

You can go up to 110 dB and still be OK for half hour.

So no, that is a myth. You don't go deaf at 96 dB SPL peak. Average will be at 76 which is not even in the above table.

No what I was saying is to reproduce a dynamic range of 96dB, which is what your review said should be the capability of the amplifier to do justice to a CD resolution.

96dB above the noise floor of a quiet room would be peaks of around 116dB. Neighbours??

Of course this is all academic as despite the capability we don't find 96dB dynamic range recorded on our CDs. Much less in fact, which could allow for an even more relaxed sinad spec without losing peaks or introducing unwanted low level noise.

 

[GaryMnz]

Let me respond in kind. This is not true. Sound engineers do this. Prove me wrong. Because I've seen it. But the proof is up to you.
-------
See how non-confrontational that is?

Its easy to prove you wrong. Above peak digital nothing can exist. There is no way to encode it. A digital medium has a very hard and unbreachable limit

 

[Vasr]

You have to read the article I wrote. Here it is again: https://www.audiosciencereview.com/forum/index.php?threads/dynamic-range-how-quiet-is-quiet.14/

Ah, I see. This is what Gary isn't addressing.

Since the noise floor is spread across the frequency spectrum, the value at one frequency dominating in the region where our ear is perhaps less sensitive or the music content is absent in that region does not require the lows at another frequency to be above it to hear it. And it does not have to be simple additive depending on our ear sensitivity to specific frequencies.

So if I my noise floor is dominated by an earthquake rumble (which is known to happen in my parts) in low frequencies, I can happily ignore it without having to raise my music to enjoy it ... unless my music also has an earthquake rumble in it as part of the music. But if my floor noise is from a chainsaw murder going on nearby....

I can see both sides (not to say they are equivalent).

I do think floor noise could be in the region that interferes with music content but may not just be a linear addition on top of the noise floor depending on frequency, so not sure what the right resolution from a measurement perspective to this conflicting goals is. Will leave it to the experts to hash it out.

 

[Colonel7]

Its easy to prove you wrong. Above peak digital nothing can exist. There is no way to encode it. A digital medium has a very hard and unbreachable limit

You're not paying attention. Sound engineers too often clip the recordings now. They do put it in there

 

[pozz]

@pozz posted this at some point when discussing the DAC histogram.

It was in this thread. https://www.audiosciencereview.com/...t-compared-to-other-sources.14435/post-447811

The version posted by @F1308 above was actually the original proposed by @martijn86. The one in the link was somewhat modified to organize some of the charts for a while.

Please don't take them too seriously. It's a good way to gauge engineering quality for a few devices (DACs, ADCs, preamps, headphone amps) but not for others (speaker amps, phono preamps) on an individual basis. As soon as these devices are in a chain you have to take into account gain staging, loads, levels and so on.

 

[Racheski]

Please don't take them too seriously. It's a good way to gauge engineering quality for a few devices (DACs, ADCs, preamps, headphone amps) but not for others (speaker amps, phono preamps) on an individual basis. As soon as these devices are in a chain you have to take into account gain staging, loads, levels and so on.

I wholeheartedly pledge I will not

 

[GaryMnz]

You're not paying attention. Sound engineers too often clip the recordings now. They do put it in there

I didn't say it couldn't be clipped. I said that clipping cannot encode a level higher than peak digital. Its a hard level limit. It doesn't result in high transient peaks in the recording, just grunge.

 

[Colonel7]

I didn't say it couldn't be clipped. I said that clipping cannot encode a level higher than peak digital. Its a hard level limit. It doesn't result in high transient peaks in the recording, just grunge.

You did...

it would just be hard clipped, and the engineer won't have put it in the recording.

 

[LTig]

Again someone who doesn't understand the methodology.

He starts from the perspective of the recording. It cannot contain anything useful above digital clip, right?

So he then creates a test signal some dB below that. Just so it won't be too loud when you play it back.

You then play music you would normally listen to, at the loudest level you would normally use.

Then you play the test signal and measure the voltage.

From the data you can definitively say how much amplifier power you need to satisfy your listening in your system.

Those high peaks you talk about are incorporated and allowed for, but will not exceed the defined number of dB above the test tone. Why? Because you can't have recorded signal beyond peak digital, it would just be hard clipped, and the engineer won't have put it in the recording.

Very simple, deceptively clever.

The problem with this approach is that the real life crest factor of music varies a lot. As we have seen in the Harbeth video there is a factor of more than 20 in max peak power between the two pieces of music. If I've read correctly the sinus is at -12 dBFS, which should result in a crest factor of 4 (12 dB)[EDIT: the factor is 16, we're talking power, not voltage]. This is certainly sufficient for compressed pop, but not for e.g. symphonic music or very slightly compressed music (e.g. like this old classic: Tricycle by Flim & the BBs):

 

[GaryMnz]

I didn't mix the two. The complaint came from me measuring dynamic range:

View attachment 77343

This raises more questions on a second look. With reference to your dynamic range pictures for this product... What follows is that the guy with sensitive speakers is going to be sorely disappointed as the 5 watts or so average power he needs will only deliver 86dB dynamic range.... and yet were a peak to come along he would be surprised to find that all was ok... it was reproduced ok and the dynamic range he had been told was inadequate all of a sudden was ok.

I think the more useful number is the noise floor of the amplifier. Leave out the harmonic distortion as a factor in dynamic range as it plays no part. The dynamic range capability is the difference between the noise floor and clipping....

The test I find useful on production amplifiers is to listen to each channel with my ear about 1 inch from a loudspeaker... a faint residual noise is OK, this will never be heard in a system and so will not limit enjoyment. This is much more direct and easy to evaluate than a spectrum is, at least on a production line.

 

[GaryMnz]

The problem with this approach is that the real life crest factor of music varies a lot. As we have seen in the Harbeth video there is a factor of more than 20 in max peak power between the two pieces of music. If I've read correctly the sinus is at -12 dBFS, which should result in a crest factor of 4 (12 dB). This is certainly sufficient for compressed pop, but not for e.g. symphonic music or very slightly compressed music (e.g. like this old classic: Tricycle by Flim & the BBs):

Please read what I wrote again, you are not understanding what was written.

 

[LTig]

I didn't say it couldn't be clipped. I said that clipping cannot encode a level higher than peak digital. Its a hard level limit. It doesn't result in high transient peaks in the recording, just grunge.

Nope. Search here for a thread about intersample overs by @ayane.

 

[LTig]

Please read what I wrote again, you are not understanding what was written.

Just mark in red in my posting where I was wrong, please.