• Welcome to the all-new HUG. All content has been converted from the old system, and over the next days we will re-style HUG in a more familiar way.

"Fluidity"? >>> clipping & amp sonics

Status
Not open for further replies.

A.S.

Administrator
Staff member
Assumptions

Assumptions

I can't help but imagine a 20W amplifier would be more than sufficient for a tweeter, although I suppose even 5W could work. But, I'll err on the side of caution and guess that a 20W amp on a tweeter would not lead to any noticeable clipping to a studio engineer in the normal course of his work. What did Dudley Harwood find out in his investigations?
That's a well reasoned comment. Working from memory as I don't have his paper here, I recall that his conclusion was that (with certain caveats about type of music, degree of clipping, repetition rate of clipping and so on) that a low powered tweeter amp could be driven into more or less continuous measurable clipping, none of which was audible.

Once again, that's perhaps counter intuitive if we assume that the ear behaves like a technical instrument. It doesn't. To make sense of the pressure wave entering the ear in a timely fashion that enables an appropriate fight-or-flight response, the ear (and brain, mostly) has to make lots of 'assumptions' and take various 'short-cuts', such as averaging the pressure wave. Can you guess why?

Here is an excellent video of the physical reality of the human organ of hearing. Having seen that, do you think the organic ear with it's terrible age related decay or this precision sound measuring instrument is more likely to give consistent, year on year, accurate results?
 

Pluto

New member
Imprecision in language?

Imprecision in language?

First and formost I only ask if 'fluidity' can be measured or seen in a spec sheet of an amplifier. Never asked for anything more nor am I selling anything. Couldn't care less if you like it or not really.

Your heard it stunningly good, but what are you using? And why?
This comes across as ranting.

Fluidity of an amplifier cannot be measured unless and until you define 'fluidity' with rather greater precision. We can measure most of what matters in amplification but, obviously, cannot relate such measurements to vague terminology. My personal interpretation of ‘fluidity’ might be entirely different from yours, so the use of such terms really isn't helpful.

Earlier you asked...

For what its worth, do you have a member here that uses a 150$ amp to drive his harbeths?
to which I answered, yes. I have tried an amplifier that, give or take, costs $150 and found that there were no audible distinctions between it and many other (considerably more expensive) amplifiers I have tried so what, precisely, is your point?

Normal people would not invest 4000$ for a speaker to use in a 150$ system, its downright crazy to even consider it
Does this logic make it equally crazy to use a $150 amplifier with Harbeth speakers? What price tag has to appear on the amplifier to avoid your questioning my sanity?
 

mhennessy

Member
Gross clipping and benign clipping

Gross clipping and benign clipping

We don't want to be the source of the audio virus 'I need a new $$$ amp to escape constant clipping...' which, on the face of it, is the advice we have unintentionally implied.
Oops! Apologies - I didn't intend to cause panic or alarm :)

The subject of clipping is an interesting one, and one that I've researched a fair bit over the years. It's perhaps worth expanding slightly...

First of all, it should be stated that gross clipping is very obvious, and "a bad thing". Many loudspeaker manufacturers warn against it - declaring warranties null and void if you've damaged a 'speaker with a clipping amplifier - and for many hi-fi enthusiasts, this is the limit of their knowledge - and perhaps that's why my previous post could have caused Alan to worry. It should be stated straight away that no-one who appreciates music enough to have chosen Harbeth loudspeakers will never choose to listen under those circumstances - the sound is horrible. Does it actually damage loudspeakers? Maybe, maybe not. I's a complicated answer, and one that I'll leave for now...

The type of mild clipping I'm referring to is benign. It doesn't sound like distortion. Without an oscilloscope across the output, you wouldn't see it, and you certainly wouldn't interpret it as distortion. So why is it a problem?

Because it changes the harmonic structure of real instruments, giving them a different timbre. I mentioned brass earlier, because the effect is very obvious - to me, at least - with brass instruments. And it's this that hi-fi reviewers pick up on, and this is why amplifier designers might wish to pay attention to clipping behaviour.

I'm pretty sure that I've reported all this before, so if you've heard this already, please forgive the repetition!

I first noticed this effect when listening to a Diana Krall track ("From this moment on"). About 10 years back I built a simple 4-channel amplifier that gives 50 watts per channel, but it can be "bridged", which turns it into a stereo amplifier of around 150 watts per channel. One evening, having recently been reunited with all my hi-fi gear after a house move, I compared normal to bridged mode and was disconcerted to hear a slight difference in sound quality. Just 5 minutes with an oscilloscope proved that while operated in the 50 watts per channel mode, it was clipping. When working in bridged mode, each amplifier only needs to produce half the voltage swing for the same level, so was some way clear of the clipping point.

The two modes of operation simply sounded slightly different. Neither was obviously wrong; neither sounded better or worse. Just different. If I was a hi-fi reviewer, then no-doubt I could have invented a couple of paragraphs describing these differences in some meaningless way, which a consumer could easily have been greatly influenced by. But ultimately, the difference was pretty subtle, and only really affected the brass parts of the track. I quite was happy to listen to either rendition. The mild clipping in "50 watt mode" was not a problem, and would not have damaged the loudspeaker in any way.

Just for background, the loudspeakers used are around 84dB/W sensitivity, so very similar to the LS3/5A or the Monitor 20, and only a little behind the Monitor 30 (if I've remembered the specifications correctly). The room is small, at 12 foot square, and the level was set to a level that was just above conversational - not too loud as we have attached neighbours. And, I was pretty surprised when I saw what the 'scope was telling me!

To expand further, it is worth thinking about the character of well-recorded music. I like to quote something called PMR, which stands for "peak to mean ratio". In other words, how much louder are the peaks, compared to the "average" level of the audio signal. I've seen numerous sources over the years, but 20dB is a good nominal value to have in mind for non-classical works. Now, let's just pick apart what 20dB means: audio professionals use decibels as a measure because they do a fair job of approximating the non-linear human hearing mechanism, but we can readily convert a quantity in decibels back to volts or watts, which of course electronic engineers prefer. And, a change of 20dB means a power change of a factor of 100. Believe it or not!

So, if you set the volume control to give an average electrical level of around 1 watt, this means that the average levels coming from each loudspeaker, measured at 1 metre and ignoring room effects, would be around 84dBA. Which is quite loud - not deafening, and not loud in rock-concert or night-club terms, but about as loud as you'd want for short bursts of serious listening.

But, the peaks will be 20dB louder - around 104dBA - assuming the amplifier could provide the power without clipping. And how much power would that need? 100 watts.

Now, I really don't want to get hung up on the numbers, and I really don't wish to cause panic. After all, an average level of 84dBA is pretty high, and most people listen at much lower levels for most of the time. And as I've said, mild clipping is pretty benign.

By way of an example, let's suppose that we only have a 50 watt amplifier. Again, we adjust the volume such that the average level is close to 1 watt per channel. Taking the same notional recording, with peaks that should require 100 watts per channel, how much difference will we really observe?

If the peaks are limited to only 50 watts, it does seem like a pretty serious problem, no? Well, in decibel terms - which is how our ears perceive sound, of course - the difference between 100 watts and 50 watts is only 3dB - a pretty small change in the scheme of things. So, the momentary peaks in the recording will 3dB less than expected. Will we hear this limiting? Probably not, especially as our ears tend towards an average-response anyway. If we hear any difference at all, it's because the harmonic structure of the instruments has been changed (clipping adds odd harmonics), and instruments like brass sound the way they do because they have a rich and complex harmonic structure.

It's important to note that my tests were done at the same level - very simple and repeatable with my gear, where the volume control is calibrated in decibels. I think that this test is pretty unique, and would be hard to reproduce unless you had 4 identical channels of amplification and the means to bridge pairs of them - but I also believe that it removes all other variables from the equation, leaving just clipping as the changing parameter. Such a test ought to be simple to reproduce using digital audio editing software, but I've never really tried this - there's a thought for someone... But anyway, if you try this test at home, using a stereo amplifier and an oscilloscope, do bear in mind that you'll be listening at two different levels for unclipped and clipped, and that the differences in your hearing response at these two levels will probably "swamp" the changes brought about by just clipping.

Of course, we are very used to a diet of heavily clipped music if we listen to contemporary music, especially on the radio. Putting aside what radio stations do to their audio output, a lot of recordings are deliberately messed around with to make them sound louder - the so-called "loudness wars" is a massive subject in its own right. I stumbled across this the other day, which is well worth watching: http://www.youtube.com/watch?v=u9Fb3rWNWDA

Regarding the question about the LS5/8, the tweeter level was limited to around 12 watts in the last sample I measured. This value is very nominal and varies a fair bit from sample to sample. My official paperwork is at home, but I seem to recall a slightly lower minimum value being quoted. The later Chord amps limited the power by changing the peak current limiter built into the amplifier - it was restricted to around 2 amps, which is around 16 watts into 8 ohms.

These values might seem small for such large system, but it's really not outrageous. I'll let Alan expand if he wants to - this is his territory, not mine :)

I would love to know more about Hardwood's research regarding clipping. I live in hope that one day details will surface from the archives.

I hope the above helps,

Mark
 

npoguy

New member
Thanks

Thanks

Mark,
Thank you for that wonderfully detailed explanation. I know someone suggested earlier that the thread be closed, but I'm actually glad that it stayed open long enough for your contributions!
 

A.S.

Administrator
Staff member
Looking at clipping

Looking at clipping

Oops! Apologies - I didn't intend to cause panic or alarm :)

The subject of clipping is an interesting one, and one that I've researched a fair bit over the years. It's perhaps worth expanding slightly......Because it changes the harmonic structure of real instruments, giving them a different timbre. I mentioned brass earlier, because the effect is very obvious - to me, at least - with brass instruments. And it's this that hi-fi reviewers pick up on, and this is why amplifier designers might wish to pay attention to clipping behaviour.

I first noticed this effect when listening to a Diana Krall track ("From this moment on"). About 10 years back I built a simple 4-channel amplifier that gives 50 watts per channel, but it can be "bridged", which turns it into a stereo amplifier of around 150 watts per channel. One evening, having recently been reunited with all my hi-fi gear after a house move, I compared normal to bridged mode and was disconcerted to hear a slight difference in sound quality. Just 5 minutes with an oscilloscope proved that while operated in the 50 watts per channel mode, it was clipping. When working in bridged mode, each amplifier only needs to produce half the voltage swing for the same level, so was some way clear of the clipping point....
What would illustrate this perfectly would be to record onto video the waveform displayed on an oscilloscope, one channel the input to the power amplifier, the other channel the signal across the speaker, with the axis adjusted for best comparison. Could that be done?

The sonic nature of an amplifier as heard over the loudspeakers will, as you say, include a sonic component added to the music depending on how the amplifier behaves in clipping, detectable according to the sensitivity of the listener to the sonic consequence. A formal A-B comparison of amplifiers then should, if they are driven into this clipping region, clearly expose sonic differences between amps. By implication, a 50W amp can be expected to be operating closer or beyond its clipping point more often than a 100W amp on the same music at the same volume setting (measured with a sound meter at the listener's ears), as you have said.

That surely implies that a bigger amp (higher output rating) would have a sonic advantage for 'neutrality' over every smaller amplifier, without exception, and that with real-world music under controlled A-B conditions where you indicate clipping due to the energy in the music is inevitable, the smaller amp should be readily detectable by listening if clipping changes instrumental tonality. Is that or is that not the case? How have formal amp comparisons handled the differences in power potential (and hence, clipping likelihood) when selecting amps for comparison? Is it only reasonable to compare 50W amps against 50W amps? Or 150W against 150W? Wouldn't a 50W when compared against a 150W immediately reveal itself under blind conditions because it is unwittingly being driven into clipping rather often?

In other words, if clipping behaviour defines the sound of an amplifier, a small amp will always, and without exception, contribute more 'coloration' (incl. changing the nature of instrumental tone, such as brass) to the output than a larger amp with real-world music in the real-world home listening setup. If that is so, then there must be a minimum, sensible, clipping-free power output for typical speakers, and all amplifiers of significantly less power potential than that sensible minimum (such as 300B tube amps rated at a handful of watts) must, as you said, be running more or less permanently in clipping mode to the apparent delight of their users.

Alternatively .... the actual amount of wattage needed to make a decent sound at home is so low (say, 3 or 4 watts) that even small amplifiers of, say, 30W, are, in practice, not driven to clipping. In which case, there is a sensible practical upper requirement for power (say, 100W) and anything rated more than that is a completely unnecessary reserve that is never used (but is paid for).

Or are we in the comfortable zone of real-world music with real-world speakers in real-world listening rooms drawing only moderate power from moderate amplifiers that, in practice, the sound of a clipped amp is just not encountered during the normal listening day? In which case, we are no closer to giving credence to the popular view that amps do have strong sonic personalities.
 

Kumar Kane

New member
Small amps in clipping

Small amps in clipping

Alternatively .... the actual amount of wattage needed to make a decent sound at home is so low (say, 3 or 4 watts) that even small amplifiers of, say, 30W, are, in practice, not driven to clipping. In which case, there is a sensible practical upper requirement for power (say, 100W) and anything rated more than that is a completely unnecessary reserve that is never used (but is paid for).

Or are we in the comfortable zone of real-world music with real-world speakers in real-world listening rooms drawing only moderate power from moderate amplifiers that, in practice, the sound of a clipped amp is just not encountered during the normal listening day? In which case, we are no closer to giving credence to the popular view that amps do have strong sonic personalities.
Maybe there is a clue here to why some of this sonic signature thing persists - we read here about many people that use amps, usually the valve kind, with power outputs a lot lower than 30wpc. For these, even a load as benign as Harbeths may drive them to clip. And while the output signal may not then be true to source, some people may prefer the sound produced, as compared to a "truer" sound from a more powerful amp that doesn't clip for the same music/volume levels.

And they would justifiably be puzzled/indignant at this - "all amps sound the same" - thing. Now, I know this isn't exactly what you say, but that is how it gets heard and circulated.

PS: and if I have understood correctly, two different amps, when both are clipping, will sound different from each other.
 
Last edited:

A.S.

Administrator
Staff member
Simulating clipping, and great engineers

Simulating clipping, and great engineers

Maybe there is a clue here to why some of this sonic signature thing persists - we read here about many people that use amps, usually the valve kind, with power outputs a lot lower than 30wpc. For these, even a load as benign as Harbeths may drive them to clip. And while the output signal may not then be true to source, some people may prefer the sound produced, as compared to a "truer" sound from a more powerful amp that doesn't clip for the same music/volume levels.

And they would justifiably be puzzled/indignant at this - "all amps sound the same" - thing. Now, I know this isn't exactly what you say, but that is how it gets heard and circulated.

PS: and if I have understood correctly, two different amps, when both are clipping, will sound different from each other.
A number of interesting points here. Yes, human preferences are not predictable, and it must be a fact that 7W amplifiers are being driven at clipping point almost continuously, and it must be that their satisfied listeners like and prefer whatever sheen or glare that adds to the music. There can be no other explanation since it is an observable fact (using an oscilloscope) whether or not an amp is at clipping point.

I have never said 'all amps sound the same' as you know. I have said that, 'under regulated, controlled, structured test conditions, which must logically mean in the linear region of operation and away from clipping (because, the amp is out of control during clipping and even for a fraction of a second after clipping), the differences that are claimed for amplifiers in unstructured listening diminish or vanish.'

Two amps will potentially take on a different sonic nature when at clipping point if their circuitry is different, but that does not mean that the sonic consequences of clipping is at all audible unless it is really gross. Harwood amazed me* when I first read his report years ago by illustrating just how poor the ear is at hearing clipping, and it's easy to demonstrate just how much observable clipping of a waveform can be present, yet still completely inaudible, in the right circumstances. Now, we can arrange to take a sound waveform through a clipper (such as Adobe Audition) and chop the tops and bottom off the waveform, and we are unlikely to hear the clipping in the right circumstances. Whether any conclusions about clipping audibility would equally to a power amplifier driven to clipping (where it is still expected to deliver volts and current), as opposed to just CD-like voltage, would have to be explored.

Don't forget, real world music is not at all like sine waves in the laboratory. If you look at the waveform of music, you don't see nice, clean sine wave tones. You see a complete jumble of sounds, phases and tones. Musical instruments do not generate perfect tones. Far from it. So we must be careful to relate any theoretical observations about power amps (or loudspeakers) to music, not just to the lab tones.

*Harwood was the most realistic audio engineer I've ever met. His pragmatism could effortlessly crush ones pet theories because he constantly introduced the question 'but can you really hear it?' into subjective discussions. Spending time with him, you'd very soon appreciate that ones accrued knowledge was built on a few demonstrable facts, layered with a thick carpet of wishful thinking and erroneous assumptions. It taught me to bite my tongue before advancing some half-baked pet theory, and in the pause, just think through how I'd answer his inevitable killer question. If I couldn't, without a lot of bluster and a bead or two of sweat, it was a worthless comment from an inexperience mind which could only leave me feeling, rightly, foolish.

All a great engineer has to say to his subordinates is 'prove it' to get to the facts.
 

Kumar Kane

New member
Designed or not?

Designed or not?

I have never said 'all amps sound the same' as you know. I have said that, 'under regulated, controlled, structured test conditions, which must logically mean in the linear region of operation and away from clipping (because, the amp is out of control during clipping and even for a fraction of a second after clipping), the differences that are claimed for amplifiers in unstructured listening diminish or vanish.'

Two amps will potentially take on a different sonic nature when at clipping point if their circuitry is different, but that does not mean that the sonic consequences of clipping is at all audible unless it is really gross.
Yes, I do know what you say and it is a precisely worded statement. But it gets condensed to the soundbite of - all amplifiers sounds the same. This happens on other audio sites too.

Anyway, this post is to ask the follow up question, in the context of your "amp is out of control during clipping" statement above. Can an amplifier designer design/build an amplifier that will reflect a desired sonic signature when clipping, or is that resultant sonic signature that some people end up liking, a matter of luck?
 

A.S.

Administrator
Staff member
Can an amplifier designer design/build an amplifier that will reflect a desired sonic signature when clipping, or is that resultant sonic signature that some people end up liking, a matter of luck?
Excellent question.

I have my own view on that, but let's see what Mark, an amplifier designer, has to say.
 

mhennessy

Member
Realities of clipping

Realities of clipping

Some excellent points and questions raised. My time today is limited, so I'll return to this later...

Yes, I agree that is should be possible to produce some kind of demonstration - perhaps a video - that shows the clipping at the output of the amplifier. Something I definitely want to do when I get the chance. Leave that with me for now...

Regarding the audibility of clipping, it's definitely the case that it is surprisingly hard to hear when listening to music. With pure tones, it's a different story - we do a simple demonstration here at work using a pure 1kHz tone which is slowly increased in level until an amplifier starts to clip. We can clearly hear a pure 1kHz tone clipping because suddenly a signal at 3kHz becomes audible. With a bit of practice, you can hear the clipping before you see it on an oscilloscope - even on a mediocre loudspeaker. When measured, the distortion at this audible onset of clipping is typically less than 1% - in other words, the 1kHz fundamental is 100 times bigger than the 3kHz harmonic.

Interestingly, if - in some parallel universe - clipping introduced second-harmonic distortion at 2kHz, we would barely hear it - it would have to be much higher in level (loudness) than the 3kHz case. And that's because of the masking effect - a fundamental way our hearing works that allows data-rate reduction schemes like MP3 and AAC to work. But that's another story...

BUT, having run this test, and demonstrated that extremely mild clipping makes a pure 1kHz tone sound "harsh", we have to accept that this isn't quite what happens with real music. As Alan has pointed out, Harwood has worked on this (although I've never seen any documentation of these experiments - perhaps none was published outside of R&D at the time. Funnily enough, working for the BBC's training department does not entitle me to riffle through R&D's remaining internal archives).

So, turning up an amplifier until it sounds obviously distorted with real music actually represents gross clipping. Typically, it's modulated by the kick drum and bass instruments, as these normally represent the largest signal components in contemporary music. If not, then what you might notice is this change of timbre I've mentioned previously, or you might notice a generally "hardening" or "sheen" to certain instruments, but this can be hard to hear - and separated from the hearing system, as don't forget that as you turn up the volume, your ears will respond very differently. On more than one occasion I've been in a studio control room watching a mix take place that sounded hard and distorted. But, by simply inserting ear plugs to bring the level down by about 20dB, the sound was transformed into a clean and very clear mix - i.e. the distortion came from my ears! Presumably the engineer had severe hearing loss, but given enough level from loud monitors, he was still able to produce a sensible mix. Fortunately, hearing loss is normally a very gradual process, so this is not as outrageous (or uncommon) as one might think.

The final point for now (and yes, I'm sorry I've not had time to address each point as it was raised), is that yes, amplifiers that clip can sound OK. Especially valve amps. And people become very used to that sound, and that is what they expect from a hi-fi. When they hear a larger amplifier which isn't clipping at their preferred listening level, they can indeed find the sound "sterile" or "lifeless" or whatever. An amplifier that is clipping is compressing the signal. When you push beyond the point of mild, inaudible clipping and get ever closer to "ouch, that's distorting" region, the audio signal is being progressively robbed of its dynamic range. And what do recording engineers or mastering engineers do to their recordings to make them louder? They compress. So, perversely, a powerful amplifier might not sound as loud as a small amplifier that is being over-driven, because we as humans respond more to the average levels than the peak levels.

It's all quite complicated when you start peeling back the layers. I really must work on a better way of explaining and demonstrating some of these aspects.

All the best,

Mark
 

Pluto

New member
Far below clipping at home

Far below clipping at home

Mike's suggestion that we are routinely listening in the amp clipping region more than you might imagine is something I find a little far-fetched. Now I have no idea how loud any of you listen at home - perhaps many of you are head bangers who expect realistic concert rock'n'roll levels, but I listen to jazz and similar stuff peaking, typically, around the mid-eighties dBA. So I turn it up a bit, making it uncomfortably loud for me, but sounding clean. Under these conditions, I've 'scoped around 10V absolute peak (ref to 0V), so total amplitude could be described as 20V peak to peak.

So unless mainstream listening habits are vastly different from mine, I really don't see how I'm supposedly getting anywhere near clipping, of any kind, on a 100W-ish amplifier.

Now for those using an amplifier that has a nominal maximum output in the 25W region....
 

A.S.

Administrator
Staff member
The dynamics of music

The dynamics of music

I think we really do need to nail this clipping issue. Like so many audio issues, it is not as simple as 'gross clipping' or 'light clipping'. We need to consider how musical instruments actually behave - assuming that the objective is to reproduce music at home, not test tones - and how loud they can become over what time period.

For example, is it intuitive which instrument of the orchestra is capable of the widest dynamic range (the range between the quietest recognisable note it can emit and the loudest)? We also need to consider the reality of recorded western music where there is not a huge dynamic range; once the performance is under way, the light and dark shades of dynamic range is very predictable. Then we need to consider the repetition frequency of loud elements in the music. Are loud transients once or twice a performance or more often? What is the spectral content of the energy in those transients that could lead to clipping - in the high frequencies or low ones? Would that impact on our perception of transients? Can a symphony orchestra really jump in loudness in just a note or two, or does it have to wind itself up? Can a cello or horn really be made to play much louder on demand?
 

A.S.

Administrator
Staff member
Examples of chronic (gross) clipping....

Examples of chronic (gross) clipping....

Just to keep our feet firmly on terra firma so that we don't get into a neurotic audiophile spin about whether or not we are drowning in a sea of constant clipping, I've spent a few minutes very crudely replicating Harwood's experiment. I can't lay my hands on Harwoods papers on clipping in the LS5/8 (it seems that I have not digitised them, so I need to find the paper copy) but this should set the scene.

He said, (if my photographic memory recalls correctly), that when an audio signal is even severely clipped, depending on a number of factors, that clipping may not advertise itself to the human ear. The test equipment (which we are right to have confidence in), being of far higher resolution than the ear, can immediately see evidence of clipping, which is nothing more than cropping the tops and bottoms of the sound pressure (loudness) wave. So, I've taken a dynamic piece of piano music which peaks at a couple of dBs under absolute maximum level for a CD, and destroyed it by scything off the tops and bottoms of the transients. If you imagine that the sound wave was created not by a piano, but by a very energetic person waving a blanket up and down, the effect of clipping is analogous to us placing a restraining grip on his arm far before he reached the maximum upper or lower extent of a normal wave. His range of motion is restricted, so he cannot produce as great a waft of air.

Here is an example of a grossly clipped waveform and its unclipped source one after another. The picture shows the clipped version followed by the original, natural version after a short gap. But that may not be true for the sound file here: I might have reversed the order. Listen and decide: are you certain which comes first: the clipped or natural version? Is your audio memory sufficiently reliable (and unlike the goldfish memory), that you could leave the room for a few seconds and return with absolute confidence that you are or are not listening to the original source? I couldn't: it's a very close call indeed.



The point to make is that we must be very careful indeed not to put too much emphasis on just one parameter of amplifier performance, as being the golden arbiter of sonic quality.

>
 
Attachments only viewable to members

Kumar Kane

New member
Value amp sound? An explanation?

Value amp sound? An explanation?

Now for those using an amplifier that has a nominal maximum output in the 25W region....
And that is the playing field for a large majority of valve amplifiers. Perhaps it is how they sound when clipping that is part of the reason for the claimed warm, smooth, liquid valve sound, which may get reinforced by some pyschoacoustics at play. If so, plenty of valve amp enthusiasts are going feel vindicated.
 

anonymous

New member
Listen and decide: are you certain which comes first: the clipped or natural version? Is your audio memory sufficiently reliable (and unlike the goldfish memory), that you could leave the room for a few seconds and return with absolute confidence that you are or are not listening to the original source? I couldn't: it's a very close call indeed.

>
Admittedly, I only listened on my laptop speakers, but they sounded virtually the same to me.{Moderator's comment: no preference at all? One clearer or cleaner than the other?}

ADDED following moderator's comment - I suppose I was a bit hasty. I listened again. I would guess that the first half of the clip was the original. It sounded much brighter, almost as if there was more high frequency information. I preferred the second clip because it did not sound quite so shrill.

To be perfectly honest, the first part of the clip sounded so bright and shrill through my laptop speakers, I think my attention turned away. Thus, my appraisal of virtually the same.

But, forcing myself to remain attentive, I definitely prefer the second half of the sample, clipped or not.
 

A.S.

Administrator
Staff member
You have the means ...

You have the means ...

And that is the playing field for a large majority of valve amplifiers. Perhaps it is how they sound when clipping that is part of the reason for the claimed warm, smooth, liquid valve sound, which may get reinforced by some pyschoacoustics at play. If so, plenty of valve amp enthusiasts are going feel vindicated.
Nope, that's far, far too big a deductive leap at this stage. Audio simply isn't that neat a problem and solution. For a start, these terms 'liquid', 'smooth', 'warm' are not defined and the assumption made is that all value amps exhibit these. Could there be a valve amp that sounds 'bright' and 'detailed'?

We have to break this one down into digestible chunks and build up (and test) a theory correlating clipping with audibility. I gave you the tools to start that process in my post #73....
 

Kumar Kane

New member
Only 'perhaps'

Only 'perhaps'

Nope, that's far, far too big a deductive leap at this stage.
Of course - that is why I used the word "Perhaps" and more words on its lines. I look forward to seeing a validation or otherwise of the proposition I made. Hopefully on this thread, which has turned out to be useful after a doubtful start.
 

anonymous

New member
Clipping demo - confusion?

Clipping demo - confusion?

Admittedly, I only listened on my laptop speakers, but they sounded virtually the same to me.{Moderator's comment: no preference at all? One clearer or cleaner than the other?}

ADDED following moderator's comment - I suppose I was a bit hasty. I listened again. I would guess that the first half of the clip was the original. It sounded much brighter, almost as if there was more high frequency information. I preferred the second clip because it did not sound quite so shrill.

To be perfectly honest, the first part of the clip sounded so bright and shrill through my laptop speakers, I think my attention turned away. Thus, my appraisal of virtually the same.

But, forcing myself to remain attentive, I definitely prefer the second half of the sample, clipped or not.
Well, curiosity got the best of me. I dug out a little dac/headphone amp I have and listened on some in-ear monitors I have. The first clip doesn't sound shrill anymore, but I'm stumped. Even scrolling quickly between the versions, they sound very close. I still prefer the second version, but I have less faith in my previous reasoning that the first half was the natural version. With the shrillness of my laptop speakers out of the equation, it strikes me that the second half is the natural version. It seems every so slightly fuller sounding in the higher frequencies.

Now I'm really eager to see other people's votes and have the truth revealed.
 

A.S.

Administrator
Staff member
Getting to grips with 'clipping' audibility

Getting to grips with 'clipping' audibility

If we achieve nothing else here ever on the HUG, let's run this one down to the wire. It's an excellent illustration of psychoacoustics which, like it or not, distrust it or not, fear it or not, loathe it or not, predicts with remarkable insight how we will emotionally respond to different sound events.

Of course, there are no absolutes when it comes to preferences. A preference is personal: it is valid only to the beholder. And nobody wants to ridicule a preference, merely to highlight how preferences cannot be universal. Because of that, so much comment about audio equipment really should not be voiced, because whilst wholly and absolutely valid to the beholder, it is unlikely to be fully valid for another individual. So you are are right as the next chap if you subjectively prefer one clip over another.

But as to which one is clipped and which isn't, is a matter of objective fact.
 

mhennessy

Member
Brass clipping example

Brass clipping example

The clipped excerpt is obvious to me - but I'm accustomed to listening for this sort of thing.

I did get a brief amount of time to experiment with Audacity last night - am I able to upload pictures and audio here? I couldn't immediately see the option...

I'm a bit short of time today, so hope to get back to this later tonight or tomorrow, but in the meantime, take a look at a quick grab of the Diana Krall track I keep mentioning ("From This Moment On", from the album of the same name). Notice the large positive-going "spikes"; that's the brass instruments that are very forward in the mix - I'd guess that the peaks are roughly 12dB above the rest. In power terms, 12dB is 16 times greater! Compared to the piano excerpt, there is a lot of dynamic range and HF in this track.

http://www.flickr.com/photos/[email protected]/9416738789/

All the best,

Mark
 
Status
Not open for further replies.
Top