Why is the term "warm" such a controversial subject?

[Robin L]

That's why I argued grainy is more controversial than warm. I never heard the term grainy being used in a professional context, except when referring to an underlying technical problem (e.g. clipping). I expect such problems to be measurable. However, I see "grainy" being used in situations where's there's no indication (incl. via measurement) of a problem.

The following situation gave me the impression of "Grainy" sound---a somewhat worn LP played back through a solid state amp into a speaker with wonky treble. That's a situation I've encountered many times. However, as you point out, in this circumstance there is measurable cause for the effect. I wonder how many "grainy" solid state amps appeared after CDs took over? I know my Marantz 8b [a "warm" sounding amp if ever there be] was grainless, though also effectively "topless as well, which explains part of its syrupy smoothness [though I suspect I'm mostly hearing the "sound" of the output transformers]. So a lot of treble detail was rounded off along that odd midrange presence I notice with transformer coupled vintage tube amps. Fun, but one can miss out on a lot of detail the amp glides over.

 

[egellings]

I have an 8B. Mr. B is a venerable amp! So polite!

 

[Robin L]

I have an 8B. Mr. B is a venerable amp! So polite!

Exactly. Ideal if you only listen to Ella or Sarah.

 

[clearnfc]

i.e., we're talkin' steady state

Yup, steady state...... So, you can see in a really really tiny forum....

 

[clearnfc]

200 active members at a given moment...

Yes, less than 200 active members at any moment (ok, more than 200 once in a blue moon)..... its really very very few compared to bigger forums which has thousands of active members online (Even more guests).

So, if possible I would prefer folks here to use a bit more time and effort on how you all can make this forum more appealing to others. There are reasons why many pple don't even come here anymore. Go check out the old threads from a few years back. There are really many knowledgeable pple who post really good information. But vast majority of them no longer come here...

 

[theREALdotnet]

But even my responses aren't good enough for you. And no matter how many times it's explained that "warm" can legitimately mean different things depending on the context (e.g. warmth in terms of harmonics/distortion as well), it seems you will want to use this to say "see, people use it to mean different things, which means nobody can agree on what it means!" rather than just seek to understand the meanings in context.

You’re responses are as good as anyone’s, and just as different. That’s my only issue with them.

I had assumed the context of “warm” we’re talking about is “sound”, are you saying we need to consider sub-contexts? That we’re not trying to define “warm sound”, but things like:

• warm tonality
• warm distortions
• anything else?

What does it mean when someone claims that a particular playback chain sounds warm? Can we take our pick from the list above, or should we ask follow-up questions? What if an engineer was asked to make a recording sound warmer? Does that mean dialling up the mid-bass? The H2? Anything else, or all of the above?

I don't see any use in trying to convince you of anything at this point. But I'm sure there are lurkers who are getting something out of a bunch of the replies.

It doesn’t matter whether I’m convinced that the term “warm” is well-defined or not, if I’m not comfortable with it I’ll just avoid using it. But the question of this thread was why the term is controversial. So far the respondents that claim it shouldn’t be controversial because it has an obvious meaning that is easy deduced from (sub-)context have not been able to put that into words they themselves all agree on. It doesn’t matter one iota whether I’m happy with their definition, as long as there is one. Or one per sub-context and a way of choosing between them.

 

[xaviescacs]

Yes, less than 200 active members at any moment (ok, more than 200 once in a blue moon)..... its really very very few compared to bigger forums which has thousands of active members online (Even more guests).

So, if possible I would prefer folks here to use a bit more time and effort on how you all can make this forum more appealing to others. There are reasons why many pple don't even come here anymore. Go check out the old threads from a few years back. There are really many knowledgeable pple who post really good information. But vast majority of them no longer come here...

1000 users at a given time, 20% - 80%. Are you aware of other audio forums with more than instant 1000 users? I'm not. Can you give an example please?

 

[MattHooper]

My point is; a tick or crack are nouns, in this case direct references to distinct sounds. The meaning can be found in any dictionary. Warm is an adjective, it describes the character of sound in our case. A warm sounding tick and a warm sounding crack are different sounds.

Which implies "warm" would be just as valid a descriptor to add when characterizing such a difference.

That's why I argued grainy is more controversial than warm. I never heard the term grainy being used in a professional context, except when referring to an underlying technical problem (e.g. clipping). I expect such problems to be measurable. However, I see "grainy" being used in situations where's there's no indication (incl. via measurement) of a problem.

From one of the many threads on the meaning of grainy sound on other forums: "we've discovered that "grainy" as applied to sound means very different things to different people. This is a problem".

It depends on what your point is in calling "grainy" "controversial." If you mean you've seen it applied in dubious ways - e.g. where they may be no measurable difference to even describe - then, sure. But as I mentioned before, that doesn't entail "grainy" can't or isn't ever used to describe something real. Just like the fact some audiophiles will describe a new AC cable as "bright" - it's their imagination - doesn't entail that "bright" isn't also a valid sonic description for real audible differences (e.g. tipped up high frequencies).

Grain isn't a technical term - it's just an attempt to put sonic impressions in to words to try to communicate. It's up to someone else if they want to try and understand how anyone is using that term.

Take a pure tone and start adding intermodulation distortion. Let's say you don't know intermodulation distortion is being added, you only recognize the sound is changing. How would you convey this change in sound from the pure tone to the one with increasing intermodulation?
"Distortion" is one word, but it's also incredibly broad since there can be tons of VERY different sounding distortions. "Grainy" might actually come to mind (or "fuzzy" or "buzzy" or "harsh" or "dissonant" depending on the exact example). Sometimes we have access to direct technical terms, sometimes we don't. But we can still communicate.

If my friend said his record playback was starting to sound "grainy" or "fuzzy" I'd have a general idea of what he means.

 

[MattHooper]

An amplifier company insider contradicts the "mixing engineer" official definition (mid bass boost) claiming the mixing engineers and most everyone else is confused and "warm" sound is really caused by low order distortion masking higher order "harsh" distortion (so "harsh" is the opposite of "warm" not "cool" or "cold"). While other "informed" individuals say "warm" has to do with FR "tilt" (vs a "boost") at certain frequencies. Yet another explanation is that since tube amps are literally warm (temperature) they some how must have a warm sound. While there are worse audiophile adjectives than "warm" it still is a counter productive adjective because it's meaning is not defined or agreed on even by audio industry insiders much less consumers.

Words can have different meanings in different contexts.

Why is this so hard to understand?

Are you going to cherrypick atmosphere's post to conclude "warm" doesn't have a relevant meaning in certain contexts (e.g. in mixing?)
That would be like cherrypicking a man-on-the-street's version of "theory" (to imply "just a hunch"), point out how this "contradicts" the way scientists use "theory," to say "see, nobody can agree, the word is therefore useless."

You’re responses are as good as anyone’s, and just as different. That’s my only issue with them.

I had assumed the context of “warm” we’re talking about is “sound”, are you saying we need to consider sub-contexts? That we’re not trying to define “warm sound”, but things like:

• warm tonality
• warm distortions
• anything else?

What does it mean when someone claims that a particular playback chain sounds warm? Can we take our pick from the list above, or should we ask follow-up questions?

Look at the context in which the term "warm" is being used. Familiarize yourself with the context - e.g. if a mixer is saying he's going to boost eq to give an instrument some more "warmth."

And, yes, if you don't know what any individual is referencing by "warm," ask him/her for more detail. Maybe they will have a nonsense reply, maybe a cogent reply.

Again: the point isn't that "warm" doesn't have some different meanings in audio, or even that some people apply the term vaguely.
The point is that it nevertheless DOES have some meaning depending on where how how it's used. You can either appeal to the different ways it's used to say "well, then it's a useless term" or you can instead try to understand the contexts in which it has meaning. Up to you.

 

[levimax]

Words can have different meanings in different contexts.

Why is this so hard to understand?

What's hard to understand is what "warm" means in any audio context. Mixing engineers and amplifier designers can't even agree what "warm" refers to technically. Is it a specific "FR boost" or a "FR tilt" or "low order distortion masking" or something else? If audio professionals in the same context don't have a consistent definition how is there any use for the term for the consumer where "warm" is mostly used to mean "sounds good" or more usually "sounds better than what you have now so you better buy something new and warmer".

 

[ahofer]

So far the respondents that claim it shouldn’t be controversial because it has an obvious meaning that is easy deduced from (sub-)context have not been able to put that into words they themselves all agree on.

..or testable measurements/observations.

 

[MattHooper]

What's hard to understand is what "warm" means in any audio context.

I've already given links to show the context in which mixers/engineers may talk about audio "warmth."

Mixing engineers and amplifier designers can't even agree what "warm" refers to technically.

Again...this seems a flat refusal to consider context.

Back to my analogy: The layman casual use of 'theory' is different from how scientists use the term 'Theory.' Since you can say those two contexts don't agree...well I guess it's a useless term that doesn't really mean anything. I mean...this is the road you are going down. Can you imagine why a scientist might be tearing her hair out if she's been trying to explain their use of the term 'theory' and you keep pointing to the word used in another context to say 'but since those are different, it's all contradictory'??

It's the same when you say "well an amp designer is talking about warmth in terms of added distortion/harmonics, but a mixer is talking about warmth in terms of adding or subtracting a certain frequency range..therefore nobody agrees and it's all contradictory."

The term can be used for either...in their own context. (And a sound engineer for instance, can be talking about adding warmth in terms of harmonic or distortion 'warmth,' or the 'warmth frequency range' for a particular instrument...depending on context.

Is it a specific "FR boost" or a "FR tilt" or "low order distortion masking" or something else? If audio professionals in the same context don't have a consistent definition how is there any use for the term for the consumer where "warm" is mostly used to mean "sounds good" or more usually "sounds better than what you have now so you better buy something new and warmer".

Again...it's both somewhat exact, depending on the context (e.g. certain frequency ranges) and less so in others, but it doesn't mean it "doesn't mean or communicate anything."

It's like the term "thin" or "thick" sound. Clients, mixers, sound editors will often remark that a track may sound "too thin" and we can generally understand what someone is getting at. And you can't point to any particular frequency or "fix" simply from the word "thin" or "wanting X to sound thicker." That will depend on the specific sound being referenced. It could be an entire combination of tracks being mixed. Or for instance in music, someone could say "My guitar tone is too thin, I want to thicken it." That may have a different sonic solution than "the vocals sound too thin, I want it to sound thicker." Or a drum snare is too thin sounding, make it thicker sounding. Or strings. Or tuba. The "fix" for changing the sound in any particular context will be different, but that doesn't mean describing a sound as "thin" or "thicker" has no reference, or meaning, or usefulness. It will depend on the context: what it is you are referring to as "thin" or "needing to sound thicker."

As I've said, people resisting these terms likely don't work in professional sound, where we communicate with subjective terms all the time.
It's not an exact science, but it's useful.

 

[MattHooper]

..or testable measurements/observations.

A boost or dip of 7dB at 150Hz can make a measurable and audible change in the character of many recorded male voices.

A sufficient addition of second harmonics will be audible, often as a slight thickening of tone.

Are you actually disputing this?

If not, the question is whether some people find this gives a character they'd describe as differences in the "warmth" of a sound.

Turns out...yes, they do. That too is indisputable, since examples abound.

 

[pseudoid]

... Through the process of bifurcation at the feedback node (which usually has its own non-linearities) feedback would add distortions of its own (Norman Crowhurst and Peter Baxandall both wrote about this issue). This has given feedback an undeserved bad rap in high end audio...

Thank you for your informative post/link and I hope you won't mind me to continue considering feedback as a band-aid fix.

With your help, I've come to the conclusion that the descriptive audio-word "warm" is just a generality indicating a not-so negative subjective listening evaluation; especially in comparison to the negative remarks, such as "grainy" or even "sounds like sh*t" (SLS).

 

[Geert]

Which implies "warm" would be just as valid a descriptor to add when characterizing such a difference.

...

I never rejected the meaning of warmth in the context of sound. I can even attest that it's commonly used by musicians and sound engineers

Grain isn't a technical term

I see, that explains why it wasn't mentioned in my EE studies.

"Distortion" is one word, but it's also incredibly broad since there can be tons of VERY different sounding distortions. "Grainy" might actually come to mind (or "fuzzy" or "buzzy" or "harsh" or "dissonant" depending on the exact example).

And theirin lies the problem. Because we also know "harsh" in the pro audio world, and it's something different than grainy. Grainy is like "fractured sound on a micro level'. Speakers don't do that, unless when overpowered. But they can sound harsh for sure. Also unpleasant, but without the fractured (grain) aspect.

 

[Axo1989]

In case its not been mentioned directly, 'warmth' is not a frequency response error, its caused by (since it is audible) the 2nd and/or 3rd harmonics, usually profound enough that they are able to mask the higher orders (if they were not, 'harsh' might be the descriptor instead). The ear converts all forms of distortion to tonality, which is why audiophiles developed this term. If you are not familiar with it, you might construe it as a bit of extra mid bass or the like. But IME audiophiles describe that as a 'mid bass hump' or the like, thus drawing a distinction.

All single-ended discrete circuits whether tube or solid state will generate a 2nd harmonic (which is an indication of a quadratic non-linearity; fully balanced/differential circuits tend to suppress the even orders throughout the circuit and so are better described as having a cubic non-linearity). I suspect a lot of people aren't used to the idea that distortion is often audible (where the signal isn't being clipped). Feedback is often used to suppress innate distortions such as this. The problem has been that up until about 25 years ago or so, the devices needed to design an amplifier circuit where you really could run the high amounts of feedback necessary didn't exist. Through the process of bifurcation at the feedback node (which usually has its own non-linearities) feedback would add distortions of its own (Norman Crowhurst and Peter Baxandall both wrote about this issue). This has given feedback an undeserved bad rap in high end audio.

The second and 3rd harmonics can mask higher orders, but if the former are suppressed by feedback this can result in the amp being bright and harsh instead of 'warm'. The reason is the ear uses the higher orders to sense sound pressure (this is readily demonstrated with simple test equipment) and so is keenly sensitive to their presence even in tiny amounts. This is what has kept tubes in business since being declared obsolete 60 years ago.

So for decades audiophile have had to chose their poison- warmth, or more detail and more neutral, but with harshness and brightness. Both are colorations (if we're being honest). You need a lot of gain bandwidth product to support high amounts of feedback; if the GBP is insufficient, feedback will fall off as frequency is increased (the two are very directly related, see the link below), resulting in distortion rising with frequency. That describes all tube amplifiers and most solid state amps made. Its only been recently that this problem could be overcome, but some still haven't got the memo.

Bruno Putzeys wrote a nice primer on this topic:
https://linearaudio.net/sites/linearaudio.net/files/volume1bp.pdf

That was informative, and fleshes out the "analog warmth" usage suggested early in the thread.

I don't think it obviates usage relating to deliberate FR manipulation in music production, or voicing of speakers (setting aside any argument about desirability or otherwise of the latter). I don't see it as problematic, controversial or even surprising that audio producers employ different vernacular idioms (and different usage of the same word) compared to electrical engineers, for example. Different meanings in different working contexts.

Putzeys certainly rehabilitates negative feedback via his novel work on class D amplification. Unless I missed it, he doesn't use the term warm in his paper however (apart from one instance of 'warm and fuzzy'). Which doesn't take anything away from your description, just interesting.

 

[Axo1989]

What's hard to understand is what "warm" means in any audio context. Mixing engineers and amplifier designers can't even agree what "warm" refers to technically. Is it a specific "FR boost" or a "FR tilt" or "low order distortion masking" or something else? If audio professionals in the same context don't have a consistent definition how is there any use for the term for the consumer where "warm" is mostly used to mean "sounds good" or more usually "sounds better than what you have now so you better buy something new and warmer".

I sympathise with the ideal of clear, unambiguous terminology. But insisting on absolute consistency for terms of art it is tilting at windmills all the same. We might wish pragmatically that English was linear/deterministic, but in practice it is fuzzy/contextual. Maybe it's better in German?

I wouldn't say that electrical engineers designing amplifiers and music producers (along with artists and mixers) laying down tracks in the studio are audio professionals in the same context. That's too broad, for the purpose of analysing terminology or terms of art, especially.

Edit: I don't really get the a priori difficulty imposed by context either, at least it isn't insurmountable. An amp designer, or a guitarist or synth player warming up their sound is most likely referring to the distortion profile. A speaker designer is more likely referring to FR tonality. A music producer could be referring to either, depending whether they are playing with individual instrument tracks or overall mix. More or less detail could refer to overall tonality or more specific bands, depending on the focus. We resolve these fairly easily in idiomatic conversational exchanges. We might go into definitions more in a written paper. And so on.

 

[MattHooper]

And theirin lies the problem. Because we also know "harsh" in the pro audio world, and it's something different than grainy. Grainy is like "fractured sound on a micro level'. Speakers don't do that, unless when overpowered. But they can sound harsh for sure. Also unpleasant, but without the fractured (grain) aspect.

I'm unsure as to whether you are rejecting the description "grainy" for reference to anything in audio. Or whether you accept it could be a valid sonic description, but reject that "grainy sound" actually occurs in specific examples (like from speakers).

?

I mean, we could have the same conversation about "gritty" sound or "fuzzy" or "diffuse" or "clean" or countless other sonic descriptions.
Any non-technical attempt to put impressions in to words.

So for grainy, an example:

I play with acoustics in my room, usually by moving thick curtains along parts of the walls, especially side walls, to modulate reflectivity.
There is nothing controversial about the fact you can make audible changes to sound via introducing more reflections.

I find that cutting down reflections makes for a cleaner, more pure sound of the recording (so long as the recording sounds clean, especially). I was just playing a track with a trio of closely mic'd low woodwinds and the sound was very pure and clean of obvious distortion and richly differentiated in timbre between the players.

I started opening up the curtains allowing more reflections and this had two general effects: Made the sound a bit more present and in the room vs in the studio reverb. But in doing so it also "lightened up" the tonality and, in intermingling the room reflections, I perceived a sort of deterioration of the purity of timbre, like the "aural image" of the woodwinds was now being overlayed by a type of lightened "grain."

The most obvious analogy is to grain in a photograph:

That's exaggerated to make the point, but more examples:

If you just imagine taking out and putting in the grain in those photos, you will get an almost exact visual analogy to what I'm perceiving.
I personally could not find any more perfect examples of what is happening in my mind's eye to the purity and timbre of the sound as more
room reflection mixes with the original recording. Like I'm seeing it through a speckled overlay that starts to mute and homogenize the
clarity of timbral differences.

Now, you could still reject the description "grain" and say either you don't hear it that way or whatever. But there'd be no grounds to tell me that is not the impression it leaves on ME when I'm listening. That really is the most accurate analogy to my subjective perception I could produce.
I'd describe the effect of several of my "low-fi" distorting audio plug ins as adding a granular/grainy quality to sound. I'd describe "brown noise" as having a "grainy quality" relative to a pure tone.

Does that make it hopelessly subjective and useless for communicating? Of course not. If you don't see it that way, I can talk to others who do, and who do understand when I or others use terms like "grainy" to describe sound (which I do with some of my audio pals when applicable, and we get it).

(I'd also say when I compare my tube preamp in my system vs my Benchmark LA4, the most apt description for one difference I percieve is "added graininess" from the tube preamp, much like the examples given above).

Now as to whether speakers can produce a sound where one will sound to someone "more grainy" I think it would be quite rash to reject that out of hand. For instance even changes in speaker dispersion could produce more sidewall reflections in one speaker over another, just like my introducing less or more sidewall reflections, and if I described wider dispersion Speaker A as producing more "grainy sound" in the room than narrow-dispersion speaker B, I can't see how you could say "no, that's incorrect."

And given all the different ways speakers can sound different, the prospect that "sounding more grainy to someone" is certainly plausible.

So for instance take the Devore speakers I mentioned, where I wrote: "one thing that was consistently standing out to me was the Devores sounded a bit more "grainy" - just a slight sort of granular hash slightly obscuring the purity of the tones, somewhat similar to the slight "hashy sound" that occurs when I introduce more sidewall reflections in my room."

Here are the measurements from Stereophile:

DeVore Fidelity Orangutan O/96 loudspeaker Measurements

Sidebar 3: Measurements I used DRA Labs' MLSSA system and a calibrated DPA 4006 microphone to measure the DeVore Fidelity Orangutan O/96's frequency response in the farfield, and an Earthworks QTC-40 for the nearfield and spatially averaged room responses.

www.stereophile.com

There's some funky stuff going on in there. JA mentions as well:

The cumulative spectral-decay plot (fig.9) reveals a generally clean decay in the treble and midrange, but with some low-level hash evident in the low treble and a prominent ridge of resonant energy coincident with the on-axis peak at 1730Hz.

Can you confidently say that I absolutely could not have heard any such artifacts, that didn't strike me as "grainy?" And how would you prove that?

 

[Geert]

I'm unsure as to whether you are rejecting the description "grainy" for reference to anything in audio.

I already did 4 attempts to explain what grainy sound is or what could be the cause. My last attempt to be as clear as possible was:

Grainy is like "fractured sound on a micro level'.

That description was a perfect assist for your example analogy of grain in a photo.

I find that cutting down reflections makes for a cleaner, more pure sound of the recording

Staying with your photography analogy; your auditory system does not has the resolution to hear sound reflections in the way your eyes can see grain on a photo. So 'cleaner' and 'more pure' yes, 'grainy' probably not. The same goes for hearing the decay of a speaker. As an analogy, this is what reflections and decay would do to a picture:

I would call it hazy. As a sound engineer you would emulate this effect by adding reverb, which are ... reflections.

 

[MattHooper]

Thanks Geert. I appreciate the perspective you are bringing, and also what you are looking to achieve.

I think there are two issues entangled here:

1. How any particular sound is perceived by a listener, the subjective impression, and then communicating that.

2. Trying to make a term more technical, by directly relating it to specific, limited, measurable phenomena.

Both are valid ways of communicating about sound. #2 would obviously be even more specific if it could be agreed upon (e.g. if there were some specific, measurable distortion we'd agree to call "grain," even if it didn't necessarily produce that impression for every listener).

As to #1:

Staying with your photography analogy; your auditory system does not has the resolution to hear sound reflections in the way your eyes can see grain on a photo. So 'cleaner' and 'more pure' yes, 'grainy' probably not. The same goes for hearing the decay of a speaker. As an analogy, this is what reflections and decay would do to a picture:

I would call it hazy. As a sound engineer you would emulate this effect by adding reverb, which are ... reflections.

Sure, you may call it "hazy" and since subjective impressions are imprecise we would be on the same page. Which I think is good enough.
However "hazy" does not actually capture what I perceive. "Hazy," like that picture, implies to me a sort of obscuring that leads to a softening of the image and texture. Whereas I perceive an added coarseness to the sonic texture. Just like the grain in those photographs both obscures, homogenizes the clarity, while coarsening the texture of the overall image. So "grain" still to me is the most accurate description.

But, in this realm if you listened and called the effect "hazy" then we are mostly on the same page and that's much better than nothing.

(BTW, as a sound engineer, if I were to replicate my perception of what room reflections are doing to the timbre of recorded instruments, I wouldn't just add short reflections from a reverb, I would mix in what is known as "room tone" (literally recordings of empty rooms, the "air" tone as captured by a microphone, which we use all the time). Many room tones, which amount to a very quiet low level "rush" of apparent air, have this type of "grain" I'm talking about, and slightly overlaying it on to a recording would help get at what I'm hearing.

Anyway, I'm sure we've both gone as far as we can, and want to, on this one, so thank you!