What is a "warm" sound? Options

[zielwolf]

Here's 3 lossyFLAC files in a zip.
The track in question is Alternative Ulster by Stiff Little Fingers (and they've all been through WavGain - 89dB).

The 1st file is the CD version, the 2nd is the vinyl and the third is the vinyl version processed with a plugin called CLAS by Refined Audiometrics.

I find the CD release too harsh and thin ("cold") and the vinyl too muddy ("warm"? - as per what Ed Seedhouse and tfarney said). The 3rd version lies somewhere between the two and that's where I'd have preferred the CD version to have been mastered. I'd say with the CD version they've castrated it for the sake of "digital clarity" i.e. high-end boost and sucked out too much low-mid - IMO it's just a poor mastering choice and obviously nothing to do with the format.

C.

I disagree with the idea that CDs have a harsh sound. It's all very subjective though. Perhaps because a lot of people have grown up with the "warm" sound (in my opinion, "muffled" and "muddy" are better expressions) of analogue devices people prefer it because it's what they're used to. The digital clarity of CDs is awesome I reckon. I still remember the first time I ever played a CD and it just totally blew me away. I had no idea recorded music could sound so incredibly good. If people don't like hi-end frequencies, apply an equalizer. I'd never go back to vinyl because they just sound like crap compared to digital.

[KnobTwiddler]

Contrary to popular belief, "warm" sounds are actually very high frequencies. See, once the oscillation passes 10 kHz the motion becomes rapid enough to raise the temperature of the air by 0.085 degrees Farenheit every hour. That's assuming a constant tone like a pure sine wave at 85 dB at 1 meter, so obviously the real world results vary.

[analog scott]

I agree that all this talk about warm sound and wet sound is ridiculous. While were at it what about this nonsense about "bright" sound? Does the sound change with the lights in the room? Huh? We definitely should stay away from anything but technical terms to describe any sound quality. People start using words like warm or bright the next thing you know audiophiles are having fun. Slew rate is a much better way to describe the sound one hears for sure.

This post has been edited by analog scott: Sep 7 2009, 03:26

[Gag Halfrunt]

I hear alot the term "warm" applied to sound yet I have no idea what is meant. I listen to both digital (CD, files) and analog (vinyl, cassette) music and often hear the term "warm" applied to analog (and "cold" applied to digital). To my ears neither sound "warm" nor "cold".

Can anyone explain what is meant by "warm" and hopefully post or refer to an example.

I don't really get these meanings in terms of audio sound, but I certainly understand them in the way different versions of the same musical instrument. Here's a good example of how the choice of tonewood in a guitar can shape the tonality:

Three almost identical Taylor guitars

I guess the same could apply to a loudspeaker, and lot of even-order harmonic distortion from a tube amp - or tracing distortion from a record player - would make the sound appear 'thicker'. I think that means 'warm' is a polite way of saying 'distorted'

[krabapple]

Keep in mind that a lot of this is a perception problem with early CDs. "Cold" and "brittle" were adjectives tossed around that made some musicians highly suspicious of digital recording vs. analog. I don't know whether or not some of the early CDs capitalized on the wide dynamic range of CDs as compared to typical sound source media and mastered albums with less compression or exaggerated highs--but damage was done to the reputation of digital recording as a result.

For an analog rock recording all of the analog compression that was used during mix and mastering, was there on the early CDs too...captured more faithfully, in fact, than LP ever did.

The only thing missing was distortion added by LP playback itself.

Deficits of early CDs, such as they were (and I believe they are exaggerated) seem to have originated from 1) use of the wrong source tapes (tapes equalized to compensate for deficits of LP playback, will not necessarily sound good if played digitally), 2) lack of dithering and 3) subpar ADC or DAC filtering

[analog scott]

Keep in mind that a lot of this is a perception problem with early CDs. "Cold" and "brittle" were adjectives tossed around that made some musicians highly suspicious of digital recording vs. analog. I don't know whether or not some of the early CDs capitalized on the wide dynamic range of CDs as compared to typical sound source media and mastered albums with less compression or exaggerated highs--but damage was done to the reputation of digital recording as a result.

For an analog rock recording all of the analog compression that was used during mix and mastering, was there on the early CDs too...captured more faithfully, in fact, than LP ever did.

The only thing missing was distortion added by LP playback itself.

Deficits of early CDs, such as they were (and I believe they are exaggerated) seem to have originated from 1) use of the wrong source tapes (tapes equalized to compensate for deficits of LP playback, will not necessarily sound good if played digitally), 2) lack of dithering and 3) subpar ADC or DAC filtering

Why would a source tape EQed for LP mastering not sound good if played digitally? Can you cite some specific examples where a source tape made specifically for an LP wrought poor sound that was specific only to the CD that was sourced from the same tape?


This post has been edited by analog scott: Sep 7 2009, 17:53

[[JAZ]]

Why would a source tape EQed for LP mastering not sound good if played digitally? Can you cite some specific examples where a source tape made specifically for an LP wrought poor sound that was specific only to the CD that was sourced from the same tape?

You may start looking at here:
http://en.wikipedia.org/wiki/RIAA_equalization


Meaning, basically, using an RIAA equalized master needs to be played back with the inverted equalization (which is almost always done when you connect a turntable to an integrated amp or mixer table, but almost never done for line level inputs, like those for CD's).

[analog scott]

Why would a source tape EQed for LP mastering not sound good if played digitally? Can you cite some specific examples where a source tape made specifically for an LP wrought poor sound that was specific only to the CD that was sourced from the same tape?

You may start looking at here:
http://en.wikipedia.org/wiki/RIAA_equalization


Meaning, basically, using an RIAA equalized master needs to be played back with the inverted equalization (which is almost always done when you connect a turntable to an integrated amp or mixer table, but almost never done for line level inputs, like those for CD's).

No the *RIAA* EQ is not on any of the tapes. That EQ is done in the electronics of the cutting lathe. I am quite confident, as bad as so many CDs are for whatever reasons, it has never been because the RIAA EQ has been accidentally applied.

This post has been edited by analog scott: Sep 7 2009, 20:14

[Arnold B. Kruege...]

Here's 3 lossyFLAC files in a zip.
The track in question is Alternative Ulster by Stiff Little Fingers (and they've all been through WavGain - 89dB).

The 1st file is the CD version, the 2nd is the vinyl and the third is the vinyl version processed with a plugin called CLAS by Refined Audiometrics.

You deserve kudos for actually gettting this stuff together and posting it.

I didn't get around to downloading it and listening to it until last night.

I find the CD release too harsh and thin ("cold") and the vinyl too muddy ("warm"? - as per what Ed Seedhouse and tfarney said). The 3rd version lies somewhere between the two and that's where I'd have preferred the CD version to have been mastered. I'd say with the CD version they've castrated it for the sake of "digital clarity" i.e. high-end boost and sucked out too much low-mid - IMO it's just a poor mastering choice and obviously nothing to do with the format.

After a sighted listening evaluation, I agree that the vinyl is too muddy and the CD version is a bit hot. I would call the LP version castrated. I'd prefer a high end balance that was in-between the two. Easy enough to do

I ran FFTs of the CD and LP tracks in CEP, and no surprises. The LP and CD versions were very similar from about 100 to 2500 Hz. The LP version was rolled-off pretty severely below 60 Hz. The CD version was 3-6 dB hotter from 5-12 KHz.

It is a clear case of different mastering.

These tracks measure differently enough that its easy for me to say those often anathmetic words: "These are so obviously different that there's no need for a DBT". ;-) So tried one anyway and of course got 16/16.

My "correction" was simple. I needed to process the CD version with a compensating dip. The geometric mean of 5 and 12 KHz is 8.5 KHz is about 7.8 KHz. I tried the CD version with a 2.2 dB Q= 1.07 dB dip at 7.8 KHz using CoolEdit, and it sounded about right to me.

[audioapprentice]

Deficits of early CDs, such as they were (and I believe they are exaggerated) seem to have originated from 1) use of the wrong source tapes (tapes equalized to compensate for deficits of LP playback, will not necessarily sound good if played digitally), 2) lack of dithering and 3) subpar ADC or DAC filtering

Why would a source tape EQed for LP mastering not sound good if played digitally? Can you cite some specific examples where a source tape made specifically for an LP wrought poor sound that was specific only to the CD that was sourced from the same tape?

AFAIK this means that it was EQed to account for technical things like stylus tracking. So it wouldn't sound worse than the vinyl unless the deficits of points 2 and 3 were greater than the deficits of the vinyl format. It just wouldn't sound as good as if it didn't have to be EQed so that the stylus wouldn't jump out of the record groove.

According to mastering engineer Greg Calibi, in the early days of CD the vinyl master was often used for all 3 formats (vinyl, tape, CD).

[audioapprentice]

In summary, I think most pop/rock music engineers, when they refer to "warmth", are discussing a perception of a combination of well-applied multiband compression, slight even-series analog distortion, and a slight attenuation of highs with a slight boost of low-mids. At least that's what I would mean by that term.

That's a very useful summary from the recording engineer's perspective. Cheers.

[Canar]

I've always associated the "warm" sound with harmonic distortion, specifically the harmonic right above the root. That being said, what sounds "warm" about analog formats is the subtle pitch variations and spectrum response variations that occur naturally on analog media. These have names like wow and flutter. Vinyl doesn't necessarily wear evenly, and tape even less so. The more you play a vinyl, the more worn it becomes, and the wear sounds "warm" to me as well.

"Warm", to me, is a catch-all term roughly describing the ways analog fails at accurate signal reproduction.

[analog scott]

I've always associated the "warm" sound with harmonic distortion, specifically the harmonic right above the root. That being said, what sounds "warm" about analog formats is the subtle pitch variations and spectrum response variations that occur naturally on analog media. These have names like wow and flutter. Vinyl doesn't necessarily wear evenly, and tape even less so. The more you play a vinyl, the more worn it becomes, and the wear sounds "warm" to me as well.

"Warm", to me, is a catch-all term roughly describing the ways analog fails at accurate signal reproduction.

I have always associated warm sound with the sort of sound one hears with mid range heavy live instruments (cellos, bassons, guitarss etc) played in a fairly reverberant envirement but from a mid hall seat or a bit further back. I don't associate it with any distortions that can be identified on a cold listening with no reference as distortions.

[Speedskater]

Way back in July 1993, J. Gordon Holt wrote an article in Stereophile magazine "Sounds Like? An Audio Glossary" and a book "Audio Glossary". Included were words like "warm" and it's equivalent "dark". See the Glossary at the end of the article.

www.stereophile.com/reference/50/

[Richard Greene]

If you ever have the opportunity to hear an orchestra
from really close, like a conductor, the sound has much
more treble energy than it does when you sit far back in the
auditorium seats I can afford.

The sound quality heard by the conductor is detailed and analytical.

The sound heard in the middle of the auditorium. or further back where
I sit. has much less treble energy (air absorbs treble) and much more room
reverberation -- it is a much "warmer" sound quality than the conductor
hears.

In recordings the use of multiple close mikes and digital recording
allows a very flat treble frequency response that can be much more
like what a conductor hears, than like an audience member hears.

If a typical close-miked recording is played at home through speakers
that are not full range, and placed near field, the sound quality
can be very bright, detailed, "analytical" -- maybe that's a good thing
for amatuer conductors who want to hear "details", but it's not the
warmer less detailed more relaxing sound quality you would
normally hear in the audience in the middle of an auditorium.

[dextronaut]

I apologize for necroing this thread! but damn, this discussion made me join this forum. Havent read the last page, but quite interesting. one other word i have to say, methoxetamine.

[db1989]

When bumping threads from so long ago, it’s generally courteous to have something specific to say and to contribute some actual feedback or contribution, especially if one claims to find the subject so interesting.

[Arnold B. Kruege...]

Warmness is also ascribed to recordings with more reverberance or echo to them - in comparison to a raw recording in an acoustically damped environment, which may sound unnaturally silent (especially on headphones).

It has also been used to describe low-order harmonic distortion.

The spectral balance of reverb can vary all over the place. If its tipped towards the bass, then it adds warmth. If it it tipped towards the upper midrange then it makes the sound colder and often harsh.

Nonlinear distortion tends to be higher at both ends of the spectrum. If present in present audible amounts at low frequencies, then it may be perceived as added warmth. Again, if present in audible amounts at high frequencies, then it is perceived as anything but warmth.

Linear distortion that raises response at low frequencies from about 200 Hz down also increases the perception of warmth.

It is probably the low frequency rise in the spectral contents of the sound that creates the perception of warmth, with nonlinearity or reverb being means to that end.

[alanofoz]

Just stirring but...

In order to have a meaningful scientific discussion about warmth we need to first establish the units it's measured in.

Degrees celsius? No - Kelvin. (Degrees Kelvin would be wrong of course).

Actually degrees Fahrenheit might be more appropriate here, given that one of Fahrenheit's reference points was supposed to be the temperature of his wife's armpit...

[prufrock]

Can a sound that is dry ( i.e a person talking in a forest), also be warm? ( I'm using dry in the usual audio terminology. Wet the opposite, i.e sound in a cave or canyon.)
If you take warm in audiophile to be the end product of manipulation, i.e added (pleasing to some) distortion, EQ to taste and other tech tricks, then warm and wet seem better matched.
A solo acoustic guitar played in a forest verses one played in a cave. Which sounds warmer? Richard Greene a few posts back made related comments about auditoriums.

[dhromed]

In order to have a meaningful scientific discussion about warmth we need to first establish the units it's measured in.

intensity in dB and frequency bandwidth in Hz.

[LithosZA]

Instead of a glossary of ‘audiophile’ words with descriptions of the sound in English I think it would better to create a glossary with the each ‘audiophile’ term like this: (So that most people know what the audiophiles are actually talking about)

Warm:
[Click to hear unmodified version]
[Click here to hear modified WAV with a warm sound]
[Technical description over here. Like frequency graph, slew rate, distortion etc.]
[Audiophile description using English]

[db1989]

So that most people know what the audiophiles are actually talking about

A necessary precondition is that they know what they’re talking about…

[mjb2006]

It is probably the low frequency rise in the spectral contents of the sound that creates the perception of warmth, with nonlinearity or reverb being means to that end.

I mostly agree. Listen to the Thomas Köner - Teimo album (from the label's SoundCloud account, don't worry) — it's mostly bass, with no high frequency content whatsoever. So as "dark" ambient drone music goes, it's "warmer" than most, but the metallic dissonance in the lower midrange really makes it difficult for me to categorize the music as "warm". So I would say bass content alone isn't the sole determiner of overall "warmth."

[hellokeith]

Warm is the colorization of sound that a tube gives. Now before you get your DBTs in a knot, this is very simple: tube = warm. Whatever that ends up sounding like, it's warm.