How is a "warm sound" created in analogue components?
Jan 27, 2016 at 3:47 AM Thread Starter Post #1 of 14

seanwee

Headphoneus Supremus
Joined
Jun 9, 2015
Posts
4,808
Likes
3,381
Hey guys , as you all know at head fi , we often say how warm /bright an amp or DAC sounds but how is this warm sound "acquired"?

What happened in the process of converting from digital to analogue and/or amping the analogue signal?

I hope someone would be able to give a clear-ish answer. THX!!!
 
Jan 27, 2016 at 3:59 AM Post #2 of 14
A nebulous question.  Generally warmer sound is tilted so bass is a bit up and treble is a bit down. 
 
Analog processes aren't always warm.  See early solid state gear which is generally called cold and harsh and sterile. 
 
Jan 27, 2016 at 5:49 AM Post #4 of 14
the idea that a sound is "warm" is basically the feeling that we're hearing more low frequencies than trebles. often it's only a matter of trebles being rolled off a little. so the resulting sound is very much the same as taking the flat signature and applying an analog EQ to it. even the phase shift would be similar as there are only so many ways to alter a frequency response in the analog domain.
 
 in some cases, it can also be helped by distortions. some massive distortions in the low frequencies seem to feel like I'm hearing a little more bass than what the frequency response is showing sometimes. also the kind of distortions can sound harsh or soft(even or odd THD for example), I would imagine harsh sound to be easily associated with bright sound, and veil feeling sound to be associated with warmth. but you need a rather high quantity of distortions for this to matter in any way(like some old tube amps often reaching a few percent of distortions).
 
 
the last possibility comes from headphones that don't have a flat impedance response. in some cases, the impedance of the amp can alter the signature of the headphone. in such a case, if the headphone has a bump in impedance in the low frequencies, then any amp with a high impedance output would "sound" warmer than the same amp with low impedance output.
but if the bump in impedance on the headphone is in the trebles, then the reverse would occur and the high impedance amp would this time be the one feeling colder, making it clear that we're hearing the result of amp+headphone and not just a warm amp vs a cold amp.  but people often don't realize that and end up calling an amplifier or a DAP warm by mistake because they tested it with only one given headphone/IEM.
 
Jan 27, 2016 at 7:49 AM Post #5 of 14
Quote:


Hey guys , as you all know at head fi , we often say how warm /bright an amp or DAC sounds but how is this warm sound "acquired"?

What happened in the process of converting from digital to analogue and/or amping the analogue signal?

I hope someone would be able to give a clear-ish answer. THX!!!

 
You basically roll off treble in favor of midrange and bass. See how you can scratch a blackboard and everyone will cringe just because it hurts? Compare that to when a black Escalade with spinnerZ drives past with enough bass to wake up the whole neighborhood - some will actually cheer, not everyone will cringe, and those who do cringe don't necessarily do so because it hurts their ears but because they're thinking "dude that is soooooo totally ghetto." There's that difference in how higher frequencies are perceived vs midrange and bass. 
 
Basically, some amp designs are designed so that their distortion patterns are biased into producing that kind of response curve - whereas a low distortion, low noise amplifier like the O2 will be perceived as "bright" and "lean" by those who think that "high fidelity" comes from a euphonicating boombasticator with high THD and noise levels. 
 
As to exactly how that's done, you'll have to get into a specific engineering discussion, something I already forgot the explanation for when my friend's professor hung out with us once, but basically it's actually a lot easier to come up with a design that deliberately euphonically boombasticates the signal than it is to make one that can amplify the signal into a high power one where the distortion and noise don't rise as fast as the output level (but then again you can calibrate the graph in any manner you want and that curve will take on any shape I guess, and be convincing unless whoever is looking at it starts comparing it to something else).
 
http://www.head-fi.org/t/20139/euphonic-distortion
 
https://sites.google.com/site/sssglossary/Home/E/euphonic-distortion
 
http://waltcrawford.name/Pioneer.htm
 
http://www.post-gazette.com/business/finance/2013/07/14/Sound-Advice-What-produces-euphonic-sound/stories/201307140216
 
http://www.stereophile.com/reference/406howard/#gOueO6KSOV3iJl7x.97
 
Jan 27, 2016 at 8:29 AM Post #6 of 14
(sniff) Wow thx guys it always seems to be you two answering these kind of questions. Such vast knowledge. Sometimes i wonder if you guys actually have to use this info to make a living instead of it just being a hobby. (sorry for my sloppy-ish english if it came out wrong)
 
Jan 27, 2016 at 11:33 AM Post #7 of 14
  (sniff) Wow thx guys it always seems to be you two answering these kind of questions. Such vast knowledge. Sometimes i wonder if you guys actually have to use this info to make a living instead of it just being a hobby. (sorry for my sloppy-ish english if it came out wrong)

 
If you know how to use PEQ, then you add warmth to the entire chain by adding a PK at 220 Hz of about 1-2 dB with a Q of 1.5.
 
Then you can buy your electronics to be neutral, instead of trying to use electronics as tone control.  Tone should be handled at the transducers and in EQ.
 
Jan 28, 2016 at 9:42 AM Post #8 of 14
   
If you know how to use PEQ, then you add warmth to the entire chain by adding a PK at 220 Hz of about 1-2 dB with a Q of 1.5.
 
Then you can buy your electronics to be neutral, instead of trying to use electronics as tone control.  Tone should be handled at the transducers and in EQ.

Wow ... never knew you could do that with a PEQ. and ive never really understood what the Q ratio does.
 
Jan 28, 2016 at 10:02 AM Post #9 of 14
  Wow ... never knew you could do that with a PEQ. and ive never really understood what the Q ratio does.

 
Q ratio is a measure of how sharp the slope of the peak is and how wide the base is.  High Q is sharp and narrow, low Q is shallow and wide.
 
It also relates to perception.  Higher Q ratios in the bass are less audible, but our ears' sensitivity to Q goes up with frequency.
 
Jan 28, 2016 at 10:07 AM Post #10 of 14
  Wow ... never knew you could do that with a PEQ. and ive never really understood what the Q ratio does.

 
Q-Factor refers to the "width" of the effect. If you use a Peak EQ effect for example a high number means a narrower width of effect, ie it affects the surrounding frequencies less; lower value means wider effect.
 
Light grey - center frequency; blue - narrow Q, purple - wide Q (a cut in both cases)

 
Jan 28, 2016 at 8:53 PM Post #11 of 14
As an additional note, Q is also used to refer to bass alignments of boxes, too.
 
Jan 28, 2016 at 9:52 PM Post #12 of 14
Rane has a handy note here explaining the Q-factor, and its relation to the octave. Even more on filter bandwiths here.
 
One octave, in case it isn't clear, represents one doubling in frequency, so something like 440Hz to 880Hz (A4 to A5 in ISO 16) or 10kHz to 20kHz. The bandwith of our sense of hearing, if defined to go from 20Hz to 20kHz is log2(20,000/20)=9.966, or roughly ten octaves. (or a Q-factor of ~0.03)
 

Users who are viewing this thread

Back
Top