dev builds
themes manual
device status forums
mailing lists
IRC bugs
dev guide

Rockbox mail archive

Subject: RE: rockbox-recorder-20061122 - volume representation messed up

RE: rockbox-recorder-20061122 - volume representation messed up

From: Christopher Woods <>
Date: Tue, 28 Nov 2006 02:31:58 -0000

Ok, so the unit of one 'bel' is nothing more than the representation of an
abstract value. WhatIs defines the Bel as 'the logarithmic ratio between two
levels of signal, power, voltage or current.' dB (Decibel) is commonly used
though, as a unit of 1 Bel is a rather large amount.
Due to its abstract nature, you can effectively use decibels to describe
anything - the temperature, the weight of something even! When my lecturer
told me that, I went '.... buh?' But yes, you just have to do the maths
Two sites of use:
A great resource is Jim Price's 'Understanding dB' article, - it has comparisons between the
various types of dB used in the audio arena. Whenever I think of dB I still
think of dBFS though, nothing beats full-scale imo. ;) However,
unfortunately some of the mixing desks at uni are of the +4dBU variety and
others are of the -10dBV, so you're constantly having to do the mental maths
to make sure you're not oversaturating your buses and you have enough
headroom left. Oh well, one of the 'perks' of the job I guess.
To me, using decibels for the volume scale makes much more sense (provided
that it's one of the recognised decibel scales, as it peaks at 0 I'm
guessing dBFS?). Having an arbitrary linear unit of measurement of say, 0-50
is a bit stupid for audio gear, especially audio gear and/or software
tailored to the more expert user, Sony realised this a long time ago and
used decibel scales for all their audio gear, and my soundcard's third-party
drivers use dBFS faders throughout the control surface. It just... Makes
sense. At least I can be more sure that -6dB on my H140 is most likely going
to be roughly equivalent to -6dB from my sound card's line level out!
instead of having to bring up a set of VU meters and balance the faders each
time myself.
Totally in agreement with the move to decibels for gain representation.
Chalk one up for the audionerds :D


From: Paul Louden []
Sent: 28 November 2006 00:14
To: Rockbox development
Subject: Re: rockbox-recorder-20061122 - volume representation messed up

Well, firstly decibels are a relative scale. So 0 would be "not reduced or
amplified." In that context, 90% could actually be considered the same as
-10%, as they're both "10% less than the full level" but since you don't
have an absolute point of reference with dB, the negative numbers make
perfect sense.

Secondly, -10 is still less than 0, so I still don't really understand where
it's confusing. Higher numbers are still louder, right?

The comparison to -10K is irrelevant because you're comparing an absolute
scale to a relative scale.

On 11/27/06, Nix <> wrote:

On 23 Nov 2006, Linus Nielsen Feltzing told this:
> mat holton wrote:
>> You see, this is the problem when you let audiophiles or programmers
>> design User Interfaces.
> Actually, the dB scale solves a very old issue on the Archos, namely
> the question what volume to set to avoid clipping (0dB). The MAS chip
> can set the volume above 0dB, you see.
> In the old Rockbox version, the answer was 92% (if I recall
> correctly). I think "0dB" conveys that information much better than
> "92%".

Personally, being a mere programmer and not any sort of audio geek, I
don't understand why 0dB doesn't mean `dead silence'. I mean, that's
what decibels are, right, a unit of sound intensity? So how can you have
a negative sound intensity? -10dB reads to me like -10K on an absolute
temperature scale would (and, yes, I know that -10K really does have a
meaning, but it's a rather obscure one that doesn't relate to
*thermodynamic* temperature, i.e. to what most people understand as

I'm an extreme geek compared to pretty much everyone else I know
off-net. I don't think I know *anyone* who wouldn't be confused by a
negative volume. (It confused me enough that I hunted through the source
to fix the bug, saw that it was intentional, and left it alone, shaking
my head over the apparent bizarreness of this scale.)

`The main high-level difference between Emacs and (say) UNIX, Windows,
or BeOS... is that Emacs boots quicker.' --- PdS 
Received on 2006-11-28

Page was last modified "Jan 10 2012" The Rockbox Crew