function y = dbspl(insig,varargin)
%DBSPL Calculates the SPL (in dB) of a signal
% Usage: y = dbspl(insig);
% y = dbspl(lvl);
% y = dbspl(insig,'ac');
%
% `dbspl(insig)` calculates the sound pressure level (SPL) in dB of *insig*
% considering the AMT default level convention, which is per default SPL of
% 93.98 dB for an RMS of 1.
%
% `dbspl(lin)` outputs the SPL in dB for the linear level of *lin*.
%
% `dbspl(insig,'dboffset',dboffset)` computes the SPL (in dB) of *insig*
% for a level convention different than the AMT default by considering
% an SPL offset given by *dboffset* such that $dbspl = 20*log10(rms(insig)) + dboffset$.
% Some commonly used offsets are:
%
% * $dboffset = 0$. With this offset, $dbspl(insig) = 20*log10(rms(insig))$.
% This convention was used for the development of the
% |lopezpoveda2001| and |breebaart2001|.
%
% * $dboffset = -20*log10(20e-6) = 93.98$. This corresponds to the
% convention of the SPL reference of $20 \mu Pa$. This addresses
% signals representing pressure in Pascal.
%
% * $dboffset = 100$. This convention was used for the development of |dau1996|.
%
% The AMT default level convention can be obtained by::
%
% dboffset = dbspl(1);
%
% In addition, DBSPL takes the following flags at the end of the line of
% input parameters:
%
% 'ac' Consider only the AC component of the signal (i.e. the mean is
% removed).
%
% 'dim',d Work along specified dimension. The default value of []
% means to work along the first non-singleton one.
%
% See also: scaletodbspl
%
% References: moore2003introduction
% #Author: Hagen Wierstorf (2009): original implementation as rmsdb
% #Author: Peter Soendergaard (2009-2013): improvements and adaptations as dbspl
% #Author: Piotr Majdak (2021): adaptations for the AMT 1.0
definput.keyvals.dim=[];
definput.flags.mean={'noac','ac'};
definput.keyvals.dboffset=93.98;
[flags,kv]=ltfatarghelper({'dim','dboffset'},definput,varargin);
% ------ Computation --------------------------
% The level of a signal in dB SPL is given by the following formula:
% level = 20*log10(p/p_0)
% To get to the standard used in the toolbox.
y = 20*log10( rms(insig,flags.mean) )+kv.dboffset;