Level and Delay for Auralisation Files
How can I handle in EASE/EARS signals (WAV files) which have different levels and onset times (delays) ?
EASE 4.0 Only: EASE 4.0 supports an extended version of the WAV file format. It is completely compatible to the standard WAV definition, but allows to store level and delay information directly inside the file.
EASE 3.0 Only:In the DOS versions EASE 2.x and EARS 1.x signals were stored in TIM files. The TIM files, if created by EARS, contained data about their SPL and delay. This information was used when
- TIM files were mixed
- TIM files were converted to WAV in the "Joint" option.
TIM files contain, like standard WAV files, 16 bit integer numbers for the description of sound pressure amplitudes. Their inherent signal/noise ratio is (except 1 bit used for the sign) S/N = 20 log (2^15) = 90 dB.
Upon mixing, the amplitudes of the input files are added with weights due their SPL values and shifted in the file position due to their delay diffrence and the sampling rate. The resulting set of amplitudes is scaled down then such that the maximum amplitude does not exceed 2^15 - 1 = 32767.
In the "joint conversion" process, all input TIM files were searched for the one with maximum SPL among them. This TIM file was converted to WAV without altering the signal. All other TIM files were scaled down due to their SPL difference to the "loudest" one, and then converted into the WAV structure.
As a result, the created WAV files appear with their "true" relative volumes when played through a sound card one after another.
The main disadvantage of this method is that all those WAV files which have been scaled down are damaged in their SN ratio in an irreversible way. If scaled up again to full amplitudes of 2^15, they carry a significantly increased numerical noise level.
In the present EASE 3.0 Ears Modules, TIM files are no longer actively used, instead, signals are stored in standard Windows WAV files. These do not carry explicit SPL and delay information. In order not to interfere with the common WAV format, EASE stores level and delay separately in a WAV extension file, WAX. WAX files are created by EASE automatically when the EARS convolver is used.
(To do this, start EASE EARS, goto menu Tools, Auralisation, which brings up the Auralisation window. On its top right, activate the big "EARS" button in the "Convolver" frame.)
WAX files are automatically used by EARS without notice. If there is no or a wrong WAX file, EARS assumes SPL 0 and delay 0.
The mixer function is implemented in EASE 3.0 in a very similar way as it is in EARS 1.x. Additionally to its predecessor, however, you can use several input channnels simultaneously, and you can modify their SPL and delay info manually. The mixer is available in EASE EARS, menu Tools, WAV File Mixer.
We have refrained in EASE 3.0 from a "Joint Conversion" because i) there is no conversion to WAV anymore, ii) the EARS WAV player produces the same acoustical impression without damaging the files themselves.
To use the EARS WAV Player, start EASE EARS, goto menu Tools, Play WAV File, Ears. The player comes up after input file selection.
Activate "[x] Set Volume to File Level" at the window bottom.
Enter the reference level into the "Max dB" field, and the current volume to be played into the "Volume dB" field. When playing, the volume is down by the difference between "Max" and "Volume".
When you open another input WAV file now, its level information is automatically copied into the "Volume dB" field. If, and only if, this level exceeds the current "Max dB" number, this value is incremented to the current max SPL.
Once you have played all files you intend to compare, the "Max" field contains the highest level occuring among them (unless you have entered a suitable Max value manually from the beginning).
When you now play the files in any succession, they are audible as loud as their relative SPL prescribes.