Dynamic Range = Difference between loudest and softest portions
Crest Factor = Ratio of peak values to average values
The DR database is not dynamic range even though they call it that. It is crest factor. Crest factor has little to do with the dynamic range of the music. A metric similar to dynamic range is the
R128 Loudness Range.
R128 Loudness Unit Algorithm Description
The input to the algorithm is a vector of loudness levels, computed as specified in ITU-R BS.1770
[2] with different relative-threshold gating, using a sliding analysis-window of length 3 seconds for
integration. An overlap between consecutive analysis-windows must be used in order to prevent
loss of precision in the measurement of shorter programmes. A minimum block overlap of 66% (i.e.
a minimum 2 s of overlap) between consecutive analysis windows is required; the exact amount of
overlap is implementation-dependent.
A cascaded gating scheme is employed which uses an absolute threshold of very low level, in
combination with a relative threshold of higher, signal-dependent, level.
The purpose of the relative-threshold gating is to gate out any periods of silence or background
noise, using a method that is independent of any level-normalisation of the input signal. The lower
edge of Loudness Range should not be defined by the noise floor (which may be inaudible), but
should instead correspond to the weakest ‘real’ signal. The relative threshold is set to a level of
-20 LU relative to the absolute-gated loudness level. The purpose of the absolute-threshold gate is
to make the conversion from the relative threshold to an absolute level robust against longer
periods of silence or low-level background noise. The absolute threshold is set to -70 LUFS,
because no relevant signals are generally found below this loudness level.
I have 3 versions of Elton John's
Madman Across the Water. Here are the DR and LRA of each album for the Madman Across the Water song:
DR 13, LRA 13.6 (DJM Records W. Germany)
DR 10, LRA 11.7 (MCA Records USA)
DR 8, LRA 13.3 (HD Tracks)
Another metric that is useful is R128 dBTP (db True Peak). There are intersample peaks which can be higher than 0 dB. These can lead to distortion if the digital levels are not attenuated (i.e. using analog volume control). Here are the dBTP for the same 3 albums:
+0.5 dBTP (DJM Records W. Germany)
-0.4 dBTP (MCA Records USA)
+0.4 dBTP (HD Tracks)
You can see that while the HDTracks version has the lowest DR rating, it maintains is Loudness Range (or dynamic range) and is the only one that won't clip or cause distortion if one isn't using Volume Leveling or digital attenuation. I think the reason the DJM album has a higher DR number is because it has a higher noise floor. This is probably similar to MadFloyd's suggestion that the surface noise of records causes a higher DR number.