Recording technology improved gradually from the commencement of electronic recording in 1926 until the end of the shellac age, after 1950. This involved among other things better cutting tecniques, increasing the frequency response of the records. At the same time gradual improvements in microphone technology also increased the range of the records. To demonstrate this we focus on two examples of the Edad de oro, D’Arienzo’s 1935 Hotel Victoria and Pugliese’s 1944 Recuerdo. As before we will be using the TangoTunes’ versions as definitive copies, seeing how we can equalize other versions to approximate them.

1. D’Arienzo’s Hotel Victoria from 1935,

We will be comparing two versions of Hotel Victoria with the TangoTunes edition, the first from Audio Park and the second from El BAndoneón – Blue Moon. Both are reasonably good transfers that require different methods of equalisation. 

1.a TangoTunes vs Audio Park

This is a comparison of two versions of D’Arienzo’s Hotel Victoria. In blue, the TangoTunes version, in green the one from Audio Park. We see that the higm mids follow a lower curve in the Audio Park version and that the bass drops off too suddenly below 100 Hz.

This is how we would try to address most of the differnces, with an increase in the high-mids by about 5 dB, centered around 3.5 dB and enhanced bass by about 4 dB centered around 80 Hz.

Here is the result of the above adjustment, bringing the balance closer to the TangoTunes version.

1.b TAngoTunes vs El Bandoneón – Blue Moon

In the El Bandoneón – Blue Moon version the bass follows a similar line as TangoTunes, while the mids and highmids drop of gradually after 500 Hz.

Her we see that we do not need to adjust the bass at all, except to shlf the lower frequencies. The mids and high-mids we increase by a wideranging rise from about 500 Hz with a peak of 6 dB at 3.5 kHz.

Here we can observe the result of the above curve: a gradual enhancement in the mids and high-mids.

2. Pugliese’s Recuerdo from 1944

As we’ve seen before, most standard versions of the popular Recuerdo by Pugliese from 1944 seem to stem from the same ill-balanced transfer. We will be comparing the TangoTunes version of the song to the EMI version, where the disparities are the most obvious. Following that we will also examine a more balanced version, from CTA. Note that by now the frequency range has expanded, especially in the upper levels, where there is by now usable information in the lower part of the treble range, up to about 8 kHz.

Here is a comparison between the frequency response of two versions of Recuerdo, from TangoTunes (in blue) and EMI (in green). The great disparity between the two versions is obvious.

To adjust the response of the EMI version to approximate the one by TangoTunes, we need quite drastic measures. First we raise the mids and high-mids by a total of 12 dB, in a wide spectrum with the peak around 4 kHz. The bass needs to be raised a bit less, by around 8.5 dB peaking at 75 Hz.

Here we see the result of the above equalisation, with both the bass and upper registries raised a great deal.

Above is a sample from the TangoTunes version, for comparison.

Here we have the EMI version. Notice the imbalance of the sound, it’s lacking in strength with the highs sounding muddy.

In the third example, we have the EMI version with equalisation applied. Now the music sounds much sharper and clearer, even though the original is quite faulty. Of course the sound quality of the TangoTunes version is superior, because although we are able to repair some of the faults a lot of information has been lost in the original transfer process, something we cannot possibly heal.

The CTA version is noticably a lot better balanced than the EMI one. It has slightly stronger bass compared to TangoTunes with the high-mids falling off a bit earlier. One can speculate that the CTA engineers are not adjusting for the increased upper range. Non-the-less  only a subtle adjustment is needed to align the two versions.

Here we see how we can adjust the CTA to similar levels as the TangoTunes version. We reduce the bass a bit, from 100 Hz, and put a 10 dB spike in the treble range, around 8 kHz.

We can see here that with the adjustment the files are now better aligned.

Above is the unequalised version of the CTA Recuerdo.

And here it is with the above equalisation curve applied. Notice a slightly more open sound in the upper registries.