Tango Tunes

2013 February 20
by Simba


The first results are trickling down from the TangoVia project. Tango Tunes is a Vienna based organisation, digitizing music with the help from TangoVia. They recently provided a preview of what they will offer, and I have to say I think it sounds very good. The Troilo sample in particular was very promising. The first songs available for purchase are 100 Biagi songs.  The price is EUR 100, which is expensive compared with commercial releases, but cheaper than the Japanese labels.

What it comes down to then is selection and transfer quality. The really great thing is that they will offer lossless files in high resolution (96 kHz, 24 bit) for those that have time to spend restoring music. We others can just downsample to 44.1kHz, 16 bit. Unfortunately, Biagi is not high on my prioriy when it comes to repurchasing tango music. The commercial releases are already ok quality. Troilo on the other hand, will put me first in line. Same goes for De Caro, and probably also Di Sarli. D’Arienzo will also be interesting for many DJs, I think.

I don’t know how they plan on rolling this out later, but making complete discographies available would be another thing to make them stand out. I’m crossing my fingers for complete Troilo later this year.



8 Responses leave one →
  1. 2013 February 21

    Thanks for the information. I have to confess: I am not convinced of the sound-quality yet. The frequency distribution is unbalanced and the Biagis tend to have very aggressive high-frequencies. Another point is the distribution of 96kHz material. Either a DJ knows exactly what he is doing or he has a good chance to experience some strange behaviour of his equipment.

    I think I have to write an article too.

    Kind regards from Germany.


  2. 2013 February 21

    Thanks for commenting, cassiel.

    In what way do you find the frequency distribution unbalanced? (too much bass, low pass?)

    Personally, I think 96kHz only makes sense if you are going to do restoration, change tempo etc. I would certainly downsample to 44.1kHz to use at milongas and for my home stereo. I guess the lossy option (with low pass?) is supposed to be for direct playback, but I would prefer lossless there too, as you could get 44.1kHz lossless at approximately the same bitrate and avoid all lossy artefacts. Anyway, it seems that getting the 96kHz version (with no low pass) and downsample resolves this issue, so I don’t think it’s a big problem.

    I look forward to reading your comments. Please add a link to your article here when you got it up.

  3. 2013 February 21

    In the meantime I have published my article:

    Vielleicht so? Höchst subjektive Anmerkungen zur Klangqualität in der Milonga und zur Frage nach hochaufgelösten Audio-Daten der klassischen Tangos

    (in german – Google-Translation is linked in the page footer)

    I would like to go a step further: nobody really needs 96kHz sampling frequency and 24 bit resolution. In fact this causes additional difficulties. You need an accurate product for changing the sampling frequency (audacity won’t do that good enough). A friend of mine converted some files for me with Saracon by Weiss Audio (maybe the only suitable software solution).

    Sure! The first impression of these new files is breath-taking. But you can’t listen it for longer than 30 minutes (at least that holds for my person) – these files were destroyed by to much trebble (and therefore to few parts in the bass an middletone-area). It’s a pitty.

    I have listened to the new files carefully and talked to other DJs to be sure. I have contacted Christian Xell and we are discussing the reasons for the observed phenomenons.

  4. 2013 February 21

    Thanks, that was interesting reading. I think the commenters DrR and Chris summarize my position well. I prefer lossless because I have the space, and I encode to lossy for portable use (e.g. on my phone), and want to avoid repeated lossy encoding of the same material. I also removes any worry (theoretical as it may well be) that there will be audible artefacts.

    That said, when you already got the right music, and many djs have already purchased the music several times over, if we are to purchase the same music one more time, it is for the transfer quality, and that means well done transfers from mint condition shellacs.

    Did you purchase the Biagis already? If so, are you experiencing trouble with the flacs or the m4as? I would imagine too loud high frequencies could be easily resolved with an equalizer, no?

    I stand by my assertion that downsampling is not a big issue. I know there have been some problems with resampling in some versions of Audacity, but there are other high quality resamplers available for free, e.g. SoX and SSRC, see http://src.infinitewave.ca/ for test results from a wide range of resamplers (Saracon indeed performs very well).

    Most of these distortions are way below audible levels, not to mention the noise level of material from lo fi shellacs, so I wouldn’t worry too much. There are reasons for making 96kHz/24 bit available for people that know what they are doing, but that is for remastering and restoration only. For normal playback it introduces more problems than it solves, like you say.

  5. 2013 February 22

    I heard the Biagis with a friend of mine. I haven’t bought them yet. But Christian Xell send me some files (including one original-transfer-file). So I listened carefully and I have tried to process the data with a mastering processor with different settings. But I am not an audio-engenieer I think I know my personal limits (hopefully 🙂 ). There is much more possible with this files. That’s sure.

    I think a complete re-mastering is necessary. But let’s wait how the discussion is developing…

  6. 2013 February 22

    Simba wrote:

    The really great thing is that they will offer lossless files in high resolution (96 kHz, 24 bit) for those that have time to spend restoring music. … Personally, I think 96kHz only makes sense if you are going to do restoration, change tempo etc.

    I’d be interested to hear why anyone would think 96kHz, 24 bit would advantage restoration, change tempo etc. And how any advantage could apply to restoration, change tempo etc. but not to listening.

  7. 2013 February 22

    @Cassiel: Yes, it will be interesting to hear when they get everything ready.

    @Chris: Doing restoration in higher resolution avoids some rounding errors, gives you more headroom, higher sampling frequency makes sure you can downsample to delivery sample rate (and not upsample). Higher sampling frequency can possibly be taken advantage of by declicking algorithms, I’m not a DSP expert, but that wouldn’t surprise me.

    How much degradation you actually avoid is another matter, but when you’re restoring, not adding unnecessary additional noise and distortions is a good start, methinks.

  8. 2013 September 17
    Chistopher Everett permalink

    High sampling rates help a lot with audio restoration, at the cost of vastly increased processing time. Overclocked processors and solid state drives are almost necessary here. Even when working with commercial CD releases, I invariably upsample to 96K/32 bit format as an initial step.

    The reason these transfers show excess treble is because that’s what is on the shellac/vinyl … the playback equipment has an EQ curve built in that results in the correct sound. In this case, excess treble is a blessing in the hands of someone with good DSP skills … we can always take away, but we can never put it back once it’s gone.

Leave a Reply

Note: You can use basic XHTML in your comments. Your email address will never be published.

Subscribe to this comment feed via RSS