Wikipedia:Featured sound candidates/Rusalka

From Wikipedia, the free encyclopedia

Rusalka[edit]

One of my favourite operas - I named my cat after it - and I thought it was time I tried another Destinn piece. Unfortunately, the recording was somewhat degraded, and there was only so far I could improve it without losing Destinn herself - indeed, this was a particularly difficult one, and I was using it to make a fairly comprehensive teaching tool on sound restoration! Oh, dear! Wrong choice! Still, I think it's a superb performance, and a wonderful example of both Dvorak and Destinn's skill. Also, the original can be found at File:Emmy_Destinn_-_Antonin_Dvorak_-_Rusalka_-_Song_to_the_Moon.ogg. It's my belief that the rareness of free-licenced Dvorak recordings, combined with the excellence of Destinn make up for the problems; however, there can certainly be reasonable disagreement on this point. By the way, Destinn's Czech, Dvorak's Czech, the opera's in Czech. This recording is in German. Oh, those crazy early 20th century people!

  • Nominate and support. Shoemaker's Holiday Over 206 FCs served 07:43, 20 September 2009 (UTC)[reply]
  • Context is important here. It might be featurable as an illustration of the singing of Emmy Destinn, but I think the fact that it is in German disqualifies it from being featurable as an illustration in Rusalka (opera). Thatcher 13:37, 20 September 2009 (UTC)[reply]
    • Unfortnately, it's fairly typical of the period: I have a news clipping somewhere announcing that Wagner's new opera will premiere in London soon: they're already hard at work translating it into Italian. See, there was this belief in the Victorian period that Italian was the best language for singing... Shoemaker's Holiday Over 206 FCs served 20:46, 20 September 2009 (UTC)[reply]
      • Not just Italian. Opera was often performed in the local vernacular around the turn of the century. Back then they didn't have the fancy subtitles that we do today, and the libretti weren't as well-known (especially with new operas). I don't think the language is an issue for making this a featured sound, although it should be noted that this is a performance in translation. ThemFromSpace 18:17, 25 September 2009 (UTC)[reply]
  • Oppose This version is interesting, different rhythm of German translation significantly influenced the singer's phrasing and "Song to the Moon" became rather clumsy. However, this is not a problem - the file is an illustration of usual practice, mostly abandoned over the past century. (Btw, the translating of opera libretti was not crazy, but logical step, and we, early 21th century people are maybe more crazy :)) I don't like the unpleasant and distorted sound of this recording. --Vejvančický (talk) 12:54, 21 September 2009 (UTC)[reply]
Unfrtunately, the record was severely degraded. Shoemaker's Holiday Over 206 FCs served 15:12, 21 September 2009 (UTC)[reply]
  • Comment I was looking for some excuse to take a break from a work-related project and decided to play with some audio restoration on a random featured sound candidate, and this was the one. What I ended up with is a pretty drastic change from the original, so maybe it's too far. But I figured I'd mention it here to see what you guys thought. It's File:Emmy Destinn - Antonin Dvorak - Rusalka - Song to the Moon-Edits By Kevin McCoy.ogg. Thanks. kmccoy (talk) 20:40, 23 September 2009 (UTC)[reply]
    • Comment Good work Kevin, you completely removed the unpleasant hiss. Your restoration is nice, but there is a new element, a strange echo - it sounds like Emmy singing in a cave. Is it possible to reduce that? I'm sorry for being too pedantic. --Vejvančický (talk) 06:53, 25 September 2009 (UTC)[reply]
      • You're not being too pedantic at all. The "strange echo" you're hearing is an artifact of the noise removal. It's an unfortunate tradeoff between hiss and this effect you're hearing (many people describe it as a "digital" sound, because it sounds a little like the artifacts you get from a heavily compressed mp3 file.) I did a few things to alleviate the sound, including a low-pass filter (there was not really any useful audio information above 4500 Hz or so, only hiss, pops, and artifacts), and several hours doing spot fixes, where I removed specific frequencies in small sections of time (rather than a song-wide EQ). I still play with it in my free time, but there's certainly a point of diminishing returns, and I think I may have reached it (or maybe even passed it :P). If I come up with something better, I'll upload it and leave a note. Thanks a lot for the feedback! kmccoy (talk) 17:49, 25 September 2009 (UTC)[reply]
Thank you for your effort and also for interesting and clear explanation, now I understand quite well. --Vejvančický (talk) 10:27, 26 September 2009 (UTC)[reply]
  • Oppose Weakly and with regret. As I noted above the language issue isn't a concern to me. The overall fidelity is lacking a little and her final notes blasted quite a bit. I also oppose the restored version; it takes way too much music out of the performance. It's better to leave some hiss in, which gives the illusion of a greater frequency range, than to remove too much which makes the performance sound like its done underwater. ThemFromSpace 18:17, 25 September 2009 (UTC)[reply]