Wednesday, October 28, 2009

Poor Sound Quality in E-R ??????

There is currently a discussion on the Progressive Ears forum about the quality of the sound of the new Emehntehtt-Re CD. There has also been some discussions on the French Magma forum about the same topic.

Mr. Stigglesworth took a small sample of the CD and created a WAV file and a snapshot of the sound.

http://www.stigmusic.com/fk.wav


And, Udi Koomran (who is the sound engineer for bands like Present and others) said:

[Quote removed by Udi's request]

I am not an audio engineer, so I don't know what "-9.3 RMS" means. But, some people said that the clipping is very noticeable when you listen to it through headphones (which I do).

Maybe Udi and Akoustikus can explain this a bit more to the people (like myself) who are mere mortals.

4 comments:

  1. I will try to explain to you what it means.
    Human ear do not perceives volume in terms of peaks (the brain is too slow to calculate how much is high an instantaneous transient as a kick drum hit, or vocal plosives such as P, S, F or T sounds that are very powerful due to the high amount of air moved) but only a "loudness" that means sensation of volume of the entire sound program (the song). This is called RMS or Root Mean Square and is about to be the square root of the arithmetic mean of the squares between highest and lowest peaks values.
    Who mastered this album have lifted the RMS to -9.3 dB because today, in digital era, is common use to have the higher volume possible in CD music, such as a punch in your stomach, because they studied that the higher is the RMS volume, the better is the reaction who is listening the song to appreciate it. Notice that lifting RMS volume and making all to sound higher don't means to preserve original dynamics (that's the difference between softer and harder sounds, silent and strong sounds, pianissimo and fortissimo), this is due to the fact that in digital world the max scaling reachable is no more than 0 dB, the max level possible in digital CD support in example, then if you want to lift the lower sounds, you can lift also the higher sounds above than this 0dB. In this terms, dynamics are compromised.
    Getting back to the answer: he's telling that there is an high RMS value THEN low dynamics range (piano and forte are very near) and there are other 2 problems:
    first, there is a difference of 2 dB between the channels, this means that the right channel is about to be the half less powerful than the left channel; the second is is an error in a mastering operation called "normalization", which is when you bring the higher peak to a threshold you can choose. They choose to bring the higher peak to the threshold of 0 dB, and is another error, because not all CD players can read above 0.1 or 0.2 dB, so it's better to bring the highest peak to 0.3 to escape every distortion probability. So if you ear noticeable distortion is due to this problem. I hope this will be helpful, sorry for bad english!
    Best regards

    ReplyDelete
  2. sorry please correct:
    ..then if you want to lift the lower sounds, you CANNOT lift also the higher sounds above than this 0dB. In this terms, dynamics are compromised.

    ReplyDelete
  3. ...because not all CD players can read above -0.1 or -0.2 dB, so it's better to bring the highest peak to -0.3 dB to escape every distortion probability...
    the digital Fullscale is always negative until 0dB, which is the maximum amount of sound phenomena "volume" you can describe into this margin :) still sorry

    ReplyDelete
  4. I believe it's related to this:
    http://en.wikipedia.org/wiki/Loudness_war
    I never thought Magma would fall victim to it!

    ReplyDelete