Loudness-Controlled Audio Previews
GarsuMene
28 May 2012 19:45
Let me start by emphasizing that the object of the discussion I’m proposing here is the overall loudness of audio previews on P5 and not the inner dynamic loudness of a piece of audio (i.e. the level differences between loud and quite parts and/or passages within an audio track).
The overall loudness of an audio track does not constitute an artistic property of its content. Instead, it is more of an objective physical condition under which the listener (customer) is able to properly evaluate the artistic content of a given audio track.
It is a well known fact that humans have a tendency to mistake the overall loudness of an audio recoding for its qualitative superiority when it is compared to another recording that may in fact be superior in terms of quality but is less loud.
I do not wish to delve into the many whys and hows of this human tendency, fascinating and interesting though it is. Suffice it to say that this tendency, and dare I say, weakness of human appreciative powers is well research, documented, and is the main reason why the audio industry for a few decades now has fatally entangled itself in what is known as “loudness war” or “loudness race.”
I am not to preach here what overall loudness levels authors should impose on their audio tracks, but I am here to argue for the case that not controlling the overall levels of audio previews on P5 by making them comparatively equal is misleading and possibly harmful to the customer as well as unfair to the author.
- It is misleading to the customer (listener) because the drastically varying levels (up to 10 dB!) of audio previews on P5 hinder the listener from properly concentrating on the content of audio tracks as his or her attention is likely to be overpowered by a psycho-physiological reaction to the mere fluctuations of comparative loudnesses between audio previews (mistaking louder for better). It is like trying to determine what painting to purchase in a gallery where some of the paintings are hung up in proper lighting, distance, and position and others moved away in an underlit dusky corner. The customer is misled because he or she is not provided with more or less equal conditions under which to evaluate items of potential purchase.
- It may be harmful to the customer because playing back a preview of, say, a more dynamic piece of music at a higher level (e.g., to appreciate its dynamic and spatial characteristics) and forgetting to lower the volume slider on the soft player or the knob on the amplifier before hitting a preview of a highly limited and thus very hot audio track may indeed hurt or even impair the customer’s audible sense.
- It is unfair with respect to the author because it makes harder to communicate his or her artistic intentions to the potential customer without being forced to participate in the loudness race which is actually prone to degrade the quality of end products by introducing into the audio signal all sorts of fatiguing digital distortions.
An audio preview should inform the customer of the artistic content of an audio track as a deciding factor while making sure that the loudness of the previewed item is not taken by a non-specialist listener for its quality.
We’re in the digital age. There are solutions. Let me suggest this one:
1) Add a replay gain tag to the preview audio file when creating it; 2) enable your preview player to read replay gain tags; 3) allow users (by adding something like a check box in the settings window) to choose whether they want replay gain services enabled or disabled when previewing audio on P5.
What do you think?
The overall loudness of an audio track does not constitute an artistic property of its content. Instead, it is more of an objective physical condition under which the listener (customer) is able to properly evaluate the artistic content of a given audio track.
It is a well known fact that humans have a tendency to mistake the overall loudness of an audio recoding for its qualitative superiority when it is compared to another recording that may in fact be superior in terms of quality but is less loud.
I do not wish to delve into the many whys and hows of this human tendency, fascinating and interesting though it is. Suffice it to say that this tendency, and dare I say, weakness of human appreciative powers is well research, documented, and is the main reason why the audio industry for a few decades now has fatally entangled itself in what is known as “loudness war” or “loudness race.”
I am not to preach here what overall loudness levels authors should impose on their audio tracks, but I am here to argue for the case that not controlling the overall levels of audio previews on P5 by making them comparatively equal is misleading and possibly harmful to the customer as well as unfair to the author.
- It is misleading to the customer (listener) because the drastically varying levels (up to 10 dB!) of audio previews on P5 hinder the listener from properly concentrating on the content of audio tracks as his or her attention is likely to be overpowered by a psycho-physiological reaction to the mere fluctuations of comparative loudnesses between audio previews (mistaking louder for better). It is like trying to determine what painting to purchase in a gallery where some of the paintings are hung up in proper lighting, distance, and position and others moved away in an underlit dusky corner. The customer is misled because he or she is not provided with more or less equal conditions under which to evaluate items of potential purchase.
- It may be harmful to the customer because playing back a preview of, say, a more dynamic piece of music at a higher level (e.g., to appreciate its dynamic and spatial characteristics) and forgetting to lower the volume slider on the soft player or the knob on the amplifier before hitting a preview of a highly limited and thus very hot audio track may indeed hurt or even impair the customer’s audible sense.
- It is unfair with respect to the author because it makes harder to communicate his or her artistic intentions to the potential customer without being forced to participate in the loudness race which is actually prone to degrade the quality of end products by introducing into the audio signal all sorts of fatiguing digital distortions.
An audio preview should inform the customer of the artistic content of an audio track as a deciding factor while making sure that the loudness of the previewed item is not taken by a non-specialist listener for its quality.
We’re in the digital age. There are solutions. Let me suggest this one:
1) Add a replay gain tag to the preview audio file when creating it; 2) enable your preview player to read replay gain tags; 3) allow users (by adding something like a check box in the settings window) to choose whether they want replay gain services enabled or disabled when previewing audio on P5.
What do you think?
FxProSound
6 Jun 2012 14:40
Hello! Other places automatically normalize audio to equal level, but it still differs, because of many things (mostly compression level). IMHO more annoying is listening 64 kbps mp3.
GarsuMene
6 Jun 2012 14:59
Hi and thanks for your reply!
No, I'm not talking about normalization because that's about wave peaks and not average loudness.
And yes, the low-fi compression here does at times remind me of underwater life :)
No, I'm not talking about normalization because that's about wave peaks and not average loudness.
And yes, the low-fi compression here does at times remind me of underwater life :)
FxProSound
6 Jun 2012 19:41
Your idea is good indeed, but I suppose that many authors leave this option blank or overuse to have louder sound. Now I'm not filling some unnecessary things like creation date, place etc. Another thing is that these previews are usable only to show it on laptop speakers or checking how audio works with video, but not showing sound value of songs.