Okay - had more time to catch up on this thread, so hopefully no more making egg-on-muh-face comments [today]. Some purely objective thoughts -
Assigning song, album, or even genre will skew too many aspects of responses. I understand what you’re trying to do, but from an analytical standpoint it probably won’t work that way.
Myself as case in point:
Let’s assume all elements of the universe are aligning for the ideal “listening session”.
Still:
I will not enjoy playing Arthur Lyman (avg dBA 67-69) or CocoRosie (~70-73) nearly as loudly as I’ll enjoy playing Lily (78-82) or Pearl Jam (80+…)
Some OST’s, especially massive classical / electronic scores, can go louder than all the above (what can I say, good soundtracks are
fun).
As was suggested, I’m talking about tunes I can enjoy at
album-length (Steely Dan generally not full album-length, no offense).
If you’d suggested an Eagles song, my avg music dBA would = my avg noise floor dBA
There are more resilient ways to account for a “blank” outliers issue in data points.
Conversely, duration of listening could be one of your data fields. It would be interesting to know if there’s a relationship between length of “listening session” and average playback volume (dB).
Again - one specific song (or album, or genre) is not the way to standardize the design. Put simply, people’s preferences for average listening level is not always ( or often? ) independent of music song / album / genre. If you assign these to be independent variables, you might not get meaningful results.
This bullet point is in contrivance to the second bullet point. And “harsher” file / master etc. will probably not make listeners want to increase the volume to their loudest preference. It probably
will make listeners want to reduce volume.
It also seems to be a question about enjoying quiet music.
It won’t work to prescribe a standard song to derive average listening level (dBA), then ask for preference based on duration of a full album. This begs a level of omniscient input that people can’t reliably give about themselves.
Zero feasibility to standardize the measuring gear in a way that would make results robust for analysis.
I wouldn’t worry about it - just accept there’s a margin of error here as wide as the horizon outside the window in your picturesque prairie music-listening home.
Yeah, ‘bout that - I lived in prairie habitat for some years.
No thanks. The constant wind and frequent thunderstorms made too high a noise floor.
HT “measurements” should skew results quite a bit. Most people listen to movies/tv with a different set of underlying criteria, so those data would be for different questions.
Very good to include this reminder! That 1w / 1m speaker rule can be pervasive!
The biggest hiccup for me is that average listening SPL will be a reality within certain albums / genres more than it’ll be among them. (To me) some music is better loud, and some quiet.
There’s no way to standardize personal preference in track / album / genre in order to balance your design for music’s sonic characters an SPL meter will be analyzing.
Listening environment (speaker quality, room quality) is also highly suspect of influencing listener tolerance thresholds (dBA), but these do not ( =cannot) factor in here. It is probably not an insignificant (though necessary) omission, unfortunately.
Gotta get at the answer(s) by asking the question(s) a little differently. And no way to make results anything but anecdotal. That’s
not me suggesting anecdotes can’t be just as interesting as rigorously analyzed study results!
So all that diatribe in the rearview, this is a great topic to broach - thank you
@killdozzer .