Why boosting 30 Hz by 30 dB sounds so good...

[DiscoProJoe]

Here are some fun "equal-loudness contour" charts. The first one is of the Fletcher-Munson curves, created in 1937. The second is a more up-to-date chart from 2003.
 
If you follow a contour line, the tone pitches appear to be the same loudness.
 
No wonder the higher bass increments on my personalized EQ presets in Rockbox sound so good! And this is combined with the additional boost of 5.5 dB of sub-bass on my Cayin C5 portable amp, and also combined further with my V-Moda M-100s naturally boosting the sub-bass region by about 8 dB on top of that.
 
Music to my ears.... 

 
 

 

 

[MindsMirror]

The reason you like a 30dB boost in bass has nothing to do with the equal loudness contours. Music is already made to sound correct on neutral speakers/headphone. If any EQ was needed, it was already applied when they made the track. Sounds like you are just a bass head who likes 30dB more bass than your average person.

 

[DiscoProJoe]

  Music is already made to sound correct on neutral speakers/headphone. If any EQ was needed, it was already applied when they made the track.

 
On newer music, you're right. A smaller bass boost is needed.
 
On older music from the '70s & '80s, the recordings haven't been boosted as much, so a huge bass boost is needed.

 

[limpidglitch]

   
On newer music, you're right. A smaller bass boost is needed.
 
On older music from the '70s & '80s, the recordings haven't been boosted as much, so a huge bass boost is needed.

 
That sounds like a reasonable observation.
When music is being mastered for vinyl there are limits to how much bass can be included, while still making all the program material fit, and the record even playable. So the cut in bass frequencies is a technical compromise.
If a more modern remaster where this has been corrected hasn't been released, you might have to try to remedy it yourself.

 

[MindsMirror]

   
On newer music, you're right. A smaller bass boost is needed.
 
On older music from the '70s & '80s, the recordings haven't been boosted as much, so a huge bass boost is needed.

If anything I would say the opposite. Modern music sometimes takes a hit in the bass frequencies to help push the volume higher. That didn't happen pre-loudness wars. Different music styles and preferences probably give us different opinions.

 

[Steve Eddy]

That sounds like a reasonable observation.

When music is being mastered for vinyl there are limits to how much bass can be included, while still making all the program material fit, and the record even playable. So the cut in bass frequencies is a technical compromise.

You're aware of RIAA equalization, yes?

se

 

[Don Hills]

I listen almost exclusively to rock/prog/metal from the 60s to the present day. I've looked at the spectrograms of a large number of tracks. There are exceptions, but the majority have 2 defining characteristics:
1. They cut off sharply below the frequency of the lowest bass note. Typically, this used to be 40 Hz or so, but with the increasing use of 5 and 6 string basses 30 Hz is becoming more common.
2. The logarithmic response curve is a fairly straight line, highest level at the bass cutoff and sloping down at about 6 dB / octave as the frequency increases.
 
For most such tracks, attempting to boost the bass below the cutoff is a waste of time, especially with modern tracks that have been cut off with steep digital filters.
Some tracks have their highest levels at 80 or 100 Hz or so, dropping off towards the lower cutoff, and can benefit from boosting the lower frequencies to restore the "straight line". This applies more often to older tracks, and may indicate that the bass was reduced to make it easier to cut to vinyl.
Some tracks are simply pathological, such as Led Zeppelin's "Kashmir". "Boosting 30 dB at 30 Hz" is relevant there, but even then is largely a waste of time - there's little music there to be boosted.

 

[Skooter]

 .

 

[Steve Eddy]

The Fletcher Munson curves, and subsequent studies, are interesting estimates for audio. Two important take aways:

I'm afraid your take aways are completely off base.

1. The Fletcher Munson curves indicate that the sensitivity of average ears varies over frequency. This is one reason speakers or headphones with a "flat" frequency response don't sound realistic. As noted above, low frequencies are generally perceived at lower spl (we can call that volume) than higher frequencies are. That is, the bass may sound "light." Our perception varies by venue, but the effect exists regardless if it is in a theatre, a concert hall, or with a home stereo system.

This is absolutely incorrect and represents a complete misunderstanding of what the curves represent.

They represent how we perceive sound with respect to loudness. It is NOT something that needs to be "compensated for" except in one instance that I will get to later.

If a loudspeaker or headphone doesn't sound realistic, IT IS NOT BECAUSE THEY HAVE A FLAT FREQUENCY RESPONSE. If you make a recording using microphones with a flat response, you want it reproduced with a flat response. That's the only way it can possibly sound real to us.

Item 2 has a special practical implications for audio. When listening at home at low volumes, a "bass boost" can compensate for the steep curve, and makes one perceive the music sounds more balanced. Using this same "bass boost" at higher volumes will make one perceive the music to be excessively "boomy."

Audio equipment manufacturers have been incorporating "loudness" and "bass boost" circuitry for decades that may adjust dynamically or via a button to compensate for item two. Using bass boost at low volumes can be pleasing. At higher listening volumes, that bass boost should be reduced.

Yes, that's the one instance where you want to compensate for the curves, so that a recording played back at low volume will have roughly the same spectral balance as it would have when played at realistic volumes.

se

 

[DiscoProJoe]

  For most such tracks, attempting to boost the bass below the cutoff is a waste of time, especially with modern tracks that have been cut off with steep digital filters.

 
In my music collection (in terms of older vs. newer music), I've experienced the opposite. The majority of the songs I have from the '70s & '80s sound like they have no sound below 50 Hz, although there are some exceptions to this. (One exception off the top of my head would be the "Top Gun Anthem" by Harold Faltermeyer & Steve Stevens.)
 
In my EQ presets in Rockbox, when listening to older music, I usually use the "Bass" increments, which boost 32 Hz and 64 Hz together in lockstep. I almost never use the "Sub-Bass" increments with older music, which boost 32 Hz a lot while raising 64 Hz by 1/5th as much.
 
On newer music I tend to use the "Bass" and "Sub-Bass" increments equally as much, but mostly just use the "Bass" ones with older stuff. If there's nothing below 50 Hz in a track, then the "Sub-Bass" increments certainly don't seem to help very much!

 

[DiscoProJoe]

If a loudspeaker or headphone doesn't sound realistic, IT IS NOT BECAUSE THEY HAVE A FLAT FREQUENCY RESPONSE. If you make a recording using microphones with a flat response, you want it reproduced with a flat response. That's the only way it can possibly sound real to us.
 

 
With music playback, I guess some may define "realistic" by the way the recording engineer intended it to sound, while others would define that word by how well the playback follows the equal-loudness contours.
 

They represent how we perceive sound with respect to loudness. It is NOT something that needs to be "compensated for" except in one instance that I will get to later.

 
This is purely a matter of personal taste. Some would prefer it to sound the way the studio engineers made it, and others (like me) like it better with it EQ'd.

 

[Steve Eddy]

Ok, I can see that this discussion isn't going to go anywhere but downhill.

I'm out.

se

 

[limpidglitch]

You're aware of RIAA equalization, yes?

se

 
I'm talking about the sub-bass, which is often reduced to a level below what the RIAA curve would suggest.
This is done primarily to make the grooves narrower, so that each side can hold a longer stretch of music.

 

[Steve Eddy]

I'm talking about the sub-bass, which is often reduced to a level below what the RIAA curve would suggest.

This is done primarily to make the grooves narrower, so that each side can hold a longer stretch of music.

I'm not sure what you mean by "a level below what the RIAA curve would suggest."

se

 

[limpidglitch]

I'm not sure what you mean by "a level below what the RIAA curve would suggest."

se

 
Below what the negative RIAA curve would cause.
So when you play it back, you'd have to add a further few dB boost to the very low end, because the filter in the pre-amp won't compensate entirely on its own.