I was referring to the change in sensitivity, not absolute sensitivity.
Yes you're right, the ear is very insensitive to the lower end of the sound. But as the volume goes up, the ear gets a little less insensitive to the lower end. 😀 This change in sensitivity vs volume will cause speakers to sound very different depending on the volume it is played it, and correcting this difference is at the core of a good equal loudness contour.
didnt check that ''hobby'' for a long long time...
CRAIG BUTLER'S SETUP AND METER 182.6DB - YouTube
Wow man, they're as crazy as before., 182.6db
what is the theoretical maximum reachable in such an enclosed space ?
CRAIG BUTLER'S SETUP AND METER 182.6DB - YouTube
Wow man, they're as crazy as before., 182.6db
what is the theoretical maximum reachable in such an enclosed space ?
Fletcher munson curve is an approximation, an average. It changed over time and still is an average.
You sure can make your own curve based on your own ear's ''pain threshold''
You sure can make your own curve based on your own ear's ''pain threshold''
Last edited:
I was interested to know whether bcodemz thought it was a good idea since he mention equal loudness curve. There was a trend for it but seems people have seen the light?
I personally think equal loudness curve is the key for the absolute best sounding/hifi system possible. But that means you have to find the exact curve (say 0.3db max diff). For everyone.
Not realistic.
Not realistic.
Not really. I don't think doing such detailed frequency response changes is ideal. The reason is because I don't trust the study enough to get that kind of accuracy of the little bumps in the frequency response in the middle. Jon is correct, it is an approximation and average. There are ear to ear variations that will likely change the frequency that happens for most people.
What I think is ideal though, is to do a broad stroke correction at the lower range and upper range. This is universal for everyone. The magnitude shouldn't be what fletcher munson curve dictates, but less. The reason is because again there are variations between ear to ear. The goal is to improve subjective listening experience for all cases. If you go for a detailed correction, you will get it to sound better for many situations, but you are bound to make certain situations worse. A broad stroke correction allows you to achieve an across the board improvement, and it is certainly better than doing nothing. However, venture too far into optimization, and you're bound to make certain things worse. I'm not saying optimization isn't good. But there are too many unknowns when doing equal loudness curve, and the biggest one that no one can control is the source. Therefore all of this correction needs to work with a variety of unknowns, and therefore the best we can do is for an as much across the board improvement as possible, which means sacrificing a few percentage of absolute perfection.
Last edited:
It depends. They saved lot of money on phone communications limiting bandwidth.
It is the only good thing they did with their curves. The rest is, horrible attempts to "correct" hi-fi audio system misunderstanding the meaning. 😀
Excuse me if I'm misunderstanding you - this is where I think the idea of correction is flawed, this is how we hear and if it is "corrected" it will sound wrong.
This does make sense to me. What might be of value is correction of the difference between the curves as level changes
The meaning being there is no or very little correction needed because there is nothing (much) wrong?
The ear drum is a flawed transducer.
We need USB implant so we can EQ correct our brain's signal. A mute button option? Oh yes please.
We need USB implant so we can EQ correct our brain's signal. A mute button option? Oh yes please.
Yeah I think you misunderstood me, but understood my general idea. I agree that we shouldn't correct that part because that's just how we hear and it will sound wrong if corrected. I think equal loudness corrections get a bad rap is partially because people simply invert the curve. Then of course the results will sound quite bad.
The only thing that should be corrected is, as you said, the difference between the curves as level changes. But even for the differences, I think the changes in the mid band should be ignored, or only corrected a little bit.
I do think the correction for lower and upper ranges should be corrected on an absolute level too (i.e, outside of the differences of curves). However, this is problematic because it cannot possibly apply for a variety of music. What works for compressed music will be too much for well recorded music. This is because well recorded music's average volume is much lower than compressed music in order to have dynamic range. Therefore, the amp has to be turned up much more to listen at the same average level as compressed music. But by doing that, it thinks we are listening at a far louder volume than we really are, and will apply too much of the equal loudness correction. If we make the equal loudness correction based on the SPL at the listening position, then it will apply equal loudness correction during the loud parts of the music when no or much less correction should be applied. This is why equal loudness correction is difficult to implement well. The solution requires knowing the reference volume, like on all movie soundtracks. However, there is no such thing as a reference volume for music (which unfortunately led to the loudness war), and it would require a completely new audio format, which isn't happening.
I thought I did misunderstand but wasn't sure 🙂 I recently built the Doug Self pre amp with tunable bass and treble which helps, not least with my HF hearing loss. Thankfully I don't listen to much compressed material, wonder if an expander would help there
There is absolutely no meaning in F-M curves for audio reproduction, except how much power from a system with flat frequency response you need for bass and high frequencies to be audible. No, rather, recognizable, because we perceive wider range and lower levels than consciously hear.
- Status
- Not open for further replies.