If you are happy with the way you play music that's perfect.
Don't tell that if others are happy by doing it different from you that it is wrong!
Don't tell that if others are happy by doing it different from you that it is wrong!
Are you talking to me? The OP asked a question, which is an invitation for people to provide well reasoned answers!
He is asking if he should bake in a specific non-linear frequency response. I'm saying no, and I'm explaining my reasoning for my answer!
You commented that you use digital bass compensation, so your answer would also be no....because you use an EQ to get the bass you want instead of altering the design of the speakers/ subwoofer!
Yes.
All I was saying is that you tell others that other are wrong:
They are not wrong , and you are not right.
There are more ways to enjoy music and both are right if you think you enjoy it.
There's hardly any truth in recorded music I fear. It would be quite a thing to have universal EQ, dealing with the EQ in the control room, in the production chain and in the playback system. Level dependent (and I mean SPL at ear drum) and regarding system limits.
Thanks for the continued discussion. I will say that I think the music you listen to is of a frequency range and timbre that is not noticeably affected by the response curve of the ears as shown by the equal loudness curves. I mostly listen to pipe organ music which can have soft timbres extending below 20Hz. When I think subwoofer, I don't think about covering 40-200Hz, rather anywhere below 40 in most cases, and only extending up as high as needed to blend with the higher range speakers.
I do somewhat agree with what you say, because I can still perceive sub bass to a good extent with the volume turned down, but for me it begins to fade out of balance with the treble, and even fades out completely at some point. I don't know if it is due to the decreased efficiency of the air and the driver or our ears to transmit those low frequencies at such low volumes or what.
I honestly wouldn't mind adding some distortion to compensate for the perceived roll off of our perception of soft timbre sub bass, as long as it benefits the subjective experience of the music.
My opinion is primarily rooted in the belief that it will be problematic to build a sub that has a built in, steeply sloping response curve. That is what you call a broken subwoofer!
The thing to do is build the sub with proper response (because physics), and then you can EQ as desired. This way, if you don't like the settings you chose, you can un-do them without building a new subwoofer.
If you design the sub well, size it for your desired output, and operate it within it's capabilities, you won't have any distortion problems.
The other point to make is that it is always better to have 2 (or 3) subwoofers. This improves the in room response, and can be optimized with DSP. This alone might address the preceived roll off--it might have something to do with room modes.
It's funny...once upon a time I thought I had a nice system with stereo subwoofers. But then I started building my own subwoofers, and I realized the subs I had were literally the most terrible things ever created by man! The benefit of the custom subs is not being louder. They are far more clear and detailed, no horrible resonances, etc. I felt like I had never "actually" heard good bass before!
Tell me what you've learned about building subs and why/how yours turned out so much better than the factory ones. I don't doubt you, I just want to learn more.
I have flattened the response of an undersized sub with a passive filter, and the result was good enough that when I learned about the equal loudness curve, I thought that's what was going on, when in reality I was just flattening the roll off curve.
I have flattened the response of an undersized sub with a passive filter, and the result was good enough that when I learned about the equal loudness curve, I thought that's what was going on, when in reality I was just flattening the roll off curve.
Neat trick is to place the sub in listening position first, play some well recorded music, go around the room and listen to most pleasant balanced bass.
Then place sub there.
SPL meter says.." yes, but, as far as i know, most usual sonometers aren't they weighted "curve A" ? ... This A curve mean a non flat response, more or less following the human ears sensitivity curve ... "on risque de tourner un peu en rond" ...
The main thing I learned is you need a good sub driver, and then build the cabinet out of 3/4" / 19mm material.
The "affordable" manufactured subs are very cheap, using thin material and a very cheap driver. And high quality manufactured subs seem extremely expensive considering what it costs to make a good one yourself.
The other thing I learned is that you can do nearly anything you can imagine, as long as you match the parameters from a modeling software. I'm surprised with what I've gotten away with. One was an "L" shape (part of a big 4-way speaker), and I had two 35"/ 890mm by 4"/ 100mm diameter ports running vertically, venting out a notch in the bottom of the speaker. It works great!
But I always build a quick and dirty prototype from cheap materials first to confirm the design.
"A weighted" is just an averaging method. There is also a C weighted, which typically yields a higher value. But most things are measured A weighted.
I think you are referring to the decibel scale being a logarithmic scale, which corresponds to how we hear. But the averaging method is "linear" on the decibel scale. So we don't have to do anything special to interpret an A weighted SPL measurement.
Yes, there are always details and nuance. But I think we are still drawing the wrong conclusions from the information!
Yes, SPL meters do not measure the full audible frequency range! They are not supposed to! The specs on the meter will tell you what frequency range it can measure.
DBa is a standardized, representative measurement used for pretty much everything. Someone decided that a limited frequency band represents the entire freqency range well enought that it is not necessary to measure the entire audible range!
Now, if a person calibrated their subwoofer with an SPL meter that only measures dBa, their result would be hilariously wrong! But nobody does that, right?
I hope we all use a calibrated measurement microphone? With a calibrated microphone, you know that the measurements are correct relative to each other, but by default we don't know if the actual SPL is correct. To find out the actual SPL, we reference a dBa SPL meter...computer plays a tone specific for a dBa SPL meter, and we input the value. The computer makes an adjustement, and now our measurement mic is calibrated to give us accurate SPL across the entire range, even at frequencies the SPL meter can't detect.
When I posted about listening at low levels, I also switched my SPL meter to dBc. It gave me a different number, but the relative difference between the music level and the ambient noise was the same! Everyone is accustomed to dBa measurements, so that is what I used!
Yes, SPL meters do not measure the full audible frequency range! They are not supposed to! The specs on the meter will tell you what frequency range it can measure.
DBa is a standardized, representative measurement used for pretty much everything. Someone decided that a limited frequency band represents the entire freqency range well enought that it is not necessary to measure the entire audible range!
Now, if a person calibrated their subwoofer with an SPL meter that only measures dBa, their result would be hilariously wrong! But nobody does that, right?
I hope we all use a calibrated measurement microphone? With a calibrated microphone, you know that the measurements are correct relative to each other, but by default we don't know if the actual SPL is correct. To find out the actual SPL, we reference a dBa SPL meter...computer plays a tone specific for a dBa SPL meter, and we input the value. The computer makes an adjustement, and now our measurement mic is calibrated to give us accurate SPL across the entire range, even at frequencies the SPL meter can't detect.
When I posted about listening at low levels, I also switched my SPL meter to dBc. It gave me a different number, but the relative difference between the music level and the ambient noise was the same! Everyone is accustomed to dBa measurements, so that is what I used!
That actually makes more sense. I like that one.
Mixing recordings can be frustrating. Basically trying achieve a good balance or as you say everything is equal loudness.
You have to review on many different monitors because what sounds great on one, might sound strange or something will stand out or be to low on another.
Not to be confused of course with the whole actual "curve" thing too.
Agree was always confusing, or to many solutions or theories being tossed around to fix a problem we were told that exists.
When way before we knew about it or at least in my life.
I agree, with most songs or music you might not need to change the EQ or need more bass.
Other recording maybe sometimes a little boost. At low levels. Then if you turned it up, you had to lower the bass.
Really can be whatever speaker your using.
Big learning thing with mixing a recording. A lot of guys do make the mistake of boosting the bass to much.
It does play into equal loudness curve. It tends to happen because of the levels of the monitor.
And you will mix different at high or lower levels. So you have to be aware of it and test at different levels and speaker types.
This track? https://lespecial.bandcamp.com/track/jackwise
There's enough information in the harmonics from the aggressively plucked / slapped bass guitar, that you don't need any actual bass to discern the individual notes being played on the bass guitar.
I didn't follow the link. Believe it or not, I can actually tell when I'm hearing the fundamental. I don't know if you can, but it doesn't matter.
If you want to prove something, propose a controlled scenario. But it doesn't matter because it doesn't change the answer to the OPs question: Design a sub with a flat response. Then EQ however you want.
If that's how it is actually tuned like (with the 25dB peak at 26Hz, the room gain probably adds 10+dB more), then it's the absolutely perfect one-note subwoofer.
Physics doesn't make any exceptions for certain manufacturers. Not even if they pay for it. 😏
That simulation looks like a random high-Q driver was thrown into a heavily mismatched enclosure. Or you have entered the wrong TSP.
If you want to compensate for the loudness curve, keep in mind music is mixed for mostly linear speaker reproduction or popular cheap mass consumer electronics (and matched mass-music material). The former does not have a 25dB peak at 26Hz, the latter does not even get down close to that frequency range.
I see you don't believe what so many ppl already said. So do an experiment yourself. Take a somewhat linear subwoofer, use a parametric EQ (EQ APO, App on your computer, cell phone , plug-in for your player software etc) and crank up 26Hz by 25dB. Most music will not contain that music range but action movies often do. Watch one at your proposed loudness (whatever that may be) and see if you like it.
I dare to predict you either like it because you listen to it super quiet or you don't like it. And that's also the solution for it: At very low listening levels you can still use the parametric EQ, at higher level that specific curve doesn't apply anyway. But don't build such a sub, save some trees.
Would be nice if you could make also your claim more controlled.
You are suggesting you can hear the fundamental below the hearing threshold?
So could you find out what frequency and level that fundamental was that you could hear?
Then after that play the same fundamental as a pure sine wave at the same frequency and level and verify you can hear it.