16-Jul-2019, 10:12
@Confused
Re your Fletcher Munson widget, I was guessing that you were talking about something like the old loudness switch. AVRs which use Audyssey room correction routinely offer a DEQ (digital EQ) feature which provides a similar sort of thing but it doesn't work well with music because while there is some standardisation of the processes involved in mastering soundtracks for movies and video, there isn't for music and the differences in mastering that we get with music recordings is actually quite variable in some ways so people who use their AVRs for both music and movies often seem to turn DEQ on for movies and off for music, simply because they don't like the result. Those switches are actually an automated tone control of sorts so you can turn your bass tone control up to get a result you think is right for your ears and the music you're playing. I don't think it would be easy to come up with an automated control that worked well for music.
I won't even try to guesstimate the level generated in the room in the video. I don't know how well it was recorded, what the characteristics of the mic were and what it's doing to the sound, and I'd be listening on a very different system with no idea what volume I should set it at in order to get some kind of idea. In addition I find it hard to estimate SPLs, our ears aren't measuring instruments. They're great for qualitative assessment, they're not good for quantitative assessment unless perhaps you're doing that on a regular basis and I'm not.
Quick story to illustrate: I once had to measure the level of a letter opening machine to see if it posed a hearing risk. I stood next to the operator with the meter's mic next to her ear in order to measure the level at her ears and I was getting measurements ranging between the low 70 dBA range and peaks of around 83 dB so it wasn't a risk but every now and then I'd see a peak of around 93 dB and I wasn't hearing anything that sounded anywhere near that loud. I started watching what was happening and realised that when the operator took the letter out of the opened envelope and unfolded the letter, if they unfolded it fast and stretched it taught, the unfolding produced a "snap" sound and that was producing the peak. When I went back to my desk I got some A4 paper, folded it, and started opening it and getting it to make that "snap" and watched the meter and I could generate a transient of around 94 dB quite easily if I tried. That's a high reading but the actual sound doesn't sound loud to us, probably because it only lasts a fraction of a second and our ear may not respond fast enough for us to register the actual level. Trying to estimate SPLs by ear isn't easy and there are several reasons for that, one being the variation in the sensitivity of our ears to frequency that is documented in the Fletcher Munson curves , but there are probably others as well and if it's not easy when you're actually present at the event then it's going to be much harder when what you're listening to is a recording made with unknown equipment in a location you're totally unfamiliar with.
In an earlier post you compared the amount of power required to produce a sound at 40 Hz that sounded as loud as one at 1000 Hz. Your 1000 times the power may not be far off the mark but if the amount of power required to produce that volume for the sound at 1000 Hz is very low, then you're going to need a lot less power than you think. You also have to recognise that when you start adding 2 SPLs together, say the level of the 40 Hz sound and that of the 1000 Hz sound, if the difference in level is 9 dB the combined level produced is only 0.5 dB higher than the level of the louder sound and if the difference is 20 dB the combined level is going to be less than 0.1 dB higher than that of the softer sound. You were talking about a difference of 30 dB and the table I've got which gave me the figures I just quoted doesn't even bother showing differences greater than 20 dB. Basically with a 30 dB difference, the level of the combined sound for your 2 tones is insignificant, the combined volume is basically that of the 40 Hz tone.
As I said earlier, the musicians are balancing the levels to produce the right frequency balance in the music as they play. We don't have to worry about how much amp power we need to generate the bass to match the level of the mids, all we need to worry about is the amount of power we need to reproduce the music at the level we want because the music contains lots of different sounds at different frequencies and levels at the one time when there's a group of musicians playing, the balance of levels is preserved and combines to produce the SPL we hear, and what we want to do is to be able to get the combined sound of all of the musicians produced at the level we want to listen to it. Identifying how much power we need to reproduce the bass player correctly ignores the need to reproduce the other musicians as well at the same time and if we want to have peaks at some particular level, say 95 dB but increase or decrease that to suit yourself, we need to reproduce the whole of the music so why bother measuring the bass player only, measure the level of the whole group of musicians.
Seriously, the Fletcher Munson curves and the varying sensitivity of our ears to different frequencies at different SPLs is important for some things and can be fascinating but it really is irrelevant to the question of how much amp power we need. Musicians were balancing the levels of their instruments and voices to produce music that made their listeners happy centuries before we got SPL meters and could make the measurements that showed us that there were variations in the sensitivity of our ears to sound and that's the way they're still balancing the sound of their music. They don't pull out SPL meters and measure themselves and each other in order to ensure that they produce the balance between instruments and voices that they want, they do it by listening to each other and adjusting their volume by ear. The audience doesn't look at a meter in order to find out whether the music sounds good. Measurements are irrelevant to the making and enjoyment of music. It can help us work out how big an amp we need but the measurements we need to work that out are measurements of all of the sounds in the music, not just the sound in one frequency range or another.
I think you're falling into a trap that many of us, including me, fall into when we try figuring some sort of quantitative question out and we get faced with new info that brings factors we hadn't thought of to our attention. We try to bring everything into our calculations and we often don't need to bring everything into them, so we start overthinking the problem and making it more complicated than it really is. I found myself doing that early in the course of this thread before I came up with the rather simple analysis I made in post #56. You have no idea how complicated some of my thinking got before I suddenly managed to see things in a way that made sense to me. Sometimes we have to do a bit of overthinking to eventually get to a much simpler view of things which makes sense to us. I think you're getting sidetracked at present with things like the Fletcher Munson curves and dBA scales. They do take us off topic from the real issue as you said. If you want to send me a personal message I'm happy to try and help where I can on that sort of question but I suspect I know a bit less on some of these things than you think.
Re your Fletcher Munson widget, I was guessing that you were talking about something like the old loudness switch. AVRs which use Audyssey room correction routinely offer a DEQ (digital EQ) feature which provides a similar sort of thing but it doesn't work well with music because while there is some standardisation of the processes involved in mastering soundtracks for movies and video, there isn't for music and the differences in mastering that we get with music recordings is actually quite variable in some ways so people who use their AVRs for both music and movies often seem to turn DEQ on for movies and off for music, simply because they don't like the result. Those switches are actually an automated tone control of sorts so you can turn your bass tone control up to get a result you think is right for your ears and the music you're playing. I don't think it would be easy to come up with an automated control that worked well for music.
I won't even try to guesstimate the level generated in the room in the video. I don't know how well it was recorded, what the characteristics of the mic were and what it's doing to the sound, and I'd be listening on a very different system with no idea what volume I should set it at in order to get some kind of idea. In addition I find it hard to estimate SPLs, our ears aren't measuring instruments. They're great for qualitative assessment, they're not good for quantitative assessment unless perhaps you're doing that on a regular basis and I'm not.
Quick story to illustrate: I once had to measure the level of a letter opening machine to see if it posed a hearing risk. I stood next to the operator with the meter's mic next to her ear in order to measure the level at her ears and I was getting measurements ranging between the low 70 dBA range and peaks of around 83 dB so it wasn't a risk but every now and then I'd see a peak of around 93 dB and I wasn't hearing anything that sounded anywhere near that loud. I started watching what was happening and realised that when the operator took the letter out of the opened envelope and unfolded the letter, if they unfolded it fast and stretched it taught, the unfolding produced a "snap" sound and that was producing the peak. When I went back to my desk I got some A4 paper, folded it, and started opening it and getting it to make that "snap" and watched the meter and I could generate a transient of around 94 dB quite easily if I tried. That's a high reading but the actual sound doesn't sound loud to us, probably because it only lasts a fraction of a second and our ear may not respond fast enough for us to register the actual level. Trying to estimate SPLs by ear isn't easy and there are several reasons for that, one being the variation in the sensitivity of our ears to frequency that is documented in the Fletcher Munson curves , but there are probably others as well and if it's not easy when you're actually present at the event then it's going to be much harder when what you're listening to is a recording made with unknown equipment in a location you're totally unfamiliar with.
In an earlier post you compared the amount of power required to produce a sound at 40 Hz that sounded as loud as one at 1000 Hz. Your 1000 times the power may not be far off the mark but if the amount of power required to produce that volume for the sound at 1000 Hz is very low, then you're going to need a lot less power than you think. You also have to recognise that when you start adding 2 SPLs together, say the level of the 40 Hz sound and that of the 1000 Hz sound, if the difference in level is 9 dB the combined level produced is only 0.5 dB higher than the level of the louder sound and if the difference is 20 dB the combined level is going to be less than 0.1 dB higher than that of the softer sound. You were talking about a difference of 30 dB and the table I've got which gave me the figures I just quoted doesn't even bother showing differences greater than 20 dB. Basically with a 30 dB difference, the level of the combined sound for your 2 tones is insignificant, the combined volume is basically that of the 40 Hz tone.
As I said earlier, the musicians are balancing the levels to produce the right frequency balance in the music as they play. We don't have to worry about how much amp power we need to generate the bass to match the level of the mids, all we need to worry about is the amount of power we need to reproduce the music at the level we want because the music contains lots of different sounds at different frequencies and levels at the one time when there's a group of musicians playing, the balance of levels is preserved and combines to produce the SPL we hear, and what we want to do is to be able to get the combined sound of all of the musicians produced at the level we want to listen to it. Identifying how much power we need to reproduce the bass player correctly ignores the need to reproduce the other musicians as well at the same time and if we want to have peaks at some particular level, say 95 dB but increase or decrease that to suit yourself, we need to reproduce the whole of the music so why bother measuring the bass player only, measure the level of the whole group of musicians.
Seriously, the Fletcher Munson curves and the varying sensitivity of our ears to different frequencies at different SPLs is important for some things and can be fascinating but it really is irrelevant to the question of how much amp power we need. Musicians were balancing the levels of their instruments and voices to produce music that made their listeners happy centuries before we got SPL meters and could make the measurements that showed us that there were variations in the sensitivity of our ears to sound and that's the way they're still balancing the sound of their music. They don't pull out SPL meters and measure themselves and each other in order to ensure that they produce the balance between instruments and voices that they want, they do it by listening to each other and adjusting their volume by ear. The audience doesn't look at a meter in order to find out whether the music sounds good. Measurements are irrelevant to the making and enjoyment of music. It can help us work out how big an amp we need but the measurements we need to work that out are measurements of all of the sounds in the music, not just the sound in one frequency range or another.
I think you're falling into a trap that many of us, including me, fall into when we try figuring some sort of quantitative question out and we get faced with new info that brings factors we hadn't thought of to our attention. We try to bring everything into our calculations and we often don't need to bring everything into them, so we start overthinking the problem and making it more complicated than it really is. I found myself doing that early in the course of this thread before I came up with the rather simple analysis I made in post #56. You have no idea how complicated some of my thinking got before I suddenly managed to see things in a way that made sense to me. Sometimes we have to do a bit of overthinking to eventually get to a much simpler view of things which makes sense to us. I think you're getting sidetracked at present with things like the Fletcher Munson curves and dBA scales. They do take us off topic from the real issue as you said. If you want to send me a personal message I'm happy to try and help where I can on that sort of question but I suspect I know a bit less on some of these things than you think.
Roon Nucleus+, Devilalet Expert 140 Pro CI, Focal Sopra 2, PS Audio P12, Keces P8 LPS, Uptone Audio EtherREGEN with optical fibre link to my router, Shunyata Alpha NR and Sigma NR power cables, Shunyata Sigma ethernet cables, Shunyata Alpha V2 speaker cables, Grand Prix Audio Monaco rack, RealTRAPS acoustic treatment.
Brisbane, Qld, Australia
