How does improved bass also improve the midrange as well?

[gks333]

Recently I started using REW to improve the bass response in my room. In doing so I have noticed a slight improvement in midrange. How does integrating bass freqs sub 100hz with mains (5 speakers in my case) increase clarity in midrange? It’s almost as if I added room acoustics and room decay has improved but nothing of the sort has happened. Listening environment stayed exactly the same.

This article touches upon what I have experienced specifically this paragraph: Fourth, a system’s bass presentation affects such seemingly unrelated aspects of the sound as midrange clarity and soundstaging. Thickness in the midbass reduces the midrange’s transparency. A cleaner midbass not only makes the midrange sound more open, it also lets you hear more clearly into the extremely low frequencies. Moreover, extending a system’s bottom end has the odd effect of increasing soundstage depth and our overall sense of the recorded acoustic, even on music with no low-frequency energy. I’ve heard an unaccompanied voice in a large hall reproduced by a pair of mini-monitors with and without a subwoofer. Adding the subwoofer revealed the full extent of the hall’s size, as well as presented the vocalist as a more tangible image within the acoustic.

A Guide To Better Bass (TAS 197) - The Absolute Sound

It does not go into detail why specifically, improving bass frequency’s can improve the midrange. Anyone have any technical facts that can explain this and have any of you noticed this in your system?

 

[abdo123]

the word you're looking for is pitch or tonality, when the system has too much (sub)bass the pitch of every sound you hear becomes lower, so the clarity of the entire audible spectrum becomes lower.

Human voices are not a tone generator, they excite a large spectrum.

 

[Jdunk54nl]

Could be lots of different reasons along with what abdo123 said above about tonality. The sub frequencies (if from seperate speakers) could be causing phase cancellations, could have been causing ringing and/or room mode excitement which has it's own set of issues not to mention distracting if something is rattling, etc.

If you posted before and after graphs (with appropriate scaling of 50db y-axis in 5db increments), not just frequency response (or just your REW .mdat file), people could help explain it more.

 

[gks333]

I wish I had saved a "before" screenshot or file. I was thinking along the lines of phase as well but since my XO is at 80hz on a 48db slope for the 4 subs I tossed that idea out since midrange is so much higher. I did phase match subs to center channel and then L&R mains. XO for L&R is 100hz while center I believe is 90hz and go figure, the surrounds showed best results at 80hz. Weird since the fronts are Revel F208/C208 while surrounds are Revel S206.
Red line is all subs and all mains. My final tune...for now. Middle line was my initial sweep before changing XO's. Bottom green is just the subs at 80hz XO, 48db slope. Graph goes up to 130hz.

 

[alex-z]

Our hearing changes with level, excess bass literally makes the mid-range quieter.

Equal-loudness contour - Wikipedia

en.wikipedia.org

There is also perceptual masking, tones which are close together blend together in our brains. If there is an unwanted frequency response peak, things on either side will be less detailed. I assume this is largely why the Harman preference rating system for speaker punishes deviations in the listening window response.

 

[gks333]

Our hearing changes with level, excess bass literally makes the mid-range quieter.

Equal-loudness contour - Wikipedia

en.wikipedia.org

There is also perceptual masking, tones which are close together blend together in our brains. If there is an unwanted frequency response peak, things on either side will be less detailed. I assume this is largely why the Harman preference rating system for speaker punishes deviations in the listening window response.

Good read. I was hoping to see specific specs like: for freqs 10-36hz one must increase amplitude by 8db for the sound to be perceived as equal loudness as freqs 37hz-180hz and for freqs 181hz -........etc etc. As a ballpark generalization.

 

[Jdunk54nl]

I wish I had saved a "before" screenshot or file. I was thinking along the lines of phase as well but since my XO is at 80hz on a 48db slope for the 4 subs I tossed that idea out since midrange is so much higher. I did phase match subs to center channel and then L&R mains. XO for L&R is 100hz while center I believe is 90hz and go figure, the surrounds showed best results at 80hz. Weird since the fronts are Revel F208/C208 while surrounds are Revel S206.
Red line is all subs and all mains. My final tune...for now. Middle line was my initial sweep before changing XO's. Bottom green is just the subs at 80hz XO, 48db slope. Graph goes up to 130hz.

View attachment 172364

Did you save the overall file? This doesn't tell us much as this is all just sub bass and just a frequency response.

 

[gks333]

Did you save the overall file? This doesn't tell us much as this is all just sub bass and just a frequency response.

I dont believe so. I was only saving 10-130hz screenshots since all I have used REW for is getting the 4 subs to work as one and then integrate them into the 5 mains. When tuning a friend system I did do a full sweep to 18Khz but the graph looked like absolute shit. Completely unusable. So much data to show in such a small space. Any suggestions on REW set up to get a usable full range sweep? Appreciate any suggestions. No PEQ curves have been applied yet BTW. I think I may apply a 2db bump at 32hz and 2db drop at 54hz and mess around with slope. Only been using REW for less than 2 months and been busy with life. I intend to do a compression test over xmas break on all four subs individually to find the weakest link. Then all together.

 

[Jdunk54nl]

There are a lot of articles on how to use REW and videos on youtube. My suggestion would be to choose one of those and follow it. Don't try to mix a bunch of them.

Your graph above definitely needs some work though. You have 15db swings between 20hz and 35hz and then again between 35hz and 55hz. That will make a lot of stuff sound boomy or lacking bass.

This was my last weekend work.

Pre eq:

Post EQ:

 

[gks333]

There are a lot of articles on how to use REW and videos on youtube. My suggestion would be to choose one of those and follow it. Don't try to mix a bunch of them.

Your graph above definitely needs some work though. You have 15db swings between 20hz and 35hz and then again between 35hz and 55hz. That will make a lot of stuff sound boomy or lacking bass.

This was my last weekend work.

Pre eq:

View attachment 172450

Post EQ:

There are a lot of articles on how to use REW and videos on youtube. My suggestion would be to choose one of those and follow it. Don't try to mix a bunch of them.

Your graph above definitely needs some work though. You have 15db swings between 20hz and 35hz and then again between 35hz and 55hz. That will make a lot of stuff sound boomy or lacking bass.

This was my last weekend work.

Pre eq:

View attachment 172450

Post EQ:
View attachment 172451

Wow!!! Post EQ looks amazing!! I want to keep the 20hz peak there since we dont hear those freqs as perceived loudness as much as bass 30 some odd Hz on up?

 

[Jdunk54nl]

This was after level setting and some minor eq work. The subs use a minidsp 2x4hd and the mains are using my yamaha rx a-1000 7 band "parametric eq". It is really a semi parametric eq due to having a select frequencies and q's to choose from. It works well enough. Crossover was set to 80hz.
If you look at 45hz, that is a phase issue between the two subs. It does not appear with a single sub playing. It isn't a fully 180 out of phase, just partially out of phase. I need to get an all pass filter on one of them to fix that phase issue.

These are also spatial averages using periodic pink noise and moving microphone method.

Here is the phase and you can see something weird is happening with the phase on the right subwoofer (red line) compared to the left subwoofer (green line). At 45 hz the red line has a big dip and the green line has a slight bump to make them about 105 degrees out of phase.

This is using SMAART and a single point in space measuring phase, so ignore the magnitude/frequency response stuff as it is not a spatial average.

 

[MRC01]

I find that proper bass, extended and linear, makes the midrange sound smoother, less edgy, more natural.
Conversely, with systems that have edgy sounding mids, sometimes the problem isn't in the mids, but it's uneven or attenuated bass response.
This is also a problem with some recordings: when they mic the instruments, they fail to capture low frequencies, or the lows are attenuated, which makes the instruments sound edgy. Some people like this, find it an artistic way to enhance the sense of detail. I find it annoying.

 

[abdo123]

Good read. I was hoping to see specific specs like: for freqs 10-36hz one must increase amplitude by 8db for the sound to be perceived as equal loudness as freqs 37hz-180hz and for freqs 181hz -........etc etc. As a ballpark generalization.

Don’t worry about that, these things should be considered in the mixing and the mastering stage of the music.

 

[gks333]

Ugh, I am COMPLETELY unfamiliar with the 2nd scrn shot. I dont even know how to read that "phasing" graph. The 10-20K graph looks great and not all squiggly lined like mine was. I am using a 2X4HD as well. Below, the brown line is roughly where I started. It was actually worse but I was not screen shoting and documenting much at first. All I did before was use a radio shack SPL meter to dial in subs. Yeah, REW has changed my system significantly for the better!

 

[Jdunk54nl]

Don't worry as much about the second graph right now. Learn how to use REW.

It was more of a learning for you as to why that 46hz dip was there and then showing an actual phase measurement (the middle graph in that second screen shot).
Basically phase/time alignment are similar (but not the same) things and those two lines should ideally have the same slope (slope of phase graph is time information) and overlap each other. Being 60 degrees (y-axis on that middle graph is degrees) out of phase gives you ~5db summation and 0 degrees gives you +6db. I am 105 degrees out of phase and that is causing the dip at 46hz in my first graph. 180 degrees out of phase would cause a HUGE dip in that first graph.

 

[gks333]

Don't worry as much about the second graph right now. Learn how to use REW.

It was more of a learning for you as to why that 46hz dip was there and then showing an actual phase measurement (the middle graph in that second screen shot).
Basically phase/time alignment are similar (but not the same) things and those two lines should ideally have the same slope (slope of phase graph is time information) and overlap each other. Being 60 degrees (y-axis on that middle graph is degrees) out of phase gives you ~5db summation and 0 degrees gives you +6db. I am 105 degrees out of phase and that is causing the dip at 46hz in my first graph. 180 degrees out of phase would cause a HUGE dip in that first graph.

Yep, I agree, I need to learn REW more. Man, REW has been one of the biggest kept secrets I have ever come across. Not as EASY as buying a new piece of gear/plug and play but MUCH cheaper and bigger rewards no doubt. Thanks for the info. I will likely refer to this post in the future as I venture further down the REW hole.

 

[Jdunk54nl]

Watch this, it may help:
Also Erin’s audio corner has a video on moving mic methods

I do moving microphone with periodic pink noise for eq work and single point measurements for time/phase work.

 

[gks333]

Watch this, it may help:
Also Erin’s audio corner has a video on moving mic methods

I do moving microphone with periodic pink noise for eq work and single point measurements for time/phase work.

I will check that out this WE. Again, thanks for the info and knowledge. Always appreciated.

 

[Trif]

Good read. I was hoping to see specific specs like: for freqs 10-36hz one must increase amplitude by 8db for the sound to be perceived as equal loudness as freqs 37hz-180hz and for freqs 181hz -........etc etc. As a ballpark generalization.

You can work those relationships out from the graphs, once you decide which one to use.

To answer your earlier question, a "thickness in the bass" is the result of distortion and the bulk of distortions occur much higher than the fundamental frequency. Remember, vibrations work in octaves, they double, like 40 80 160 320 640.... If you reduce the fundamental, you reduce the distortions, and the midrange becomes cleaner.

I wouldn't mind standing waves half so much if they didn't make everything rattle.

 

[Cote Dazur]

There is also perceptual masking, tones which are close together blend together in our brains

So much truth here and you will not see that in a graph, microphones are immune to it.
Their is so much more to the acoustic science than measuring sine waves with a microphone, except of course if your ears are microphones.

As humans are concerned, it is a little bit more complicated.