Frequency Response vs Listening Room vs Crossover Impact on Sound Quality

0 Members and 1 Guest are viewing this topic. Read 1414 times.

77SunsetStrip

  • Jr. Member
  • Posts: 35
Recently picked up a set of speakers from the 2005 timeframe.  FR from the manufacturer showed +/- 0.5 dB from 200 to 20K, without smoothing. Independent reviews that did measurements confirmed the manufacturer claim.  My in room measurements with REW show +/- 2.5 dB, 200 to 10K, gentle downward slope to 20K, 1/3 smoothing.  Probably not surprising for in room vs. ideal measuring conditions. 

Another set of speakers owned since 1996, updated (more on that later), measured +/- 2 dB from 200 to 14K, steep drop to 20K 1/3 smoothing.  Overall, a smoother curve between 200 to 10K.

Listening evaluation done with same equipment, same music, and same measured SPL. The 2005 vintage speakers sounded very smooth, clean, with excellent detail.  However, the overall impression was a bit clinical and lifeless.  Vocals were very smooth, but subdued  Overall nice, but an impression of listening to speakers.

The 1996 vintage speakers sound was very similar with regard to clean, with slightly less crisp cymbal while other percussion popped.  Vocals more forward.  Certain frequencies more forward.  Overall, a more lively presentation of music with the speakers almost disappearing.

Comparing FR to listening experience leaves me puzzled.  The 1996 vintage speakers in room FR smoother, but produced music with more life.  The 2005 vintage in room FR had a peak about 800, dip at 1K, and peak at 5K.  Yet the overall sound smooth and clean, but without life.

Back to the 1996 vintage speakers updates.  After watching several of Danny's videos improving the performance of speakers sent to him, did a crossover rebuild.  All quality parts used, nothing boutique or expensive, to rebuild the 1996 crossovers.  The 2005 vintage speaker crossovers are stock with several poly caps, a couple electrolytic in the LF section, iron core inductors in the LF section, a couple small air core inductors, and all sand cast resistors. 

Can the crossovers be the main reason there is a clinical vs lifelike sound quality?

genjamon

Can it be the main reason?  Yes, I think so.  Is it in this case?  Who knows if it's the MAIN reason, but you're certainly not comparing apples to apples there. 

Other differences?  You mention high freq. rolloff in the 2005 speakers starting around 10k vs 14k in the 1996 speakers.  I would expect that difference to present a bit less high frequency energy, and the acoustic effect could be less of a sense of air, but potentially a bit less high frequency attack too.  This could also have something to do with the perceived difference.  And of course there could be other details hidden in the morass of hills and valleys of the rest of the frequency response that would only reveal themselves if properly measured and if you shared the full frequency plots of the speakers. 

But I suspect the lack of crossover upgrades in the 2005's could be significantly holding those speakers back compared with your 1996's.

corndog71

  • Full Member
  • Posts: 1260
  • Some people call me Rob.
I would say it's a combination of crossover parts quality but also the limitations of the drivers.  I've had speakers that sounded fantastic but still missed the resolution of better designs.  You could send one to Danny to see if anything in the crossover needs improving or you could just try to replicate the crossover with better parts which will improve resolution somewhat.  If the manufacturer is still around shoot them an email.  There may have been updates to the crossover.  Also, if you can't figure out what the part values are, see if you can get the crossover circuit from them.

Tyson

There’s a difference between smoothness (frequency response) and liveliness (usually associated with efficiency) and imaging (high frequency extension plus phase response). 

But the biggest thing people don’t talk about is the room.  If you have a speaker with high efficiency and perfectly flat frequency response in a bright, untreated room, it will likely sound hard and clinical.  That’s because the room itself sounds hard and clinical.  Put that same speaker in a well treated room and it will sound amazing.

What I see happen a lot is audiophiles try to shoehorn a high performance speaker into a common area like a living room that’s poorly treated and too lively and no ability to make any changes acoustically to the room.  So they end up being forced to use a very polite sounding speaker like Dynaudio or any Scan Speak based speaker because the WAF prohibits a true high performance speaker from sounding it’s best.  My advice is to treat the room and screw the WAF.  Or at the very least, get the spouse involved in the room treatment process.

Early B.

What I see happen a lot is audiophiles try to shoehorn a high performance speaker into a common area like a living room that’s poorly treated and too lively and no ability to make any changes acoustically to the room.  So they end up being forced to use a very polite sounding speaker like Dynaudio or any Scan Speak based speaker because the WAF prohibits a true high performance speaker from sounding it’s best.  My advice is to treat the room and screw the WAF.  Or at the very least, get the spouse involved in the room treatment process.

That's not practical for most people. Let's face it -- room treatments don't belong in shared spaces. No matter how well you dress them up, they still look terrible.

Oftentimes, audiophiles have to make due with what we have. For instance, probably 90% of us have "bad" rooms to start with, and most are too small. Some people have their main rigs in a spare bedroom -- uugghhh!!!       

For those where room treatments aren't a viable alternative, speaker designers should offer more speaker tuning options.  I love my OB servo subs for this reason -- you can make the bass sound however you want. I want that same freedom for the mids and highs, as well.


corndog71

  • Full Member
  • Posts: 1260
  • Some people call me Rob.
Some people have their main rigs in a spare bedroom -- uugghhh!!!

I feel attacked. lol

Tyson

That's not practical for most people. Let's face it -- room treatments don't belong in shared spaces. No matter how well you dress them up, they still look terrible.

Oftentimes, audiophiles have to make due with what we have. For instance, probably 90% of us have "bad" rooms to start with, and most are too small. Some people have their main rigs in a spare bedroom -- uugghhh!!!       

For those where room treatments aren't a viable alternative, speaker designers should offer more speaker tuning options.  I love my OB servo subs for this reason -- you can make the bass sound however you want. I want that same freedom for the mids and highs, as well.



Ah yes, you are so right.  I should have pointed out that OB speakers avoid a lot of these problems very easily and naturally, due to the nature of OB design and how it interacts with less than ideal rooms.  Far easier to integrate into a common area than a box speaker, as long as you can get them 3 feet from the front wall.

HAL

  • Industry Contributor
  • Posts: 4653
That's not practical for most people. Let's face it -- room treatments don't belong in shared spaces. No matter how well you dress them up, they still look terrible.

Oftentimes, audiophiles have to make due with what we have. For instance, probably 90% of us have "bad" rooms to start with, and most are too small. Some people have their main rigs in a spare bedroom -- uugghhh!!!       

For those where room treatments aren't a viable alternative, speaker designers should offer more speaker tuning options.  I love my OB servo subs for this reason -- you can make the bass sound however you want. I want that same freedom for the mids and highs, as well.

Then an option if no room treatment is going to be used use DSP room correction to make it the best possible.  Usually would only say that for the last step in the room setup process. but it would be my second step in that case. 

One company that made very tunable speakers was VMPS.  They finally went to DSP room correction with measurements to make it easier for customers.

Those options are much better today than in the past.  Do not knock it, until you try it.

Early B.

Then an option if no room treatment is going to be used use DSP room correction to make it the best possible. ....  Do not knock it, until you try it.

Good point. What are the benefits and limitations? Which brand(s) do you recommend? Are they easy to set up and use?

HAL

  • Industry Contributor
  • Posts: 4653
The easiest one to setup is MathAudio RoomEQ and it is only for digital replay with programs like Foobar2000 and JRiver MC27.  I can also use RoomEQ VST style plugins with an ADC for analog sources.

You use a calibrated mic with the cal file loaded and it will make as many room measurements as you would like and then averages them together for a total response curve.  You can also create a house curve if you do not like flat response.

Trying out the very inexpensive Dayton Audio iMM-6 mic now that is a powered condenser that uses an iPhone style connection to the computer.  Most laptops have that as does my MS-6 Music Server.  Get a TRRS extension cable and mic clip and you start measuring.

Takes some practice, but easy instructions to follow. With each trial you can then listen to the results and decide to keep them saved or start over.

I believe in room setup and room acoustic treatments first, but some cannot do that for their space.  I have a dedicated listening space, so I do all of it.


Tyson

If you use Roon, then Roon has a very high quality digital EQ built in.  I use it in my upstairs system where I have my box based speakers. 

Sonicjoy

  • Full Member
  • Posts: 347
 In my mind there is something just not right about using DSP to intentionally screw up the frequency response that the speaker designer worked so hard to achieve. A few designers including Richard  Vandersteen and Danny Richie if I remember correctly say that DSP should only be used for low frequency corrections. No matter what you do with DSP you are not fixing the room.  If the room is very reflective it's still going to be reflective after you have applied DSP and so on. It's just a band aide at best.

And as has been mentioned, it's an unfortunate reality, but small rooms are just hard to make work well. Small rooms should have small speakers. Large speakers just over load the room with too much energy. Keeping the volume lower helps as well.

HAL

  • Industry Contributor
  • Posts: 4653
A parametric EQ as in the servo amps like the A370PEQ are no different than the DSP version.  The DSP version actually has more dynamic range with modern processing than analog circuits.  The difference is that you can do many more PEQ bands to work on more room modes.






Early B.

In my mind there is something just not right about using DSP to intentionally screw up the frequency response that the speaker designer worked so hard to achieve.

A speaker designer isn't designing speakers to your particular tastes, style of music, and room. The "last mile" should be under your control. It's analogous to buying a stock Harley Davidson -- it's a great starting point, but now you gotta personalize it. However, I don't think DSP is the answer yet.

A major limitation with DSP is it ain't plug & play. Calibrated mics, software, room measurements, room modes,....geez. It's overwhelming just thinking about it.  Even the A370 amp is complicated and there's still no clear instructions on how to use it.

HAL

  • Industry Contributor
  • Posts: 4653
The DSP solutions are actually easier to use as they have measurement capability to help with the adjustments. 

You can start as easily as installing Room EQ Wizard on a computer and use the built in soundcard with an external mic and cable to the system.  As I posted, the Dayton Audio iMM-6 calibrated mic is about $20.  Add a longer TRRS cable for $10.  My new laptop has the TRRS old iPhone style connector for mic and headphones.  You need a 3.5mm stereo to RCA cable adapter to go to the system.  You just download the calibration data from their website with the serial number of the mic.  REW asks for the mic data on setup.
 
There is a tutorial in REW for how to set it up and use it for measurements. 

Other than techniques like that, you are listening to music and making adjustments by ear.  That takes more time than with simple measurements.

REW also can output the PEQ settings for different EQ systems after the measurements are made.  This is much easier than calculating them based on the measurements.  For some systems it outputs the file needed to EQ the room. 

It has a lot of other capabilities that can be explored, like RT60 measurements for room reverberation time if you are going to acoustically treat it.


Rusty Jefferson

  • Full Member
  • Posts: 620
Well, this thread went sideways quickly.  :D

 
...Can the crossovers be the main reason there is a clinical vs lifelike sound quality?
Sure they can, but there's too many variables you haven't mentioned.  What brand of speakers? Same model? Same drivers just different year? Have you done impulse response measurements? Phase vs frequency response?

The best way to demonstrate whether the new crossover is making the difference is to compare your original crossover to the new one on the pair in question, not comparing 2 potentially different speaker pairs based on FR. I'm solidly in the camp that believe time domain behavior is as or more critical than FR measurements.

HAL

  • Industry Contributor
  • Posts: 4653
Time and frequency domains are just transforms of each other.

The flatter the frequency response, the cleaner the impulse response.  You can prove that easily by taking one sample of 1 in a file of zero's and do the FFT transform.  You will see it gives perfectly flat response over the Nyquist sample rate theorem's 1/2Fs range.

The ringing in the impulse response is the waterfall rolloff in frequency.  The shorter the impulse the shorter the waterfall response.  Long tails in the impulse response are long duration tails in the frequency response. 

Just different aspects of the same signal with different math.

Rusty Jefferson

  • Full Member
  • Posts: 620
Rich, it's not uncommon for a speaker (or electronics for that matter) to have good measured FR and have problems in the time domain with group delay and phase issues. Is one pair of his speakers a 4 way dynamic driver with a complicated passive crossover, and the other a full range electrostat?

The OP asked a question that's hard to answer with the information provided. It didn't sound like he was trying to fix the pair that sounded lifeless, but may be trying understand why it sounds lifeless. (?)

HAL

  • Industry Contributor
  • Posts: 4653
Rusty,
Yes, and if you unwrap the phase of an FFT of the test signal as REW will do, the phase response of the speaker can be seen and compared to others with the room removed if done as a gated time measurement.  There is a check box in REW to see it. 

The factor not discussed is how does the original recording actually sound?  If it is anything other than an acoustic recording with stereo micing, it will all depend on what the mastering engineer wants it to sound like in his facility. An acoustic recording direct to stereo mics and not processed is the best second best way to judge system sound in a room.   Getting the response in the room as flat as possible below 300Hz is a large part of the better sounding systems I have heard.  Most of the room correction systems will let the user adjust the upper end cutoff for correction if wanted.   

Once folks realize that the room is the biggest part of the sound quality of the system, and no designer can account for all rooms in a single speaker design, then finding specifics of sound quality of a specific speaker in a room will be different in every room for the same speaker. 

Tyson

As usual, HAL is absolutely right :)