Skip to main content

Topic: DSP and room treatments vs speaker responce when making buying decisions (Read 5462 times) previous topic - next topic

0 Members and 1 Guest are viewing this topic.
DSP and room treatments vs speaker responce when making buying decisions
Folks,

I'm looking at two speakers models while i try to decide which to set up for engineering and playback.  I want a near flat response of course.  One model has a frequency response variance of +-0.75dB and one has a frequency response variance of +-6dB.

Can one make up the difference with DSP and room treatments?  This article: http://www.soundonsound.com/sound-advice/q-can-i-improve-my-monitors-frequency-response-eq leans towards "no", but that was 5 years ago and things are always advancing in the DSP realm.

What are your current thoughts on this?
Music lover and recovering high end audiophile

Re: DSP and room treatments vs speaker responce when making buying decisions
Reply #1
I'm looking at two speakers models while i try to decide which to set up for engineering and playback.  I want a near flat response of course.  One model has a frequency response variance of +-0.75dB and one has a frequency response variance of +-6dB.

First, how do you know?  Are you under the illusion that the manufacturer's claimed frequency response necessarily means anything?

Second, even if the measurement is done by a reliable third party,  as stated the difference is meaningless even if correct.  Any frequency response needs to be specified over a frequency range and at a certain loudness.  Without that information there is no basis for comparison.

Ed Seedhouse
VA7SDH

  • greynol
  • [*][*][*][*][*]
  • Global Moderator
Re: DSP and room treatments vs speaker responce when making buying decisions
Reply #2
I guarantee he doesn't really even know this for sure:
I want a near flat response of course.
http://seanolive.blogspot.ca/2009/11/subjective-and-objective-evaluation-of.html
  • Last Edit: 30 October, 2016, 07:24:14 PM by greynol
Is 24-bit/192kHz good enough for your lo-fi vinyl, or do you need 32/384?

Re: DSP and room treatments vs speaker responce when making buying decisions
Reply #3
I'd assumed that manufacturers measure their speakers in anechoic chambers.  I don't know how accurate the numbers are, but that wasn't the point of the question. I was giving a hypothetical. What i'm really asking is: If we can make efforts to flatten a speaker responce plot via DSP and room treatments, how much does the variance that the speaker itself has matter, and if it does matter what is the threshold that we can't correct for with these tools?

I'd meant to put a "suppose" at the beginning of the post, but i missed it as i was thinking it out in my head.
Music lover and recovering high end audiophile

  • ajinfla
  • [*][*][*][*][*]
Re: DSP and room treatments vs speaker responce when making buying decisions
Reply #4
What i'm really asking is: If we can make efforts to flatten a speaker responce plot...
Speakers radiate 3 dimensionally, so they don't have "a" response plot, they have thousands. Which one and where? For what purpose?
Loudspeaker manufacturer

  • ajinfla
  • [*][*][*][*][*]
Re: DSP and room treatments vs speaker responce when making buying decisions
Reply #5
This has been posted before, should be mandatory reading here for all speaker/acoustics questions:
(open access) http://www.aes.org/e-lib/browse.cfm?elib=17839
http://www.audioholics.com/room-acoustics/room-reflections-human-adaptation
Loudspeaker manufacturer

Re: DSP and room treatments vs speaker responce when making buying decisions
Reply #6
I'd assumed that manufacturers measure their speakers in anechoic chambers.

Some do, but anechoic chambers are really really expensive and most manufacturers can't afford one.  They can do a close approximation using impulses and time gating.  But you don't listen in an anechoic chamber, do you?

Even if they do all that testing how do you know their adds aren't lying to you?  You think advertising people are scrupulously honest?

And as already pointed out, no loudspeaker has "a" frequency response.  Every loudspeaker has many frequency responses since a truly omnidirectional speaker is virtually impossible to create and then they will have other severe problems.  And even a perfectly omnidirectional loudspeaker will have a different frequency response in any given point in a reflective room from all the other points, and all these frequency responses will change when the loudspeaker moves even a small amount.
  • Last Edit: 31 October, 2016, 07:19:18 PM by Ed Seedhouse
Ed Seedhouse
VA7SDH

  • DVDdoug
  • [*][*][*][*][*]
Re: DSP and room treatments vs speaker responce when making buying decisions
Reply #7
Quote
One model has a frequency response variance of +-0.75dB
I don't believe that, especially if that's over the 20-20kHz range.  Is that spec from a reputable studio monitor manufacturer?

Quote
Can one make up the difference with DSP and room treatments?
To some extent, yes.    But, DSP can't fix reverb or a standing-wave node (where the sound is canceled) because it takes (nearly) infinite power and infinitely large woofers to overcome the cancelation.  So, normally you do room treatment first and then maybe EQ if it's necessary. 

And of course, you can't EQ a 5-inch woofer to sound like a 12 or 15-inch woofer/subwoofer.

Quote
I'm looking at two speakers models while i try to decide which to set up for engineering and playback
You don't need "perfect" speakers and a perfect room.   You need reasonably-flat and reasonably-smooth frequency response, and then you have to learn to make good mixes/masters on your system.   Every speaker/monitor sounds different and most mixing/mastering engineers wouldn't want "better" monitors once they've learned to get great results on the ones they've got.

  • greynol
  • [*][*][*][*][*]
  • Global Moderator
Re: DSP and room treatments vs speaker responce when making buying decisions
Reply #8
You need reasonably-flat
Quote
A flat in-room target response is clearly not the optimal target curve for room equalization. The preferred room corrections have a target response that has a smooth downward slope with increasing frequency. This tells us that listeners prefer a certain amount of natural room gain. Removing the rom gain, makes the reproduced music sound unnatural, and too thin, according to these listeners. This also makes perfect sense since the recording was likely mixed in room where the room gain was also not removed; therefore, to remove it from the consumers' listening room would destroy spectral balance of the music as intended by the artist.
http://seanolive.blogspot.ca/2009/11/subjective-and-objective-evaluation-of.html
Is 24-bit/192kHz good enough for your lo-fi vinyl, or do you need 32/384?

  • augustine
  • [*][*]
Re: DSP and room treatments vs speaker responce when making buying decisions
Reply #9
Digital Room correction can make a big difference but get the flattest speakers you can first.

  • greynol
  • [*][*][*][*][*]
  • Global Moderator
Re: DSP and room treatments vs speaker responce when making buying decisions
Reply #10
I'll stand by Sean Olive's conclusions that are based on actual test data, thanks.
Is 24-bit/192kHz good enough for your lo-fi vinyl, or do you need 32/384?

  • KozmoNaut
  • [*][*][*][*][*]
Re: DSP and room treatments vs speaker responce when making buying decisions
Reply #11
Which is why you should use something similar to JBL's or Brüel & Kjær's target response curves for your room correction, instead of aiming for 100% flat.

  • ajinfla
  • [*][*][*][*][*]
Re: DSP and room treatments vs speaker responce when making buying decisions
Reply #12
Which is why you should use something similar to JBL's or Brüel & Kjær's target response curves for your room correction, instead of aiming for 100% flat.
No one aims for flat in room mid/far field/at LP, as discussed ad infinitum in prior threads. That's either strawman or misunderstanding of what "flat" is being referred to, which is the native, onset response of the speaker, usually anechoic.
As for the JBL and B&K curve "targets", that too is questionable. There is no EQ police, EQ to whatever floats your boat. A well designed speaker should need no EQ above the transition frequency in Olives link...and minimum below, with exception of reducing large peaks. The bass "curve" should be whatever you feel best for your tastes in music. JBLs (Synthesis) boosted bass preference curve is based on HT and music, of unknown provenance. They are included in the Toole paper I linked earlier and is not what trained listeners preferred (thought it might help sell Synthesis subs).
This is still the best bass article I know of to date, as it is a compilation of many studies: https://secure.aes.org/forum/pubs/conferences/?elib=17270

Loudspeaker manufacturer

  • greynol
  • [*][*][*][*][*]
  • Global Moderator
Re: DSP and room treatments vs speaker responce when making buying decisions
Reply #13
I'm not an AES member and don't plan on purchasing the paper, but does it explain why so many people are impressed with Bose speakers?
Is 24-bit/192kHz good enough for your lo-fi vinyl, or do you need 32/384?

  • ajinfla
  • [*][*][*][*][*]
Re: DSP and room treatments vs speaker responce when making buying decisions
Reply #14
I'm not an AES member and don't plan on purchasing the paper, but does it explain why so many people are impressed with Bose speakers?
Nope. Summary of 30+ LF studies,
Quote
Use of a mono signal, replayed via multiple subwoofers with suitable signal processing, allows one to present more uniform bass over a listening area. However there is still no opportunity to present spatial information. Use of two subwoofers placed to the left and right of the listener and playing left and right LF signals allows presentation of spatial information. The system which has been studied least in the literature is the one where all 5 (or 7) main channel loudspeakers have good low frequency response. This
system in combination with appropriate signal processing would seem to present the greatest opportunity for controlling room modes, using both modal equalisation and the interaction of multiple drive units, and presenting spatial information.
Harman advocates the former, JJ used the latter, I'm in between, since I'm stuck with stereo recordings, for the most part (although I do have MCH music).
Loudspeaker manufacturer

  • emo_hp
  • [*]
Re: DSP and room treatments vs speaker responce when making buying decisions
Reply #15
Folks,

I'm looking at two speakers models while i try to decide which to set up for engineering and playback.  I want a near flat response of course.  One model has a frequency response variance of +-0.75dB and one has a frequency response variance of +-6dB.

Can one make up the difference with DSP and room treatments?  This article: http://www.soundonsound.com/sound-advice/q-can-i-improve-my-monitors-frequency-response-eq leans towards "no", but that was 5 years ago and things are always advancing in the DSP realm.

What are your current thoughts on this?


Yes, I use Math Audio Room EQ. It'a s freeware for foobar. I use foo_record and VB Cable "driver" to stream all my Windows sound thru foobar.
I'm with studio monitors. They're really one of the best. I though they're amazing... All people think though that...
Today if I stop the DSP, I'm wondering how is that possible:
1. Stereo effect is immediately reduced
2. High become strange
3. Sound is no more crystal clear, I loose details
4. Most instruments sound like a mix... more like a single distorted sound... unclear... instead each string as perfect distinguishable detail.

I don't like the fact I lost so many years with my audiophile friends with passive corrections, when the true is one: only digital correction could be perfect. Of course only EQ without phase correction always makes sound worse.

  • knutinh
  • [*][*][*][*][*]
Re: DSP and room treatments vs speaker responce when making buying decisions
Reply #16
... Of course only EQ without phase correction always makes sound worse.
I disagree with this statement.

The magnitude errors that we want to correct are usually not linear-phase. When applying a (usually minimum-phase) equalizer to flatten the magnitude, the change in phase response is hard to predict without measurements.

"Worse sound" implies listening tests, where it is usually found that moderate phase errors are quite inaudible.
 
-k
  • Last Edit: 29 May, 2017, 04:31:23 AM by knutinh

  • 2tec
  • [*][*][*][*][*]
Re: DSP and room treatments vs speaker responce when making buying decisions
Reply #17
Wouldn't DRC be most beneficial in an acoustically "unfriendly" room? Wouldn't DRC account for a speaker's dynamic  frequency response? Isn't the goal of DRC  to hear a "flat" room?
Quis custodiet ipsos custodes?  ;~)

  • knutinh
  • [*][*][*][*][*]
Re: DSP and room treatments vs speaker responce when making buying decisions
Reply #18
Wouldn't DRC be most beneficial in an acoustically "unfriendly" room?
DRC and acoustic behaviour tends to go hand in hand, affecting different somewhat aspects of sound.

If one really cares about sound, one should probably utilize both DRC and acoustic measures.

I imagine there are some cases where you really care about sound, but architectural/design/spousal requirements means that you are stuck with a highly suboptimal acoustic design. Perhaps DRC is the solution for such cases, but I guess that you will put in a lot of complexity/cost and still be far from a really good solution.
Quote
Wouldn't DRC account for a speaker's dynamic  frequency response? Isn't the goal of DRC  to hear a "flat" room?
What is a "dynamic frequency response"? Loudspeakers can be seen as linear time-invariant devices with a spatially (angular) variant transfer function for small-scale signals, and gradually altered transfer function for large-scale signals.

What is a "flat" room? A flat frequency response from signal output to the microphones of a Kemar? A flat direct-wave response from loudspeaker to an omni measurement microphone? A flat summed response from loudspeaker via room reverberation to an omni measurement microphone? A "smooth" magnitude response that has no large-scale magnitude tilt compared to the room/loudspeaker combination used to mix the recording?

Floyd Toole, as I recollect, advocates anechoically measuring the loudspeaker magnitude response in a limited angular "window" around the front-facing direction. This would be a compromise between "on-axis" anechoic measurements, full power response measurements in special reverbrant chambers and measuring the in-room response for some "reference" room using one of the suggested strategies above.

-k
  • Last Edit: 30 May, 2017, 04:56:20 AM by knutinh

  • 2tec
  • [*][*][*][*][*]
Re: DSP and room treatments vs speaker responce when making buying decisions
Reply #19
What is a "dynamic frequency response"? Loudspeakers can be seen as linear time-invariant devices with a spatially (angular) variant transfer function for small-scale signals, and gradually altered transfer function for large-scale signals.

I'll try to clarify my question, perhaps you could clarify your response for me? Speaking in strictly non technical terms, doesn't DRC software compensate for any given loudspeaker's less than ideal "frequency response"?

What is a "flat" room? A flat frequency response from signal output to the microphones of a Kemar? A flat direct-wave response from loudspeaker to an omni measurement microphone? A flat summed response from loudspeaker via room reverberation to an omni measurement microphone? A "smooth" magnitude response that has no large-scale magnitude tilt compared to the room/loudspeaker combination used to mix the recording?

Again, speaking as a layperson, I meant a tonally neutral space, neither too bright nor too soft. So, once more, isn't an effective DRC system supposed to compensate for a less than ideal room? Therefore my actual question is, given an effective DRC system, shouldn't the speakers and the room sound close to ideal loudspeakers of equal power in an ideal space of equal size?
  • Last Edit: 30 May, 2017, 07:10:03 PM by 2tec
Quis custodiet ipsos custodes?  ;~)

Re: DSP and room treatments vs speaker responce when making buying decisions
Reply #20
What is a "dynamic frequency response"? Loudspeakers can be seen as linear time-invariant devices with a spatially (angular) variant transfer function for small-scale signals, and gradually altered transfer function for large-scale signals.

At some level of detail speakers are constantly changing, probably more so that any other audio device. Voice coils heat up and cool, changing their resistance for example. But, the performance variations due to that are not all that noticeable for high quality devices run well within their capabilities.

Quote
I'll try to clarify my question, perhaps you could clarify your response for me? Speaking in strictly non technical terms, doesn't DRC software compensate for any given loudspeaker's less than ideal "frequency response"?

Digital Room Correction is IMO a horrible misnomer because correcting 3 dimensional rooms via the output of a small number of co-located speakers is  mission impossible.  Digital Audio System Response Adjustment (DASRA) might be a better term.

[quite]
What is a "flat" room? A flat frequency response from signal output to the microphones of a Kemar?

That may sound intuitively good, but I'm intimately familiar  with the SQ of a number of audio systems based on that criteria, and they tend to make everything sound like Hip Hop.

Quote
A flat direct-wave response from loudspeaker to an omni measurement microphone?

Ignores the influence of speaker off-axis response, which is known to be a very significant omission.

Quote
A flat summed response from loudspeaker via room reverberation to an omni measurement microphone?

Ignores the stronger infuence of direct sound.

Quote
A "smooth" magnitude response that has no large-scale magnitude tilt compared to the room/loudspeaker combination used to mix the recording?

Known to be harsh sounding, especially in larger rooms.

I think you know all these things and formed this list for an important rhetorical reason - to point out that there is no simple generally agreed-upon goal for audio system response. The ones that work are generally not simple.


Quote
Again, speaking as a layperson, I meant a tonally neutral space, neither too bright nor too soft. So, once more, isn't an effective DRC system supposed to compensate for a less than ideal room? Therefore my actual question is, given an effective DRC system, shouldn't the speakers and the room sound close to ideal loudspeakers of equal power in an ideal space of equal size?

If wishes were fishes...