Phase Shift Audible?

[B_Pay]

When using a parametric EQ, what factors contribute to a more noticeable and audible phase shift.

Is there any disadvantage to using a lot of filters besides phase shift?

Is phase shift one of those things that everyone agrees is bad but normally is not much of a noticable issue. For example, everyone agrees that total harmonic distortion needs to be as low as possible. But a device that has a THD of .05% is really no better than one with .06% because the difference is not audible?

[brucek]

I don't believe phase shift is audible from a single source (subwoofer). When the sub is mixed with a main speaker, then phase shift can cause uneven response at the crossover. That is certainly correctable though.

brucek

[Kevin Haskins]

It depends on how much you are talking about and at what frequency.

All the research shows, that the phase shift of your typical LR 4th order filter is only audible with very specific test signals. We are talking a square wave type signal, not music. With normal program material there is zero correlation in listener preference between a DSP corrected crossover with zero phase distortion, and a good ole trusty LR 4th order with a mild phase wrap from 100Hz-20K.

Will that stop people from pursuing "perfect phase"? Nope.... people will still chase it because the audio market is fundamentally irrational.

If you are using an equalizer in the bass you even have less to worry about. The room contributes far more out of phase issues than does the electrical filter.

Kevin Haskins
Exodus Audio

[lcaillo]

Can you point out some of that research? I think that is a fundamentally rational request, don't you?

[Kevin Haskins]

lcaillo wrote:

Can you point out some of that research? I think that is a fundamentally rational request, don't you?

Are you an AES member?

Audibility of phase shift above the transition zone:

Toole 1986
Greenfield & Hawksford 1990
Hansen & Madesen 1974a, 1974b
Lipshitz et al, 1982
Van Keulen 1991

Related: Audibility of Group Delay:

Bilsen & Kievits 1989
Deer et al., 1985
Flamagan et al., 2005
Krauss, 1990
Lipshitz et al. 1982

Greenfield & Hawksford give some guidelines that kind of summarized their studies and others. If you want one quick and dirty go there or buy Toole's book.

Kevin Haskins
Exodus Audio

[lcaillo]

Thanks.

[B_Pay]

This site is great....knowledgeable and really helpful people...thanks.

Gotta love the scientific method. Any research on how all amplifiers sound the same?

My plan is to use as many filters as I need to get the freq response as flat as possible. Do you guys have any tips on how to go about leveling the freq response? I was gonna just start from left and right using whatever filter is necessary to get it as close to reference as possible and work my way to the higher frequencies. Your advice is much appreciated.

[Kevin Haskins]

B_Pay wrote:

This site is great....knowledgeable and really helpful people...thanks.

Gotta love the scientific method. Any research on how all amplifiers sound the same?

More than you want to know about. ;-)

There are some things that obviously make them sound different. The really high output impedance of some tube amps, especially single ended triodes, plays havoc with the FR due to the changing impedance of most loudspeakers. This just doesn't occur on the bottom-end, it often is also an issue whereby you can get 2-3dB FR changes across the spectrum when used with loudspeakers that are not a stable load. That is the norm too, most speaker designs have peaks near crossover and it is not unusual to see 20-30 Ohm impedance peak near crossover. When they say nominal 8-ohm, it is pretty much a lie. ;-)

Moral of the story, is that if your going to use one of those amps, use a loudspeaker designed such that the impedance doesn't vary much with frequency. That is certainly a factor that will be audible as you switch loudspeakers and amplifiers.

The other differences are more difficult to nail down. I'll stay out of it other than to say that if you go looking for a difference, you will often find one.

My plan is to use as many filters as I need to get the freq response as flat as possible. Do you guys have any tips on how to go about leveling the freq response? I was gonna just start from left and right using whatever filter is necessary to get it as close to reference as possible and work my way to the higher frequencies. Your advice is much appreciated.

Ha... there are books on the subject. I'll punt....

Kevin Haskins
Exodus Audio

[brucek]

Do you guys have any tips on how to go about leveling the freq response?

Most everyone here uses Room EQ Wizard (REW) software.

Read and become familiar with the REW HELP files and the REW Cabling and Connections Basics.

Also note the REW information Index and the Download Page.


brucek

[B_Pay]

I've already downloaded REW and have been playing with the EQ function on there. I just wasn't sure if there was a particular way to go about implementing the filters...for example, does it make sense to keep the bandwidth of the filter as narrow as possible?

I read somewhere to cut narrow, boost wide, but I think that pertains to corrections to sound imperfections, not leveling out the freq response...not sure.

[brucek]

I just wasn't sure if there was a particular way to go about implementing the filters...for example, does it make sense to keep the bandwidth of the filter as narrow as possible?

REW determines the filters for you. The only filters you would require entering manually would be gain filters (as REW only creates cut filters). The bandwidth of the gain filter should match the bandwidth of the dip, and generally very narrow filters should be avoided.

brucek

[Andysu]

brucek wrote:

I don't believe phase shift is audible from a single source (subwoofer). When the sub is mixed with a main speaker, then phase shift can cause uneven response at the crossover. That is certainly correctable though.

brucek

It will display as a slight deviation in the frequency shift on an RTA when sliding the variable phase though (0° 180°).