Should we correct to Schroder, or full range?

[Keith_W]

I would like to gather opinions from ASR members as to whether we should apply room correction to the Schroder frequency, or whether we should do full range correction from 20Hz - 20kHz. There appear to be two schools of thought:

- Correction to Schroder: aims to only correct bass peaks and dips, leaving everything above Schroder uncorrected.
- Full range correction: bass peaks and dips are corrected, and an overall target curve at the listening position applied.

In my own system, I only correct to Schroder. I have tried a full range correction before, but it sounds awful. And this is with using very generous smoothing for upper frequencies. This made me ask myself whether there is something I am not measuring which, when correction is applied, makes the upper frequencies sound shrill (even with the Harman curve). The most obvious thing I am not measuring is the speaker's off-axis response and sound power, because it is not easy for me to do so. My speakers weigh 110kg and are physically massive, so it's not easy to take them outside and hoist them without hiring a forklift or something similar.

I have seen debates between proponents of both approaches. My theory for people having different approaches is that different systems need different types of correction. For one, most of us do not know what the directivity of our speakers are, and even if these specs are published, we can't exactly load them into our DSP software for correction. Also, different rooms may modify the spectral response in different ways.

So, I would like to know: which approach do you use, and why?

Hoping to hear responses from @mitchco and @UliBru

 

[Purité Audio]

Just to where it gets less wiggly.
Tried full range corrections with Dirac, Trinnov and Illusonic but I sort of feel ‘less is more’, unless of course you are living in a Turkish bath.
Keith

 

[sigbergaudio]

Correct for peaks (not necessarily full dampening) below Schroeder, and then gradually more careful through the transition range. Rule of thumb: No EQ above 500hz, rarely above 250-300hz.

Don't use DSP to force a different target curve than what the speaker would naturally reproduce in your room.

 

[SSS]

When DSP EQ then only below say 300 Hz. Above the correction will probably be incorrect due to speaker and room influence. For me it seems wrong to try correction of every dip and peak with narrow bandwidth allover the audio band. I do not belive that this will make the sound better. But if somebody likes it, why not.

 

[Blumlein 88]

If you are using multiple measurement points I've found full range to be good with less good speakers. With good controlled dispersion speakers seems useful to me to go an Octave above Schroeder.

 

[hvbias]

Schroeder. My assumption is any speakers I'm interested in are objectively good enough to not need correction above that.

 

[tmtomh]

My $.02:

1. Don't auto-correct full-range; go to Schroeder (or Schroeder + 1 octave per @Blumlein 88 - very cool idea, and worst-case I imagine it would do no harm).

2. If your speakers have nonlinear response or flawed dispersion, or your room is unusually weird/difficult, then manual PEQ at the specific trouble frequencies is going to produce a result as good as, and usually better than, full-range auto-correction.

3. I think a target curve can and should be separated from auto-correction. GLM does this: it auto-corrects more or less only to Schroeder (maybe a little above but not much). Then a separate, optional step allows you to overlay a sonic profile aka a target curve. But that target curve is just a couple of shelves - so the sonic profile/target curve just provides a smooth, gradual boost and/or cut to tilt the overall response. The sonic profile aka target curve doesn't implement filters to correct peaks and dips - it just tilts the whole response based on how you set the low and high shelves.

 

[dfuller]

Schroeder. Above that is pretty much a fool's errand, especially if there's any substantial amount of directivity error.

 

[digitalfrost]

With Frequency Dependent Windowing you can easily influence how much correction is applied in the higher frequencies. I agree that one should one not simply apply full range correction without regard for anything else, but in my tests, when done right full range correction is easily better than stopping at Schröder.

 

[DVDdoug]

If it doesn't sound right, of course it's OK to correct by-ear. And if EQ "correction" makes it sound worse to you, don't do it just because it "measures better". Music is for you to enjoy!

You can measure to get a more-clear idea of what the problems might be, but if you try to perfectly-flatten the response it will only work if never move your head!

 

[Spocko]

I would like to gather opinions from ASR members as to whether we should apply room correction to the Schroder frequency, or whether we should do full range correction from 20Hz - 20kHz. There appear to be two schools of thought:

- Correction to Schroder: aims to only correct bass peaks and dips, leaving everything above Schroder uncorrected.
- Full range correction: bass peaks and dips are corrected, and an overall target curve at the listening position applied.

In my own system, I only correct to Schroder. I have tried a full range correction before, but it sounds awful. And this is with using very generous smoothing for upper frequencies. This made me ask myself whether there is something I am not measuring which, when correction is applied, makes the upper frequencies sound shrill (even with the Harman curve). The most obvious thing I am not measuring is the speaker's off-axis response and sound power, because it is not easy for me to do so. My speakers weigh 110kg and are physically massive, so it's not easy to take them outside and hoist them without hiring a forklift or something similar.

I have seen debates between proponents of both approaches. My theory for people having different approaches is that different systems need different types of correction. For one, most of us do not know what the directivity of our speakers are, and even if these specs are published, we can't exactly load them into our DSP software for correction. Also, different rooms may modify the spectral response in different ways.

So, I would like to know: which approach do you use, and why?

Hoping to hear responses from @mitchco and @UliBru

It depends on the correction algorithm's design objective. If the "correction" seeks to eliminate unwanted interference between speakers and the room in order for clearer imaging of surround effects, then I would absolutely correct through full range - however, this is very dependent on the capability of the room correction software. Comparing the before/after SQ of movies, it was very clear to me that Trinnov definitely cleaned up surround clarity and imaging; this is especially notable as I had already run Genelec's GLM on all my connected SMART speakers (8351, 8350, 8331, 7380, etc.). GLM made all the necessary target curve corrections and fixed some of the bass peaks/dips to my speakers as a whole, but Trinnov cleaned up the time alignment and room interactions to an even higher (audible) level - that chaotic race scene near the beginning of Ready Player One is the perfect test track for this.

 

[Blumlein 88]

I've found the old Tact system to work better than some full range. Maybe it has to do with sweeps vs impulse measuring. I know one can be converted to the other, but perhaps windowing is more effective with impulse measurement. The Tact used two different impulses. Filtered for high or low frequencies if I recall. You also could set it to do multiple impulses. You needed that as impulse measurement is more polluted by noise than sweeps. Five impulses was sort of a minimum useful amount in my experience, but doing 25 was better and more consistent.

 

[Keith_W]

Hahah, I have tried everything With DSP, the only thing it costs me is time! This is what I tried:

- extremely generous smoothing (FDW setting of 1 in Acourate)
- spatial averaging over the entire listening position with generous smoothing
- quasi anechoic using beamforming technique (link and link)
- different target curves

In the end, what sounds best is doing nothing above Schroder. Every time I correct the full range axial response to look nice, it sounds awful. So I have stopped doing it.

Like I said in my first post, I suspect it is system dependent. I have a few friends who use DSP, and I am the exception in that all of them are doing full range correction, whereas I am the only one correcting up to Schroder. I am somewhat surprised to see responses in this thread with so many people only correcting up to Schroder and not above, I was expecting to see more people doing full range correction!

 

[Gringoaudio1]

Correct for peaks (not necessarily full dampening) below Schroeder, and then gradually more careful through the transition range. Rule of thumb: No EQ above 500hz, rarely above 250-300hz.

Don't use DSP to force a different target curve than what the speaker would naturally reproduce in your room.

Why not? Please cite your references.

 

[ozzy9832001]

I find that correcting up to about 700hz-1000hz works best, but it will all ready depend on what the "issue" is and what one is trying to accomplish. Flat frequency to me from 200-750ish sounds awful. Like you in a wind tunnel, but different genres of music have a different effect.

A gain mismatch would be another thing you may want to correct for up until about 3000hz.

I think it really just depends.

 

[sigbergaudio]

Why not? Please cite your references.

No references, this is my opinion and recommendation. I was also mainly thinking about manual EQ, but the same goes for automatic calibration unless you have very high end stuff like Trinnov.

 

[Sokel]

Why not? Please cite your references.

What is your favorite house curve

OK. Next question for you folks. As far as I can tell there is no spinorama for the Focal Aria 926's, which happen to be my main speakers. However, Amir did do a spinorama for the Aria 906 standmount speakers. They use the same tweeter and mid driver as the 926, but lack the two woofers as they...

www.audiosciencereview.com

 

[HarmonicTHD]

Hahah, I have tried everything With DSP, the only thing it costs me is time! This is what I tried:

- extremely generous smoothing (FDW setting of 1 in Acourate)
- spatial averaging over the entire listening position with generous smoothing
- quasi anechoic using beamforming technique (link and link)
- different target curves

In the end, what sounds best is doing nothing above Schroder. Every time I correct the full range axial response to look nice, it sounds awful. So I have stopped doing it.

Like I said in my first post, I suspect it is system dependent. I have a few friends who use DSP, and I am the exception in that all of them are doing full range correction, whereas I am the only one correcting up to Schroder. I am somewhat surprised to see responses in this thread with so many people only correcting up to Schroder and not above, I was expecting to see more people doing full range correction!

So stick with it and don’t worry what others prefer.

 

[Balle Clorin]

Above 1000hz you you at betting more of the tweeter direct sound and the correction becomes more speaker correction than room mode correction . I find that I like both full range correction and below 500-1000. I now use use smooth 1/3 correction above 1000hz and more aggressive below. Too aggressive correction above 1000hz does not sound right, loose the air and space around vocals.
I cannot longer listen without my Trinnov Optimizer running, It is just not good without . I hate the flat default correction-sounds thin and wrong , I lift the bass 3-5db , that resembles the Toole subjective curves

 

[krabapple]

Gee, it's like Floyd Toole has never been quoted on this topic here before.

Yes, above Schroeder you are 'correcting the speaker'. If it needs that -- and you won't know without a Klippel-like set of measurements -- then try it. If not, why mess up your nice speaker performance?

Below Schroeder -- room modes -- that's fair game. Peaks mostly, though there are tricks to treat troughs too. Ask Amir!