A collection of speaker target responses in .csv/.txt format

[Le Concombre]

Diminishing results ? Trying to correct Amplitude of Speakers with poor directivity, even very expensive (ever seen any anechoic on off axis measurement of Wilson Audio speakers ?) is more failure risk prone.

Difference is not subtle with my very well behaved speakers

 

[ernestcarl]

Difference is not subtle...

Perhaps in some systems the difference may be more obvious.

Below is the time domain performance of the Neumann KH120s at nearfield listening position:

From this zoomed-in vantage point, at first glance, the time-domain performance in the mid-range looks god-awful in comparison. But actually, most other speakers are in-fact like this... regardless of price or pedigree. I doubt most good speakers would have as much benefit when linearizing the time-domain as can be so obviously heard when equalizing gross errors in amplitude of the frequency domain, esp. by just simply fixing large modal peaks in the bass -- granted, you can't fix directivity issues.

The marketing department of Presonus for my S8 monitors exaggerate certain things well (but all marketing departments do that). At least they start with the phrase "articulates nuances" -- yes, nuances, that's the word I was looking for!

I'm not anti-DSP, of course. Only skeptical of hyperbole of any kind...

 

[Le Concombre]

Illustrated consequence of optimized Time Domain on Amplitude ;
When I would have picked a Trained Listeners bass level, I choose the filters with bass level a good dB lower (referenced as #Toole's in my previous postings) because enjoying the beautiful fast transients beats enjoying heavier bass level

 

[Le Concombre]

I have merged, at 140 Hz, my actual summation of L+R outputs for Trained Listeners with HATS target for it seems obvious IMO that it's its idealisation.
I suggest its usage in lieu de the Trained Listeners kindly posted before.
IMO it's the number one target to have. Yes it's about 2 dB above BK @ 50 Hz or so but it's 1dB below around 150 Hz, making it snappier and very acoustic music compatible while bass extended.
Once again, players such as HQPlayer allow rapid switching between filter sets and I also have at hand BK, HATS, and an intermediate bass boost between Trained and HATS

 

[ernestcarl]

IMO it's the number one target to have. Yes it's about 2 dB above BK @ 50 Hz or so but it's 1dB below around 150 Hz, making it snappier and very acoustic music compatible while bass extended.

Curve looks good, but, this does not seem to account for sub bass extension at all.

An interesting phenomena: L+R measurements boost the bass region a bit.

Measured individually, the bass between 100-400 Hz looks anemic -- but it doesn't sound that way to me.

And if I applied any more negative shelving to the HF of the S8, it would sound too dull and lifeless... so personally, I'd rather apply an increase or reduction in the treble or bass subjectively rather than to some fixed target.

 

[Le Concombre]

Curve looks good, but, this does not seem to account for sub bass extension at all.

View attachment 105806
An interesting phenomena: L+R measurements boost the bass region a bit.

View attachment 105808
Measured individually, the bass between 100-400 Hz looks anemic -- but it doesn't sound that way to me.

And if I applied any more negative shelving to the HF of the S8, it would sound too dull and lifeless... so personally, I'd rather apply an increase or reduction in the treble or bass subjectively rather than to some fixed target.

Neither does the original Harman Test target posted by the OP.... Even if your room is 8m long you have uniform pressurization but no distinct notes below 21.5 hz, below 17 if it's 10m long.... Actually my latest efforts aimed at reducing my bass response.... Linkwitz-Riley 24 dB @ 15 Hz helped .

 

[thewas]

Target curves most match the room reverberation and loudspeaker directivity functions as well as the chosen listening distance, for example in my current listening room above target sounds too mid shy and too bass heavy like also most Harman ones. Its usually more expedient to use the desired loudspeaker (with smooth directivity) anechoically linearised to measure at the LP how the target should be below the statistic region (approximately in the middle between the room induced peaks and dips) to minimise the room influence. Above transition frequency a linear direct sound or LW.

 

[Le Concombre]

Target curves most match the room reverberation and loudspeaker directivity functions as well as the chosen listening distance, for example in my current listening room above target sounds too mid shy and too bass heavy like also most Harman ones. Its usually more expedient to use the desired loudspeaker (with smooth directivity) anechoically linearised to measure at the LP how the target should be below the statistic region (approximately in the middle between the room induced peaks and dips) to minimise the room influence. Above transition frequency a linear direct sound or LW.

Above the transition frequency, around 300, way below Mids then, I only use a handful of eQ points, no higher than 2.87 and with less than 1 db amplitude. And my in room response fit BK and Toole's predictions very well above the transition frequency. I aim to match targets only below the transition frequency, following the excellent recommendations by Toole that you reproduced, simply to manage the bass response.

 

[thewas]

Above the transition frequency, around 300, way below Mids then, I only use a handful of eQ points, no higher than 2.87 and with less than 1 db amplitude. And my in room response fit BK and Toole's predictions very well above the transition frequency. I aim to match targets only below the transition frequency, following the excellent recommendations by Toole that you reproduced, simply to manage the bass response.

Excellent, if it came wrongly across I also didn't refer my reply to your post and target but to the responses of others that it didn't work well for them.

 

[ernestcarl]

Neither does the original Harman Test target posted by the OP.... Even if your room is 8m long you have uniform pressurization but no distinct notes below 21.5 hz, below 17 if it's 10m long.... Actually my latest efforts aimed at reducing my bass response.... Linkwitz-Riley 24 dB @ 15 Hz helped .

For most music it’s not so important, but for me it’s much more about the physical sensation gained rather than hearing distinct notes. I’ve tried cutting it off earlier, but something definitely was lost — that occasional infrasound feeling of impact in some films and sub bass heavy music soundtracks.

 

[ernestcarl]

I only use a handful of eQ points, no higher than 2.87 and with less than 1 db amplitude.

Wow... that’s some super strict rules you have. I agree about keeping it low Q and keeping the boost max low — but I also think 5-6dB to fill in wider dips is fine as long as you have the headroom available without adding too much distortion, and if it doesn’t affect the adjacent listening area around too much.

 

[Le Concombre]

Wow... that’s some super strict rules you have. I agree about keeping it low Q and keeping the boost max low — but I also think 5-6dB to fill in wider dips is fine as long as you have the headroom available without adding too much distortion, and if it doesn’t affect the adjacent listening area around too much.

You may find a guru to tell you that he threw his speakers wherever they looked nice in the room and that with his magic and magical wand (DSP room eQ program) you can do the same and he will turn whatever speakers into wonders of the world by targeting a global downward tilt extracted from Harman's research.

The excellent collection of Floyd Toole's writing gathered here https://www.audiosciencereview.com/...e-targets-room-eq-and-more.10950/#post-307665 by @thewas tells different : we should not confuse predictions and measurements of well behaved speakers with targets. I use "targets" to have at hand a substitute for the bass tone control I don't have. See attached Zoom attachment. The lowest curve (green) aims flat up to 2K and the one above BK, upper come Trained, Boosted Trained and HATS. They all connect around 300 Hz with my (almost) untouched in room response.

Let s take 2 excellent trio jazz albums recorded 50 years apart : Corea's Now he sings will remain bass shy even with the upper correction while Bad Plus' Never Stop II is (nicely) bass heavy even with a flat to 2K correction

In Synthese you can see how Toole's and BK's predictions fit my responses though I only use a handful of eQ points, no higher than 2.87 and with less than 1 db amplitude. Nothing in the reading of the FDW 3 presentation ( L R aiming BK pair appearing in the middle) would justify filling the Steady State 440 dip and I prefer to match the corrections ( the lowest pair displays my LR filters) to the LEAST common tweaking rather than correcting the gap around 2K though in theory with FDW 3 we are looking at Direct Sound at that frequency. The map is not the territory

 

[ernestcarl]

When you apply FDW 3 and use it as reference, do you vector average multiple sweep measurements from slightly different spatial locations?

 

[Le Concombre]

When you apply FDW 3 and use it as reference, do you vector average multiple sweep measurements from slightly different spatial locations?

I RMS average multiple sweep measurements from slightly different spatial locations

 

[ernestcarl]

I RMS average multiple sweep measurements from slightly different spatial locations

I was under the impression you mainly used the static measurements and applying FDW 3 afterwards to inform your corrections. From what you said above, I gather you still use FDW, but you mainly use regular spatial averaging of swept sine measurements for your EQ correction.

 

[Le Concombre]

I was under the impression you mainly used the static measurements and applying FDW 3 afterwards to inform your corrections. From what you said above, I gather you still use FDW, but you mainly use regular spatial averaging of swept sine measurements for your EQ correction.

I correct based on steady state below transition ("the room is in command"), lightly and to least common ("the speaker is in command") FDW 3 above
Might be of interest to mention SPLs at LP

as low as 78 dB A for a string quartet as high as 96 dB C with some excellent DVD Audios, piano peaking around 85/86 dB C, symphonic orchestras up to 90 +-2, movies up to 92 for exceptionally loud sounds. I use the upper (HATS) correction for movies, a bit bass heavy but

 

[Tom.O]

Are these curves for single or dual channel? In case they are single channel then when measuring L+R you will see 6db boost in bass region. (if phase is alligned between left and right channel)

 

[ernestcarl]

Are these curves for single or dual channel? In case they are single channel then when measuring L+R you will see 6db boost in bass region. (if phase is alligned between left and right channel)

Where did you get that huge figure?

If you mean my post, phase is aligned well enough. Boost is more like 1-3dB with MMM depending on frequency -- swept sign measurements show less ~1.5 or 2 dB (?) in the single mono sub region -- if I remember correctly.

*You probably are thinking the overall volume level when summing L+R measurements? If so, yeah. But only about 5dB in my room/space.

 

[Tom.O]

Where did you get that huge figure?

*You probably are thinking the overall volume level when summing L+R measurements? If so, yeah. But only about 5dB in my room/space.

Exactly, when measuring both channels together I observe 6dB increase especially below 100Hz. mid and heights are less boosted as phase vary a lot with shorter waves. There is still open question whether I should tune each channel separatelly to harman curve or when they play together....

 

[Le Concombre]

@Soniclife Sorry for the long wait. Here's Toole's curve for trained listeners only:

Spoiler: Comparison

View attachment 89076

I think Harman has created the attached RR1 target as an optimisation of the Trained Listeners only (see comparaison ). I'm not a big fan of its beefed circa 160 Hz region, never found it correct with classical/acoustic jazz etc, but sometimes it's the good match with the mastering. I have reduced my "quiver" to 2 pairs to choose from depending upon the content: BK/RR1 in dotted lines in the attached picture, and Harman/Toole.