An experiment in "extreme" toe-in for the Sceptre S8 coaxial/coactual monitors

[ernestcarl]

Room dimension: 3.3 x 4.9 m / 11 x 16 ft

The couch is 11-inches from the back wall. The Speakers are about in the middle of the room's length almost touching the sidewalls, say, an inch.

This room is clearly over-damped (haphazardly), and I would have preferred to do my measurements with/out the foam absorbers -- but I can't without A LOT of work and mess as they are already spray glued (felt sick and nauseated for two days from all the flammable adhesive spray fumes -- never again!) When I remove them, it will have to be final & permanent (replaced with "strategically" placed broadband absorber panels):

*BTW, in case you are wondering, there is a way to remove them without damaging the drywall/paint: use Goo Gone. It's a mess, but it works.

*According to some people, a room as dead as this is absolutely not a good candidate for doing the extreme-toe in speaker placement experiment. Too bad. We'll have to make do.

Not having the luxury of enough space to a put a centre channel in the middle of the room (cables on the floor is a tripping hazard too), we will have to make do with just left and right channels.

I cannot make use of the front-half of the room because it's currently occupied by my desk/office station. And there is an electric drop-down screen 3.3m from the back wall.

That's the LSR305 placed conveniently above the S8 to act as surrounds for the KH120s.

Cleaner days (some old pics):

*I did not install the in-wall speaker cabling (previous owner). He probably had an AVR and some passives. I probably will never own one.

I was inspired to do this "extreme" toe-in speaker placement test after reading reviews about the HSU CCB-8 and the design behind it. What the Sceptre S8 and CCB-8 have in common is their coaxial driver alignment. In reality, other than that fact, little else.


SOME MEASUREMENTS:

MLP with extreme toe-in

Compression in the sub bass region starts above the red line.


MLP with extreme toe-in

MLP with extreme toe-in

I was interested to see how the KH120 would sound in the far-field (3.4m actually):

S8 with extreme toe-in

Apart from the huge hole in the FR, not bad. But I don't think it's suited for that purpose. Sitting at the edge of the big ottoman instead of the couch sounded best.

Desk MLP vs Couch MLP

Desk MLP vs Couch MLP

The horn flare up increases as you go off-axis the S8s.


My conventional crossover doesn't work well (who knew without doing the measurements!)
Yeah, it sounded good, then, but it wasn't exactly the best choice, was it?

Overlapping it a bit improves the the response in ALL SEATS:

Unfortunately, the sub output drops significantly above 30Hz due to placement. I know where the sub works best for the couch (closer to the middle of the room), but cannot use that location as it's a horrible place in light of my shared front desk setup in this tight space.

So what is the FR in the three couch listening positions I tested for regular toe-in?

regular toe-in

The bass build-up in the left speaker in this position is overbearing.


regular toe-in

While not that great, the balance is much better in the right listening position.

Sidewall boundary effect from the speakers coupled with the side-walls is quite prominent in the bass region as you get closer the far left and right areas.

I'm not sure why the right speaker in far left listening position have two large dips in the mids/upper-mids.


Okay, now below are the FR for the same three listening positions BUT with extreme toe-in:

More linear balance overall, except for the mild horn flare up.

In the left listening area, the volume difference (at least) between left and right speaker is better with extreme toe-in.

Linear balance looks somewhat worse in the right listening position with extreme toe-in.

Not only is the sidewall boundary-effect is just as prominent, at 45 degrees off axis the flare up is even worse.

As expected, the opposing speakers from the far left and right LP on axis are much more linear now with this extreme toe-in (relative to the MLP).


Maybe the large ottoman/leg rest is causing a lot of the problems in my measurements?

Nope. The leg rest in the middle of the couch doesn't affect the FR by much.


Some conclusions coming up after.

 

[ernestcarl]

I tried to find some directivity performance measurements for the Sceptres and could only find some badly scanned ones from a magazine:

Note the two large dips in the mid-region as you increase the angle.

So I cleaned up those scanned images a bit.

I don't find the vertical directivity graph useful so didn't bother with it.

*The S8 horn (significantly cheaper) looks very similar to the RM28ac horn (same designer Dave Gunness):

RM28ac measurements from manual (compare with S8 horn, 2.4kHz and up):

sourced from RM28ac manual

*above 2.4kHz horizontal off-axis response looks uncannily similar to the S8 horn response I measured in my own room.


SHORT CONCLUSION (after some subjective listening):

Let's be frank, the S8's aren't really designed for this type of setup so I wasn't really expecting much. Moreover, the room is also not conditioned to get the best results from such extreme angling of the speakers.

In the MLP, having the speakers on-axis sounds better. The soundstage, sense of depth, and vocals is more concrete, solid, and realistic. Even though the FR graphs look not much worse when the speakers are 15 degrees off-axis, the sound becomes more indirect and diffuse -- cloudlike instead of a solid object in space. This is not meant to say that the sound is bad with extreme toe-in. No it doesn't. How to describe it... Simply, it just is not as good as having it aligned centrally on axis. I can find myself sitting all day, lost in the experience listening to music with this normal setup far more often than when I have the speakers toed-in too much. BTW, the opposite is true when I'm in the far left and right positions.

In the far left and right listening positions when the speakers are toed-in the regular style, the sound is also more concrete. However, the speakers closer to you are also more distracting and loud. Constantly calling attention to themselves. This is okay, when you don't have a picture or video screen in front of you to focus on. But when one watches movies, one does not want to focus on the speaker closest to you -- rather focus on the screen! Here, the extreme toe-in helps move the vocals and sound about a foot or two closer to the screen. Still not perfect, but it's good enough. When you only have yourself in the room sitting at the far left or right listening position, you can additionally increase the delay and decrease the volume of the speaker closest to you. I already have presets for this and it improves the balance even more. Unfortunately, I don't like to do this often because I find myself forgetting to change the DSP settings back to normal. Overall, the sound is far more laid-back and may be best when entertaining guests. You can talk more easily without the music/audio in the background intruding too much. Also, this setup works best when reading a book on the couch etc.

So what now? Do I fix my speakers at extreme angles or the normal way? Hmmmn...

Actually, luckily, I can switch between both setups in a matter of seconds:

As it turns out, these stands are adjustable. And by default, I will have them angled normally. However, when watching movies with three or more people, I can just as quickly change the angle with no fuss. I don't want to add too much EQ and DSP for the most part -- as I would just frequently find myself forgetting about putting the settings back to normal.

So, I just use the setup that works best at any given moment. No need to fix the angle in one position. Same with my desk setup (height).

 

[digicidal]

I realize this is a garbage post with nothing to contribute regarding the sound (though it was an interesting read), but I have to ask:

You crawl over the couch to get into and out of the room?

Other than that, it does look cozy and being completely sealed off it makes me jealous of that aspect considering all of my listening spaces are open to everything and very hard to pressurize to any decent level. Measurements definitely look better than what I see in my theater. Tape for quick adjustments is a great low-tech solution to that problem as well!

 

[garbulky]

I don't think your room is "overdamped." That acoustic foam isn't quite as effective as they are advertised by itself. So though it makes a difference it isn't as massive as it may seem just by looking at the spread. I still haven't heard an "overdamped" room.

 

[ernestcarl]

I realize this is a garbage post with nothing to contribute regarding the sound (though it was an interesting read), but I have to ask:

You crawl over the couch to get into and out of the room?

Other than that, it does look cozy and being completely sealed off it makes me jealous of that aspect considering all of my listening spaces are open to everything and very hard to pressurize to any decent level. Measurements definitely look better than what I see in my theater. Tape for quick adjustments is a great low-tech solution to that problem as well!

Thanks!

Nope. That door actually is a second (redundant) door for the utility room. It's closed permanently. The black and red black-out curtains open to a hallway.

I've tried subs in my main living room a number of times, but could just not get the bass right. I would have to add a lot of heavy, thick bass "traps" and that simply would not do.

 

[ernestcarl]

I don't think your room is "overdamped." That acoustic foam isn't quite as effective as they are advertised by itself. So though it makes a difference it isn't as massive as it may seem just by looking at the spread. I still haven't heard an "overdamped" room.

Sorry, I should have been more clear: in the mids and high frequencies. With extreme toe-in, you are supposed to want more of the reflections -- that's what I read, anyway.

 

[digicidal]

Thanks!

Nope. That door actually is a second (redundant) door for the utility room. It's closed permanently. The black and red black-out curtains open to a hallway.

I've tried subs in my main living room a number of times, but could just not get the bass right. I would have to add a lot of heavy, thick bass "traps" and that simply would not do.

That exactly is the hell I live in. Now that I'm aware of the other access... I'm completely jealous of your space.

 

[ernestcarl]

That exactly is the hell I live in. Now that I'm aware of the other access... I'm completely jealous of your space.

It is quite cozy.

I would have opted for a bigger room/separate office desk space but the other rooms in the house are "squarish" and acoustically even worse. I think this was a good compromise. In the meantime, I suggest headphones? I know I used to use headphones more often back then. Now, not as much.

 

[Severian]

Interesting info. I've long been curious about these speakers due to their high output capabilities and the measurements look pretty good.

 

[ernestcarl]

Interesting info. I've long been curious about these speakers due to their high output capabilities and the measurements look pretty good.

My measurements here actually look a lot worse because of the room and my placement. These speakers have a unique sound -- more akin to an excellent PA monitor than 'audiophile' speakers. They sound 'big' and the stereo imaging is superb -- better than the KH120s. They aren't as clean and neutral as the Neumanns. Quite forward and in-your-face with vocals -- but not harsh -- to me at least. Excellent especially for getting that big cinema/home theatre sound. The horn's directivity is not as good as I would want it to be though, especially past 30 degrees off-axis. But it's the cheapest decent coaxial I could afford at the time. I would like to do an AB listening comparison test with the Genelec 'The Ones'... though I doubt those sound PA-like.

 

[ernestcarl]

Just a small update...

I EQ'd these speakers well before I knew about the moving microphone method (MMM). Finally I've been able to apply it to these monitors now. Setup is more or less the same, but I made a small adjustment to the speaker crossover: Sceptre S8 (xo 40Hz BW 48dB/oct) + F12 Sub (xo 60Hz LR 48dB/oct) -- I was remembering incorrectly, I meant from 100Hz or 120Hz when the sub was placed in a different (more optimal) place.

Applied the ff. high shelving -1.5dB @900Hz beforehand -- although, to be honest, I can't remember if I also engaged the speakers' internal -1.5dB HF shelving @2kHz before taking these measurements. My bad.

a) CENTER SEAT AREA: pre-correction

b) CENTER SEAT AREA: post-correction

c) DSP correction OVERLAY LEFT

d) DSP correction OVERLAY RIGHT

e) ALL SEATS with MLP DSP (correction applied at central axial point of L&R speakers)

f) I saved two DSP presets which allows me sit on either end of the couch with improved FR as well as center imaging by using some delay for a more balanced interaural time difference (ITD):

Even with a few additional 'modest' PEQs, the far left seating area still exhibits (2) significant peaks and a giant dip between 90Hz to ~280Hz -- may have largely to do with the thick concrete behind the dry wall.

In contrast, the far right seating area easily achieved a more linear response -- very likely to do with the single thin sheet of drywall (no concrete) as well as open hallway directly behind the back and on the right side of the speaker.

 

[ernestcarl]

I don't think your room is "overdamped." That acoustic foam isn't quite as effective as they are advertised by itself. So though it makes a difference it isn't as massive as it may seem just by looking at the spread. I still haven't heard an "overdamped" room.

They were "left overs" from a previous house. I didn't want to throw them away in the dump so I just brought them over here, but will remove/replace these when I find the time to repaint the whole house.

*haha! I replied twice. duh

 

[ernestcarl]

g) miniDSP output stage 'fine PEQ' correction curve (8 additional filters)

h) EQ generated in REW by summing Center L&R avg + ALL seats avg MMM RTA together (relatively broad/low Q & mild peak adjustments)

i) Final Results: MLP vs Entire Couch MMM RTA

j) before fine correction

I think I'm done for a while with all the 'tweaks'.

Sceptre S8 @ ~2 meters (8 fine PEQs 1-2dB both full-range correction) vs Neumann KH120 @ ~1 meter (3-4 fine PEQs 1-2dB either left or right narrow Q targeted correction [barely audible, if at all, probably] at sitting MLP)

The Neumann looks much more uneven from 50Hz - 300Hz (magenta colored curve VS yellow curve) —but that is deliberate as I felt it sounded thinner if I evened it out any further. And filling the null between 60Hz - 80Hz caused other audible issues with the Sub at higher volumes. In the end, with listening tests, the Sceptre S8 still looks 'bumpier' overall and Neumann still sounds smoother & cleaner — but not by much.

Nah, that's okay with me... It's all good. Still happy with both systems either way.

 

[ernestcarl]

Hmmn... Upon closer examination, the last graph looks rather inaccurate to me now, and appears to be more 'flat' due to the summing/averaging I did -- I also think I may have held the mic towards the ceiling -- again, god, I really should take down more rigorous notes!

The individual PEQs I used for each speaker could definitely be made more flat/precise, however, I needed/wanted changes that minimally impacted the frequency response of both the far right and left seating areas across the entire couch. It's really very hard to achieve that balance -- apart from work in doing all the numerous (re)measurements each time I did a change -- all seats and not just the center seat area had to be taken into account!


The ff. graphs should be more accurate:


1) 'Forward' EQ 'house curve' presentation (minimal tilt manipulation and closest sound to using the speakers with almost no EQ). This is what I prefer (or the third one, but it's a toss up) for critical listening and watching Blu-ray discs with the least amount of audio compression and most dynamic range. With video streaming services e.g. Amazon prime/Netflix/Youtube, and its accompanying lossy compression, the next EQ mode (second) sounds better overall to me.

2) 'Relaxed' presentation (most used EQ curve setting for general music listening e.g. Spotify and streaming video watching e.g. Netflix, as well as closest in similarity to the tilt of the KH120s I have setup right on my desk)

Let's call this 'extra' relaxed EQ, that is, IF we chose to add a bit more shelving

3) 'Mild Forward' (rather, a little less aggressive/milder bass)

*additional side-by-side comparison between the two curve presets

And here's how closely the KH120 and S8 slopes now align:

'relaxed' EQ mode preset slope

At least there's consistency in both listening spaces -- and the sound isn't as drastically different as before.


mild 'forward' EQ mode preset with internal -1.5dB HF shelf

mild 'forward' EQ mode preset IF internal HF shelf set as FLAT

AND sometimes blockbuster movies just seem to need more boost in the sub bass ~9dB above the rest of the FR:

*The more presets to change the sound as desired, the better... in fact, I would've loved to have even more buttons in the remote, but, Alas!

There's simply not enough.

With some PC-based software solutions (e.g. JRiver), though, you could save as many presets as desired:

 

[ernestcarl]

Decided it best to perform several measurements at a more standard 90dB SPL at 1 meter setting. This should be a more accurate representation of the S8's %THD distortion at a normal listening level. *taken without the 100Hz +9dB peaker filter for the left speaker

(old post)
Distortion measurement update:

'forward EQ'

My +9dB Peak Filter at 100Hz for the left speaker adds quite a considerable amount to distortion when pushing the monitor at max volume. Since I haven't noticed it degrading the sound even with bass heavy tracks, I've kept the peak filter turned on for a while. I've also tested it turned off, and listened to some test tones and bass heavy music/movies. The drop in volume doesn't seem to significantly affect overall FR balance, so I've decided remove it.

Measured in MLP at 2 meter distance (100Hz +9dB Peak Filter)

Yep, a significant, sharp ~31% distortion peak around that 100Hz area. Somewhat audible, sure, but I've never really noticed it to be unpleasant.

*Issue has been resolved. Sharp spike in distortion corresponds with narrow null. The actual peak distortion in the S8's bass (below 100Hz) shown above at my predifined maximum volume is ~12%

*In the following graph +9dB peak filter at 100Hz is disabled. The peak distortion here now is ~7% instead of ~12% under 100Hz

There's also a +5% distortion peak around 900Hz which you don't see in the right speaker -- strange. Thinking way back, I actually purchased these monitors on discount as they were a store demo pair; there's also some minor cosmetic damage, so who really knows what kind of abuse they were exposed to and if that's somehow related to the differing results seen next...

Measured in MLP at a 2 meter distance

Even weirder still, there's a 10.8% narrow distortion spike at 350Hz in the right speaker, but again, only at my max volume boost. Again, in normal use, I've not noticed these causing any audible listening problems/fatigue or whatever else degradation in listening pleasure.

*Actually, it's an artefact caused by a deep null at that FR in the MLP, duh! See below :

Now, I've also got the LSR305s monitors, headphone holders, a bunch of headphones, and one super compact Blu-ray player (at left monitor) on top of the S8s -- quite a stack -- but I'm too lazy to remove them to see if that alleviates some of the weird differences in distortion we've been seeing here.

Presonus advertises the speaker's max SPL peak at 1m to be 116dB but gives no detailed measurement graphs when it comes to maximum SPL at the ff. distortion levels e.g. 0.5%, 1%, 3% and 5% etc.

I could try to put my mic at ~1m and really push the amp & volume up to 116dB -- as I've barely increased the amp adjustment knob above zero...

Eh, No! Sorry, but I have no interest in measuring or pushing or testing my monitors at their 'advertised' limit. I'll leave that (potentially damaging?) task to someone else.


*Log sweep (post-removal 100Hz +9dB pk filter in left speaker) vs MMM RTA view (pre-removal 100Hz +9dB pk filter in left speaker)

The MMM RTA certainly looks better (smoother and more linear), and seems more aligned with what I'm actually perceptually hearing across the main listening area.

 

[QMuse]

Distortion measurement update:

'forward EQ'
View attachment 69372

My +9dB Peak Filter at 100Hz for the left speaker adds quite a considerable amount to distortion when pushing the monitor at max volume. Since I haven't noticed it degrading the sound even with bass heavy tracks, I've kept the peak filter turned on for a while. I've also tested it turned off, and listened to some test tones and bass heavy music/movies. The drop in volume doesn't seem to significantly affect overall FR balance, so I've decided remove it.

When you use filter with +9dB gain you also have to apply attenaution of input signal by 9dB otherwise you will get clipping. If you haven't done that the distortion you're measuring is coming from clipping and is not being produced by speakers.

 

[ernestcarl]

When you use filter with +9dB gain you also have to apply attenaution of input signal by 9dB otherwise you will get clipping. If you haven't done that the distortion you're measuring is coming from clipping and is not being produced by speakers.

That is indeed quite possible, though I have checked the levels beforehand and made sure there was some headroom left. Hmmmn... I'll just double check again to make sure. Thanks!

 

[QMuse]

That is indeed quite possible, though I have checked the levels beforehand and made sure there was some headroom left. Hmmmn... I'll just double check again to make sure. Thanks!

Every software (either standalone convolution engine or the one built into your player, like Jriver or Roon) that is processing EQ filters has attenuation parameter and you have to set it for the maximum gain of the filter you're using otherwise clipping of the signal will occur. If you haven't set attenaution that is surely what is happening, especially with so high gain at the LF range.

 

[ernestcarl]

Every software (either standalone convolution engine or the one built into your player, like Jriver or Roon) that is processing EQ filters has attenuation parameter and you have to set it for the maximum gain of the filter you're using otherwise clipping of the signal will occur. If you haven't set attenaution that is surely what is happening, especially with so high gain at the LF range.

So I double checked my settings in JRiver:

There are no PEQs applied for my S8 listening couch setup in that program. The -5.2 dB volume attenuation above is simply to match the extra headroom I've added to my KH120s as you'll see HERE, which pretty much exactly mirrors the LUFS EBU R128 analyzer results.

In the miniDSP that I'm using, I've added the ff. volume attenuation:

And turned to increase the volume adjustment knob in the speaker's amplifier itself and reapplied my +9dB peak filter.

Unfortunately (and maybe fortunately!) the results came out the same as before.

The actual culprit of the aformentioned ominous distortion figure may become much clearer below:

There was a null in that specific area where I placed my pointer at...

This may be why it wasn't really all that much noticeable as an audible issue in the first place.

The distortion peak around +900Hz is still there, though not much; and only appearing in the left speaker.

Yeah, I think we've "solved" that particular imagined problem -- and it's probably fine to add the peak filter back. @QMuse, what do you think?

 

[QMuse]

So I double checked my settings in JRiver:

View attachment 69403

There are no PEQs applied for my S8 listening couch setup in that program. The -5.2 dB volume attenuation above is simply to match the extra headroom I've added to my KH120s as you'll see HERE, which pretty much exactly mirrors the LUFS EBU R128 analyzer results.

In the miniDSP that I'm using, I've added the ff. volume attenuation:

View attachment 69405

And turned to increase the volume adjustment knob in the speaker's amplifier itself and reapplied my +9dB peak filter.

Unfortunately (and maybe fortunately!) the results came out the same as before.

The actual culprit of the aformentioned ominous distortion figure may become much clearer below:

View attachment 69406
There was a null in that specific area where I placed my pointer at...

This may be why it wasn't really all that much noticeable as an audible issue in the first place.

The distortion peak around +900Hz is still there, though not much; and only appearing in the left speaker.

Yeah, I think we've "solved" that particular imagined problem -- and it's probably fine to add the peak filter back. @QMuse, what do you think?

Filter you showed in your previous post has gain of +9dB, not -5.2dB.