what I measure VS what I hear

[aakkaaii]

Dear all, these are my first measurements taken today in my music listening room.
I'm fairly surprised how well they (at least) look in terms of flatness of the curve. I thought my starting point was worst than this honestly.
I know that I need to treat in some way, to improve RT60... but need to consider this is not a dedicated room.

Apart from measuring, while listening, I know my room suffers one main defect that gets evident in rock/prog music: when big dist guitars or drums cymbal kicks in, the sound gets really messy and unpleasant.
Can anyone 'see' this awful behavior in my measurements and advice how to try to correct?
In general, how would you approach correcting this room?

many many thanks in advance
Angelo

 

[Sonnie Parker]

Welcome to AV NIRVANA ... glad to have you with us.

You response doesn't look that bad overall. It looks a bit flat and likely sounds about the same. Most will prefer a tilted response with 5-15dB more bass from 80-100Hz on down.

There is some question as to what you measure at the listening position being what you actually hear, since the measurements will include reflections. If you have access to the anechoic measurements of your speakers, you can get a starting point as to what you should be hearing. Otherwise, the best you can do is fix the lower end, since it will measure fairly true to what you are hearing. You can fix most of the low-end with EQ fairly easy. Not sure what your HK AVR has EQ wise, but a miniDSP with PEQ filters is very reasonable.

 

[aakkaaii]

Thanks Sonnie for your kind reply.
as of the frequency response, one think that it's still not clear to me is if I have to jump directly to electronically treat my signals before treating my room: is it that my speakers aren't able to produce those frequencies?
I agree that the anechoic response would help, but I have a collection of quite old speakers (from the 80s') and have never come across those information.
As you see the L speaker is giving the worst contribution to the bass section... does it mean that it all has to do with placement and not with anechoic response?

as of reflections, I recently moved to this new house and remember very well my old listening room. I recently realized I was really lucky because my previous *untreated* room was really impressive. I think it was because of the kind of cealing (irregular stones, barrel shaped) and furniture (mostly half full cd/book shelf). The new room instead does that messy/unpleasant effect on mid/high frequencies I was talking about... and this is my first concern today.
Are those reds in the very low end of the spectrogram room modes?


This is the room simulator run for my current room. As you see, I'm forced to not simmetrical speaker placement.

 

[Sonnie Parker]

You could certainly try some panels to deal with any serious reflections you have ... some use absorption and some use diffusion. As for the lower bass area, that is harder to deal with using panels and electronic EQ may work better in most cases.

Placement indeed will affect the response, which is something you could experiment with, but in most cases you want to start with the placement that gives you the best soundstage and imaging, then try to deal with the response from there. Having the anechoic response off-axis can help you determine what the room is doing to your response.

 

[aakkaaii]

Your suggestion really makes sense
I'll keep the community updated, the journey just begun

 

[bobolix]

Your listening room seems to be quite "live". If the volume of the room is 100 cubic meters, then the optimum RT60 is 0.25 s, at 50 m³ it is 0.2 s approx. The electrical correction doesn't solve the problem, as it is valid for only one point in the room. Further, if the loudspeaker response is flat under anechoic condition (a reference), then in the a semireverberant room certain lift in the bass region usually appears because of the underdamping there, which is normal and well accepted generally.

 

[Tomeh]

Hi,

Looking at your RT60 graph, you might have what Mr. Linkwitz would have called a "lively room" = highly reflective = more complex music passages might be lost in a high % of reflected music??

See his info at linkwitzlab.com/rooms.htm

sections D5 & D6

As Sonnie has said "you could certainly try some panels" which would also lower your RT60 average.

As always, the bass region will be mre of a challenge, but it will be very helpful on the clarity, % of direct sound vs % of reflective.

Using the ETC graph of your rom can help with placement since a before and after measurement will allow you to see if you have dampened the primary reflections.

see info at gikacoustics.com/unpacking-etc-time-domain-measurements-early-reflections/

Lowered down there is asection called "smooth spikes" and they explain how the delayed reflection spikes can help you identify where the reflective spike comes from and if you hit the mark with placement of absortion panels, they "after" measurement will lower these spikes (as measured from your listening position, the one that matters most).


Cheers,

Tom eh

 

[BenToronto]

The ear is not a mic. What is needed is a plot which takes into account the factors used by your brain to sense tone colour, of which instantaneous FR is only one.

 

[bobolix]

The ear is not a mic. What is needed is a plot which takes into account the factors used by your brain to sense tone colour, of which instantaneous FR is only one.

This is too cheap objection. Certainly the ear is not a mic, but some connections can be revealed by ear as well as by mic. The excessive reverberation buildup at low frequencis is disturbing and in fact it can even destroy the localisation, what can be used (and misused) at subwoofer installation. The physiological effects at mids and highs can be respected simply by using reasonably short time window for the FR measurement, what, conversely, unfortunately doesn't work at low freqs, where it is better to believe the mic unwindowed.

 

[BenToronto]

Although I am having some trouble with your "word salad", you seem to have some sense of the issue.

My point is that some assembly of different physical measurements might lead to a better depiction of human tone.

A related point is the way Toole used "standard" listening rooms both in Ottawa and S. California*. He also seems to have sort of standardized on certain recordings - esp female singers; granted these were chosen for a variety of good reasons, not as a standard music stimulus.

You can't just play sine tones in any sort of room and say, "that's the FR" and expect the sound human experience in other rooms to be the same even if the FR is the same.

*Toole's main testing at Harman was in a large industrial-space room... odd.

 

[ELOJR]

Hi Angelo.

Your measurement results do look good and the suggestions regarding reflections may be right on target. One question I have...before you do all sorts of room treatment...are you sure that the audio "messiness" is due to acoustic issues rather than equipment ones?

If the sound problem is related to your equipment, I guess you could determine that by switching out the amp, something not doable unless you have another AVR handy.

In lieu of that...do you have enough power to drive your speakers to as loud a level as you want without distortion?

I'm not saying that your issue is not room-related. Just floating the equipment thought out there for consideration/elimination.

Good luck & keep us posted.

 

[PEB]

Headphones are also a good tool in these situations. There are no room acoustics with headphones, and whether or not you are pleased with the tonal balance of your particular headphones, you can still identify if the harshness you hear is in the recordings, and not your speakers, and not your room.

 

[dhai]

I'd suggest the upper midrange is too hot. While the SPL plot is absurdly scaled, it seems the 600 - 1k3 Hz range is about 5 dB louder than the flanking bands. This usually sounds like a thin midrange and nasty presence range.
RT60 and Spectrogram with same tendency hint at the room, but +5 dB due to room characteristics would be extreme.

Is the SPL measurement done with one or all speakers active?

 

[PEB]

I'd suggest the upper midrange is too hot. While the SPL plot is absurdly scaled, it seems the 600 - 1k3 Hz range is about 5 dB louder than the flanking bands. This usually sounds like a thin midrange and nasty presence range.
RT60 and Spectrogram with same tendency hint at the room, but +5 dB due to room characteristics would be extreme.

Is the SPL measurement done with one or all speakers active?

Yes, please change your scale from 10dB to 5dB.

Alternatively, I'd suggest a low-Q filter (~0.5?) to raise the 200-600 range.

 

[aakkaaii]

Hi Angelo.

Your measurement results do look good and the suggestions regarding reflections may be right on target. One question I have...before you do all sorts of room treatment...are you sure that the audio "messiness" is due to acoustic issues rather than equipment ones?

If the sound problem is related to your equipment, I guess you could determine that by switching out the amp, something not doable unless you have another AVR handy.

In lieu of that...do you have enough power to drive your speakers to as loud a level as you want without distortion?

I'm not saying that your issue is not room-related. Just floating the equipment thought out there for consideration/elimination.

Good luck & keep us posted.

Hi Elojr,
you have to know that the Harman/Kardon AVR in the signature is currently being used with the projector in a room -for which I do not have plans about measuring and treating for the moment- that is not the one in subject.

THE room we're talking about is my 'music listening room' -not dedicated unfortunately- that hosts a big desk (smartworking!), a 3 seats sofa, a music keyboard and a full wall cabinet, the opposite wall full of paper posters, parquet floor, no carpets for the moment.

I just moved to this house and the equipment is more or less the same I used in the old house, where the listening experience was -by luck- really extraordinary.
During the recent years I collected a few german made hifi components and speakers from the late 70' to mid '80 that to my hears have a very pleasant sound with a perfect sound stage... at a very reasonable price, but at the cost of having to repair them more frequently than I really wanted to.

I may try to switch the amp, and see if I have any improvement as I have a spare (sometimes easier than repair is replace).

Enough power? I can try a different amp too.

Thanks a lot for the clues
A

 

[aakkaaii]

Headphones are also a good tool in these situations. There are no room acoustics with headphones, and whether or not you are pleased with the tonal balance of your particular headphones, you can still identify if the harshness you hear is in the recordings, and not your speakers, and not your room.

I'm pretty sure the headphone has a different amp than the one serving the speakers...

 

[aakkaaii]

I'd suggest the upper midrange is too hot. While the SPL plot is absurdly scaled, it seems the 600 - 1k3 Hz range is about 5 dB louder than the flanking bands. This usually sounds like a thin midrange and nasty presence range.
RT60 and Spectrogram with same tendency hint at the room, but +5 dB due to room characteristics would be extreme.

Is the SPL measurement done with one or all speakers active?

Can't remember but I believe two,
as the picture in the following post has the L and R speakers plotted separately.

One of my doubts: I understand one speaker measurements help you understand better what you need to treat, but then you switch both speaker on and need to treat the treatment...
So, why not to treat the sum of the L+R contributions?

 

[aakkaaii]

I hope a plot scaled like this is more intelligible.
A different measurement though, as I'm experimenting/moving speakers and some furniture.

A

 

[aakkaaii]

and, by the way, thanks to all for the replies and the helpful discussion.

 

[aakkaaii]

just to let you know: if I listen to jazz music or pop, easy-listening music, the results in my room are quite satisfactory
but when it comes to rock/prog (my favorite unfortunately) and big dist/overdriven guitars or drums come in, the magic disappears, and the messiness rules

to figure out what I mean, try to listen A perfect circle, The package (at about 4:00) from the album Thirteenth steps