Skip to main content ↑ ↓

Notice

Please note that most of the software linked on this forum is likely to be safe to use. If you are unsure, feel free to ask in the relevant topics, or send a private message to an administrator or moderator. To help curb the problems of false positives, or in the event that you do find actual malware, you can contribute through the article linked here.
Topic: Effect of room reflections (Read 4445 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

Effect of room reflections

Being an occasional reader, and noticing that many folk here do seem to understand sound, would like to ask about an aspect of room reflections.  In a room with solid brick walls overlaid with plaster, unpapered, and a solid wood floor on concrete, audio playback (from 16/44k PCM) of orchestral violins sound quite 'scratchy'.   The effect is not exclusive to violins; while most 'pop' material does sound ok, if a little 'bright', the harmonica on one well-known song sounds piercingly loud.  (I've done a number of trials with different amplifiers and sources and the problem 'stays with' the speakers and the room.  I haven't moved the speakers to another room to try so, at the moment, the cause seems to be speakers or room or the mix of both.)  May I, please, say I'm not posting here expecting folks to solve my problem for me - I'm going to do that - but instead I would like to ask, because the sound seems consistent with a non-linear response somewhere in the system (perhaps around 4KHz or so), whether room reflections could be having that sort of effect?  I don't know much about reflections and waves, and I'd like to understand better so that I can make some sensible decisions about solving the heard quality issue.

On the room details, there is a thick rug on the floor, a couple of bulky 'sofas' in soft fabric, and on the wall opposite the speakers a lot of books (none, though, on audio engineering, sadly).  I would have thought some of the higher frequency signals (perhaps more the harmonics) will be absorbed or diffracted though, because a lot of the signal paths to hard surfaces ( and which are most of the primary ear-level paths, in fact) would remain unimpeded, overall not much attenuation due to the fabrics might occur.

Look forward to hearing any thoughts on reflections, especially about possible effects above 1KHz or so.  I don't even know whether lower or higher frequencies reflect more, I really do know very little.  Hoping that folks here know rather more.

Seafarer

Re: Effect of room reflections

Reply #1 –
Quote from: Seafarer on
I've done a number of trials with different amplifiers and sources and the problem 'stays with' the speakers and the room. 
That's symptomatic of audiophilia. Not good.

Quote from: Seafarer on
On the room details, there is a thick rug on the floor, a couple of bulky 'sofas' in soft fabric, and on the wall opposite the speakers a lot of books (none, though, on audio engineering, sadly).
Pictures of setup would be helpful. Well designed speakers in furnished domestic spaces, don't exhibit such audible problems for normal folks. So we have a lot ifs. And many existing threads discussing contentious issues like room reflections and common belief.

Loudspeaker manufacturer

Re: Effect of room reflections

Reply #2 –
In my experience room reflections, or room boundary interactions, primarily affect the lower frequencies where the combination of room dimensions (width, length, height) can create standing waves which add emphasis to particular bass notes.  This is most common when two or more of the room dimensions are the same or somewhat close in value.

If you didn't tell me about the thick rug, bulky sofas, and wall of books opposite the speakers I'd have suggested that your room is a bit "lively" but you've got this covered.  Also, frequencies above 1kHz travel more or less directly to your ears depending on the baffle design of your speakers; you shouldn't get much in the way of reflections.  You can try temporarily hanging some blankets on the side walls or moving your speakers further inward if possible given your space constraints.  You can also try toe-ing in your speakers a bit so that when you're in your favored listening position they "point" directly to your head or slightly in front of you.  This should cut down on side-directed energy but may take away some apparent soundstage width.

Try these things and if still you're unhappy with the sound you may need to switch to a speaker with a less pronounced upper midrange.  What speakers do you currently have and how are they mounted (floor, stands, wall, bookshelf)?  And, yes, a photo or two would be very helpful.

Re: Effect of room reflections

Reply #3 –
Quote
(I've done a number of trials with different amplifiers and sources and the problem 'stays with' the speakers and the room.  I haven't moved the speakers to another room to try so, at the moment, the cause seems to be speakers or room or the mix of both.)
It's probably worthwhile to try the speakers in a different room (or even outdoors).   

Does it get better or worse if you get very close to the speaker (for more direct speaker-sound and less room sound)?

You can also try turning the balance one way or the other (or disconnecting one speaker at a time) to see if one of your speakers is bad or damaged.   

Quote
(perhaps around 4KHz or so), whether room reflections could be having that sort of effect?
It could be.   Have you tried experimenting with an equalizer?   You can try pulling-down one EQ slider at a time (assuming a graphic equalizer) to see if you can find & fix the problem frequencies.   Almost every software player has an equalizer if you don't have a hardware EQ.

Note that equalization can work well to reduce the signal where reflected waves are added (standing wave anti-nodes) but it doesn't work where reflected waves are subtracted (nodes) because it can take an almost infinite amount of energy to overcome the cancelation.  And of course, EQ won't remove unwanted reverb.

Does it change as you move around a few feet?    If it's really around 4kHz, the effect of high-frequency room reflections will change a lot with a small location change.  You'll notice that if you try a 4kHz test tone.*

The good news is that high frequency treatment is easier (and probably cheaper) than bass traps.

The first thing to do before trying to treat the room is to measure/diagnose the room (together with the speakers).   RealTraps.com has some good information about how to measure your room, as well as links to some free software.   (You will need a calibrated microphone.)





* Audacity can generate test tones as well as white and pink noise.    (Be careful with high-power high-frequency test tones because you can fry your tweeters.)

Re: Effect of room reflections

Reply #4 –
Quote from: DVDdoug on

Quote
(perhaps around 4KHz or so), whether room reflections could be having that sort of effect?
It could be.  Have you tried experimenting with an equalizer?  You can try pulling-down one EQ slider at a time (assuming a graphic equalizer) to see if you can find & fix the problem frequencies.
[..] try a 4kHz test tone.*

*Audacity can generate test tones as well as white and pink noise.

I hadn't seen that.  I use Audacity to transfer from LPs & singles, etc and, as part of trying to get to the bottom of this problem, have been scouring the 'net for test files.  I'll just generate them, instead.  Much obliged, Doug.

The amp is a Yamaha AVR and it has an EQ.  I've tried it.  The distortion goes away with settings such as:
1k ... -1 click
2.5k .. -3 clicks
6.3k .. -4 clicks
16k .. -1 click (I wanted - as much as possible - to keep the harmonics so that instruments would remain  differentiated and recognisable)
but the sound does result in sounding more 'muffled' and less 'here'.  And on non-classical sources, it really sounds quite 'dull' with these settings.

Quote
Well designed speakers in furnished domestic spaces, don't exhibit such audible problems for normal folks.

The system is brand new.  I'd tested with the alternate amplifier - normally used in another room - to see if there was a problem with the Yamaha; seems not. 

I agree with you; the speakers sounded really good at the retailer - and we'd taken the exact same piece (same CD!) with us for listening.  I imagine the speakers are well-designed - as much as is practicable for a volume manufacturer.  And because they are volume manufactured I think a fault is unlikely - possible, of course, but rare, really.  Hoping so, anyway.

Quote
What speakers do you currently have and how are they mounted (floor, stands, wall, bookshelf)?  And, yes, a photo or two would be very helpful.
AVR - Yamaha RX-S600
Sources - USB DAC (2 x RCA); CD (2 x RCA); Network player SP/DIF; PC HDMI
Speakers - Tannoy Revolution DC-6SE (floor standing)

I'll get some pictures of the room tomorrow.  No curtains, yet; but the material is draped over the sofas, for now.

Very grateful for the comments, and some useful things to try.  It's been tricky to test with a dynamic piece of music, that's why I was searching for test files.  But I'll use tones, and check placements, listening position, and such.

Seafarer

Re: Effect of room reflections

Reply #5 –
Quote from: Seafarer on
Speakers - Tannoy Revolution DC-6SE (floor standing)

I agree with you; the speakers sounded really good at the retailer - and we'd taken the exact same piece (same CD!) with us for listening.  I imagine the speakers are well-designed - as much as is practicable for a volume manufacturer.  And because they are volume manufactured I think a fault is unlikely - possible, of course, but rare, really.  Hoping so, anyway.
Well, Tannoy is a good brand and has made many good speakers over the years, but that doesn't eliminate the possibility of a dud. I couldn't find any measurements of yours, but the (non SE) DC6T measurement is shown below:

That type of response could indeed sound a bit sibilant with voices and "edgy" with some instruments, due to the slight excess of sound power in the 5-9k region, under some conditions.
Do yours emphasize the Sss and Shh sounds with vocals?

Quote from: Seafarer on
I'll get some pictures of the room tomorrow.  No curtains, yet; but the material is draped over the sofas, for now.

It's been tricky to test with a dynamic piece of music, that's why I was searching for test files.  But I'll use tones, and check placements, listening position, and such.
Pics will be good, but better would be measurements of the speakers themselves, both near and far field. For $25-$90 you can measure using any smartphone or pc.

cheers,

AJ

Loudspeaker manufacturer

Re: Effect of room reflections

Reply #6 –
If your walls are bare, then the bright sound you're getting could be due to untamed reflections of high frequencies from the side walls. An easy experiment to do would be to (temporarily) hang large absorbant materials (eg. quilts, duvets, curtains) on the side walls at a position approximately equidistant between speakers and listening position. If that reduces the problem, then you can think about what might be a domestically acceptable permanent fixture that achieves the same effect.

Re: Effect of room reflections

Reply #7 –
Quote from: Seafarer on
Being an occasional reader, and noticing that many folk here do seem to understand sound, would like to ask about an aspect of room reflections.  In a room with solid brick walls overlaid with plaster, unpapered, and a solid wood floor on concrete, audio playback (from 16/44k PCM) of orchestral violins sound quite 'scratchy'.  The effect is not exclusive to violins; while most 'pop' material does sound ok, if a little 'bright', the harmonica on one well-known song sounds piercingly loud.  (I've done a number of trials with different amplifiers and sources and the problem 'stays with' the speakers and the room.  I haven't moved the speakers to another room to try so, at the moment, the cause seems to be speakers or room or the mix of both.)  May I, please, say I'm not posting here expecting folks to solve my problem for me - I'm going to do that - but instead I would like to ask, because the sound seems consistent with a non-linear response somewhere in the system (perhaps around 4KHz or so),

As you describe the room and your system, you probably have less nonlinear distortion than most.

Quote
whether room reflections could be having that sort of effect? 

It is most likely that audible linear distortion, particularly time and frequency distortion, is what you are hearing.

Quote
I don't know much about reflections and waves, and I'd like to understand better so that I can make some sensible decisions about solving the heard quality issue.

Two likely effective routes to that kind of understanding. One is reading books like:

Acoustics,  L. Beranek

Sound Reproduction The Acoustics And Psychoacoustics Of Loudspeakers And Rooms, Floyd Toole

The other is to set up a personal room and loudspeaker measurement system (probably PC - based)  using (free) software like:

Room Eq Wizard

Holme Impulse

The best is a combination of the two. They will probably cost you less out-of-pocket than any one of the several major audio components that you have, but can easily improve your mind and life forever.

Re: Effect of room reflections

Reply #8 –
Quote from: Arnold B. Krueger on
Two likely effective routes to [...] understanding. One is reading books like:

Acoustics,  L. Beranek

Sound Reproduction The Acoustics And Psychoacoustics Of Loudspeakers And Rooms, Floyd Toole

The other is to set up a personal room and loudspeaker measurement system (probably PC - based)  using (free) software like:

Room Eq Wizard

Holme Impulse

The best is a combination of the two.



Thanks, indeed, for these;  it is understanding that I need most, at this stage.  Something is not quite right, and I hadn't expected there to be any issue.  I have a couple of microphones but they date from 1/4 inch mono days and I've ordered some adapters.

Quote from: DVDdoug on
a 4kHz test tone.
Audacity can generate test tones as well as white and pink noise.    (Be careful with high-power high-frequency test tones because you can fry your tweeters.)

I checked with a friend who plays orchestral violin and the violin frequency range during (most) pieces is lower than I had realised.  Violins (other than solo) will often play between 196 Hz and, say, 1,320 Hz.  A harmonica may play between 130 Hz and 1,200 Hz.  I hadn't realised that the fundamentals for even relatively high-sounding notes were actually that low.

I want to check for issues - whether the room reflections or some other problem (response, distortion, timing, all those) in drivers, crossover, or even the amplifier - throughout the frequency range.  I'll need test signals way above those fundamental notes.  I have started creating some test tone files in Audacity.  Audacity offers square and sawtooth, as well as sine waves; I'm using sine waves, for now.  For frequencies much above 1,500 Hz, and certainly above - I'm guessing - 4k, I will use a lower signal level.  Audacity doesn't seem to offer me a dB scale, merely a 'level' in the range 0 to 1.  In case (for example) the volume is mis-set on the amplifier, I thought I'd record tones at something akin to -20 dB, because I had understood that these much-higher frequencies are the harmonics and are usually at a much lower amplitude than the fundamentals.  Assuming Audacity's 'level' offering is linear, -18dB (which will do) would be equivalent to a 'level' of 1/64th, or a linear level of 0.015625 .  That is fairly low, but I don't want to damage the tweeters (or my hearing).

Lastly, I found AJ's picture of the speaker response useful.  May I ask what the three colours represent?

Seafarer

Re: Effect of room reflections

Reply #9 –
It is 100% speakers and listening environment. Don't even waste a second fretting over any single piece of electronic componentry (including the crossovers, assuming they're passive).  Whatever led you to the idea that your amp is to blame needs to be expunged.

Also, -18 dB is 1/8 of full scale. -3dB is a halving when working with power which is not what is happening when talking about signal level prior to driving a load.

As much as this will annoy AJ, I would take cliveb's lead and hang something on your side walls. I wouldn't spend too much time with Ethan Wiener's site either. Common "wisdom" based on sighted testing isn't going to lead you down the path to enlightenment; this is not to say that the general concepts are worthless (I'm already suggesting you apply some treatments, if only to steer you away from the idea that your amp is to blame).

Re: Effect of room reflections

Reply #10 –
Quote
I have a couple of microphones but they date from 1/4 inch mono days and I've ordered some adapters.
The problem with a random uncalibrated microphone is that you don't  know its frequency response so it has an unknown effect your measurements.   Measurement microphones are not perfectly flat, but they come with a calibration curve (or file) so you can take the variations out of your measurements....    It's like knowing the speedometer in your car reads 3MPH faster than your true speed.       (Microphones are generally mono and you want one mono mic for measurement.)

Quote
I checked with a friend who plays orchestral violin and the violin frequency range during (most) pieces is lower than I had realised.  Violins (other than solo) will often play between 196 Hz and, say, 1,320 Hz.  A harmonica may play between 130 Hz and 1,200 Hz.  I hadn't realised that the fundamentals for even relatively high-sounding notes were actually that low.
Your friend is forgetting about harmonics & overtones (the things that make a piano playing an 'A' sound different from a piano playing an 'A').    If you filter-out everything above 2000Hz, the violin will sound "dull".

Quote
The amp is a Yamaha AVR and it has an EQ.  I've tried it.  The distortion goes away with settings such as:
1k ... -1 click
2.5k .. -3 clicks
6.3k .. -4 clicks
16k .. -1 click (I wanted - as much as possible - to keep the harmonics so that instruments would remain  differentiated and recognisable)
but the sound does result in sounding more 'muffled' and less 'here'.  And on non-classical sources, it really sounds quite 'dull' with these settings.
Speaking of "dull"...  When you do cure this "harshness" problem, the sound may seem little dull.    You'd get the same impression of you get used to boosted highs and then switch to a flat system...   You might try living with the EQ for a day or so to see if it still sounds dull after your ears & brain adjust to the new sound. 

Quote
Audacity doesn't seem to offer me a dB scale, merely a 'level' in the range 0 to 1.  In case (for example) the volume is mis-set on the amplifier, I thought I'd record tones at something akin to -20 dB,
  

Quote
Audacity offers square and sawtooth, as well as sine waves; I'm using sine waves, for now.  For frequencies much above 1,500 Hz, and certainly above - I'm guessing - 4k, I will use a lower signal level.
Audacity isn't going to be THAT useful...  It might give you a better clue about the problem frequency ranges and if you move-around the changes in amplitude should be very noticeable if you've got refection problems.    It's probably better to mover around a bit to see if the violin harshness goes away (or gets worse in certain locations-positions (and you shouldn't have to move a whole lot).

...I forgot about something.    If you're doing the "moving around" experiment, try one speaker at a time because the distance-different between the speakers will also cause "phase issues".    And, you'll get variations as yo move around in a good room so it may be difficult to tell if you've got an unusual problem.

Quote
Audacity doesn't seem to offer me a dB scale, merely a 'level' in the range 0 to 1.  In case (for example) the volume is mis-set on the amplifier, I thought I'd record tones at something akin to -20 dB,
You can use Audacity's Amplify effect to set the (peak) dB level after generating tones.    But, Audacity's level is virtually USELESS...   The digital dB levels (dBFS = decibels full-scale) are not calibrated to the acoustic dB SPL levels....    Your computer has an unknown voltage output and a volume control...  Your AVR has unknown sensitivity/gain and a volume control...   Your speakers may have unknown sensitivity/efficiency...   Distance from your speakers affects loudness....   Room reflections affect loudness...

Re: Effect of room reflections

Reply #11 –
Quote from: DVDdoug on
Quote
I checked with a friend who plays orchestral violin and the violin frequency range during (most) pieces is lower than I had realised.  Violins (other than solo) will often play between 196 Hz and, say, 1,320 Hz.  A harmonica may play between 130 Hz and 1,200 Hz.  I hadn't realised that the fundamentals for even relatively high-sounding notes were actually that low.
Your friend is forgetting about harmonics & overtones (the things that make a piano playing an 'A' sound different from a piano playing an 'A').    If you filter-out everything above 2000Hz, the violin will sound "dull".
Quote
I hadn't realised that the fundamentals for even relatively high-sounding notes were actually that low.

Re: Effect of room reflections

Reply #12 –
Quote from: Seafarer on
Lastly, I found AJ's picture of the speaker response useful.  May I ask what the three colours represent?
Blue is direct on axis. Green and red are off axis, 15 and 30 degrees, though it's from a non-english site so I can't fully decipher.
It is classic coax behavior. The peak/dip/peak on axis due to symmetric diffraction from the horn/cone interface, surround and frame, the very smooth off axis from the controlled directivity of coincident coaxial format.
If it is representative of your model, the last thing you would want to do is screw up that smooth off axis reflection with non-linear bandaids on the sidewalls, unless fully broadband (thick), which has issues of its own.
A better solution would be a scalpel, EQ of the on axis peaks. I generally don't recommend EQ > 500Hz or so, but in this case I would.
Foobar has EQ capability and a MiniDSP is around $100 this side of the pond. Problem is you have many sources and modern AVRs like yours don't have pre-main couplers or tape loops.
Loudspeaker manufacturer

Re: Effect of room reflections

Reply #13 –
Quote from: greynol on
Also, -18 dB is 1/8 of full scale. -3dB is a halving when working with power which is not what is happening when talking about signal level prior to driving a load.

Tst.  I'd better do my homework:

Power is V^2 / R, and

A dB ratio is Power / Power which, when R is the same in both top and bottom, becomes the ratio of

Va^2 / Vnom^2  [Va being the Audacity 'selected level', Vnom being whatever Audacity's level of '1' is]

Vnom can be assumed to be Audacity's 1, so the power relationship is

-18dB = Va^2

-18dB approx = 1 / (2^6), giving

1 / (2^6) = Va^2

Since the left side and right side are equal, the square roots of each side will be equal, as well, so

1 / (2^3) = Va

This means Va will be 1/ 8.

Was that right, Grey?

I thought I'd post what I think is the explanation because there are a lot of people finding these threads through web searches, and maybe I will not be the only person to make this mistake of forgetting that dB are a power relationship and then end up just assuming volts and power are the same thing.

Testing update.

1.  Moved speakers to a different, smaller, more carpeted room (but fairly hard walls), and noted two effects.  Much less bass.  Some 'peakiness' between (I'm guessing) 400 Hz and ~1,200KHz.  (Same test source, Mozart CD transcoded to flac, budget hifi amplifier.)  Speaker placement was much constrained, though, placed not far from centre of room.  Smaller speakers are usually used in that room (KEF Q300, 1 on a bookshelf, 1 on a stand) and those sound fine with this and other material.  In particular using the KEFs, the violins sound very 'clear'.  So, I think these Tannoys are perhaps a litle brighter and, maybe, mounted in a way that adds some susceptibiliy to reflections, I'm not sure.  The retailer mentioned something about the 'SE' (Special Edition) but I don't recall what he said, whether it was simply a marketing push, or whether they had been engineered towards a particular music type. 

The next test, back in the main room, was much more revealing.

2.  Moving position of speakers, and listening position.

Moved the speakers further from the corners, now probably 2 to 2.5 ft away from the back wall, and 1.5 ft from the side walls, and pointed them (very) slightly more into the room, while keeping the listening point very much on-axis.  That seemed to improve the clarity of the sound, but there was still the more-intense (heard) response of the violins.  I also tried a CD of a female singer, and her voice also seemed to intensify, almost 'boom' or 'ring'or sustain, perhaps (she wasn't shouting, she is accomplished and is mostly employed as a backing singer).  I would have thought that what I heard was much to do with the room, and not very much at all to do with the electronics, as others have indeed said.

Better still was moving my listening position.  Not yet tried blankets on the side walls, to see any effect (and I note AJ's again-insightful comments) but I will try this because I want to know what effect that'll have, either way.  What I did instead was to sit on the floor, with my ears below the levels of the backs of the sofas, so that I wasn't hearing (very many, but still some from one side) reflections.  The sound was very much clearer, very much.  For the first time, I had the sense I could 'hear' the instruments as though they were 'just there'.  The speakers were around 10 or 11 ft away from me, but I was a little off-axis because I was maybe 2 ft below the tweeter axis, around -11 deg, I think.  Doubtless there's an element of confirmation bias because I do want everything to perform well but nevertheless I felt there was a difference.

Seafarer

Re: Effect of room reflections

Reply #14 –
You might also want to try heavier toe-in, i.e. cross-firing them where the tweeter axis intersects around 1' or so in front of you
Loudspeaker manufacturer
 

Re: Effect of room reflections

Reply #15 –
Quote from: Seafarer on
Was that right, Grey?
10 (-18/20)

Glad to see you are experimenting with speaker placement.