• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Wait... science AND audio? At the same time?

nerdstrike

Active Member
Joined
Mar 1, 2021
Messages
258
Likes
311
Location
Cambs, UK
Greetings from the UK.

I'm a scientist and a musician, so in retrospect it's no surprise that I've sunk many hours into reading content on this forum. Ever since I was inspired to take some in-room measurements, I realised it would be rude not to register. What a fantastic resource/seething pit of rationalist vipers.

Hello to you all, please don't ask me to double-blind test my conviction that I am a scientist or some part of my soul will implode!

Gear:

B&W P7 headphones - Very nice, very comfy on my head. Best headphone I could audition within budget.
Audioquest Dragonfly Red - Turns out to be a poor choice for the money, shouldn't have bought it before reading ASR...
B&W 602 s3 - My first, best purchase back in 2004. Having a "proper hifi" really opened my ears to enjoying music outside of playing it.
Rotel RA01, RCD02 - Yes. I was cheap. They've served me well, just wished I had sprung an extra £50 for the remote volume control option...

and latterly:

Arcam SA30 - Best one-box package with Dirac that I could find within budget, and got a discount I couldn't refuse. I was tempted by talk on here of some sweet class D Hypex/Purifi gear, but couldn't find a combo with a receiver/pre-amp+DAC that didn't cost twice as much as the integrated option.

Looking to replace my speakers with some floor-standers soon... I'm compiling a shortlist of graphs for when I'm allowed to audition without catching the plague.
 

Wes

Major Contributor
Forum Donor
Joined
Dec 5, 2019
Messages
3,843
Likes
3,788
Welcome

vipers?

pit vipers can see in the infra-red...
 
OP
N

nerdstrike

Active Member
Joined
Mar 1, 2021
Messages
258
Likes
311
Location
Cambs, UK
My current shortlist includes:

KEF's R5 and R7 (by extension from the fantastic R3 measurements, haven't seen anything for the R5/7, only their predecessors)
Focal Chora 936
Revel F35

Reckon I'm going to struggle to audition any Revels, and the good ones are well out of budget.

I took in-room measurements with Dirac.live for my old 602 s3's and got this with the Arcam-provided mic: (averaged over 15 measurements around the sofa)


Screenshot 2021-03-02 at 11.51.09.png


The corrected plot looks a bit supernatural... and of course is predicted rather than measured.
Screenshot 2021-03-02 at 11.52.24.png


but the corrected spread (one channel only for clarity) I think is more honest:

Screenshot 2021-03-02 at 11.55.18.png


In any event, the correction has done a lot (I believe) to balance the lows and highs, and I like the sound better. I did add a couple of dB to the bottom end of the target curve, because the default target sounded really clear in the mids but bland everywhere else. Some of the fun was missing.

I can't do much with the room (L-shaped, right channel is much too close to the curtains when they're open), but my hope is that a better speaker will be able to provide more clarity lower down, and a more balanced response higher up. Dirac has worked a little miracle with my old speakers, but it cannot come without some compromises.

My other priorities are:
No sub - nowhere to tuck one
Difficult to knock over - hence floor-stander
Looks pleasing in a non-ultra-modern decor

So that's the story!
 

Attachments

  • Screenshot 2021-02-20 at 15.31.03.png
    Screenshot 2021-02-20 at 15.31.03.png
    285.1 KB · Views: 91
  • Screenshot 2021-02-20 at 15.31.34.png
    Screenshot 2021-02-20 at 15.31.34.png
    287 KB · Views: 84

DSJR

Major Contributor
Joined
Jan 27, 2020
Messages
3,317
Likes
4,431
Location
Suffolk Coastal, UK
Do please be careful in tying measurements to perceived sonic results. One should guide the other I feel - dispersion in a domestic environment is also vital as it can react - or not - with room furnishing and so on in my experience. In the UK, we don't all live with large listening spaces sparsely furnished...

I heard the £1100 Chora and it was a nice listen and nothing like the boom-plus filling-removing/paint-stripping tizz of older stand mounters they made at similar money. The tilt-back visuals didn't find favour with herself though when I showed them to her...

Don't dismiss Q Acoustics at this price level. They're popular and are well finished and presented, so should hold value well as your 602v's have almost certainly done. Quad also have some interesting models at around a grand or so I believe. They look good but I haven't heard them (i have old lingering doubts regarding cone-ribbon hybrids). The Wharfedale Linton is also around this price and knowing the designer since his Heybrook days forty odd years back and the extensive experience he's accumulated since, I suspect they're a great of trad-looking speaker as the cheaper Diamond range seems to be (I only know the 220 which does exactly as reviewed here - rather nice when tweeter covers are removed).
 
OP
N

nerdstrike

Active Member
Joined
Mar 1, 2021
Messages
258
Likes
311
Location
Cambs, UK
Do please be careful in tying measurements to perceived sonic results. One should guide the other I feel - dispersion in a domestic environment is also vital as it can react - or not - with room furnishing and so on in my experience. In the UK, we don't all live with large listening spaces sparsely furnished...

Don't dismiss Q Acoustics at this price level.

Duly noted, and thank you for your recommendations! In truth, my expectations were that Dirac would make only a small sub-placebo difference, but I was genuinely shocked by the change. I couldn't do it blind, but flipping it on and off was extraordinary. Perhaps that is an indication that my room is terrible. While my lounge is 31m^2, it's still a bit cluttered and I can only begin to speculate what the shape does to the acoustics.

I do like the look of the Concept 500, but it's out of reach, and the Concept 40 looks are a bit brutal to my taste, and there's no way those Lintons would pass the stringent LPAT (life-partner-approval test). Shopping for speakers is tough!
 

DSJR

Major Contributor
Joined
Jan 27, 2020
Messages
3,317
Likes
4,431
Location
Suffolk Coastal, UK
I've had long PM's with a chap on another forum, who wants a gentler (yes, I know) sound for female vocals and chamber music. He's smitten with antique Spendor BC1's driven by an even older Quad 33/303 design. Set on higher stands, the BC1 'bass honk' is less noticeable and a good 33/303 offers a great bandwidth limited signal to them, rolling the bass off below 35Hz or so, but th emidrange is still wonderful at low to medium volumes. Nothing whatever to do with the strictest 'highest fidelity standards regardless' approach here (which I do approve of by the way) but enchanting for a domestic system nonetheless. What I do find highly alarming is the prices asked for 'serviced' examples of these now half century old products, but that seems to be it at current high labour rates and the skills needed to restore them properly.

Me? I was impressed by the ATC 11's (I'm a confirmed ATC fanboy and deeply miss the 100A's I used to own in a former life) but they're more 'boxes on stalks' so maybe not suitable. They do seem rather more neutral than the mid-forward 19 and 20 models with more bass weight and 'authority,' but they do need some power as the brand demands.
 
Last edited:
OP
N

nerdstrike

Active Member
Joined
Mar 1, 2021
Messages
258
Likes
311
Location
Cambs, UK
Follow-up!

I auditioned the KEF R5 against the Focal Aria 926. So the fun began...

The audition room was rigged with a Tidal setup and no CD player... Alas that meant I could not use my carefully collected selection of well known tracks to do comparisons. I was also unable to easily flip from A to B, which in turn meant no blind listening. Did Tidal have the majority of the tracks I wanted? Not really. No Rachmanninov played by Rachmanninov, wrong performance of Tchaikovsky and so on.

The two speakers also had a different efficiency (4dB at reference). The Focals were noticeably louder, and I didn't have the means to match volume.

A total failure on objective speaker testing, I think you'll agree!

So I bought the R5... We liked it better after a good few trials. The two were different but close competition. I liked furniture of the Focal better, but somehow there was less clarity where I wanted it, and the bass was a bit over the top at times. Could have been the room... Anyhow, I have this forum to thank for giving me opinions and evidence to support that either speaker would be pretty good in class provided I like an even response.

In room measurements will follow as I'll have to redo my Dirac config, and we'll see what a 6x price increase over my old speakers gets once it hits the real life (subject to measurement error).
 
OP
N

nerdstrike

Active Member
Joined
Mar 1, 2021
Messages
258
Likes
311
Location
Cambs, UK
As promised... room measurements for the R5

R5 both channels.png

Left channel with variance around wide target area:
R5 left channel with dispersion.png

and right channel also (this one is nearest the wall and some curtains)
R5 right channel with dispersion.png


and finally impulse stuff. Not entirely sure how to interpret it!
R5 Impulse response.png


I'd welcome any guidance on how to interpret the impulse response. I haven't gone and inverted polarity on one speaker by accident have I? I tried pretty hard not to but turning on Dirac had a dramatic effect: There was a marked change in the stereo imaging as I turned a filter on. I've been having slight doubts about stereo imaging in my early listening, and the turning on of the filter really improved things.

As before I found that the default target curve sounded anaemic (perhaps I am in fact a bass junky). This time I have had the most success with a target curve that aims to "fix" the speaker response without taking away from the existing broad trends. For example, the default target seems to make the speaker sound very bright/brittle sounding. I'll be playing around some more, but I need to do some serious listening first!

Comparing my venerable 602's in-room graph to these is interesting. I observe:

1) 602s3 has a 5 dB hole from 2-3 kHz
2) R5 has a 5-8 dB hole around 150 Hz
3) R5 responds very strongly down to 50 Hz, and in general there is a slightly steeper slope from low to high frequencies

I'm thinking that my R5's might benefit from a bit of port bunging to tame some of the reinforcement, although it doesn't sound totally over the top as-is.

Suffice to say that I'm enjoying myself! I find it interesting that such dramatic differences in tonality are only found in relatively small differences in the respective measurements. I wonder how much of it is just in that 2kHz hole in the B&Ws, or just eyeball sensitivity to logarithmic scales? Ramble over...
 

RayDunzl

Grand Contributor
Central Scrutinizer
Joined
Mar 9, 2016
Messages
13,204
Likes
16,987
Location
Riverview FL
I'd welcome any guidance on how to interpret the impulse response.

I interpret it as SPL vs Time.

Not frequency, just pressure (or rarefaction) at the mic.

Ideal might be a single peak, the rest flat.

You won't see that, so you interpret what you see.

Direct sound in the first bump.

Reflections and noise bumps after that.

REW allows normalization - make the first bump 100% and compare from there - a ratio of direct to reflected/delayed/extraneous sound.

Initial "impulse" of JBL LSR 308 (red) vs MartinLogan reQuest (black), with the initial peaks "normalized", first three milliseconds.

ML has a higher Signal (direct sound) to Noise (reflections or other extraneous sound) than the JBL, here at my place.

1619968558403.png




Then, the first 40 milliseconds, showing many bounces off the room from the wide dispersion JBL, not so much from the narrow dispersion ML.

1619968713327.png


No EQ above, raw speaker response at the listening position.



When "Room Correction" is applied - AcourateDRC via a miniDSP OpenDRC-DI), the direct impulse of the JBL is cleaned up a bit, becoming more similar to the ML, but the later reflections remain largely unchanged.

1619969109891.png



1619969186575.png


Measurements are at the couch 3m from the speakers.




Zooming in on the time after the impulse shows 10 to 15dBSPL differences in the ratio of direct to reflected sound level after the direct. (the ratio of percentages can be converted to relative SPL )


1619969469944.png



I am not an Audio Scientist, so the above is my observation/opinion.
 
Last edited:
OP
N

nerdstrike

Active Member
Joined
Mar 1, 2021
Messages
258
Likes
311
Location
Cambs, UK
Then, the first 40 milliseconds, showing many bounces off the room from the wide dispersion JBL, not so much from the narrow dispersion ML.
Thanks! So in my case, the big echo at 3ms after impulse is travelling 90cm further than the direct noise, and is likely a ceiling reflection or something similar, owing to its strength and comparatively short delay. At 2ms after impulse I think you can see the

The first impulse seems to be mirrored between right and left channels, suggesting anti-phase. On that basis I think I've got my wires crossed! One way or another Dirac seems to be fixing that but I guess what I really need to do is reverse one channel's terminals and measure again to see what happens. *sigh* there goes another 30 minutes...
 

RayDunzl

Grand Contributor
Central Scrutinizer
Joined
Mar 9, 2016
Messages
13,204
Likes
16,987
Location
Riverview FL
Measure left, right, and both speakers playing simultaneously at the listening position.

If using REW, apply "Acoustic Timing Reference" to the measurements.

Then you can see if the speakers are time aligned, and if they are phase reversed.

Here, two speakers playing the sweep simultaneously, microphone is off-center by a foot or so, you see two impulses, but the phase is not reversed.

1619980215540.png
 
Last edited:
OP
N

nerdstrike

Active Member
Joined
Mar 1, 2021
Messages
258
Likes
311
Location
Cambs, UK
Right you are. The plot above is left and right in isolation, but not the interference pattern of both. I'll see if I can get REW going with the provided mic.
 

abdo123

Master Contributor
Forum Donor
Joined
Nov 15, 2020
Messages
7,425
Likes
7,941
Location
Brussels, Belgium
there is a lot of messiness between 1K and 20K, do you happen to have anechoic measurments of the speakers? looks weird.
 
OP
N

nerdstrike

Active Member
Joined
Mar 1, 2021
Messages
258
Likes
311
Location
Cambs, UK
There's a window nearest the right channel that is probably doing a bang up job of reflecting high frequencies, and nothing to the left side. As regards the R5... we've only got KEF's published measurements for those as far as I know. It'll be not completely dissimilar to the R3 measured on this site given it's the same uni-q driver doing the treble.
 
Top Bottom