Harbeth SHL5 specific

[A.S.]

... I mean that the SHL5's remain articulate in the midrange (I can hear detail in the mids and understand sung lyrics) and maintain a full sound that does not lose bass tonal balance or dynamics.

In actual fact, the performance of a Harbeth is uniform over a wide operating range - the speaker itself doesn't 'know' how loud it is playing, so it doesn't change its character with volume level. What does dramatically change with level is the human ear's perception of sound. You can read about it here: http://www.webervst.com/fm.htm Curves like these have to be normalised at some 'pivot' frequency, and since the ear is sensitive at 1000Hz, this was chosen as the reference frequency. The graph looks confusing I know, but all you need to grasp is that the left hand vertical scale is the loudness of sound ranging from 0dB (absolute and total silence) up to 110dB+ (rock concert levels) and that the horizontal scale is in Hertz, from low frequencies to high ones. The wiggly lines (the equal loudness contours) describe the sensitivity of the ear to particular frequencies versus listening level.

Example: Let's see what happens if the loudness (defined at 1000Hz) is 60dB in our listening situation.

1. Find 1000Hz (1kHz) on the horizontal (x) axis and follow it up until it meets the 60dB rule on the y axis. Note that it touches a particular contour line. Keep an eye on that contour line alone: ignore all the others. I've coloured it for you.

2. Now find, say, 30Hz on the x axis and follow it up until it touches the same contour line we found in step 1. Read-off from the left scale the loudness at that point - I estimate that it is about 80dB. Agree?

Now the interesting bit. What can we deduce from this observation? We can make a statement which is intuitive but numerically rather shocking .... we can say "At a listening level of 60dB (normalised to 1000Hz), we would need to electrically boost the 30Hz region by [80dB - 60dB =] 20db for 30Hz to sound equally loud as the 1000Hz tone."

Note from the graph how as the contour lines flatten out at low frequencies as the overall listening level increases. That implies that as volume increased, less boost is needed to make those low frequencies sound equally loud as 1000Hz. That agrees with what we hear doesn't it. This explains why that when you turn the volume down on any hi-f the bass appears to drop away rapidly: it's not the speaker, it's our perception of sound balance. Also note the very interesting situation in the 3000-4000Hz region and at the top end too above 10kHz.

Appreciating the way the ear/brain works is what led to the invention of the bass and treble control and later the loudness contour switch. It's also the basis of Dolby A/B/C/S/SR/AC3 technology, and underpins MP3. In my humble opinion the deletion of these very useful and pragmatic tone shaping controls from audio amplifiers in the 1980s flies in the face of common sense and encourages the listener to play his hi-fi, TV, Walkman, car audio louder to hear the bass he wants. The consequence: needless volume, irritation to the neighbours, more room-resonances excited, wear and tear on the ears and hi-fi and noise pollution. I'm sure it does wonders for the sale of big beefy amplifiers sans tone controls though!

The great example of the art of tone shaping to match the ear's well documented characteristics is the Quad amplifier's tilt control which provides a good approximation to the optimum level-correction and is the same circuit solution that I have incorporated into the Monitor 40 Active. Peter Walker really did understand music, rooms and the ear.

Suggested Google search "Equal loudness contours". Note: These curves were developed to represent the situation with pure tones, and the test subjects were wearing headphones and presumably the subjects were in good health and with young ears: the results may not perfectly translate into music listening through loudspeakers. I suspect that the situation shown in this headphone-listening graph may somewhat overstate the boost needed at low frequencies when listening to speakers, but that a boost is needed is undeniable.

Thought: if you were a speaker designer attempting to make a natural-sounding speaker, how would you shape the frequency response curve? Clue: What may be 'right' for a diet of rock and roll may not be for chamber music. That's just one reason that there are countless thousands of speakers available to the consumer all serving a market niche and satisfying someone's tastes and needs.

[Casaross]

Hi Alan,

I am sorry to have delayed my response. Thanks very much for your thoughtful reply to my inquiry regarding speaker performance at varying sound pressure levels. I recall Yamaha receivers and integrated amplifiers in the late 70's employing a "loudness" control similar to the Quad "tilt" control. Yamaha had you turn the volume to a level with which you were comfortable, then adjust the "loudness" knob to turn the amplifier volume up or down, but it would maintain the same emphasis on each of the high and low ends of the frequency spectrum. That sort of control, of course, became unfashionable. Now I listen to "fashionable" gear omitting the bass, treble and tilt/loudness controls. I just turn the music up to a level that sounds "right" for the tonal balance, recording and what I plan to do at the time. For instance, I listen at softer levels if listening to music in the background, somewhat louder if listening only to music. Sadly, as the years pass, I bet I listen louder and louder as my hearing becomes decreasingly sensitive. Sigh.

My aging is not your worry, though, and I appreciate your insight.

Scott

[Chayro]

Alan - I was fooling around with a Stereophile test CD and, when playing the high-frequency warble tones, the 12.5K sounded significantly lower in perceived volume than the 10K and I was totally unable to hear the 16K, which was next after the 12.5k. I realize it could be me, as I am in my early 50's, and high frequencies are allegedly curtailed, or could it be another reason?

I don't usually get hung up on stuff like this, but as long as I have access to the designer, I figured I would ask. thanks.

[David Schalkwyk]

Dear Chayro

From a fellow early fifty-something: I can't hear anything above the 8KHz test tone! Sad, but true. It doesn't change my enjoyment of music, though, so after some harrowing months I've stopped worrying about it. While the family gets very agitated about pool pumps, insect noises, refrigerator warning signals, I carry on oblivious... So it has its compensations. I'm told that one doesn't miss any of the fundamental tones of orchestral instruments, but the overtones are another thing. At least it means one doesn't have to fork out extra money for supertweeters.

David

[A.S.]

I think that, sadly, David is probably on the right path: in middle age high frequency acuity does diminish and if you can hear out past 12kHz or so you are doing well. 16kHz is really going to be a struggle and I'm not sure that I can hear it any more unless I increase the volume. But that doesn't diminish my enjoyment of music one jot.

DO NOT TRY THIS AT HOME!

However, let's have another look at those Equal Loudness Contours that we discussed here: http://www.harbeth.co.uk/usergroup/s...=1291#poststop

Follow along any of the contours and note that the high frequencies (towards the right hand side of the horizontal axis) need to be massively boosted to sound as loud as the reference 1kHz. In other words, if the test CD is cut with all (high) frequencies recorded at the same level (I bet it was), you would expect that as frequency increased you would struggle to hear the tones. This is because our ears are insensitive to high frequencies relative to middle frequencies, exactly as predicted in the Contour curves of hearing performance.

Now the CD engineers could have approached this a different way. They could have built-in the necessary boost at HF following that of a typical person's hearing acuity (say a +10dB boost at 10kHz would be necessary) but with one potentially disasterous consequence: burned-out tweeters!

Another thing: when the FM broadcast system was being designed some 40 years ago, one of the key questions that needed to be resolved was that of settling upon the absolutely minimum audio bandwidth that could still be called high fidelity sound: 20Hz to 20kHz simply wasn't achievable in stereo. At the time, a good quality FM receiver was the highest quality audio source available at home so this was an important issue. After much testing, it was decided to set the upper end at 15kHz with a very sharp filter as 'there is nothing useful in music above 15kHz'. That situation remains today: leaving the studio the signal bandwidth is limited to 15kHz whether it goes to the FM transmitter or into the digital distribution chain. (Easy to prove: just feed the signal into the spectrum analyser and you'll see there is a brickwall filter at 15kHz). Despite that, those concerts on Radio 3 sound completely natural to me.

If you can recall the Celestion HF1300 tweeter, you may remember it had a steep mechanical cut-off at 14kHz.

[phil]

I am using these w/ an arcam 23cd player. The system seems a little lean also softer
passages of music seem not to be projected enough. Wondering if I need more power.
I listen about 8 feet from the speakers in a 15 x12 x9H ft room.

[Mank]

Sorry for the inconvenience, I just left half a message sent to Alan, and I would like to repeat the question short as possible, because the matter created real anxiety for learning, hope not only to me. This:

When disconnecting the LF feeding cables (on bi-wiring mode) from my SHL5s, leaving only the tweeters/supertweeters playing, all I could hear was just tiny little sounds -high freq ones- like hearing a portable radio from 10m. away! This while listening to a violin concerto, actually I wanted to hear what tweeters produce, after reading this discussion, which of course got me a bit depressed being at my early 50s... So, Alan, 2 questions:

1. Is this really all that the tweeter/supertweeter play? It seems, when isolated, so insignificant and poor, then why they are measured to have such a crirical contribution to the program? Especially the supertweeter, which, besides being middle-aged, I almost never hear! I know I must be somehow wrong, so please light up my poor knowledge! Then,

2. If, as I heard, the RADIALS really carry out more than 95% (IMHO) of what I actually listen to, then why not creating a driver with so excellent properties, that it would cover the reproduction as a single point of radiating to our ears. Theoretically, even without two speakers (remember the stereo illusion you mentioned), we would have one speaker-one driver (full range) working, with a flat and perfect response.

As you can see, your advice will surely illuminate a lot about the role and blend of drivers, I mean especially the Harbeth design topology.

Thanks in advance so much for your reply,
Warmest regards from Greece,
Thanos

[A.S.]

... 1. Is this really all that the tweeter/supertweeter play? It seems, when isolated, so insignificant and poor ...

2. If, as I heard, ... why not creating a driver with so excellent properties, that it would cover the reproduction as a single point ...
Thanos

An excellent question - and so interesting that I have just set off the smoke alarm because instead of watching my dinner cooking, I decided to read your message - dinner is a charred mass! Oh well.

Ok, first question: what should it sound like without the tweeters connected? Answer - like a mediaum-wave (AM) radio. And that is exactly what you have found for yourself. Agreed?

Question 2: Actually, out of curiosity some months ago I did try this. I disconnected the tweeter and then using my Behringer EQ box, I hugely boosted the RADIAL woofer's naturally rolled-off response and gave it 'tweet'. It worked astonishingly well, far better than I would expect conventional drivers to work* ..... but (there is always a but) ... if you ask a 200mm (8") woofer to pretend its a 25mm (1") woofer it can only do this when the listener is exactly on axis. I mean, the sweet spot is extremely narrow - just a few cms. here or there. Not much use for illuminating the entire listening room with the freshness of natural high frequencies. In other words, the sound is rather dry.

Remember: there is a mathematical relationship between the drivers diameter and its (minimum and) maximum operating frequency. So, woofers are big and wide (but can't reprosuce HF), midranges are a bit smaller, (so they can't reproduce bass but they can reproduce a little more HF) and tweeters are tiny - they can't reproduce bass or middle but they are very good at reproducing HF. Does that help?

Now to make the most of black potatoes!

* I must do some more work on this. The quality of the high frequenies on the 8" RADIAL driver was so extremely good that I think I could use this to promote and explain just how clean and smooth Harbeth's RADIAL is.

[heiner1960]

This thread specifically relates to the SuperHL5

Hi guys,

in a few months I probably will buy a speaker I have never listened to: the SHL5

As I live in Thailand (no distributer) a friend will take them over when he moves here from Hong Kong.

My only concern: my room is 4.8 x 7.8 m (16 by 26 feet) and 4 m high in the center, at the side walls 2.3 m. A lot of cubic feet! And I do not listen only to chamber music but also to blues, jazz and rock.

Will it work and/or do I need a subwoofer?

I like to listen into the music, the house doesn't need to shake.

By the way: I have never seen an internet group with so good non-nonsens information. This compliment goes to AS as well. I have the feeling it is been running buy music lovers and not by hifi freaks.


Greetings
Heiner

[David Schalkwyk]

Heiner, I have a room almost as large as yours and I find that the much smaller Harbeths, the HP3s, fill it more than adequately! I don't listen to rock, but I do listen to blues, jazz and classical music of all descriptions, including Mahler and Shostakovitch symphonies! So, although I've never heard the SHL5, I can't see, given its specifications, why it shouldn't provide you with more than satisfying music. And if it's any consolation, I also bought my Harbeths without having heard them, for the same reason: no distibutor. I've never been happier with a loudspeaker, or with my music.

David

[Chayro]

Just a follow-up to the story: I loaned the test disc to a person I work with in his late 20's and asked him to take it home and listen on his system. He was able to clearly hear the 16K tone and even the 20K tone, albeit down somewhat in volume. This leads me to one inescapable conclusion - My friend's $400 Wharfedales have better HF response than my Harbeth SHL5s! Or it could be my ears.

[A.S.]

With respect, you are not making a solid technical comparison. I have already explained in detail with graphs how the sensitivity of the human ear varies tremendously with frequency, and thus "equal loudness" depends upon the level that a signal is played at.

To make a proper evaluation of A v B you have to measure the frequency response, confirm that they are equal and then draw conclusions. My guess would be that the other speaker has a rising response at the top end which gives the illusion of enhanced calrity etc. etc..

We covered this in detail here here - take a close look at the HF end of these curves .... http://www.harbeth.co.uk/usergroup/s...=1291#poststop

[Chayro]

Sorry Alan - I was kidding. It's sometimes hard to accept the consequences of age.

[A.S.]

I guessed you were, but it was a serious point you raised and gave me the opportunity to address it. The gradual run-down in our hearing with age has nothing whatever to recommend it. Funny how when you are in your 20's the future seems infinite and you just can't wait to embrace it.
A

[Chayro]

I The gradual run-down in our hearing with age has nothing whatever to recommend it. Funny how when you are in your 20's the future seems infinite and you just can't wait to embrace it.A

I think I actually coined the phrase "The only good thing about being 50 is that you aren't 60". Pretty clever on my part, I think.

[Alastair MM]

Hi Phil . If you're still out there , I have had a similar experience with the Compact 7 and an A21a . The solution in my case was more power . Once power was applied the voices and instruments have more body and a greater sense of vitality . The low frequencies solidify and the overall balance makes much more sense !

[Rich Arthur]

I also had a similar situation using the Quad 909 amp with 99 CDP cd player with my Compact 7ES2 speakers. The 99 CDP was driving the amp direct through it's variable outputs. The sound was easy to listen to, but was a little lean. This was remedied by inserting a preamp in between the CD player and the amp. The sound remains as nice as it was except that I have more drive, more bottom end, and improved image. In my case, the preamp is a Lehmann Linear headphone amp that also functions as a single input preamp.

[Vangelis]

I read on the Soundstage web site that after " the Compact 7ES3, Harbeth will release more new celebratory speakers" as well. I've had a pair SHL5 on order for some time. Has anyone heard if the SHL5 are to receive "celebratory"update in the coming months? I would be very disapointed if a new celebratory-30th anniversary version of the SHL5 came to market just months after I take home a pair of SHL5s home.

[Vangelis]

I read on the Soundstage web site that after " the Compact 7ES3, Harbeth will release more new celebratory speakers" as well. I've had a pair SHL5 on order for some time. Has anyone heard if the SHL5 are to receive "celebratory"update in the coming months? I would be very disapointed if a new celebratory-30th anniversary version of the SHL5 came to market just months after I take home a pair of SHL5s home.

[sonofcolin]

What about the 50th anniversary models? I heard they're going to be quite special ;)