Is that the crossover to the super tweeter? I can see no technical reason for continuing with this 3 way arrangement, when using a good modern cheap tweeter would be better.Stereophile measured the SHL5+ XD.
Can anyone explain what’s going at the FR >10kHz???
View attachment 359648
-
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!
Harbeth Super HL5+
- Thread starter Roy_L
- Start date
That is an M40, first of all, second, the meters are measuring apparent power*, so it really is not nearly as bad as advertised. Indeed, they do NOT drop below 2 ohms.
* the track being used has a loud pedal tone right around the frequency where impedance and phase angle are most acute in this graph. There’s a thread where we go back and forth about this on this very site.
Again? I know they did that with the SHL5+ and there was quite a back and forth with Alan. I think, honestly, the super tweeter integration isn’t what it should be. It doesn’t really come into play for me.
And you aren’t comparing directivity, which has room for improvement in Harbeths.
Sure, I was only comparing the two graphs, and offering a hypothesis for the differences. I know Harbeth are very intent on a flat response, so a deterioration is unlikely. As a research scientist, my first reaction would be to look at the measurement protocol. Measuring speakers consistently is not easy.
Why is that a insufficient measurement protocol?
can you explain , please?
do you have a sufficient one?
- Joined
- Jan 15, 2020
- Messages
- 6,903
- Likes
- 16,918
They might measure quite flat a specific vertical angle but having that nowadays superflous supertweeter (the original Spendor BC1 needed it because the Celestion HF1300 tweeter didn't extend above 13 kHz) creates lobes at already small vertical angles so even if someone would find the angle where they add up nicely it doesn't mean much when few degrees higher or lower they don't:
Fig.5 Harbeth Super HL5plus XD, vertical response family at 50", normalized to response on super-tweeter axis, from back to front: differences in response 45–5° above axis, reference response, differences in response 5–45° below axis.
Source: https://www.stereophile.com/content/harbeth-super-hl5plus-xd-loudspeaker-measurements
Fortunately the other Harbeth models don't have a supertweeter which can be positively observed in the vertical direcitivity measurements, exemplary:
Fig.5 Harbeth Monitor 30.2, vertical response family at 50", normalized to response on tweeter axis, from back to front: differences in response 45–5° above tweeter axis, reference response, differences in response 5–45° below tweeter axis.
Source: https://www.stereophile.com/content...-anniversary-edition-loudspeaker-measurements
I said inconsistent. You measure the same thing twice, and yet you get slightly different results. In this case, there was a quite considerable period of time between the two measurements. Maybe the speaker was in a marginally different position in the room, or the room was not quite the same because objects had been moved. There are umpteen possibilities. As for the high frequency peak, that is well above human hearing and therefore irrelevant. The new active Harbeth speakers with dsp controlled electronic crossovers no longer have a super tweeter, and have a superbly flat response.
- Joined
- Jan 15, 2020
- Messages
- 6,903
- Likes
- 16,918
Thankfully also all their past models except the ones of the HL5 family which had it as a kind of tribute/modernised BC1.
Will be interesting to see their off-axis / directivity measurements as making a loudspeaker flat just on one reference angle is not difficult nor enough to make it a well measuring neutral transducer.
What’s interesting about the Harbeths is that, while they don’t have the infamous Gundry Dip on-axis, their off-axis power drops in the same region (see the + in the graph above), which would create a sound power flattening at the same spot. I wonder whether people sort of like that.
I enjoy my Harbeths - they sound right through the midrange, and instrument timbre feels very accurate to me. Where they lag in comparison to my Revels is with orchestral music. You want to be closer to the speakers for that. I can spin a narrative that relates that to directivity, but I don’t know for sure.
- Joined
- Jan 15, 2020
- Messages
- 6,903
- Likes
- 16,918
I would say definitely, many well sounded loudspeakers without waveguides have their crossover frequency to the tweeter relatively high so there is a sound power dip at the presence region.
DSJR
Major Contributor
Stereophile measured the SHl5, 5+ and 5+XD on the axis of the 'super tweeter,' yet in at least one of these reviews, they mentioned that the main tweeter is the axis as designed. Like the dear old now ancient Bc1, the response is a bit of a mess where the two tweeters overlap (deliberate on the BC1 though according to a long-ago chat with Spencer Hughes, who was another designer happy to talk - and share - to peeps like me while he was alive [r.i.p. sir!]) - Dear Dudley Harwood who started Harbeth was too shy sadly and when visiting us, stood quietly and anonymously in the corner waiting to speak to our sales manager and you'd never know he was there. The current boss/designer has stated many times now that almost NO reviewer ever checks or cross checks with them to find out more about the company, products or the ethos behind it all. Heck, in that position of reviewing as honestly as I could, I'd LOVE to find out more about the gear I was evaluating/measuring to build up proper knowledge and experience (which I did as much as I could when on my side of the counter!) and to me, that's why some honest, caring and communicative makers/designers of speakers and electronics who post here on ASR should be encouraged and respected so they stay and contribute more - in my opinion of course. I feel we all gain then and learn a lot from their gentle 'education'...
So it’s impossible to compare different speaker measurements?
do you know how Stereophile do arrange their measurements?
I just returned from AXPONA having spent 20+ hrs in the Fidelis room comparing the latest 5 with the P3 on the new Nelson powered stands for many hundreds of listeners. Despite increasing the gain almost fully on the Nelsons' 5" woofer, the feedback was unanimous that the 5 was more spacious, timbrally accurate...especially in the lower mids...and preferred. It wasn't even close. I was careful to adjust gain for the 3dB delta. Several folks already owned the P3 and use separate subwoofers. All agreed that the Nelsons were seriously under-engineered poor value, and that the 5 was a vastly better speaker...especially as the P3+N cost $6.7k vs the 5 at 7.9k.
I've read only a few pages of this 37p (!) thread, but am taken by all the commotion re off-axis measurement anomalies. I too would have thought that designing an 8+1+0.75" to be foolhardy, thinking that perhaps a larger 10+5+1, like the new Dovedale, might be better engineered. But many times I was relegated to listening to both the 5s and "supported" P3 WAY off axis, as the room was used on its long wall with just a single row of 6 chairs in the relative nearfield. That there was STILL great tonality and decent soundstage, with airy detail, I found beguiling. So it's clear that the room response of the 5 is superb. This experience supports my similar view of performance at a previous show where the 5's were used in the more traditional geometry along the short wall.
The 5 apparently has been Harbeth's best-selling speaker for a long time. I could hear why. I'm also guessing that the great and seamless midrange is partly due to the cross being up at 3.3kHz.
On another note I'd just recommended to a very dear friend that his last speaker should perhaps be a MoFi8 instead of his baby Spendors (like P3) for a larger painting studio where listening angles would vary. Unfortunately the MoFi8 seem to have a searing and fatiguing quality in the upper-mids/low treble that room treatment doesn't seem to alleviate. I'm smelling that it's about the 1.6kHz cross.
So I'm going to borrow these same Harbeth 5s used at the shows and provide a demo. I don't expect them to boogie like the MoFi's, but maybe be MUCH less fatiguing and tonally clean c. 1-2kHz. He's using a big old Hegel, so there'll be enough push for an 86dB sens.
Throughout the show I quickly (and perhaps naively) redesigned the Nelson to be more effective...especially in filling in the 80-200Hz cavern of the P3 in open space without boundary support. Harbeth has designed this powered and DSP-tweaked stand to support most BBC-type and other small 2-way standmounts. Maybe that discussion is for a different thread? You tell me, as I have to exercise caution to not step on associates' toes.
I've read only a few pages of this 37p (!) thread, but am taken by all the commotion re off-axis measurement anomalies. I too would have thought that designing an 8+1+0.75" to be foolhardy, thinking that perhaps a larger 10+5+1, like the new Dovedale, might be better engineered. But many times I was relegated to listening to both the 5s and "supported" P3 WAY off axis, as the room was used on its long wall with just a single row of 6 chairs in the relative nearfield. That there was STILL great tonality and decent soundstage, with airy detail, I found beguiling. So it's clear that the room response of the 5 is superb. This experience supports my similar view of performance at a previous show where the 5's were used in the more traditional geometry along the short wall.
The 5 apparently has been Harbeth's best-selling speaker for a long time. I could hear why. I'm also guessing that the great and seamless midrange is partly due to the cross being up at 3.3kHz.
On another note I'd just recommended to a very dear friend that his last speaker should perhaps be a MoFi8 instead of his baby Spendors (like P3) for a larger painting studio where listening angles would vary. Unfortunately the MoFi8 seem to have a searing and fatiguing quality in the upper-mids/low treble that room treatment doesn't seem to alleviate. I'm smelling that it's about the 1.6kHz cross.
So I'm going to borrow these same Harbeth 5s used at the shows and provide a demo. I don't expect them to boogie like the MoFi's, but maybe be MUCH less fatiguing and tonally clean c. 1-2kHz. He's using a big old Hegel, so there'll be enough push for an 86dB sens.
Throughout the show I quickly (and perhaps naively) redesigned the Nelson to be more effective...especially in filling in the 80-200Hz cavern of the P3 in open space without boundary support. Harbeth has designed this powered and DSP-tweaked stand to support most BBC-type and other small 2-way standmounts. Maybe that discussion is for a different thread? You tell me, as I have to exercise caution to not step on associates' toes.
I assume that the Nelson is a well designed product but for the asking price I don't understand whats the advantage over a standard sub…
Beside that may I ask you what was your role at the Axpona?
A couple of weeks ago I heard the 5XD at the Sydney HiFi show. I was floored by the fabulous sound. It was balanced, non-fatiguing, open and impactful. The Ayre amplifiers may have had something to do with that. I will usually listen for a good vocal performance, and these nailed that while having real drive on all types of music.
There were many expensive systems there, but the only speakers I enjoyed - a little - more were the Devore O/96 at about two and a half times the cost.
Perhaps I’m biased as my Stirling LS 3/6 are based on the same original design as the 5.
There were many expensive systems there, but the only speakers I enjoyed - a little - more were the Devore O/96 at about two and a half times the cost.
Perhaps I’m biased as my Stirling LS 3/6 are based on the same original design as the 5.
interesting to read that you enjoyed the Devore although they measure much worse than the Harbeth….
The purported advantage of the Nelson is that it's DSP phase-corrected to mate with the P3. Unfortunately it's too little product for a sizeable ask.
Sporting only a 110mm (4.5") woofer with a wall-wart driven amp it operates only down to -3dB at 35Hz. But that's not the musical problem. It rolls off at 75Hz...far too low to fill in the upper bass/lower mids trough of the P3 when operated in open space. This was obvious when played in the room...even at relative nearfield. Bassos became tenors, metzos thinned to soprano, and my piano references decidedly took a size down in scale.
Steinway Bs and D's sounded like baby grands.
Again, this is a Super 5 thread, so I'll go more into Sir Nelson in a new thread?
The golden ears around Fidelis were so impresssed by my solid silver wire concoctions that I developed a tiny startup in retirement called ArgentPur, and have been wiring AXPONA and CapFest rooms for a couple of years. Speakers used for R&D include my ref Parsifal Encore 25th, Maggies, Harbeth 3P, 5, and 40.3, as well younger ears using Einsteins in a highly curated system. Solid Ag DOES inform soundstage detail without brightness, so certainly benefits small monitors as well larger transducers.
A hotel room wall bad ground prevented use of vinyl until discovered. An AMP-ed pre was swapped in to see if the Lab12 pre was the culprit. I had to listen to this pre for 90 minutes of hell...transistory ala 70's receivers...before the day was saved by a cheater plug and sublime return to the Greek beauty. Indeed, the AMPed's shrillness helped to mask the delta between the P3 and 5, as it was more difficult to compare their 80-200Hz relative weights.
Next year I hope to power a shared room with my AgPur loom AND new Gan-FET Monoblocs, which garnered nice interest in a static display last week. I'm so impressed wih the Harbeth 5 that I'd love to power them, maybe with the same Lab 12 pre and DAC. Would have to secure a streaming partner by then...or borrow an Aurender.
Anyone have any experience comparing the Harbeth 5 and the new Wharfedale Dovedale?
Ze Frog
Addicted to Fun and Learning
- Joined
- Jan 4, 2024
- Messages
- 633
- Likes
- 724
To be fair, especially for the old design these look decent. Wether they are worth the cost though is another story, 16 Kgs for that size of box, would have me worried despite in being the BBC type thin wall design type.
Similar threads
