John Curl's Blowtorch preamplifier part III

Status
Not open for further replies.
Personal story. (Totally subjective ;-)

After long decades in various recording studios* and personal quest for the best system i could afford at home to verify the quality of my work, I was in the 1990-2000 years in a quasi definitive position to avoid any horn for my own use.

*JBL, ALTEC, UREI, Westlake, Electrovoice, Tannoy, Jensen, Lowther etc.

... Until a great professional speaker designer, friend of mine, had-me listen to a big 2 ways enclosure he designed with spherical waves horn in plain wood and a JBL 2" driver.
I was so impressed by the natural and quality of the reproduction and measurements that I immediately ordered a pair to replace the ones of my recording studio and one for my own use.

Few years after that, I heard about the work of Le Cleac’h (RIP) and immediately arranged an appointment to listen to one of his prototype. With the same conclusion.

Here, a thread about those horns:
Jean Michel on LeCleac'h horns

Not like I dislike cones speakers, I use lot of them, as everybody else. But, if I insist on this, it is because some can loose a great opportunity to get a top speaker set because he could have the same prejudices that I had about horns and the way they sound. Of course, they are not WAF and good horns are too expensive.

Yes, the nasal sound and the multiple resonances and accidents that they caused in the past with PA systems can disappear with a good design and they can offer a precise listening experience, dynamic, very transparent, while offering a high efficiency, high levels of listening without audible distortions in a home listening room.
They are unparalleled on percussion attacks, detach micro details like nail attack or picks on guitar strings, without comparison on the sections of brass or pianos, while they can stay aerial on voices. Not to forget they are more forgiving about the quality of the acoustic treatment of the listening rooms.

Only my two cents, I do not sell anything.

More personal stories:
I was lucky enough to attend the SynAudCon in Nashville in the early 90s where Don Davis, Gene Patronis and engineers from Eminence gave speaker driver information including horn research and data. ALL agreed the data shows the lowest distortion horn shape to be spherical, with a linear taper from throat to mouth. Other tapers can give lower cutoff frequencies at the expense of distortion, and other shapes can give controlled directivity, also at the expense of distortion as well as lobing. As in everything engineering, the end design parameters determine where the compromises are made.

I was the analog engineer on a team of engineers which first put horns in cars (story here), and we learned a lot iterating different designs. The resulting horn was a custom JBL 4" driver with a special diaphragm coupled to a 2" throat relatively short horn, the mouth of which we continued into the car by shaping the underside of the dash. We ended up winning IASCA quality contests until we were banned from competition because no one else felt they had a chance. Although this horn was not spherical, proximity of the driver to the human driver and passenger gave outstanding dynamics and SPL at a relatively low powers and diaphragm excursion minimizing potential distortion.

As a side note relating to a current (pun intended) discussion; this was the first application of the eq'ed LF current servo I had designed and it really did give an improvement to the LF dynamics in that vehicle.

I concur with Tourney's statement regarding lack of compression of dynamics, especially at lower levels. Due to high throat pressures and resulting diaphragm interaction as well as potential air non-linearity, smaller horns especially begin to show stress at higher levels. When it comes to speakers, one size does not fit all, but for acoustic music and chamber classical, I think the best dynamic presentation I have heard would be from JBL 2" drivers in a full linear taper custom spherical horn.

Only my two cents, I do not sell anything either.
Howie
 
I think the best dynamic presentation I have heard would be from JBL 2" drivers in a full linear taper custom spherical horn.
I don't believe-it: It is exactly the one I use after long comparisons and measurements. :)
i use-it from 600 to 15000 (with no tweeter) and a simple passive network to linearize the response curve.

It is funny to see how close can be personal choices from people (not influenced by fashion) working in professional audio with probably very different cultures, working areas, different tastes, may-be. And reassuring to see that, at the end, trusting our ears can lead to so similar conclusions between thousands of various possible choices.

Thanks, Hhoyt, you made my day.
 
As well as us, when we try to make listening comparisons ;-)
The trade-off is to repeat the analysis with several short enough different samples, carefully chosen. Like we use to do ;-)

We have to rely on our ability to transfer information to long term memory.
It needs some time in the beginning to reach a state of mind like "awareness" but it is crucial to not concentrate first on the "whys" and "ifs" to let this transfer happen.

It is about getting everything together in a direct way; the sound quality and your emontional response to the reproduction.

As said before, if you´d forget about it within a couple of seconds (due to the mentioned memory problems) it wouldn´t matter in any case during the next replay.

What surprised-me in this article is the curve of the differences is descending. I was expecting a clear increase around 20KHz as an effect of the bric wall filters and the fact that one wave have less and less samples as frequency increase.

The diffmaker software is used for file comparison of the same format, bits and sampling rate.
When zeroing in on specifics, using short sequences of course helps to be effective.
 
Last edited:
I have not read the paper but sub-sample shifting does not mean creation of more samples. It can be done with phase rotation in the frequency domain. From a brief look at the output of the process huge FFT's are used possibly as big as the entire sample (?).

Bwaslo wrote that he uses 4-times oversampling internally when trying to compensate for uninteresting errors (as he called it) like sampling drift that leads to subsample shifts.
 
Which is not really what I have been talking about and pretty sure not what Jakob has been saying. We can't even agree what is 'good sound' on this discussion.

Absolutely (at least it seems so quite frequently), i wouldn´t say that it is the most important part to agree about the sound quality but to understand what the other means when talking/writing about his perception of sound/music.

At the end we could still diverge in our quality judgement - especially wrt multidimensional evaluations, due to intersubject differences - but we would certainly more often know that we are talking about the same things.
 
Disabled Account
Joined 2012
I don't believe-it: It is exactly the one I use after long comparisons and measurements. :)
i use-it from 600 to 15000 (with no tweeter) and a simple passive network to linearize the response curve.

It is funny to see how close can be personal choices from people (not influenced by fashion) working in professional audio with probably very different cultures, working areas, different tastes, may-be. And reassuring to see that, at the end, trusting our ears can lead to so similar conclusions between thousands of various possible choices.

Thanks, Hhoyt, you made my day.

I agree with both of you, also. My first HiFi speaker system was all JBL horns. Conical horn on midrange ...... I never forgot the dynamic range and low level detail. Now I am back there with JBL.... not a conical horn but an off-shoot -- same type driver... constant directivity, high sensitivity and with high dynamic range and low distortion.


THx-RNMarsh
 
Last edited:
I was the analog engineer on a team of engineers which first put horns in cars (story here), and we learned a lot iterating different designs.

You, Americans, are totally crazy about cars. I watch sometimes a TV program named "Car restauration" with a guy named "Chip Foose". No comment about the story telling, but what they do is just incredible and often so beautiful. Happily, they never talk about prices (let's change it for a big block, new transmissions and new modern suspensions) that could kill the poor man that i'm ;-)
And the painting !!!!
 
www.hifisonix.com
Joined 2003
Paid Member
The measured response in my system with M2 do NOT roll off.


-RNM

I think the reason they tip the response up towards the lower frequencies is to account for how they behave in a typical domestic setting - just a guess. So when you measure them in situ, you get a more flat response. I am not a speaker guy, so I may be talking rubbish here - someone feel free to correct me.
 
Sometimes we Americans grow up and realize our fathers were right it's just all tin and tires. then we can spend money on audio.
:)
When I asked my mom a record player for Christmas, she said, "What do you mean, something that turns and does not keep you warm?".
It explain this:
 

Attachments

  • laboratoire.gif
    laboratoire.gif
    80.4 KB · Views: 185
  • labotournesol.gif
    labotournesol.gif
    148.2 KB · Views: 184
Last edited:
Status
Not open for further replies.