Andrew Jones’s new speaker brand - Jones and Cerreta

[Curvature]

What is the advantage of using a field coil over using a permanent magnet in a speaker driver?

I think the primary benefit would be reduction of hysteresis distortion and perhaps more linear Bl(x) as the driver moves.

 

[Duke]

At a recent DIY event there was an active speaker that was switched between an LR2 xo and LR4. They sounded different. There is no scientific/objective "better" speaker in this case. The only way to make a decision is to listen with human ears.

I slipped in a crossover slope change without my beta-testers knowing about it. One version had better measured frequency response, and the other had better time-domain behavior but was less smooth.

My beta-testers all detected the change and all preferred the version with the improved time-domain behavior. This information has since informed my approach to system design, to enable the configuration with the better time-domain behavior without having to make a trade-off in frequency response smoothness.

 

[RickS]

A large number of DIYers do this too, you just need to be fortunate enough to have access to a variety of rooms to work in.

At a recent DIY event there was an active speaker that was switched between an LR2 xo and LR4. They sounded different. There is no scientific/objective "better" speaker in this case. The only way to make a decision is to listen with human ears.

A lot to unpack here and seems best in in a more appropriate thread (or a new one). Thanks!

I slipped in a crossover slope change without my beta-testers knowing about it. One version had better measured frequency response, and the other had better time-domain behavior but was less smooth.

My beta-testers all detected the change and all preferred the version with the improved time-domain behavior. This information has since informed my approach to system design, to enable the configuration with the better time-domain behavior without having to make a trade-off in frequency response smoothness.

Ditto.

 

[SigbergAudio]

I don't see any need for this at all.

At the end of the presentation Jones talked about judging a competition where the test songs are sent to the contestants in advance, and the contestants inevitably tailored their speakers for those songs for the win, since judging was by listening.

Speakers are general-purpose devices at the end of the day. I don't think it makes sense to compromise that by manipulating the FR in some way, on or off axis, based on a limited set of experiences or considerations, even assuming that the designer is good at controlling the circumstances of listening and coming to consistent judgments.

I'm not tailoring them to a set of rooms, I'm testing them in a number of rooms and positions in order to see what happens. Both listening and measuring here as well.

Anechoic data is useful, but it doesn't tell me exactly what happens when someone has to put it 20 cm from a wall (which we design for). Our speakers also ship with presets for boundary and room compensation. The easiest way to arrive at sensible defaults is to try them in real rooms.

 

[RickS]

This thread really heated up over the last day or so. I have issued a couple of warnings to help attempt to keep it more focused.

Some content may be relocated to more appropriate threads. Please try to avoid topic drift. Thanks!

 

[PGansz]

People should demand that only car performance figures be carried out in simulation and say "screw it" when a company's test driver provides feedback and performance direction.

For those of you who don't know, some racers like understeer, some prefer oversteer, some like neutral. Same concept. The designers give feedback to help steer engineering during the preproduction phase.

Nothing different here. There's too much hostility for such an innocent hobby. Save it for the snake oil product pushers. Not respected engineers and successful (objective measurement) product designers.

 

[amirm]

I slipped in a crossover slope change without my beta-testers knowing about it. One version had better measured frequency response, and the other had better time-domain behavior but was less smooth.

My beta-testers all detected the change and all preferred the version with the improved time-domain behavior.

They would "know it" when you send a new speaker to them. That is, they would expect a change when there could be none. Such testing needs to be done without any such knowledge on their part.

Reminds me of when we did a demo or JBL M2 to local audiophiles. A number of vocal ones kept complaining about the sound being totally wrong. Frustrated, my designer dialed in an inaudible 0.2 dB EQ and all of a sudden, they all claimed the problem was fixed!

 

[amirm]

People should demand that only car performance figures be carried out in simulation and say "screw it" when a company's test driver provides feedback and performance direction.

Such companies use measurements for 0 to 60 times instead of asking said driver how fast he went. That is the case here with tonality.

 

[amirm]

I'm not tailoring them to a set of rooms, I'm testing them in a number of rooms and positions in order to see what happens. Both listening and measuring here as well.

Anechoic data is useful, but it doesn't tell me exactly what happens when someone has to put it 20 cm from a wall (which we design for). Our speakers also ship with presets for boundary and room compensation. The easiest way to arrive at sensible defaults is to try them in real rooms.

Such compensation better be programmable as I am pretty sure that you can't control or predict other aspects of said room.

 

[PGansz]

Such companies use measurements for 0 to 60 times instead of asking said driver how fast he went. That is the case here with tonality.

That's correct, tested by a real person. Who provides feedback on the test. If the rear axle tram lines, squats too heavily, or induces vibration the real person reports it, regardless of the 0-60 time.

Tests have to be performed by humans with human feedback regardless of measurement.

 

[amirm]

A speaker designer listening to his own speaker for hundreds or thousands of hours while developing it, is however a pretty different situation than a guy walking into a showroom and somehow preferring Zu Audio over Kef.

Said designer is a perfect candidate for bias. He makes a technical change and expect a fix. And a fix he will "hear" whether it is real or not. Best to get those guys in the showroom and subject them to controlled testing. After all, he is the one that is the customer, not the designer!

Look, there is nothing wrong with the theories designers have about what could be wrong with the sound. The asks is to verify that properly. You can do it with measurements, or controlled testing. If that many hours has gone into the design, you owe it to yourself and not gamble with the design being wrong. Perform high precision measurement and don't be dismissive of what it shows. Then combine that with testing it in a group of listeners, comparing it to its competitors. If you can't do that latter then better trust the measurements.

 

[PGansz]

Said designer is a perfect candidate for bias. He makes a technical change and expect a fix. And a fix he will "hear" whether it is real or not. Best to get those guys in the showroom and subject them to controlled testing. After all, he is the one that is the customer, not the designer!

Look, there is nothing wrong with the theories designers have about what could be wrong with the sound. The asks is to verify that properly. You can do it with measurements, or controlled testing. If that many hours has gone into the design, you owe it to yourself and not gamble with the design being wrong. Perform high precision measurement and don't be dismissive of what it shows. Then combine that with testing it in a group of listeners, comparing it to its competitors. If you can't do that latter then better trust the measurements.

Andrew even mentioned this in the discussion. As did Erin. It's ludicrous that they're being ripped on for the practice, even when they acknowledge they're aware of potential flaws in the practice, so do what they can to weed them out. That's called good faith. A rarity these days. Let's celebrate that.

 

[amirm]

That's correct, tested by a real person.

No. The human was just a tool to push the accelerator. The real person did not do any measurements. A speedometer and timer measured the performance. The human could have made a mistake and hence the reason these tests are repeated to get some kind of average. They are the weak link. And at any rate, how he drove doesn't apply to how I drive.

 

[kemmler3D]

They would "know it" when you send a new speaker to them. That is, they would expect a change when there could be none. Such testing needs to be done without any such knowledge on their part.

Reminds me of when we did a demo or JBL M2 to local audiophiles. A number of vocal ones kept complaining about the sound being totally wrong. Frustrated, my designer dialed in an inaudible 0.2 dB EQ and all of a sudden, they all claimed the problem was fixed!

The flip side of this, one time I shipped a firmware update for some Bluetooth headphones. I sent a follow-up survey asking people if they liked the update, and included a question about sound quality, which had not changed even 0.1%. There is an onboard EQ but I didn't touch it.

Guess what, most people said the sound improved, many said it improved a lot. I think this shows it's basically impossible to do a valid sighted test. Marketing people can create night-and-day differences out of thin air just by talking or even sending emails. Who needs engineers when we have placebo effect?

@Duke 's experiment with the crossover is way more interesting.

 

[amirm]

Andrew even mentioned this in the discussion. As did Erin. It's ludicrous that they're being ripped on for the practice, even when they acknowledge they're aware of potential flaws in the practice, so do what they can to weed them out. That's called good faith. A rarity these days. Let's celebrate that.

I don't care what was mentioned in that video. Unless I see clear evidence of AJ's speakers putting measurements front and center, and controlled listening tests performed, I am going to criticize his approach. To wit, I don't hear him say such things in his presentation at shows.

There is no question in my mind that AJ's designs would improve if he followed these strict guidelines. He doesn't want to go there as it lowers what he is good at: selling his reputation as the reason to hire him to design. And of course, his superb skills at presenting his products at shows.

 

[PGansz]

I don't care what was mentioned in that video. Unless I see clear evidence of AJ's speakers putting measurements front and center, and controlled listening tests performed, I am going to criticize his approach. To wit, I don't hear him say such things in his presentation at shows.

There is no question in my mind that AJ's designs would improve if he followed these strict guidelines. He doesn't want to go there as it lowers what he is good at: selling his reputation as the reason to hire him to design. And of course, his superb skills at presenting his products at shows.

This seems to be way too personal for you man. I can't get on board with this level shade being thrown. Not that you care, but for real, it's too much for an innocent panel discussion at a hobbyist event.

You're still missing the point about the test driver. If the human wasn't needed, no company would use them. All cars are tested and driven by humans for months to provide feedback that the expected results are confirmed.

If you just needed an accelerator drop, surely they'd use a hydraulically pressurized system to stomp or pull the accelerator and brakes.

 

[amirm]

You're still missing the point about the test driver. If the human wasn't needed, no company would use them.

They use them out of necessity since automation is not quite there. They would be dialed out soon enough. EV cars are almost there in that they have massive amount of power and hence, are automatically traction limited in their 0 to 60 times. Averages therefore are very close to each other and no skill is required to measure them.

And before you to go handling and such, we don't ride our speakers that way. They are supposed to be transparent to the source and that is it. Decades of research shows us that. And so does the need for standardization. Heaven knows not every stick shift in a car doesn't have its own directions to operate.

 

[amirm]

Where I think listening has a solid place is assessment of distortion and playback capability. Focus your energy here instead of messing with tonality. We can fix tonality. But it is hard to extract more linear response in a broken design.

 

[SigbergAudio]

Such compensation better be programmable as I am pretty sure that you can't control or predict other aspects of said room.

It is. There are built-in tilt filters (three different ones) that can be selected by the press of a button. As an example for one model the alternatives are no tilt, 1.5dB tilt or 3dB tilt (increased bass). If this isn't exactly what works, they can be adjusted by the customer (so that the button now adjusts to -2dB and +5dB or whatever else).
In addition, there is a 9-band PEQ.

 

[SigbergAudio]

Said designer is a perfect candidate for bias. He makes a technical change and expect a fix. And a fix he will "hear" whether it is real or not. Best to get those guys in the showroom and subject them to controlled testing. After all, he is the one that is the customer, not the designer!

Look, there is nothing wrong with the theories designers have about what could be wrong with the sound. The asks is to verify that properly. You can do it with measurements, or controlled testing. If that many hours has gone into the design, you owe it to yourself and not gamble with the design being wrong. Perform high precision measurement and don't be dismissive of what it shows. Then combine that with testing it in a group of listeners, comparing it to its competitors. If you can't do that latter then better trust the measurements.

Let's say I make 100 changes to a speaker during development. A controlled blind test of each change is not realistic. But to measure each change is, and it's what I do. And listen to it too. It's an iterative process. And of course many times along the way others listen as well. So it pretty much matches what you're saying. Listen to identify possible issues to inspect, not to conclude based on listening alone.

Subject to bias isn't the same as useless. If I hear something and can't find it in the measurements, I don't just ignore it and assume I'm right. I look harder and/or bring in additional people.