Hot takes

[Martijn Mensink]

Below is the "estimated in-room response" curve of the Dutch & Dutch 8c; I posted this in the 15c thread as well. There may be a tiny bit of dippage in the region where those vertical off-axis dips occur, but it's not much.

In contrast the woofer's 1.5 dB bump @ 400 Hz kinda sticks out like a sore thumb. So, according to this curve, THAT arguably very minor bump is of much greater audible significance than those dreadful-looking off-axis dips in the vertical response.

View attachment 530689

Hi Duke,

Thank you, that’s a fair point. You’re right to question the 400 Hz bump. During early tuning of the 8c, we saw a small dip around that frequency in our anechoic measurements and corrected it in DSP. In hindsight, we shouldn’t have, because it turned out to be caused by an anechoic chamber that apparently is not completely anechoic after all.

Erin’s NFS measurements were helpful because they showed the resulting bump clearly and independently. After that, I ran ground-plane measurements, confirmed it, and we fixed it in production soon after.

8c units with serial numbers below 2300 can have this slight bump around 400 Hz. It can be corrected with a peaking filter at 400 Hz, gain -1.3 dB, Q 4. I’ve tried it, and while the curve looks a little cleaner, I think the audible difference is probably quite small.

I agree with your broader point: this small deviation around 400 Hz may be more relevant than the dramatic-looking but narrow vertical off-axis dips, depending on how they integrate into the total radiated sound.

Martijn

 

[Duke]

Hi Duke,

Thank you, that’s a fair point. You’re right to question the 400 Hz bump. During early tuning of the 8c, we saw a small dip around that frequency in our anechoic measurements and corrected it in DSP. In hindsight, we shouldn’t have, because it turned out to be caused by an anechoic chamber that apparently is not completely anechoic after all.

Erin’s NFS measurements were helpful because they showed the resulting bump clearly and independently. After that, I ran ground-plane measurements, confirmed it, and we fixed it in production soon after.

8c units with serial numbers below 2300 can have this slight bump around 400 Hz. It can be corrected with a peaking filter at 400 Hz, gain -1.3 dB, Q 4. I’ve tried it, and while the curve looks a little cleaner, I think the audible difference is probably quite small.

I agree with your broader point: this small deviation around 400 Hz may be more relevant than the dramatic-looking but narrow vertical off-axis dips, depending on how they integrate into the total radiated sound.

Martijn

That small bump in the estimated in-room response would have gone absolutely unnoticed on most speakers because most speakers would have had other issues of far greater significance. And that bump no longer even exists!

My main point in my previous posts was that the vertical off-axis comb-filter dip, which looks alarming on paper, is effectively inconsequential unless you were at the wrong listening height because it disappears in the in-room response.

 

[Ropeburn]

Heres a single transistor amplifier with a thd of .2%. Shows the miracle of feedback.

Simplest possible 1-transistor power amplifier with THD <1%

Simplest possible 1-transistor power amplifier with THD

www.diyaudio.com

Hehe. Interesting thought exercise. Even put into practice! And then...

That's not even a proper headphone amplifier.

Overall a good example of why it isn't such a good idea in practice.

 

[Cbdb2]

Hehe. Interesting thought exercise. Even put into practice! And then...

View attachment 543251

That's not even a proper headphone amplifier.

Overall a good example of why it isn't such a good idea in practice.

The only reason the power is so low is because the designer didnt want to add large heat sinks. It uses a power transistor capable of dissapating 60 watts.

HE says "so it is a power amplifier, though with very low power, if we do not want to build it big with large heatsink."

"The circuit has extremely wide bandwidth, very fast transient response, does not care about complex load or shorted output, has still acceptable distortion. But, it has very low power, in case we do not want to build a massive heater"