Horn vs. Waveguide

Status
Not open for further replies.
Jump to Latest
#161
Earl,

As a pure egotist audiophile I don't bother if published studies demonstrate that "from a statistical point of view", phase distortion (or group delay XXX whatever you name it) is not audible for most part of the population.

That the same if other studies demonstrate the inaudibility of compression drivers distortion ;-)

As an individual, I can hear phase distortion and I can also hear distortion differences between compression drivers...

I can perfectly hear the subtle changes when I modify the time alignment of my system and also when I change my "Le Cléac'h" crossover to a Linkwitz-Riley 4th order. Always I prefer when the system is aligned in order to have the lowest phase distortion under 4kHz.

I am not a professional in audio, I have nothing to sell, but I can understand that if someone have to sell something which is not perfect (well does it exist?) he may want to help the basic customer to perform his purchase using statistical arguments... (eg: MP3...)

Best regards from Paris, France

Jean-Michel Le Cléac'h






gedlee said:



But you are doing this on a subsystem and NOT on the completed EQ'd system which will be completely different. I could see "some" merit to group delay of the completed system, but not at the transition points (ie. low end where there is a crossover) in a subsystem.

Doesn't it bother you that all studies of group delay have found it inaudible? Do you just write off this data? I am not in complete agreement with these previuos syudies either as I have found situations where GD types of effects are audible, but I would certainly never conclude that GD per see is a problem just because its there.
Click to expand...
 
#162
Jmmlc said:
Earl,

As a pure egotist audiophile I don't bother if published studies demonstrate that "from a statistical point of view", phase distortion (or group delay XXX whatever you name it) is not audible for most part of the population.

Best regards from Paris, France

Jean-Michel Le Cléac'h
Click to expand...

As a scientist I question everything and am dubiuos about results that cannot be duplicated across a wide variety of people and situations. How can you be so sure that you, as an individual, are correct and the rest of the world, who aren't so dumb either, are wrong? "As a pure egotist audiophile " I guess that I know the answer.
 
#163
Earl,

This is a philosophical question and as an individual assuming his own weakness(es) I confess that I am not sure to be correct about the ideas I have in audio, but I confess also that: "I don't bother".... and I am happy to see so many friends of mine sharing in the whole world some of those ideas ;-)

Best regards and have a good weekend.

Jean-Michel Le Cléac'h, Paris, France

(PS: tomorrow I give a conference about "axisymetrical horns" at the French association Melaudia: "melomans and audiophiles associated"...)


gedlee said:


As a scientist I question everything and am dubiuos about results that cannot be duplicated across a wide variety of people and situations. How can you be so sure that you, as an individual, are correct and the rest of the world, who aren't so dumb either, are wrong? "As a pure egotist audiophile " I guess that I know the answer.
Click to expand...
 
#164

Attachments

  • csd.gif
    csd.gif
    37 KB · Views: 824
#165
jzagaja said:
Here is comparison between ESP15, low and normal noise for waveguide. I can do Stockwell transform as well:

http://unizet.net/pub/S-Transformacja.pdf
Click to expand...

Thanks - the waveguide low noise is definately what I expect. I'll have to look at my algorithms and procedures in more detail if I am going to look at this low level stuff.

Also note what Jean Michael noted, that the sample rate is not the same in the two data sets.
 
#168
Hello,

When I measure 33 samples between 2 reference points on signal "waveguide"then I measure 22 samples on signal "waveguide_00", so if sampling frequency of signal "waveguide" is 48kHz then sampling frequency of signal "waveguide_00" will be around 72kHz.


Best regards from Paris ( France) and have a good weekend.

Jean-Michel Le Cléac'h


jzagaja said:


48k and 50k does not make any difference in CSD plot.
Click to expand...
 
#169
Jmmlc said:
Hello,

When I measure 33 samples between 2 reference points on signal "waveguide"then I measure 22 samples on signal "waveguide_00", so if sampling frequency of signal "waveguide" is 48kHz then sampling frequency of signal "waveguide_00" will be around 72kHz.


Best regards from Paris ( France) and have a good weekend.

Jean-Michel Le Cléac'h
Click to expand...


Well that doesn't sound right since I have never used 72 kHz. Only 50k, 48k and rarely 44.1k and 32k, but I highly doubt its either of the last two. How do you know the reference points were of the same event. These two files were generated in different facilities with different equipment on different systems (different serial numbers same design) and measured at different distances.
 
#170
jzagaja said:
Here is comparison between ESP15, low and normal noise for waveguide. I can do Stockwell transform as well:

http://unizet.net/pub/S-Transformacja.pdf
Click to expand...


I wasn't too impressed with that paper. A lot of pretty pictures. Incorrect claims about many things and nothing really very enlightening. Its really just a spectrogram.

His example about why two frequency response that are identical can sound different is ridiculous. He's mixing up phases of signals with phases of a frequency response. Ones a signal and the other a system. If the system frequency response is minimum phase, as it mostly is, then his "example" is meaningless.
 
#172
I was asked by a friend to help him with the sound system in his church. He brought over his Yamaha 15" 2-way with compression driver and diffraction horn. I didn't copy down the model but its in their current line.

I thought that people would be interested in this complete comparison of this Yamaha with the ESP15 since they are identical in layout (15" with compression driver) and size was comparable (ESP15 about 30% larger) and cost (ESP15 would be maybe 50% more).

You can see this data at http://www.gedlee.com/Papers.htm

My friend asked me why his female vocalists always sounded so harsh when they sounded really clean on my system (he is not an audiophile). I think that the data speakers for itself once again. Good data never lies - the truth can always be seen.
 
#177
gedlee said:



I shouldn't have to say this, but they all have foam. I don't have any waveguides without foam.
Click to expand...
Well, aside from the so called "noise", decay is significantly different between the two which is cannot be otherwise explained unless they are using different foam material or foam shape. I am assuming same wave guide shape and size. But if they are different sizes, then the story would be different.

By the way, I don't get offended if I know I am clearly wrong. So please feel free to point it out whenever you feel appropriate. I rather prefer anyone pointing these out than being wrong the rest of my life.😉
 
#178
Its quite funny to think of all those years we have been fighting with removing anything in front of the drivers
These days even commercial design dont use anything in front of drivers ... is it called grilles or frontcloth ?
And now you want us to place a big chunck of foam in front of a tweeter ... and further more listen offaxiss to a directive speaker ... well, I can only say good luck with that 😉

On the other hand I do remember that the late mr.Duelund beeing quite excited about a thin piece of toilet paper in front of a tweeter
Should sound better than a resistor ... but it doesnt look nice
Today some of us know the Duelund resistor(graphite) fore tweeter attenuation
 
#179
soongsc said:

I'm guessing the noise was from the foam. Since the horn does not have the foam, it's already low noise.
Click to expand...
I'm just trying to keep some order here vis-à-vis the topic.

The Gedlee waveguide CSD doesn't look good.

Gedlee provided another dataset for the the waveguide, from another source, postulating that there was a measurement noise issue.

That didn't look good, either.

Processed "low noise," it's more what Gedlee expects.

O.K., let's look at the original data for both the waveguide and "horn," similarly processed, and we can pretend a whole bunch of stuff about the comparative results.

If the "horn" CSD remains constant, then it's a stretch to blame the air conditioner, probably.... 😛
 
#180
Just to interrupt this lovely technical discussion for a bit -- and go back a few pages.

Duke & Earl were saying ...:sing: (Duke, Duke, Duke, Duke of Earl, Duke, Duke):sing:

Sorry, Duke and Earl were talking about the large toe-in they use - like 45 degs. Radical! So I tried it for a few days. I like it! (Mikey likes it!)

First listened to some old mono jazz tracks just for reference. Center image was larger and somewhat more stable. More weight to it. Definitely a different tonal balance - no surprise there - my system was EQ'd for on axis. That's easy to fix.

In stereo it sounds very good. A wider image, better placement of sounds in space. It didn't take long to forget about listening to the speakers and start listening to the music. (That Lionel Hampton guy is good!).

Added bonus. The mid bass peak in the room is down quite a bit, maybe 6dB. It's been a real pain, that peak. Never could EQ it out. Everything now sounds cleaner and better. I just put the baffles back to the normal position for casual listening (TV). Found I like them just a little flatter than 45, so YMMV.

Small baffle (17x35") One 15". One 8" FR driver. Will have to try it with horns, soon.


OK, back to the charts and graphs. I love sonographs, BTW - all those birdsong and speech graphs I studied as a kid. =)
 
Status
Not open for further replies.
Top Bottom