|
Posted by Daniel on
03-11-2018
|
Hi,
this is my first post. This forum seems to be the hotbed for Goto discussion of some sort but there is barely any data but only a swamp of assumptions. So I bought a pair of SG-370DX drivers with S600 horns and measured them. Here's the frequency sweep from 500Hz to 25kHz:
Impedance is basically a flat line at 14 Ohm.
Before doing the sweep I did a quick impedance check with my LCR meter at 100 Hz, 1k, 10k and 100k. What I heard at 100, 1k and 10k without the Goto horn sounded very, very clear and clean. Ok, every driver sounds clean and clear with a sine wave on it unless it's not damaged. But even under that condition the 370 stood out. I then screwed on the horn and listened to 1k and 10k sine waves from the LCR meter (a simple Agilent U1733C) and noticed that something was distinctly "off" compared to the sound without horn. It's hard to describe what it was, kind of like a ringing metal horn, old Altec without damping applied. The clean and clear was gone. The Goto horn is very well damped and doesn't ring at all.
I haven't listened to any music yet as I don't feel a passion to hack something together just now but will make an adapter to attach the driver to another horn (Autotech Iwata 300 and 600) and see what that gives.
My drivers come with no supporting data from the manufacturer, not even a nominal impedance on the label.
|
|
|
|
Posted by martinshorn on
03-12-2018
|
Hi Daniel, great initiative! More of that  Can you tell us more how u measured??Like loudness, distance, settings? I know that Fuzz likes to sweep too short (0.1 sec. logsweep) and allows many crappy windowing settings...Maybe a nearfield (close to mouth) measurement would allow to see a little less smoothened (48th oct, 1400ms rectangular or flattop window).Then also interesting would be a far field from 1-2 meter (3-6 feet) with a 1msec window (to see if the heights peaks disappear). And of course, 10 seconds sweep always.Interesting observation about the sweep sound. Its not to be neglected! Check the mouth of the horn, the mount, is it properly tight / stable / aligned / rigid ? When you take the driver off, nock the mouth, is there anything? I bet on a mouth issue. Shouldnt be what you got! I saw the horns (200 & 800) live myself, theyre of very impressive quality! Even better than most TAD/Fostex massive wooden horns ive seen. Cheers, Josh
|
|
|
|
Posted by Daniel on
03-12-2018
|
The distance in the above sweep was 70cm. I did another at 230cm which gave a very ugly graph, it looked like a sawtooth wave. I can post it later if there is interest.
Peak sound pressure level was 93 dBSPL. Sweep was set to 10 seconds.
Your guess about a mouth issue is spot on. I felt yesterday that it actually is loose.
|
|
|
|
Posted by martinshorn on
03-12-2018
|
|
Hi Daniel. Sorry i meant the throat, not the mouth. But i think you mean the same Just for anyone else reading this not to scratch their head. I heard that on cheap plastic horns the mouth often rings, is loose, or not perfectly centered and then causes lot of distortion. You can measure many percent difference.One friend just repeated the sweep with his hands choaking the horn-throat tight, wou it was a lot cleaner! About your 230cm measure, if you make a 1msec window its still so ugly?
|
|
|
|
Posted by Daniel on
03-12-2018
|
Here's the graph taken at 2,30 meters distance:
|
|
|
|
Posted by Daniel on
03-12-2018
|
With "1ms window" you mean the sweep duration? I left it at the standard 5ms. I'll tighten the throat and do another sweep with 5ms and 1ms to compare what's going on.
And maybe even listen to music....
|
|
|
|
Posted by martinshorn on
03-12-2018
|
|
Hi Daniel. I mean the window, after you sweep 10 seconds, and fuzz shows the IR. You can change the IR windowing on the preferences panel... like crop/cut the analyzed part of e.g. 1ms before the peak and then duration X behind the peak (here i asked for 1msec). That can exclude reflections on long distance recording of high frequencies (on the expense of resolution, so no smoothing anymore). You can also drag that crop with the mouse depending where your room actually starts reflecting (look when the second small spike after the peak appears, that one is to be cut off).I was just curious if the 7-8k Spike is the titanium breaking up, the old fashioned phase-plug or maybe an effect of nearfield-mouth measures. But it looks like too much the same even without window/smoothing change etc... No need to try! I can see its really there. Im surprised how amazing it actually sounded in Klaus Speth' installation, crossing this at 6.6k !!!!!!Reminds me of the Goto statement "you shall not measure, you shall listen". Which sounds like stupid marketing in the first place. But for second time (first was compression driver in bass) Goto seems to bend the physics successfully. Very funny!
|
|
|
|
Posted by Daniel on
03-12-2018
|
I read about Klaus Speth crossing at 6k6 and looking at the graph it seems like the absolut wrong thing to do. My first thought would be to cross at 5k first order and make use of the rolloff there.
I'll have time over the weekend to hook them up to my Driverack and listen and play around a bit.
|
|
|
|
Posted by Daniel on
03-12-2018
|
The sticker on the driver states a frequency response from 400Hz to 10kHz. Others say 18kHz. Looking at what's going on I wonder what the condition for these "10kHz" might be.
|
|
|
|
Posted by martinshorn on
03-12-2018
|
|
Yea. In some systems, mostly trivial / conservative direct radiators in small monitor size, respond to the normal interpretation of frequency graphs very well. With more exotic and uneven dispersing systems it seems to get more difficult. I had it couple of times that i was pleased a lot with accident passive XOs in listening, then later measured horrible graphs and kept scratching my head "how come?"... Sometimes i think horns are not always logical Or our interpretation logic of measures is still very limited. Anyhow, measuring right is a lifetime task.
|
|
|
|
Posted by Daniel on
03-12-2018
|
Did you ever compare frequency response and power response plots of the same speaker system? Of course it's easy to just do a quick sweep and do a FR plot but especially with horns a power response plot might be more telling and ease some of the head scratching?
|
|
|
|
Posted by martinshorn on
03-12-2018
|
ya that’s likely where the truth goes. I had it once with electrostatics. Theyre most bitchy in regards of correct positioning in the room as we all know. They’re a lifetime task to find DPOLS. I tried to move around and found huge difference. Some crap, some magic. Then later i made a big effort to linearize the power response via complex beemforming averaging of multiple positions without any windowing. Surprisingly now all positions sounded almost the same I didn’t try the same with horns yet but expect no surprise.
Anyhow, tell us your listening findings with the crossover. Im very curious. Also with filter up down steep slow. As a reminder, K. Speth uses 4th order (!). So first order might roll off nicely, but still leaves the 8k peak to a big extend. It actually would only lower it by 6dB. Likely more of it left than crossing higher and steeper. So to get rid of that, id go for 5k and 4th order. Or the first order requires a good notch filter!Anyhow, the one i heard was as sweet as you can(nt) imagine.
|
|
|
|
Posted by Daniel on
03-14-2018
|
I wonder anyways how Goto arrives at the frequency response they give on the driver label after plotting my 370. And also if the response of one labeled 400Hz to 18kHz is any different from my "400 to 10k" version. Do you by any chance know which one Klaus Speth uses?
|
|
|
|
Posted by martinshorn on
03-14-2018
|
http://tempuri.org/tempuri.html
Looks like 370 to me at the bottom there some specs (though the crossover is not absolutely correct, i asked Klaus)
|
|
|
|
Posted by martinshorn on
03-17-2018
|
|
Hi Daniel, is there more coming up, do you have more Goto Drivers? Id be really interested to see a SG16 graph Keep us posted with ur setup, findings, listenings...
|
|
|
|
Posted by Daniel on
03-17-2018
|
I'm quite busy over the weekend but I can do measurements of the SG-16 tweeter next week. I need to make adapters first but I want to measure the 370 on a Le Cléa'ch Iwata horn and I can compare that to a Yoshimura 3500G driver on the same horn. There's more on the shelf, I'll keep posting and give subjective impressions.
Baseline for subjective in my case: I don't like it when music is "coming out of the speakers". It needs to stand in the room, a cloud, detached from anything visual. I need dynamic impact. I've been playing in bands for about 20 years and I yet have to find a stereo and a recording that gives a drumbeat close to the real thing. Cymbals are mostly dissapointing as well. Getting a bit closer to the practice room and stage experience is what I'm after. And goosebumps! And butt shaking, dancing, singing along. If this comes along with a bit of distortion somewhere or a response graph that isn't perfectly flat so be it, no problem. It does become a problem when the distortion or dips/peaks bug me. Two or three weeks ago I listened to a system with a 7k peak, mostly room induced but also a speaker issue, the highs hurt my ears after three minutes, I couldn't listen to that.
|
|
|
|
Posted by Romy the Cat on
03-17-2018
|
 Daniel wrote: | | I need dynamic impact. I've been playing in bands for
about 20 years and I yet have to find a stereo and a recording that gives a
drumbeat close to the real thing. Cymbals are mostly dissapointing as well.
Getting a bit closer to the practice room and stage experience is what I'm
after. And goosebumps! |
|
Daniel, I think you set for yourself
very faulty objectives as so many people do in audio. With compression drivers
and horn-loading topology driven by Class A SET you will be able to get highest
topologically possible dyadic range. This is given and I do not see problem in
that. However, you should not be looking to imitate with your audio sound of
life performance. The imitation of life sound is the subject the that audio marketing
people are entrenching into minds of gullible audio public for over 100 years
but this is a very faulty premise and a huge gateway to nowhere. You will be
able to build much better playback when you step away from live sound imitation
and go for similarity of your own sound consumption instead of similarity of playback
sound expression. It takes years to understand it but when you do you will be much happier, gratified and beneficial in your audio journey.
|
|
|
|
Posted by Daniel on
03-17-2018
|
Read again what I wrote. Key element "Getting a bit closer ...". Most people writing about "live" on forums have no idea what a real instrument sounds like and it becomes clear that their reference for sound quality is their mother's kitchen radio. I just want to get the most out of playback and my refernce for certain sounds is creating them myself.
|
|
|
|
Posted by Romy the Cat on
03-17-2018
|
| Daniel wrote: | | Read again what I wrote. Key element "Getting a bit closer ...". Most people writing about "live" on forums have no idea what a real instrument sounds like and it becomes clear that their reference for sound quality is their mother's kitchen radio. I just want to get the most out of playback and my refernce for certain sounds is creating them myself. |
|
Well, people do have ideas
what a real instrument sounds, it is not the point. The point is that live
sound is not a reference for playback. It is reference and it is not. Live
sound has own properties and own expressive forms. Reproduced sound has own properties
and very own expressive forms. The common denominator between them is human perception
and this perception might be moderated in many different dimensions. Now pretend
that we formulate some kind of matrix that would encompass ALL perceptional dimensions
in absolutely perfect way. Furthermore, we build a playback that would deliver an
absolutely analogues impact to listener that live sound does. So, you would be
very surprised to learn that Sound from live event and sound from playback will
be different. This reason why I say that when audio people are trying to make
this playback to “getting a bit closer” to live sound then they deceive them
and undertake objective that are mistaken to begin with.
|
|
|
|
Posted by Daniel on
03-17-2018
|
This is a bit of a childish "I'm right! - No, I'm right! - No, I'm right!" 'discussion'. I know that there are differences between acouctic live, amplified live and playback of a recording and never stated there wasn't or one could ever be mistaken for the other.
But I don't think we need to argue that certain systems fail to transport certain aspects of musical presentation. And some others fail a little less.
|
|