ProSoundWeb Community

Please login or register.

Login with username, password and session length
Advanced search  

Pages: [1] 2 3 4   Go Down

Author Topic: QSC K10 tweeter polarity  (Read 17527 times)

Rich Grisier

  • Full Member
  • ***
  • Offline Offline
  • Posts: 195
  • Erie, PA
QSC K10 tweeter polarity
« on: June 14, 2011, 10:48:06 pm »

I read recently HERE that the sound quality of the QSC K10 can be improved by changing the polatiry on tweeter.

I have a pair of K10 and thought they've always sounded great as is.  Just wondering if anyone has done this and can confirm that the sound quality improves.
Logged

Tim Perry

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1251
  • Utica-Rome NY
Re: QSC K10 tweeter polarity
« Reply #1 on: June 14, 2011, 11:36:48 pm »

Remember rule number one: if it isn't broke, don't fix it.

The post you referred to specifically mentions the use of a guitar amp "modeler" which indicates to me that it in essence is being used as a guitar amp.

For some reason most guitar amps don't have a horn or tweeter therefore is is unsurprising that doing something to lessen parts of the HF might sound better to a guitarist.

(multiple k10 owner and it will be a long time before I give them up)
Logged

Dennis Awrey

  • Newbie
  • *
  • Offline Offline
  • Posts: 37
Re: QSC K10 tweeter polarity
« Reply #2 on: June 15, 2011, 12:02:46 am »

I read recently HERE that the sound quality of the QSC K10 can be improved by changing the polatiry on tweeter.

I have a pair of K10 and thought they've always sounded great as is.  Just wondering if anyone has done this and can confirm that the sound quality improves.

A lot of various manufacturer's speakers are designed with a component's polarity reversed depending on what crossover filters were selected....reverse polarity does not necessarily mean "out of phase". If you think you know more than QSC engineers, by all means take your K10 apart and rewire it.
Logged

Jay Barracato

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 2025
  • Solomons, MD
Re: QSC K10 tweeter polarity
« Reply #3 on: June 15, 2011, 06:49:15 am »

I read recently HERE that the sound quality of the QSC K10 can be improved by changing the polatiry on tweeter.

I have a pair of K10 and thought they've always sounded great as is.  Just wondering if anyone has done this and can confirm that the sound quality improves.

I don't see anything in that post that would make me want to change things. Notice Bob Lee's response that you should MEASURE the response to confirm the correct polarity, rather than rely on a subjective listening test that you do not know the source of or the conditions.
Logged
Jay Barracato

Rich Grisier

  • Full Member
  • ***
  • Offline Offline
  • Posts: 195
  • Erie, PA
Re: QSC K10 tweeter polarity
« Reply #4 on: June 15, 2011, 05:24:05 pm »

HERE'S a link containing further explanation and plots to justify the claim.
Logged

Jay Barracato

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 2025
  • Solomons, MD
Re: QSC K10 tweeter polarity
« Reply #5 on: June 15, 2011, 06:18:29 pm »

HERE'S a link containing further explanation and plots to justify the claim.

I registered and looked at the graphs. I still don't see any evidence that this is an improvement. As a matter of fact the frequency response of the UNMODIFIED box looks better to me through the crossover region. Without a coherence trace and more details about the measurement setup, I can't say anything about the quality of the measurement.

The way I read the phase trace is that in the unmodified box (even if the tweeter is an opposite polarity to the mids) is that the slope of combined phase is smooth through the crossover. That is what you want. Flipping the polarity of the tweeter does move the 180 wrap point to match with the crossover but that is just a function of the way the data is displayed, it doesn't have any meaning as far as I know.

I am not a box designer and many people have more experience than me at using these measuring tools, but I remain whole heartedly unconvinced. It looks to me like someone believes that having a corresponding polarity between drivers in a multibox system is more important than the combined acoustic response. I think I would stick with the QSC engineers on this one.

I just finished school today, maybe I will have time to grab some measurements before heading out on the road for the summer.
Logged
Jay Barracato

Jay Mitchell

  • Newbie
  • *
  • Offline Offline
  • Posts: 17
Re: QSC K10 tweeter polarity
« Reply #6 on: June 15, 2011, 07:14:31 pm »

I registered and looked at the graphs. I still don't see any evidence that this is an improvement. As a matter of fact the frequency response of the UNMODIFIED box looks better to me through the crossover region.

Lemme make sure I understand you. Are you saying you believe this:
http://forum.fractalaudio.com/attachments/amps-cabs/5184d1308169884-qcs-k10-tips-overlaid-responses-stock.jpg looks "better" than this:
http://forum.fractalaudio.com/attachments/amps-cabs/5185d1308169930-qcs-k10-tips-overlaid-responses-tweet-polarity.jpg ?

Quote
Without a coherence trace
It appears that you are unaware of TDS (Time Delay Spectrometry). With TDS, there is no need of a (misnamed, BTW) "coherence trace."  The signal is completely "coherent," which, in SMAART-speak, simply means that it contains sufficient energy in every measured frequency range that the indicated response does not include noise. Very high S/N is an intrinsic property of swept-sine measurements. No "coherence trace" is required. IOW, the measurements are all valid and accurate.

Quote
and more details about the measurement setup,
Techron TEF-20 analyzer, B&K 4007 test mic, mic distance of approximately 1.5 meters, frequency resolution and data smoothing as indicated above the chart.

Quote
I can't say anything about the quality of the measurement.
I can. I've been doing loudspeaker testing, measurement, R&D, and instruction for the past 28 years. The data is good. Not only that, the "combined acoustic response" - which includes transient response - is bettter with the tweeter in polarity.
Logged

Jay Barracato

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 2025
  • Solomons, MD
Re: QSC K10 tweeter polarity
« Reply #7 on: June 15, 2011, 07:39:13 pm »

Lemme make sure I understand you. Are you saying you believe this:
http://forum.fractalaudio.com/attachments/amps-cabs/5184d1308169884-qcs-k10-tips-overlaid-responses-stock.jpg looks "better" than this:
http://forum.fractalaudio.com/attachments/amps-cabs/5185d1308169930-qcs-k10-tips-overlaid-responses-tweet-polarity.jpg ?
It appears that you are unaware of TDS (Time Delay Spectrometry). With TDS, there is no need of a (misnamed, BTW) "coherence trace."  The signal is completely "coherent," which, in SMAART-speak, simply means that it contains sufficient energy in every measured frequency range that the indicated response does not include noise. Very high S/N is an intrinsic property of swept-sine measurements. No "coherence trace" is required. IOW, the measurements are all valid and accurate.
Techron TEF-20 analyzer, B&K 4007 test mic, mic distance of approximately 1.5 meters, frequency resolution and data smoothing as indicated above the chart.
I can. I've been doing loudspeaker testing, measurement, R&D, and instruction for the past 28 years. The data is good. Not only that, the "combined acoustic response" - which includes transient response - is bettter with the tweeter in polarity.

We don't mind a technical disagreement around here that is backed by data, and you have convinced me we are looking at valid data, but I am not sure I understand your interpretation of that data.

I was not sure that I had correctly interpreted your labels, but yes it looks to me like the factory setting on axis is a lot smoother through the crossover, while the polarity flipped box has a large dip. The polarity flipped box looks to have the best response in 10-20 degrees off axis.

I think this response difference was clearer in the first two graphs that you posted on the other website. I still don't see evidence that the phase response is better.

Usually, in SMAART which is what I use, if I see a large dip in response, I would look to the coherence trace to try and determine the possible causes. As I said I am not a speaker designer but instead measure systems in the field where they interact with their surroundings.

Since the K10 is nominally a 90 degree box, at FOH or  even wedge distances, most listeners can be considered on axis, which appears to be better in the stock setting. A measurement at 10 m might be more indicative of what this box is really doing.

If your users are using the box differently from the way it was designed, like standing 2 feet from it and off axis, like guitar players often stand at their amps, then you can hardly blame the designers for that.
Logged
Jay Barracato

Jay Mitchell

  • Newbie
  • *
  • Offline Offline
  • Posts: 17
Re: QSC K10 tweeter polarity
« Reply #8 on: June 15, 2011, 08:14:33 pm »

you have convinced me we are looking at valid data,
Good.
Quote
I was not sure that I had correctly interpreted your labels, but yes it looks to me like the factory setting on axis is a lot smoother through the crossover,
As I pointed out, the definition of "on axis" is itself unclear with a speaker such as this. It is most common to use the HF horn axis for that definition. In the case of the K10, there is a pronounced amplitude notch at crossover on the axis of the HF horn with the stock reverse HF polarity. I was forced to identify a position in which the notch was not present, and that position did not correspond to any widely-accepted definition of "on axis."

Quote
while the polarity flipped box has a large dip.
See the overlaid data. That dip disappears within a few degrees of that position, and a dip appears within a few degrees with the stock reverse HF polarity.

Quote
The polarity flipped box looks to have the best response in 10-20 degrees off axis.
Yes. That is true in both directions, above and below axis. The response is flatter over a wider range of listening positions. Pretty much what you're looking for in sound reinforcement applications. Or what you should be looking for.

Quote
I still don't see evidence that the phase response is better.
The notion of "smoothness" of phase response is off base. What matters is the ability of a loudspeaker to preserve transient information. Reversing the polarity of one of the drivers always degrades this ability.

Quote
As I said I am not a speaker designer
I am.

Quote
but instead measure systems in the field where they interact with their surroundings.
I do that, too.
Quote
Since the K10 is nominally a 90 degree box, at FOH or  even wedge distances, most listeners can be considered on axis
Incorrect. "On axis" is, at best, a single plane. Most listeners will lie outside that plane. Many listeners will lie between 10-30 degrees off axis (the K10's coverage is 90 degrees in both horizontal and vertical), which is where the  amplitude notch is worst with the out-of-polarity HF.

Quote
A measurement at 10 m might be more indicative of what this box is really doing.
There will be no change in the behavior disclosed in my data. I have years of experience in correlating the two regimes, and I am well aware of all the considerations involved.

Quote
If your users
"My users" use products manufactured by my company. Those products are profoundly different from this type of device, and they have none of the directional anomalies in evidence here. I took the data as a favor for a friend.
« Last Edit: June 15, 2011, 09:10:34 pm by Jay Mitchell »
Logged

Ivan Beaver

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 9091
  • Atlanta GA
Re: QSC K10 tweeter polarity
« Reply #9 on: June 15, 2011, 09:42:12 pm »

We don't mind a technical disagreement around here that is backed by data, and you have convinced me we are looking at valid data, but I am not sure I understand your interpretation of that data.

I was not sure that I had correctly interpreted your labels, but yes it looks to me like the factory setting on axis is a lot smoother through the crossover, while the polarity flipped box has a large dip. The polarity flipped box looks to have the best response in 10-20 degrees off axis.

I think this response difference was clearer in the first two graphs that you posted on the other website. I still don't see evidence that the phase response is better.

Usually, in SMAART which is what I use, if I see a large dip in response, I would look to the coherence trace to try and determine the possible causes. As I said I am not a speaker designer but instead measure systems in the field where they interact with their surroundings.

Since the K10 is nominally a 90 degree box, at FOH or  even wedge distances, most listeners can be considered on axis, which appears to be better in the stock setting. A measurement at 10 m might be more indicative of what this box is really doing.

If your users are using the box differently from the way it was designed, like standing 2 feet from it and off axis, like guitar players often stand at their amps, then you can hardly blame the designers for that.
I am not on that site-so therefore cannot see the graphs to offer my own opinion.

HOWEVER-I know Jay-and he knows of what he talks and does.

I would not question his measurements.  He is NOT a beginner.

I really don't think he would post such a statement-without a through measurement.

ALSO-I do wonder how the driver got wired out of polarity.  It could be a manufacturing error (it happens) where a drawing got messed up and the assemblers just wire it up as they are told.

There are many loudspeaker designs that have a phase shift around crossover.  But if flipping the polarity of a driver can make it better (looking at both amplitude AND phase), then that should be the best "position".  But there may be something that is being missed here-hard to say.
Logged
A complex question is easily answered by a simple-easy to understand WRONG answer!

Ivan Beaver
Danley Sound Labs

PHYSICS- NOT FADS!

ProSoundWeb Community

Re: QSC K10 tweeter polarity
« Reply #9 on: June 15, 2011, 09:42:12 pm »


Pages: [1] 2 3 4   Go Up
 



Page created in 0.029 seconds with 23 queries.