Virtual SMAART training: Tell the crossover frequency and type

Re: Virtual SMAART training: Tell the crossover frequency and type

Jason Lavoie said:
the phase only rolls on the R traces. the L traces go a bit then come back. I've been puzzling over this, but given your take on it, maybe the mic is in front of the left speaker, putting it off-axis only to the right speaker?

Jason
Click to expand...

the transition zone around 2k in the phase diagramm is probably an artifact of smoothing and FPPO , multi time window or however it is called in SMAART, without these tools there should be some wild phase jumps,
so I think it doesn`t matter if it goes up or down

Uwe
 
Re: Virtual SMAART training: Tell the crossover frequency and type

Uwe Riemer said:
the transition zone around 2k in the phase diagramm is probably an artifact of smoothing and FPPO , multi time window or however it is called in SMAART, without these tools there should be some wild phase jumps,
so I think it doesn`t matter if it goes up or down

Uwe
Click to expand...

I don't care which way it goes, but if right flips by 180deg and left doesn't, that to me looks important.

*edit* Hmm.. 360deg now that I look at it..
so are you saying that because of the smoothing we may be missing some junk that might make them look more similar?

Jason
 
Re: Virtual SMAART training: Tell the crossover frequency and type

Phil Graham said:
Hello everyone,
View attachment 2149
Attached is a SMAART screenshot from yesterday's small gig, before I applied any equalization. I had time for four quick measurements of each side of the PA. These were taken with the mic on a straight stand at audience head level. The four traces shown are the L and R averages, respectively, both with and without coherence weighting. Because the box's crossover point and slope is unusually clear from these measurements, I thought it would be a good learning experience to post the measurements, and let people take a stab at figuring out these values in the data. I scaled the image down to 1280x800, but it is still fairly large, so click on the thumbnail to view full size.

Again, looking for the XO frequency, order, and type. Also how you determined these values.

Place your bets! And those who know immediately, wait a bit before chiming in.
Click to expand...

... rewired the HF...
 

Attachments

  • QSC K.jpg
    QSC K.jpg
    269.1 KB · Views: 0
Last edited:
Re: Virtual SMAART training: Tell the crossover frequency and type

All,

I've been busy, and see that there's been a good amount of chiming on this over the past couple days. There's no way for me to respond to every individual reply, but over the next week I'll try to make a few posts to unpack what this measurement is telling us, and the reason(s) I posted four average measurements and not merely one.
 
Re: Virtual SMAART training: Tell the crossover frequency and type

Phil Graham said:
All,

I've been busy, and see that there's been a good amount of chiming on this over the past couple days. There's no way for me to respond to every individual reply, but over the next week I'll try to make a few posts to unpack what this measurement is telling us, and the reason(s) I posted four average measurements and not merely one.
Click to expand...

Just letting everyone know that I haven't abandoned this, but it may end up being NEXT week before I get to it. Being busy is good, but not for finishing threads.

-Phil
 
Virtual SMAART training: Explanation Slide 3

SMAART Explanation 3.jpg

Without knowing the nature of the electrical filters, one cannot say conclusively what fraction of the apparent delay is from the physical driver offset, and what fraction is from the filters' group delay effects. If you know the electrical XO filter topology, you can figure out what fraction of the delay is from the electrical filter. The remaining will be a physical offset, and of course that will change based on the measurement axis with respect to the loudspeaker drivers.

For the sake of the thread, I remain intentionally ignorant of any details of the electrical XO filters in this powered loudspeaker. This will limit what can be said, but it is reflective of what can be determined in situ.

More later on in the week...
 

Attachments

  • SMAART Explanation 3.jpg
    SMAART Explanation 3.jpg
    189.2 KB · Views: 0
Last edited:
Re: Virtual SMAART training: Tell the crossover frequency and type

Bennett Prescott said:
Is there any external equalization applied, and what were the settings on the back of the box?
Click to expand...

There was eventually external equalization applied with a DSP, but not in these measurements.

Gains were in the 12pm position, high pass was off. Those are the only settings to change.
 
Re: Virtual SMAART training: Explanation Slide 3

Phil Graham said:
QSC HPR152i.

1/side, no subwoofers, boomy room.
Click to expand...


Hi Phil,

What I don't understand is there is a step in the phase response of 180 degrees from +90 to -90 at the crossover (2K). There also seems to be a corresponding dip in the amplitude response where things aren’t summing too well.

I would have assumed that QSC would have matched the phase and required delay nicely through the crossover. The apparent difference in arrival times of the HF and LF drivers seemed too much if it was related to your measurement being a little off axis.

I must confess I wondered if the polarity of the HF driver was wrong, but it seemed unlikely on both boxes. I also wondered if it was a QSC K series that someone had reversed the HF driver polarity as has been discussed on some websites.

Peter
 
Re: Virtual SMAART training: Explanation Slide 3

Peter Morris said:
Hi Phil,

What I don't understand is there is a step in the phase response of 180 degrees from +90 to -90 at the crossover (2K). There also seems to be a corresponding dip in the amplitude response where things aren’t summing too well.

I would have assumed that QSC would have matched the phase and required delay nicely through the crossover. The apparent difference in arrival times of the HF and LF drivers seemed too much if it was related to your measurement being a little off axis.
Click to expand...

Patience Peter,

You're thinking too much, too fast. Some of this we can be knowledgeable about, some of it, not so much. I don't have individual bandpass measurements, so we can only ascertain so much.

This thread isn't targeted at you, the experienced user/designer, but rather those desperate to make any sense of what SMAART is telling them.

Armchair quarterback less, at least for now. When I eventually get to the end of the conclusions I am comfortable in making from the data, then we can ponder over what's missing.
 
Re: Virtual SMAART training: Tell the crossover frequency and type

LaWBcUmO


Hi Jason,

this is a 12"/1" measured on axis and off axis ( means below the 12" driver, path length difference compared to on axis is about 6cm )

here you can see what Systune displays without additional windowing

Uwe


Jason Lavoie said:
I don't care which way it goes, but if right flips by 180deg and left doesn't, that to me looks important.

*edit* Hmm.. 360deg now that I look at it..
so are you saying that because of the smoothing we may be missing some junk that might make them look more similar?

Jason
Click to expand...
 
Re: Virtual SMAART training: Tell the crossover frequency and type

Hey Uwe,
For fun, lets say that your two traces are supposed to be on axis left and right PA. Can you point out, how you would conclude that the green is off axis and not a faulty speaker or measurement error?
 
Re: Virtual SMAART training: Tell the crossover frequency and type

Rasmus Rosenberg said:
Hey Uwe,
For fun, lets say that your two traces are supposed to be on axis left and right PA. Can you point out, how you would conclude that the green is off axis and not a faulty speaker or measurement error?
Click to expand...

Hey Rasmus,

In German we say "Wer viel misst, misst viel Misst"
this is a pun, which I cannot translate to English, but the meaning is "measure a lot and you will measure a lot of shit"
on the other side measuring a lot really helps, you get to know your speakers, how they behave on axis/off axis in the vertical and horizontal plane
and if you made every mistake in very small field, you are an expert :)~:)~:smile:

Faulty speaker could mean a) polarity problem like others in the thread suspected or b) transducer still operating but not like it should ( half cooked or worn out )

for a) in my example polarity of the 12" seems to be the same in both phase diagramms, for polarity of the HF driver I would look at the impulse response

for b) I think listening to the speakers and compare the supposed left to right is the easiest way

The "off axis" part:
comparing the supposed left and right means trying to measure under the same conditions or as close as possible.
From "auto delay" or "peak" like it is called in Systune I can compare distance to left/right
that would leave in my example different downtilt, here moving the microphone up and compare the results in FR helps

important is:
at that microphone position, which produces the green trace, the speaker will not sound good,
and EQ will not improve that, but maybe changing downtilt, thereby moving this problem area elsewhere


Uwe
 
Re: Virtual SMAART training: Tell the crossover frequency and type

Phil my guess would be LR 2nd order at 2KHz. My method ignores the magnitude trace entirely. I'm not sure why I would want to know this. Haven't encountered too many impulse aligned speakers in the last decade.
 
Top Bottom