Jump to content

Using REW to Determine Time Delays Between Drivers


Chris A

Recommended Posts

Try looking here for setting the plot limits and other settings: 

When you accomplish those view settings, then the spectrogram view will be useful to see impulse and decay response for your loudspeaker.

 

Chris

  • Thanks 1
Link to comment
Share on other sites

1 hour ago, Chris A said:

Try looking here for setting the plot limits and other settings: 

When you accomplish those view settings, then the spectrogram view will be useful to see impulse and decay response for your loudspeaker.

 

Chris

 

It's all about the settings. How *DO* your figure these things out? So do I have a 3.8ms delay? If so I want to apply that to the HF, correct? If I am right do we care about the zig zagging between the zero point on the left and farthest right or do we just care about the total time represented by zero and max distance right?

 

As usual, Thanks!

 

DelayGettingSomewhere.jpg.c58c9c4da973387fa3d68bce9d2809c9.jpg

 

DelayGettingSomewhere2.jpg.f9282921fef5d882ddcf637a1b91c506.jpg

 

Link to comment
Share on other sites

Very cool, all the same settings but 350 crossover point versus the 200 in the one two posts up. NOTE that the time has decreased from 3.81 to 3.27 ( well screen capture says 3.33) and it is almost at the 350 mark (331) where on the previous one it was very close to the 200 xo point (249).

 

Spectrogram350XO.jpg.421642e250161894512a353512287816.jpg

Link to comment
Share on other sites

Here is a before delay and after delay of my HF/LF sweep. That smoother 2nd line is what we are after, correct? What does the dotted line actually show? Would the ultimate speaker setup have just one dashed line straight up and down at 0ms? What is happening around 4K in the 2nd picture that is not there in the first?

 

1806227634_LFHF200BeforeDelay.jpg.99b30b4068a5a7d1f09d2896fbf4b95a.jpg

 

938944659_LFHF200After3dot812msDelay.jpg.82b2cae6de03a2473d1531f0dc7fcff1.jpg

Link to comment
Share on other sites

Yes, that's the objective: to eliminate the discontinuities in the peak energy curve, and generally, that will also minimize the group delay spikes above and below the center crossover frequency(ies). 

 

There is a way to further minimize the sweep to the right of the spectrogram tail (at low frequencies) using a Danley-style filters to look more like this:

 

Spectrogram limits and controls settings.JPG

 

...the group delay to look more like this:

 

TAD TD-4002 Jubilee vs. Danley SH-50 group delay response.jpg

 

...and the phase to look more like this (note the vertical scale):

 

TAD TD-4002 Jubilee vs. Danley SH-50 phase response.jpg

 

Chris

Link to comment
Share on other sites

  I think REW is incredible. Would love to try and align my LS ii. Need to make another donation to keep it running.

  Spent a month first sweeping the interface and evaluating issues. Then a couple amplifiers to see what they delivered. Found a 3 dB difference in one of my diy builds, plus distortion shift between channels. 

  Never played music during that month. 

  Back to listening now. I think I can spend a year diddling with the speakers. I might try the crossovers built into Pure Music. 64 bit filters with amplitude and time adjustments.

Link to comment
Share on other sites

  I think REW is incredible. Would love to try and align my LS ii. Need to make another donation to keep it running.

  Spent a month first sweeping the interface and evaluating issues. Then a couple amplifiers to see what they delivered. Found a 3 dB difference in one of my diy builds, plus distortion shift between channels. 

  Never played music during that month. 

  Back to listening now. I think I can spend a year diddling with the speakers. I might try the crossovers built into Pure Music. 64 bit filters with amplitude and time adjustments.

Link to comment
Share on other sites

On ‎9‎/‎14‎/‎2019 at 9:25 AM, Chris A said:

...the group delay to look more like this:

 

TAD TD-4002 Jubilee vs. Danley SH-50 group delay response.jpg

 

 

 

@Chris A Two questions

 

1 in the above Group Delay plot how did you get two sets of measurements? I can see how to Generate minimum phase and turn on/off the Excess group delay, but I can't seem to put two sweeps on the same screen for group delay like I can for "All SPL"

 

2 What is the flaw(s) in my thinking process below?

  • Started from scratch today, again, with single HF/LF 2-way in the center of the room (away from walls) with padding on walls/floor. Next step will be to move outside if necessary.
  • No PEQs, No delays, No XO points
  • Measure and flatten HF only with EQ wizard
  • Measure and flatten LF only with EQ wizard
  • Level gain between HF and LF based on individual measurements, SPL meter and ear
  • Set crossover points for HF and LF along with crossover type and slope
  • Take combined measurement of HF/LF, with High and Low pass crossover points and slopes all in place

This is what the measurement looks like for that last step with both the HF and LF set to a 350Hz crossover point and Bessel 6dB (I've got plans to bring down that 400-800 but sticking with EQ wizard settings for now). Also guessing that 400-800 range is in part due to 350 XO.

953571813_Combined350NoDelay.jpg.1ef7f439fef1449a836514f51cdf4c2e.jpg

 

 

Next I produced and look at the Spectrogram, it actually looked pretty decent (I think) without any delays. Am I correct that since peak energy is not jagged not too bad?

82911766_SpectrogramCombined350NoDelay.jpg.7f7cd1be95d19749f66f5f7c46236fe4.jpg

 

 

Next I generated the Excess Group Delay in the Group Delay window. Still not sure how to read it beyond the "gaps" between plot around XO point. I think I have an issue around 300Hz of about 3.5ms, is that correct? See cursor at 302Hz showing 18.58ms (orange) and ~22 (white). Is that an issue? If so since at/near XO point how do I know if it the HF or LF that is the offending driver?

 

Am I wrong about the issue at 302Hz and it is somewhere in the 115-250Hz region below?

1771741060_InkedGroupDelayCombined350NoDelay_LI.jpg.fdc275d83463de34214cf4e2b23743ec.jpg

 

 

I've got measurements with Spectrogram and Group Delay for both a 3.5ms delay on the HF and a 3.5ms delay on the LF since I can't seem to get it in my thick skull how to tell where the problem is on the GD screen and the Spectrogram does not have an obvious jagged like in previous posts. But I'll hold off posting my pictures of various graphs with delay set till we get some answers on the above.

 

I'm hoping my though process and Chris' comments will help others in particular @Ziggurat

 

Link to comment
Share on other sites

1) Look under the "Overlays" button at the very top center of the REW plot area.  You can plot multiple measurement types using this facility.

 

2) Your dialing-in approach looks like the one that I use.

 

3)  Your excess group delay curve is dominated by the movement of the acoustic center of the HF driver at lower frequencies (as I discussed in this morning's PM).  If you move the crossover frequency up to 600-800 Hz, I believe that a lot of that tail to the right at low frequencies will be corrected by avoiding that dramatic movement of the acoustic center backward in the HF driver.  You will also need to readjust the delay on the HF channel when you do this. 

 

Chris

  • Thanks 1
Link to comment
Share on other sites

2 hours ago, rplace said:
  • Started from scratch today, again, with single HF/LF 2-way in the center of the room...
  • No PEQs, No delays, No XO points
  • Measure and flatten HF only with EQ wizard

 

When measuring the HF with No XO point, it is important to set the REW test Start Frequency ___ Hz at a safe range for the delicate tweeter. 

Link to comment
Share on other sites

Those are full-range drivers that Rich is using, and there are no issues with supplying a full-range signal to them. 

 

Additionally, some tweeters have issues with low frequencies, but in general, as long as you don't boost the signal at lower frequencies, you can supply a full-range signal to them, trimmed at the LF point that you wish to cross over to the midrange horn/driver, since they will not respond below their Fc.  Just don't boost the LF to extend the response lower than it is naturally.  This is exactly how Danley does it in the SH-50, and that unit is rated at 1 kW music power, 4 kW peak, 133 dB at 1 m output.

 

Chris

Link to comment
Share on other sites

Just remember that the K-77 tweeter is only rated at 5 w with a 6 dB crest factor (...a very low crest factor I might add...).  The K-77 is not a robust tweeter.

 

Chris

Link to comment
Share on other sites

55 minutes ago, Chris A said:

 

 

3)  Your excess group delay curve is dominated by the movement of the acoustic center of the HF driver at lower frequencies (as I discussed in this morning's PM).  If you move the crossover frequency up to 600-800 Hz, I believe that a lot of that tail to the right at low frequencies will be corrected by avoiding that dramatic movement of the acoustic center backward in the HF driver.  You will also need to readjust the delay on the HF channel when you do this. 

 

Chris

 

I'm stopping where I am and refocusing my efforts on that. Nothing has changed since this morning's measurements. I'm simply going to keep the EQ REW supplied me and move the XO around to 600, 800, 1000 (keeping 6dB, Bessel) and then see what my Spectrograms and GDs say.

 

What I'm struggling with conceptionally is....I thought the point of a 2-way system was to have the majority of the "work" done by the HF section be that the 402 horn and drive in a Jubilee set up or most others. While the LF just fills in the bottom end. It is the midrange we all love with horns, right? Don't most people cross a 2-way Jubilee around 400? And you only have a 2" driver. I've got 8" drivers. Obviously I'm missing a key point or two. Please advise.

 

Thanks to all for all the help so far!!!

Link to comment
Share on other sites

22 minutes ago, Chris A said:

Those are full-range drivers that Rich is using, and there are no issues with supplying a full-range signal to them.

 

Yes, you are right, and that probably also applies to some of the professional quality drivers being used by many other forum members.

 

23 minutes ago, Chris A said:

The K-77 is not a robust tweeter.

 

So, my comment was about tweeters in general, for all the other readers. I should have made that statement in a separate post, not attached to a quote from a post @rplace directed to @Chris A.

 

:)

 

 

Link to comment
Share on other sites

30 minutes ago, rplace said:

What I'm struggling with conceptionally is....I thought the point of a 2-way system was to have the majority of the "work" done by the HF section be that the 402 horn and drive in a Jubilee set up or most others.

This reminds me of a favorite quote:

Quote

All happy families are alike; each unhappy family is unhappy in its own way. (from Anna Karenina by Leo Tolstoy)

There are a lot of ways that different drivers can have issues.  This is but one of many ways that things can be "not happy".  When you do find a driver that does everything you want without fault and at a price that you're willing to pay, that is a good day, indeed.

 

Also note that the crossover frequency of the Cornwall I and II is nominally 800 Hz (and 4.9 kHz for the midrange-tweeter).  At this frequency (800 Hz), the horizontal polars of the midrange horn approximately match those of the 15" direct radiating woofer.

 

The reason for setting the crossover frequency of the Jubilee at or just below 500 Hz is due to the dual-mouth interaction (horizontal lobing) of the Jubilee bass bin, which is a bifurcated "W" section design (just like the Khorn, La Scala, and Belle).  If there were only one mouth of the bass bin, my guess is that the crossover frequency would probably be moved higher to 600-800 Hz to take advantage of the Jubilee bass bin's extended HF response as noted in the JAES paper by PWK and Roy Delgado on its design.  I suppose that at the time of its design, the dual-mouth interaction of the "W" section bass bins was not seen as being as important as it perhaps is seen today.  So we cross at a lower frequency.  Psychoacoustically, crossing a little lower is somewhat more forgiving from a listening standpoint.  Crossing at 800 Hz means you have to do a little better job of matching polars and time aligning (which is pretty close on the Cornwall I/II woofer/midrange crossover). 

 

Chris

  • Like 1
Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...