AFMG Network Forum

Nicola · Member

What is the difference - from a conceptual perspective - when you measure the delay between drivers using the Gain/Delay Offset under Input and the method used under Virtual EQ - Delay analysis.

For example measuring two drivers, with the delay offset under Input section
- driver 1 is 12.552ms
- driver 2 is 11.653ms

Now after having take two stow, you set the reference as beeing driver 1 and the delay being driver 2 and systune gives a delay ms of 104.43

I would have assumed the delay between driver 2 and driver 1 is - because driver 1 is farther away from the measuring point or reverse driver 2 is close - 12.552 minus 11.653 since you need to "delay" driver 2.

Is my assumption correct ? Then what the delay ms of 104.43 means ?

AFMG Pedro Lima · Forum Moderator

Hi Nicola,

Could you please send the measured Overlays for both drivers so that I can evaluate this better?

Please send to pedro.lima(at)afmg.eu

Thanks,
Pedro

AFMG Pedro Lima · Forum Moderator

Dear Nicola,

I have received the measurement files. Thank you for your time.

Please notice that in order to use the Delay Analisys plug-in one needs to acquire the Transfer Function for each system under test (in your case, each driver).

From the Overlays you have sent and from the doc file, you can see for instance that in the Impulse Response tab (IR) there is nothing but noise for the Medium drive. That is reflected in the Transfer Function Magnitude tab, where you can clearly see that the frequency response displayed does not correspond to the driver response.

For the Tweeter overlay, there is some information but still that is not good. You can notice that from the very low values of Coherene in the TF Mag tab and also from the low signal-to-noise ratio in the ETC tab (notice that the peak is only about 18dB higher than the measurement noise floor).

That said, please allow me to advise you to refer to the section about Dual Channel FFT measurements in Systune user's guide. It can be found from the page 63 on.

Feel free to return the contact in case of further doubts after referring to the document.

Best Regards,
Pedro Lima

PS: Thank you for your valuable feedback with respect the procedure on the memory issue. Wink

Doug Fowler · Forum Moderator

Pedro -

Could you post the screenshots / overlays for demonstration purposes?

Thanks!

AFMG Pedro Lima · Forum Moderator

Hi Doug,

Sure, good idea. Thanks.

The first picture below shows very bad Impulse Response data. Notice that there is almost only noise in the measurement. One cannot even see the impulse. There is also no meaningfull data in the Transfer Function Magnitude (frequency response) tab.

The Impulse Response below is a bit better. But still, there is too much noise in the measurement. Notice also the low Coherence values for the Transfer Function Magnitude.

Next, a very good measurement. In fact, the Coherence values are very high. In practice they are usually lower for a high number of Averages. See the Impulse Response. There is only a rather thin line other than the Impulse. In the TF Mag tab one can see the high Coherence values. Note that for the lower frequencies the values are low, and the data is masked. That is expected since the measured system did not produce enough low frequency contents.

The measurement quality can be evaluated from the ETC tab. Note the signal-to-noise ratio.

Hope this helps.

Best Regards,
Pedro Lima

Langston Holland · Member

¡Pedro! That was absolutely beautiful. :)

Nicola · Member

Dear Pedro,
thank you for you comments, most appreciated. I've gone through the manual and the FFT pages, 63 but also the chapter on delay analysis. Would you mind me asking the following question :

If i load four STO files for different drivers (low = 0hz -> 630hz, lowmed = 630hz -> 1250hz, highmed = 1250hz -> 6300hz, high = 6300hz and up) under Virtual EQ - delay analysis i can select a list of reference and the system nicely indicates in brackets the delays. I can read this (i've not been able to insert the screen sorry)
- low (20.35ms)
- lowmed (17.50ms)
- highmed (12.49ms)
- high (11.50ms)

My understanding would be that low is the further away and all other driver must be "delayed" in order to be alligned with the LOW driver. This would mean adding 2.85ms to the lowmed driver, or 7.86 to the highmed (or alternatively 2.85+5.01). And so on.

Where i get lost, is when i select LOW as reference and then LOWMED as delay i get under
POS -30.39
NEG -31-27
FLAT -157.35

I don't know how to interpret such figures
Best regards[img][/img][img][/img]

AFMG Pedro Lima · Forum Moderator

Hi Nicola,

Please e-mail me the files again.

Pedro

Nicola · Member

[quote="AFMG Pedro Lima"]Hi Nicola,

Please e-mail me the files again.

Pedro[/quote]

Hi Pedro,
Did you receive the files ?
Were they usefull ?
Best regards

AFMG Pedro Lima · Forum Moderator

Hi Nicola,

Yes, I've answered via e-mail for now.

Regards,
Pedro