Rational Acoustics



Giant
October 2nd, 2015, 12:49 AM
Dear all,
since in my hands the auto find delay doesn't work (ASIO driver at any frequency, Operating System, and also with different, external, FireWire audio cards), I would like to know if an alternative, manual method, exists to correctly calculate the delay of the reference pathway.
Thanking you in advance for the attention
Giant

Johnny
October 2nd, 2015, 11:55 AM
Hi Giant,

Can you explain your measurement signal chain? Also, which version of Smaart are you running?

Giant
October 5th, 2015, 06:50 AM
Hi Giant,

Can you explain your measurement signal chain? Also, which version of Smaart are you running?


Hi Johnny,
sorry for the delay in replying your message, but I'm away from home. I don't remember the exact number of the version, however it has been downloaded as demo. In the settings panel I have indicated as output the Jriver driver where is the convolution engine for the FIR-based crossovers of my 4 way active loudspeakers. The input is the mic of the MOTU FireWire audio card (asio drivers). Operating systems are Seven and 8.1. In the audio path of JRiver MC it is all OK. Similarly both measurement and reference channel looks good in SMAART. However, in spite of many attempts, I have not been able to obtain a valid delay with the auto function even with high frequencies. Is it possible to find it in an alternative way?
Have a great day
GiAnt

Johnny
October 5th, 2015, 11:14 AM
What is the distance between the microphone and the speaker? The demo version does not support delay times over 50 ms.

Also, are you using pink noise as your reference signal? If not, you should use pink noise even if it's just to determine the initial delay time.

Giant
October 6th, 2015, 11:09 AM
What is the distance between the microphone and the speaker? The demo version does not support delay times over 50 ms.

Also, are you using pink noise as your reference signal? If not, you should use pink noise even if it's just to determine the initial delay time.

Hi Johnny,
many, many thanks for your quick and kind reply.
The distance is 2,5-3 mt and the reference signal is pink noise.
In any case, I think to have solved the problem. I've changed the PC (now is a sony vaio laptop with Seven 32 bit in substitution of an Dell AIO 64 bit) and checked thoroughly the ASIO routing. Now SMAART find the delay pretty good even at vely low frequencies. I found it coincides approximately with the phisical distance between the mic and the driver under test.
Taking advantage of your kindness I would like to know if, in order to align the phases of the individual drivers at the XO points, I need to run all the driver together in SMAART and then delay the individual drivers (with the exception of the sub) in the Jrivers's PEQ up to find the best combination or if is better to proceed one at the time.
Thanking you in advance for your patience,
Sincerely
GiAnt

Chris Tsanjoures
October 6th, 2015, 12:20 PM
Taking advantage of your kindness I would like to know if, in order to align the phases of the individual drivers at the XO points, I need to run all the driver together in SMAART and then delay the individual drivers (with the exception of the sub) in the Jrivers's PEQ up to find the best combination or if is better to proceed one at the time.
Thanking you in advance for your patience,
Sincerely
GiAnt

Greetings Giant -

In the case of initial crossover setting, you will find the time arrival at XO frequency of the two systems separately. Basically, the one which arrives first at XO would be delayed to that which arrives second so the point at which they hand off, they do so in time. There are several 'roads that lead to Rome' as far as initial speaker voicing Eq. You will want level, time, eq and polarity of each pass band. Reach a flat response across each passband independently, correcting for box/horn loading etc etc. Then, when you combine each together use the level, polarity, and delay controls to adjust the passbands as needed. I do not roll my own XO's regularly, so others may have a better process than myself.


Cheers!

Giant
October 12th, 2015, 06:21 AM
Greetings Giant -

In the case of initial crossover setting, you will find the time arrival at XO frequency of the two systems separately. Basically, the one which arrives first at XO would be delayed to that which arrives second so the point at which they hand off, they do so in time. There are several 'roads that lead to Rome' as far as initial speaker voicing Eq. You will want level, time, eq and polarity of each pass band. Reach a flat response across each passband independently, correcting for box/horn loading etc etc. Then, when you combine each together use the level, polarity, and delay controls to adjust the passbands as needed. I do not roll my own XO's regularly, so others may have a better process than myself.


Cheers!
Hi Chris,
thank for your reply. My interest in SMAART was related to the possibility of furtherly optimize my system, beyond the capabilities of freeware softwares. In particular my issues originate by the peculiar characteristics of my active system, mainly based on the convolution of FIR filters inside my preferred player, i.e. Jriver. This prevent some free software like REW or Holm Impulse to use a reference channel and therefore reliable measures. This is partially solved in SMAART meaning that I can choose ASIO drivers as input channel (mic) and Jriver's WDM driver as output (convolution OK). However, in this configuration auto find delay of SMAART does not work (it works accurately when I use ASIO drivers for both input and out put), so my doubt to have reliable measures when using the WDM driver of Jriver as output.
Cheers!