Rational Acoustics



Jimmy Mika
October 7th, 2015, 12:25 PM
Hi All,
I'm new to tuning sound systems and want to make sure I'm doing everything correctly or at least on the right track.

We just made the move to a digital console and installed a Yamaha CL-5. The console mains go to an older EAW MX8750 dsp and out to a set of 5 line arrays per side. We have a set of unique planer arrays made by HPV technologies MAD MTW series. We also have a pair of EAW LA128z subs.

I setup Smaart 7 to measure the transfer function with an Earthworks M30, USB pre2 and a reference signal from one of the mains from the CL-5. I placed the measurement mic in line with the left speakers about mid-way in the auditorium.

Attached is the plot of my reading. For the crossover I used the values that HPV and EAW suggest in their speaker specs. From what I can tell the 2-3K dip is a result of the crossover, the MX8750 doesn't have an option to use the 48db/oct L-R slope (like the speaker specs recommend) so I'm using the 24db/oct L-R slope between the high-mids and the highs.

Please let me know if you see anything out of the ordinary in my process.
Here are some questions:

- Is Smaart displaying the mic measurement and an average between the CL-5 and the measurement mic? It seems odd to me that the Cl-5 mains (pre eq) would be so close to the measurement mics reading.
- Is the drop in highs beyond 8k normal for line arrays?
- Lastly, my dsp tends to blink red on the output indicators and I read in the MX8750 manual that this is ok on occasion. I've tried lowering the input gain on the DSP and the output gain on the DSP and cannot seem to prevent this from happening. Any advice?

Chris Tsanjoures
October 7th, 2015, 04:19 PM
Hi All,
I'm new to tuning sound systems and want to make sure I'm doing everything correctly or at least on the right track.

We just made the move to a digital console and installed a Yamaha CL-5. The console mains go to an older EAW MX8750 dsp and out to a set of 5 line arrays per side. We have a set of unique planer arrays made by HPV technologies MAD MTW series. We also have a pair of EAW LA128z subs.

I setup Smaart 7 to measure the transfer function with an Earthworks M30, USB pre2 and a reference signal from one of the mains from the CL-5. I placed the measurement mic in line with the left speakers about mid-way in the auditorium.

Attached is the plot of my reading. For the crossover I used the values that HPV and EAW suggest in their speaker specs. From what I can tell the 2-3K dip is a result of the crossover, the MX8750 doesn't have an option to use the 48db/oct L-R slope (like the speaker specs recommend) so I'm using the 24db/oct L-R slope between the high-mids and the highs.

Please let me know if you see anything out of the ordinary in my process.
Here are some questions:

- Is Smaart displaying the mic measurement and an average between the CL-5 and the measurement mic? It seems odd to me that the Cl-5 mains (pre eq) would be so close to the measurement mics reading.
- Is the drop in highs beyond 8k normal for line arrays?
- Lastly, my dsp tends to blink red on the output indicators and I read in the MX8750 manual that this is ok on occasion. I've tried lowering the input gain on the DSP and the output gain on the DSP and cannot seem to prevent this from happening. Any advice?

Greetings - there are a few immediate red flags to address before we can get to your actual questions.

Firstly- Smaart has the ability to capture measurements (space bar or capture button). This enables you to put your traces into a format which can be shared - thusly allowing for someone else to open the exact data, settings and all, of what you are looking at, just on our own Smaart rigs. This is THE BEST way to share Smaart data.

Second best is to take a screenshot of your computer using the computer operating system command for screen shot - which varies on Mac or PC. This way we can at least see what the control strip is set to.

Your picture of your computer screen is essentially useless, as there could be any number of things right or wrong here. It is impossible to tell without knowing specific values. For example, is your coherence dropped because you are playing a live measurement without a reference signal, or did you not set your reference delay - the answer would be obvious if you can please resend in the above formats.

It is very common in hi-power PA systems (as opposed to hifi boxes for small listening rooms) to have a HF drop rather steep around 9-10k. And believe me, you want that. The only place where it is really appropriate to have seriously serious HF extension 14k+ is in a recording studio environment, or small space.

I recommend to you drop one of the hangs, and measure in the Near Field (about a meter) the response of an individual array element, and do that again for however many you feel. this way you can see what the speaker is actually doing, and how that compares to what it is doing as you get the room involved. In the NF, you will be able to see a semi-anechoic response where any holes/polarity issues will be more obvious.

That is a start. I look forward to hearing back.

Jimmy Mika
October 7th, 2015, 11:17 PM
Greetings - there are a few immediate red flags to address before we can get to your actual questions.

Firstly- Smaart has the ability to capture measurements (space bar or capture button). This enables you to put your traces into a format which can be shared - thusly allowing for someone else to open the exact data, settings and all, of what you are looking at, just on our own Smaart rigs. This is THE BEST way to share Smaart data.

Second best is to take a screenshot of your computer using the computer operating system command for screen shot - which varies on Mac or PC. This way we can at least see what the control strip is set to.

Your picture of your computer screen is essentially useless, as there could be any number of things right or wrong here. It is impossible to tell without knowing specific values. For example, is your coherence dropped because you are playing a live measurement without a reference signal, or did you not set your reference delay - the answer would be obvious if you can please resend in the above formats.

It is very common in hi-power PA systems (as opposed to hifi boxes for small listening rooms) to have a HF drop rather steep around 9-10k. And believe me, you want that. The only place where it is really appropriate to have seriously serious HF extension 14k+ is in a recording studio environment, or small space.

I recommend to you drop one of the hangs, and measure in the Near Field (about a meter) the response of an individual array element, and do that again for however many you feel. this way you can see what the speaker is actually doing, and how that compares to what it is doing as you get the room involved. In the NF, you will be able to see a semi-anechoic response where any holes/polarity issues will be more obvious.

That is a start. I look forward to hearing back.

Thanks for your response.
I stopped by the church and took a measurement from about 20ft from the speaker, closest I could get for the time being.

I didn't have the inputs setup correctly and wasn't using a reference signal. I now setup the M30 mic input to measure the room using the CL-5 main out as a reference signal. I'm also not sure about how to setup the delay for the reference signal.

I did get in contact with someone who has worked with these speakers who suggested I use 1.32msec delay on the highs to compensate for phasing between the highs and high mids.

I also plotted the crossovers by muting them on the DSP and taking a trace of each one.

Here is an updated link to the files.
https://www.dropbox.com/sh/302qkilfn93tggf/AABt7ZIC5XUs9K7ashWCjG6ya?dl=0

Chris Tsanjoures
October 9th, 2015, 02:04 PM
Okay you are almost there. As I suspected, you haven't yet set Smaart's reference delay. We can tell 3 ways, notice that the coherence is totally crashed down to the LF, and that the Phase traces are all random - and also that the delay area on the TF engines reads 0.00. use find, or track, when making live measurements to get a reference delay and you will start to see some meaningful info.

I recommend checking out the applications videos I've made on our rationalacoustics youtube channel.

Jimmy Mika
October 16th, 2015, 12:33 PM
Okay you are almost there. As I suspected, you haven't yet set Smaart's reference delay. We can tell 3 ways, notice that the coherence is totally crashed down to the LF, and that the Phase traces are all random - and also that the delay area on the TF engines reads 0.00. use find, or track, when making live measurements to get a reference delay and you will start to see some meaningful info.

I recommend checking out the applications videos I've made on our rationalacoustics youtube channel.

I spent some more time on the system yesterday. We decided to replace the older MX8750 DSP with a DBX 4800 driverack. I like the more up to date interface on the driverack software and more importantly its capable of 48db/oct crossovers which is what's recommended for our speakers.

I also found a helpful calculator (MX8750 limiter wizard) from EAW for calculating and setting DSP limiters (one of my original issues was MX8750 channel overloading). I setup the limiters on the 4800 per the results of this calculator and so far the system is operating within range without clipping. Here's the link if anyone finds it helpful.

http://eaw.com/docs/7_Tools/

The link below is my measurement from about 20' from the array. I added the delay as you suggested however am still not sure if i'm doing everything correctly. Can you please look it over and let me know what else I might be doing wrong.

https://www.dropbox.com/sh/302qkilfn93tggf/AABt7ZIC5XUs9K7ashWCjG6ya?dl=0

Chris Tsanjoures
October 19th, 2015, 01:12 PM
I think at this stage in the game, it would be in your best interest to take a step or two back. If you have access to a single powered speaker - or a simple 2-way box - practice making some measurements with that. The measurements you've loaded to dropbox are still a bit messy, I think it would be very useful for you to get to a point where you know what to expect from a properly performing loudspeaker, in a near field controlled environment. This way, when you see something that doesn't match your expectations you can start asking - why?

I fear that you may have bit off more than you can chew by jumping into a four way system right away. I look forward to seeing how you progress!

Cheers

Jimmy Mika
October 22nd, 2015, 03:02 PM
I think at this stage in the game, it would be in your best interest to take a step or two back. If you have access to a single powered speaker - or a simple 2-way box - practice making some measurements with that. The measurements you've loaded to dropbox are still a bit messy, I think it would be very useful for you to get to a point where you know what to expect from a properly performing loudspeaker, in a near field controlled environment. This way, when you see something that doesn't match your expectations you can start asking - why?

I fear that you may have bit off more than you can chew by jumping into a four way system right away. I look forward to seeing how you progress!

Cheers

Thanks. I will be looking at a simple full range system in our Gym in the upcoming weeks, will probably start a seperate thread on that.

The 4-way system is relatively stable right now and sounds amazing. I will come back to it.