Using the HAL MS-3 Music Server for speaker and room correction

0 Members and 1 Guest are viewing this topic. Read 5179 times.

HAL

Going to start this thread to discuss the use of Math Audio's Room EQ with Foobar2000 running the component on the HAL MS-3 Music Server.  This will feed the dspMusik digital crossover already running the digital crossover for the speaker. 

Starting with the Pulsing Quasar open baffle line array speaker that is a full range planar driver from the spec sheet.  The response plot of the system before Room EQ is below.  Will update the thread as I get the system running the component and measurements made.



HAL

Have Math Audio Room EQ working with Foobar2000 and some first measurements and corrections done.   

The Dayton Audio UMM-6 calibrated mic data is being used in the correction process measurements. 

Will do another set of measurement, now that I have the servo sub levels matched to the planars.


Mike B.

HAL, was the measurement taken at your listening position?

HAL

The measurement above was taken 1 meter from the line array.  It is more than the line length array from the baffle.

The Math Audio Room EQ measurement later that is not posted was taken at the listening position.

*Scotty*

Rich, is the apparent high frequency roll-off shown in graph actually occurring or is it a measurement
artifact?
Scotty

HAL

Scotty,
The FPS planar is supposed to be good from 180Hz to 20KHz from the spec sheet.  It does not detail the method of test they used.  Most likely an infinite baffle, so on an open baffle the response can be very different.

The response was after figuring out the CLIO Pocket was resetting my time gate, so this is with about a 4mSec time window.  Hence the LF response cutoff around 250Hz.  That was at the first reflection point in my room with the line array on the floor.

Will see what happens with MA Room EQ and subs running.

HAL

Here are the before and after response plots of Math Audio Room EQ running on the MS-3 and Foobar2000.  The grey plot is the original measurement and the white plot is the corrected response. 



HAL

Here is a very good explanation from Math Audio about the goals of Room EQ.

http://mathaudio.com/why-room-eq.htm

A lot of these ideas are also part of the dspMusik digital crossover using no linear phase filters and minimum phase filtering in the DAC's without pre-ringing from linear phase filtering. 


HAL

After discussion with Math Audio, MathAudio Room EQ will now be part of the Foobar2000 install on all new MS-3 Music Servers. 

It will be up to the customer to obtain the calibrated mic for use if they wish to try the process with their system.  Instructions for use of Room EQ are here:

http://mathaudio.com/room-eq.htm

HAL

For anyone interested in speaker and room measurement techniques, here is a video from RMAF that has a very good tutorial.

The Dayton Audio UMM-6 is under $90 and has the USB interface directly into the PC for very good measurements.  You download the calibration file from the website using the mic serial number.  When you start Room EQ Wizard, it sees the UMM-6 asks for the mic calibration file.  Once you point it to the file, the system now gives accurate SPL measurements vs frequency.

https://www.youtube.com/watch?v=oa2RC_DSPOk

HAL

The dspMusik RevD system will now have stereo A/D 24bit/192KHz HiRez output over USB.  This will allow the MS-3 and dspMusik to do recordings and measurements with the appropriate hardware and software loaded on the PC. 

Audacity, Room EQ Wizard, MathAudio Room EQ and other software on the PC can now use other calibrated microphones with the system when correctly setup with mic preamps for improved SNR.

One mic in trials is the Dayton Audio iMM-6 calibrated mic.  Have the correct connectors and mic preamp design to give this a try to see if the noisefloor is lower than the reported noisefloor of the UMM-6 USB mic. 

As a reference, will use the Earthworks M50 and 1021 calibrated microphone system with a noise floor of 22dBA as the check. 

HAL

Finally getting back to doing the RoomEQ measurements on The Megaliths now that they are up and running.   

Using the Dayton Audio UMM-6 with the high resolution calibration data that is used by RoomEQ. 

From folks listening to the HAL MS-3 and dspMusik setup at RMAF2016, the RoomEQ processing made a very nice change for the better in the room. 

HAL

For anyone wishing to try custom MathAudio RoomEQ correction curves, the pointing device and be used to drag the correction line.

The left button will allow you to make peaks and dips in response.  The right button will let you drag a line at an arbitrary angle.

Latest version for Foobar2000 is RoomEQ 2.6.4 and has these features available. 

HAL

Info from Dayton Audio on the UMM-6 calibrated mic.

The USB A/D converter is running 16bit/44.1KHz for the measurements.  This is a good start.

Time to try the Earthworks M550 calibrated Omni mic with the Earthworks 1021 mic preamp into the dspMusikLCD's 32bit/192KHz balanced input minimum phase ADC.  For other correction projects this has always sounded better.  Now that the HiRez ADC is available time as a USB output, time to give it a try.

HAL

Back from CAF2018 and time for some new measurements for room correction with the MS-3 running Foobar2000 and MathAudio RoomEQ.

Earthworks mic and preamp are setup.  The dspMusikLCD is being used as a 24bit/192KHz USB A/D with RoomEQ. 

Also trying a custom EQ curve to use some ideas from other measurements and listening results.


HAL

Testing out the VST plug-in version of RoomEQ with Foobar2000.  You can download the VST component for Foobar2000 and then add the VST Plug-in for RoomEQ.  It is a free trial for 30 days.

This gives access to the ASIO driver capability for use with ADC's and DAC's that have those style of drivers, like the dspMusikLCD.  Now it can become its own measurement suite by connecting a calibrated mic to the ADC balanced input.  You select the input channel in RoomEQ measurement mode.  Then add the calibration file to the folder that RoomEQ sets up in Documents.

From previous work on speaker correction, found that the shorter the impulse response of the measurement mic and preamp the better the correction sounded, so now testing with the Earthworks M550 mic and 1021 mic preamp.  Balanced output to one of the dspMusikLCD ADC inputs. 

A minimum phase room correction system just sounds really good to me.


HAL

The latest beta version of Room EQ Wizard now supports measurements with a 192KHz sample rate system. 

Just tested that version with the MS-3 running Windows 10 and it is working with the dspMusikLCD 2x8 and 2x16 systems as both the A/D and D/A for the measurements.  This gives a few new capabilities like impulse response data and parametric EQ settings for room correction. 








HAL

The latest version of MathAudio RoomEQ's VST plug-in now supports 24bit/192KHz room measurements and corrections with Foobar2000.   

The measurements worked correctly using the dspMusikLCD as the mic input and audio output for a full test system. 

Trying this with both The Megaliths and The Monoliths speakers after the crossovers have been finalized with the final hardware arriving today.

HAL

Working with MathAudio on running RoomEQ in the dspMusikLCD.  This would be like running the VST plug-in in the dspMusik to free up the PC from the processing.  The PC will still be used to run the RoomEQ measurements, but the output files would be used in the processing in the dspMusikLCD.  This would work in both the dspMusikLCD 2x8 and 2x16 systems with the digital crossover.

The first test using The Megaliths RoomEQ correction is ongoing.   If it works well, will test it with The Monoliths when they are ready as well.