Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[lillgunde]

Hey everybody.
Just wanted to share my approach to using convolution in JRiver to see if I am doing right but also to introduce others
to a fairly easy process to getting the measurements and setup right.

Software: JRiver Media Center 19, Audionet Carma 4 and REW 5 (also the calibration file used for UMIK-1 is present in Carma and REW)
Hardware: UMIK-1
               
Audionet Carma 4 is freeware, can be downloaded on the Audionet homepage.
I use this program to generate and measure a 5.1 DTS sweep signal from my surround system.
All six channels are then saved as a frequency response wav files (48khz mono wave)

These are then imported into REW as 6 independent overlays, and the EQ adjustment is calculated for each channel.
When opening the EQ window within REW, generic equalizer is selected, target level 75 db and the filter task match range is:

Subwoofer: 20-80hz (speaker type subwoofer, 12db/octave crossover and 24db/octave LF slope)

Remaining channels: 80-500hz  (speakers typ none)

**This is because I use the room correction in DSP studio with those crossover settings.

When the EQ correction has been generated for each of the channels, these are exported ( export/filter response as wave) to 6 mono 48khz wave files.

A simple convolution config file is then written to be used in JRiver DSP Studio. This is how the config file looks like:

48000 6 6 0
0 0 0 0 0 0
0 0 0 0 0 0
mono_LEFT.wav
0
0.0
0.0
mono_RIGHT.wav
0
1.0
1.0
mono_CENTER.wav
0
2.0
2.0
mono_SUB.wav
0
3.0
3.0
mono_SURROUND_LEFT.wav
0
4.0
4.0
mono_SURROUND_RIGHT.wav
0
5.0
5.0

Any thoughts?? I am not entirely sure that I got the config file right... But it works just fine

[Matt]

I don't have much to add, but just wanted to say welcome.

mojave is a bit of an REW expert, so if he sees this, he might have some more feedback.

[lillgunde]

Thank you Matt.
Hopefully someone has something to add, I am new to the whole process and convolution is
complicated to say the least.
I think my solution only scratches the surface of what is possible to accomplish.

[mojave]

I would use REW to take the measurements of all speakers. It should have a higher quality measurement. In JRiver's DSP Studio > Parametric Equalizer you can select  Order Channels. This is a neat tool that lets you route the output from REW through each speaker. Measure the front left speaker with REW and then use Order Channels to move the output to the next speaker by dragging "Left" down.

Your EQ methodology looks good. You don't need to select 75 dB as the target slope for all speakers unless you measured them close to 75 dB. I would choose a target that represents the overall level of the speaker when viewed with 1/3 octave smoothing. Use the same target for both mains and  both surrounds. For example, your mains might be 72 dB, your surrounds might be 76 dB, and your center might be 74.

You should set the distance settings in JRiver in the Room Correction DSP.

After you EQ the speakers, load the convolution file in JRiver, and set the distance settings you should level match with Tools > Advanced Tools > Volume Calibration. Basically, the REW measurement is a slice of the output of a speaker and doesn't fully represent the full output. For example, a speaker with 60 degrees of dispersion and one with 120 degrees of dispersion will both measure the same in REW, but the 120 degree one will be louder in room. By using JRiver's volume calibration you can be more accurate.

If you don't have an SPL meter, just use the one in REW. It won't be accurate regarding actual SPL since it needs to be calibrated each time to an SPL meter, but you are just worried about relative levels. With JRiver 19.0.51 and newer you can just enter the dB levels into Room Correction to automatically have the levels set. 

I've never created a convolution file, so I don't have any input there. I've been wanting to get around to doing it. I think there might be a way to even add delay and level adjustments to the file, but it is probably just as easy to use Room Correction.

[lillgunde]

Wow mojave, great reply
Lots of usable input, did not know about the spl calibration in JRiver.
What I am doing at this time is to set distance and level compensation in the receiver, but I am willing to try something new.
Dispersion from my speakers are pretty high, I have full range thx speakers from Jamo.

Once again great input, thanks.