Print

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

[newguy1]

I'm looking to take my two channel and mix the bass mono.

I want to filter the bass out (<200hz), sum it to mono and mix it back to both of my speakers a la Earl Geddes.  I have a Xonar ST card (2 channel)

How can I do this? 

[mwillems]

I'm looking to take my two channel and mix the bass mono.

I want to filter the bass out (<200hz), sum it to mono and mix it back to both of my speakers a la Earl Geddes.  I have a Xonar ST card (2 channel)

How can I do this?  

You'll need at least two extra blank channels for mixing to make this happen.  The base ST only has two analog channels, but here's something to try:  

Make sure you're outputting to the ST in ASIO mode in JRiver.  Then go to output format under DSP studio and try selecting "2 channels inside a 4 channel container" under channels.  Then try to play something normally.  If it works you've lucked out and the Asus ASIO driver will allow you to address the SPDIF outputs as extra channels.  I have an ST myself, but mine has the H6 daughter board so I can't directly test this (i.e. mine has 8 analog channels).

Tastes may vary in terms of the crossover slope, etc., but the filter bank I'd recommend for good results is:

In parametric EQ:
1. Use Mix Channels to Copy Left to SL
2. Copy Right to SR
3. Put in a lowpass with a 12dB slope set to 300 Hz targeting SL and SR
4. Put in a second lowpass with a 12dB slope set to 300 Hz targeting SL and SR (3+4 creates half of a 4th order Linkwitz Riley crossover)
5. Put in a highpass with a 12dB slope set to 300 Hz targeting L and R
6. Put in a second highpass with a 12dB slope set to 300 Hz targeting L and R (5 + 6 create the other half of a 4th order Linkwitz Riley crossover)
7. Add SL to SR
8. Copy SR to SL
8. Add SL to L with a -3dB gain
9. Add SR to R with a -3dB gain

That should work if you've got the channels for it (I don't have time to test right now; try it out and see what you think).  The places to tweak it are:

1. Adjusting the negative gain in 8 and 9.  Because you're summing the bass to mono it will be between 3dB and 6dB louder than before, so you might need more negative gain (you might need as much as -6dB for things to sound natural, but -3dB is a good place to start; measure and see)
2. Adjusting the slope of the crossovers to make them shallower or steeper (although it won't necessarily sum as nicely if you don't keep the Linkwitz-Riley structure)

[newguy1]

The 4 Chan in a 2 Chan container worked like a charm!

Is the two step 12dB filter different than a single 24dB?

[newguy1]

Also, if I have the same crossover point for both the high and low pass, will I get a bump at the crossover frequency?

Should I set the low pass at 175hz and the high pass at 225hz? (numbers picked at random)

[mwillems]

The 4 Chan in a 2 Chan container worked like a charm!

Is the two step 12dB filter different than a single 24dB?

Yes, a single 24 dB will be very different than two stacked 12dBs.  A single 24dB filter in JRiver is a butterworth 4th order, I suggested two 12dB butterworth filters stacked, which makes a 4th order Linkwitz-Riley instead, which is quite different: https://en.wikipedia.org/wiki/Linkwitz%E2%80%93Riley_filter

Also, if I have the same crossover point for both the high and low pass, will I get a bump at the crossover frequency?

Should I set the low pass at 175hz and the high pass at 225hz? (numbers picked at random)

If you use my suggested filters, you shouldn't get a bump.  The issue you're describing is exactly why I suggested a Linkwitz Riley configuration: one of the distinguishing features of a Linkwitz-Riley filter is that it creates no bump with a symmetrical crossover.  Linkwitz-Rileys solve the "butterworth bump" problem because they're 6dB down at the crossover point so two symmetrical LR's sum flat instead of creating a bump.  

Here's a chart illustrating the phenomenon: https://en.wikipedia.org/wiki/File:Linkwitz_vs_Butterworth.svg

A 4th order LR is also phase coherent (the phase on each side of the crossover is exactly 360 degrees out of phase at every point) so you don't need to mess around with inverting a channel, etc.  Using a 4th order LR takes a lot of the guesswork out of crossover design, it's very likely to "just work" without much additional fiddling (especially where, as in your case, setting delay isn't an issue)

Bottom line, if you use the exact filter bank I described there should be no bump, it should be flat across the crossover region if you get your gain set correctly.

[newguy1]

Thanks