posted 30 March 2002 13:22            

I just tested a 6-band Behringer DSP-9024 here in my lab. It performs remarkably well, but I am not an audio pro.

What I did find was that while you can tune the passband edges, you cannot tune the rolloff rates. The results of noise correlation tests show these input filters to be simple 6 dB/octave rolloffs -- which implies a simple bandpass arranged by series connection of stagger tuned 1 pole lowpass and 1 pole highpass filters. There is considerable phase bending going on inside each band.

Any thoughts? comments? criticisms?

Cheers,

posted 07 April 2002 00:36            

But I was actually using this Behringer device for something altogether different from its normally intended use...

You can pick up a DSP9024 for around $180 USD, which is cheaper than I can even buy a single DSP card. Given that many, if not most, musicians are short on excess wealth, I was looking for a cheap alternative to my Crescendo device for them. Those with substantial sums of money seem to care less about what they hear than in being "seen" at Symphony on Society Night.

So I was trying to see if one of these Behringer linear compressors with only 6 bands would work well enough to restore impaired hearing for musical listening purposes. Many musicians are getting older and the love of music lives on even if the hearing is going.

To my initial shock, these devices sport 6 dB/octave filters. That causes a lot of inter-band dependencies when setting up the compression and post-gain in each band. For hearing loss, it is not uncommon to need as much as 20 dB gain difference between adjacent bands, and with only 6 dB/octave rolloff that means a lot of the high boost in one band will bleed over into adjacent bands.

But, despite this, I found a way to tune these boxes to suit. You can't just dial in your hearing loss, or some fraction of it, because of the interband effects. But using a nonlinear optimization routine I can start with a persons audiology results and determine the best band gains and compressions to use so that despite these dependencies, the unit still functions quite admirably.

Like I keep saying, "It isn't a Crescendo", but for less than $200 how could one complain?

- DM

posted 13 April 2002 11:00            

I wanted to know the opinions of mastering processor users regarding the 6 dB/octave rolloff used by the Behringer 9024. After some more research I find that the high-end processors offer selectable rolloffs, but generally only 3 or 4 bands of compression.

Behringer took advantage of a "cheap trick" to pack as much as they did into a cheap box. I am always astonished at the chutzpa of marketeers -- in this case you have to read the fine print to find what is NOT said...