Using the Wrong DSP (split from "Peavey Versarray")

[Phil Graham]

Eric... FWIW, I hear Phil Graham already has some pretty good DSP settings for a KF650 rig (with QSC PL amps and a DR260)....

Donnie,

I'm glad that rig is out and people are experiencing good things from it. I should be clear, though, that my "pretty good DSP settings" for KF650 were/are nothing more than (properly) using SMAART to measure the output of the DSP and match the transfer function supplied by EAW. On the DR260 that first requires shuffling some of the HF parametrics to the input side filters, and then measuring what the processor is actually doing, modifying it to match EAW's curves as closely as possible. I recall getting within 1/2 dB everywhere.

I think we've FINALLY clearly established as common knowledge that all DSPs do things differently, and the transfer functions (especially at HF) can have multiple dB differences between processors with nominally the "same" settings.

I think Bennett deserves a large amount of the credit of getting that into the public consciousness of pro audio practicioners by soliciting measurements and getting an article published on it in FOH Magazine. I remember he and I having a phone conversation quite a while back discussing what sort of filter settings he could use for that article that would behave like realistic processing while serving to highlight the differences between processors as much as possible. It is gratifying to see how far he ran with that.

The folks at Fulcrum Acoustic (www.fulcrum-acoustic.com) deserve mention for providing unique, specific settings for most major DSP processors that insure their speakers are as similar as possible independent of DSP platform. This, or locking customers into a specific DSP, are the only ways to insure consistency.

Also, EAW and EV both provide transfer functions of their processing that one can match with a measurement platform (EAW -> SMAART, EV -> Systune). To be completely honest, the TFs I've seen out of EAW have been measured more properly and consistently than out of EV, but at least both manufacturers are giving it a shot. I've reached out to EV with suggestions on how to unify their Systune TF libraries, and those discussions made it to the top of their engineering team. There may be other manufacturers that do the same in terms of TF, but EAW and EV are the only two I can personally vouch for.

Finally, anyone who hasn't read Bennett's FOH article on DSP differences, or more in-depth presentation for the Parson's expo should. You'll find it eye-opening, and hopefully help you justify the expense of having a measurement-savvy person insure your processors match the desire response(s) that your loudspeaker manufacturers intended.

 

[Eric Cagle]

Re: Peavey Versarray?

Eric... FWIW, I hear Phil Graham already has some pretty good DSP settings for a KF650 rig (with QSC PL amps and a DR260)....

He does and I Have driven Kemper's KF650z rig with some of those settings. That is why I know that combination will work for what I am going to be doing.

 

[Phil Graham]

Re: Peavey Versarray?

you cannot really use a reflex loaded 12 in driver past about 500Hz if you want serious line array performance, certainly not at the frequency we are talking about, it will work but you will pay a price (L-acoustics have a paper about this that I can’t find at the moment). As good as he is, not even Phil will be able to correct for this.

Quoted and agreed with for emphasis...

 

[John Roberts]

Re: KF650 Settings for DR260

Donnie,

I'm glad that rig is out and people are experiencing good things from it. I should be clear, though, that my "pretty good DSP settings" for KF650 were/are nothing more than (properly) using SMAART to measure the output of the DSP and match the transfer function supplied by EAW. On the DR260 that first requires shuffling some of the HF parametrics to the input side filters, and then measuring what the processor is actually doing, modifying it to match EAW's curves as closely as possible. I recall getting within 1/2 dB everywhere.

I think we've FINALLY clearly established as common knowledge that all DSPs do things differently, and the transfer functions (especially at HF) can have multiple dB differences between processors with nominally the "same" settings.

I think Bennett deserves a large amount of the credit of getting that into the public consciousness of pro audio practicioners by soliciting measurements and getting an article published on it in FOH Magazine. I remember he and I having a phone conversation quite a while back discussing what sort of filter settings he could use for that article that would behave like realistic processing while serving to highlight the differences between processors as much as possible. It is gratifying to see how far he ran with that.

The folks at Fulcrum Acoustic (www.fulcrum-acoustic.com) deserve mention for providing unique, specific settings for most major DSP processors that insure their speakers are as similar as possible independent of DSP platform. This, or locking customers into a specific DSP, are the only ways to insure consistency.

Also, EAW and EV both provide transfer functions of their processing that one can match with a measurement platform (EAW -> SMAART, EV -> Systune). To be completely honest, the TFs I've seen out of EAW have been measured more properly and consistently than out of EV, but at least both manufacturers are giving it a shot. I've reached out to EV with suggestions on how to unify their Systune TF libraries, and those discussions made it to the top of their engineering team. There may be other manufacturers that do the same in terms of TF, but EAW and EV are the only two I can personally vouch for.

Finally, anyone who hasn't read Bennett's FOH article on DSP differences, or more in-depth presentation for the Parson's expo should. You'll find it eye-opening, and hopefully help you justify the expense of having a measurement-savvy person insure your processors match the desire response(s) that your loudspeaker manufacturers intended.

Not be be a broken record playing one note, but DSP platforms that deliver different transfer functions for the same control inputs are just wrong. Correcting the DSPs to agree should be done inside them, by the sundry manufacturers.

I tried to get AES standards committee to help with defining what Q means in the context of peak/cut EQ with no success. So even if the DSP platforms were accurate, we would still have this definition issue.

If I was still in that business I would surely try to do something more about it than whine... but I'm not and have other windmills to tilt at. So I will whine in passing as it keeps coming up (for years now). This may be an opportunity for measurement software sellers to sell more systems, until the DSP industry gets their act together, but if customers keep buying these products that don't work correctly, industry will keep selling them.

JR

 

[Eric Cagle]

Re: Peavey Versarray?

Quoted and agreed with for emphasis...

I hear you loud and clear Phil. I won't expect anything more out of the VRay than what I hear. By the way I bought the analog processor you recommended for the Apogee rig. Big difference over what they were using. Very nice sound but not quite loud enough for what I needed at the time. Passed on the rig for now. $60 spent on eBay. Several thousand not spent on cabinets, cases, and amps that would not do quite what I wanted. Very good advice! I will be hiring you to set up processing for what ever rig I end up with for this application.

 

[Phil Graham]

Re: Peavey Versarray?

I hear you loud and clear Phil. I won't expect anything more out of the VRay than what I hear. By the way I bought the analog processor you recommended for the Apogee rig. Big difference over what they were using. Very nice sound but not quite loud enough for what I needed at the time. Passed on the rig for now. $60 spent on eBay. Several thousand not spent on cabinets, cases, and amps that would not do quite what I wanted. Very good advice! I will be hiring you to set up processing for what ever rig I end up with for this application.

Eric,

Glad that advice saved you from making the wrong purchase. Look forward to setting up whatever you end up pulling the trigger on.

 

[Donnie Evans]

Re: Using the Wrong DSP (split from "Peavey Versarray")

Donnie,

I'm glad that rig is out and people are experiencing good things from it. I should be clear, though, that my "pretty good DSP settings" for KF650 were/are nothing more than (properly) using SMAART to measure the output of the DSP and match the transfer function supplied by EAW. On the DR260 that first requires shuffling some of the HF parametrics to the input side filters, and then measuring what the processor is actually doing, modifying it to match EAW's curves as closely as possible. I recall getting within 1/2 dB everywhere.

Hi Phil, that "pretty good DSP settings" quote was made primarily TIC by me. I haven't heard the rig for myself but I've heard really good thigs about it, and I also heard you essentially turned a DR260 into a UX8800.

 

[Phil Graham]

Re: Using the Wrong DSP (split from "Peavey Versarray")

Hi Phil, that "pretty good DSP settings" quote was made primarily TIC by me. I haven't heard the rig for myself but I've heard really good thigs about it, and I also heard you essentially turned a DR260 into a UX8800.

The statement of turning a DR260 into a UX8800 is untrue and represents wishful thinking. If Kemper actually said that, he's just being enthusiastic/hyperbolic, but I don't want people to think I would claim such powers.

DR260 got a reasonably close match for EAW non-GF factory specs, and some help from my Amp gain spreadsheet to sort out the attentuations, and that is it.

 

[Kemper Watson]

Re: Using the Wrong DSP (split from "Peavey Versarray")

The statement of turning a DR260 into a UX8800 is untrue and represents wishful thinking. I know Kemper's just being enthusiastic/hyperbolic, but I don't want people to think I would claim such powers.

DR260 got a reasonably close match for EAWs factory specs, and some help from my Amp gain spreadsheet to sort out the attentuations, and that is it.

Aww Phil, you know you can leap tall buildings with a single bound....LOL

 

[Scott Helmke]

Re: Using the Wrong DSP (split from "Peavey Versarray")

The statement of turning a DR260 into a UX8800 is untrue and represents wishful thinking.

In other words, you got lucky in being able to get the same filters into place.

I had to do something similar with an XTA processor and a newer processor with factory settings a couple times. Both times it was a bit of a puzzle to make it happen, and required very close attention to what SMAART was telling me. For one output that needed a lot of filters I had to use some input filters... then use filters on a different output to counteract the side effects of the input filters on that output (which luckily only needed a couple filters to begin with). Then I ran into a 30dB slope crossover filter which the XTA doesn't implement.

Took at least half a day... and a year later when the manufacturer updated their settings I had to do it all over again. We still saved money, but not as much as you'd think.