DBX DriveRack PX

Anything about hardware musical instruments.
RELATED
PRODUCTS

Post

AusDisciple wrote:I'm actually using the 61a capsules for my PZM designs and they actually work surprisingly well. The PZM prinicple combined with a floating power supply (mic capsule not earthed and floating between the 0 and 48v rails) and a proper compensation filter give these a decent signal to noise ratio, overcome some of the SPL limits and provide a surprising amount of detail.
I'll be posting some example recordings online soon once I'm happy with the initial tweaks.
PZM is supposed to have twice the sensitivity right? Or is it just more prone to overload?

Post

JCJR wrote:Thanks that is fascinating. Didn't know. Curious to hear your results.

I vaguely recall reading an article long ago about using the capsule for recording, that maybe included some of those strategies, but didn't pay attention. Assumed at the time (perhaps wrongly) that the author back then was probably just so deaf that he didn't mind low spl handling and s/n worse than 50 dB. :)

My narrow mindedness was not completely without cause. A lot of do it yourself articles of the time described very low fidelity gadgets. (Cough-- PAIA--) :)
The capsules in stock form with a typical power supply and preamp are quite noisy and prone to overload. There are numerous ways to significantly improve all those things though and the popular Linkwitz mod, while useful for measurement, is actually counter-productive for recording use and actually makes the s/n ratio worse despite the higher SPL handling.
camsr wrote:PZM is supposed to have twice the sensitivity right? Or is it just more prone to overload?
It's a little more complex than a straight forward gain increase. It is frequency specific and dependant on the boundary plate dimensions and design. There's a really good read on the whole thing available from the Crown website here.... http://www.crownaudio.com/media/pdf/mics/127089.pdf
The secret to getting the Panasonic capsules to work well is in the PZM frequency compensation filter, the phantom power supply design and a good, low noise, balanced preamp/buffer stage. Those are the things I've been tweaking and testing.

Post

camsr wrote:
JCJR wrote: Is your rta-m metal or plastic?
It's metal. But I think the condenser itself is plastic (pretty sure). Keep plastic out of direct sunlight and UV light, or you know what'll happen.
Thanks. My px came in a day late, today. The included M2 mic is gray. The main body is metal, and the skinny piece that sticks out a couple of inches, enclosing the mic element, is plastic.

Haven't got it hooked up yet, need to solder up some xlr to quarter inch plugs. Hopefully that will work without much hum, haven't had much hum in the system so far since I moved into the new office. Was tempted to use transformers, but dunno if line levels would saturate the little transformers found in mic transformer adapters.

If it hums, probably order some better-quality transformers some where that are made for line levels.

I can run xlr from the dbx main outs to the Jbl-urie amp's inputs, but need phone-to-xlr to feed from the patch bay into the dbx, and a single xlr-to-phone to feed the old powered subwoofer.

Post

So far been having fun with the driverack px. My Unbal->XLR cables mixer->dbx, and XLR->Unbal cable dbx->subwoofer, XLR connect to the old mains amp, no hum. Sounds clean, no noticeable hiss. Using the +4 db input level switch setting, 0 dB meter on the mixer shows -12 dB on the dbx, about perfect. Mixer meter has 12 dB headroom, red clip lights labeled +15. The dbx meters clip at 0 dB, so the old mixer and the dbx agree almost perfectly as to headroom.

DriveRack offers various starter programs for typical live powered speaker brand/models, and a couple of generic starter programs. And a wizard for auto setting levels with the pink noise and mic, and then the auto-eq wizard. Feedback notcher which I don't need. Compressors I don't plan to use. Might or might not enable the peak limiters later.

The auto-level wizard runs first. It seemed to adjust the sub in about the right ballpark relative to the mains. I could later get down on hands and knees and manually reduce the sub's gain with a screwdriver (the recessed gain knob is currently set halfway), but it seems as good to adjust the sub level from the dbx. The only benefit to reducing the gain using the gain knob on the subwoofer itself, would be to have more realistic sub metering on the dbx front panel.

For starters before trying auto-eq, I defeated the auto-sensed gain cut on the sub, set mains and sub to 0 dB. Experimented with various crossover freqs, 24 dB Linkwitz-Riley. Tried both listening and auto-eq varying the crossover from 100 hz down to 60 hz. 60 Hz seems to work pretty good.

Running the auto-eq wizard, the resulting eq had the gain cut on the sub bands about the same amount as its previous auto-gain suggestion for sub attenuation. So I manually edited the sub gain back to the suggested level, ran auto-eq again, and the resulting EQ settings are not very drastic at all. The driverack is very easy to play with.

====

Those old passive jbl 4208 speakers-- The factory specs flat +/- 2 dB from 60 Hz to 20 KHz. JBL's spec sheet picture of the factory anechoic plot is a little small for precision eyeballing, about a quarter page printed out.

Within the limits of what I can see in the little picture, it looks flat within +/- 1 dB or less from about 80 Hz to 900 Hz. Then the squiggles are a little bigger on up into the highs, but the average "straight line" level above 1 KHz is a dB or two lower than the average "straight line" level below 1 KHz.

Some old reviews of the speaker, complain that the speaker is a little too "middy" sounding, and the explanation may be in the factory chart. Two fairly flat segments, with the higher frequency line segment down a dB or two, but both segments relatively flat. Crossover is 1.8 KHz, so it is interesting that the "break" in freq response is not near the crossover freq.

===

In my room, after some experimentation with crossover freqs and such, the last-run of the auto-eq (third octave centers) settings came out--

Flat at 31.5 Hz and 40 Hz. Interesting it didn't see any need to boost or cut the two lowest bands.

-6 dB at 50 Hz, -2.5 dB at 60 and 80 Hz. Most likely a room bump, but apparently not drastic, at least at third-octave resolution.

-2 or -3 dB at 250, 315, and 630

-1.5 dB at 1 KHz.

1 or 2 dB boosts on all bands between 2 KHz and 10 KHz, except 4 KHz was left flat.

Also left 12.5 KHz and 16 KHz flat.

The room probably looks lots worse on a sine sweep, but at third-octave resolution I'd consider what the dbx auto-tweaked to be fairly minimal. It is interesting that the driverack was smart enough to raise bands that show low on the speaker's factory freq response chart, and cut bands that show high on the factory chart. Though most likely the mid/bass cut locations are also room resonances.

====

The better crossover and level matching between mains and sub improved the sound before doing the EQ. So far my ear doesn't mind the EQ. Sounds better with the EQ turned on. Only thing I hand tweaked-- Added a gratuitous +3 dB at 31.5 Hz.

The box has two parametric bands for the sub, and three parametric bands for the mains. Will try out computer analysis in the next week or two. It may be that if there are narrow low-mid room modes, that I could deal with em using narrow parametric cuts, and then remove or reduce those same cuts in the graphic eq.

Of course I might waver in my determination not to sound condition the room, depending on what more detailed testing might show. But it's not greatly offending my ears. If I was gonna do any commercial recording or mastering, would take it more seriously. :)

Post

Glad it is working for you JCJR, I would hate to have advised you to buy some unneeded equipment.
An important room acoustics point to keep in mind is, the speaker is separate from the room. The sound of both meshes over time. When EQing the speaker, you may achieve an even RTA spectrum, but the speaker AND room is telling a different story in their own way.

Post

I echo your sentiments, Camsr. :)

Post

camsr wrote:Glad it is working for you JCJR, I would hate to have advised you to buy some unneeded equipment.
An important room acoustics point to keep in mind is, the speaker is separate from the room. The sound of both meshes over time. When EQing the speaker, you may achieve an even RTA spectrum, but the speaker AND room is telling a different story in their own way.
Oh definitely. Everything has an impulse response, and the cumulative time domain impulse response of the speaker and room shows up as frequency in the pinking analysis.

There is something funky and excessively long-duration, in the room, in the lowest couple of octaves. Which may be fixable with what I'd be willing to do to the room, or maybe not. It is merely surprising that an accident of architecture has resulted in perhaps fewer problems than one might expect.

For one thing it is drastic near field. Don't even have a console between me and speakers, just a keyboard. Monitor midpoint between woofer and tweeter is exactly at ear level. The monitors are toed-in aimed at the head. Each monitor passes my head then fires straight into an open doorframe leading into halls and further rooms, so the strongest back wall near reflections one would expect, do not exist.

Because I'm so close to the monitors, side first reflections and ceiling first reflections can't be real loud, because the angles are so acute. The first reflection points are way off axis from the speakers. It doesn't sound especially good anywhere else in the room, but that is currently of little concern as long as it sounds ok at the sweet spot.

Need to listen more, do some more tests. About the only treatment can think of at the moment that wouldn't be inconvenient or ugly, would be perhaps three or four, 6' tall, 4' wide, compressed insulation movable gobos. Such might improve the look and not get in the way. I could place them in front of my shelves so it wouldn't be so ugly looking at all the tech junk on the shelves. Maybe build one sample gobo, then measure and listen with the gobo at various locations, see if it makes enough difference to encourage making more of them.

Post Reply

Return to “Hardware (Instruments and Effects)”