PDA

View Full Version : RE: DSP for the Room


Frank Stearns
April 9th 15, 08:56 PM
Thanks to all who responded to this thread...

To answer the questions/comments posed:

- The room is 1st order Eigentone, just like the last one, with double sheet rock
walls and ceiling, and a fairly dense floor. The Eigentone dimension means that LF
energy build up, while still present, is more even and tends to stay near the
boundaries. This is what the measurements show. +6 to +10 dB build-ups exist near
the boundaries (but by no means the unweildy +12 to +18 you'd find in a typical
room). No deep nulls exist at any place else in the room. The actual dimensions are
in the 11x15x8 range, but that's not exact.

- The room is well treated, with various modes of difussion, absorption, reflection
and LF trapping (19 traps of various depths and dimensions, including 6 angle traps
going up and down the rear corners and left/right upper sides -- probably the most
effective trapping considering the size of the trap).

(To Gary: no, this is not a case of overly damped HF leaving behind more LF. Plenty
of diffusion and reflection happening.)

- Neil and Scott's sub-20hz idea is interesting, and one wall-spanning rear-room
trap is quite large and deep and likely easily dipped into the sub 20 hz region.
However, I switched to a lighter-mass diaphragm to get a little higher usable range.
Saw no diff with the main issue of a slight and broad LF rise, nor in the sub 20 hz
region (but who knows if the measurements were of much use below 20hz). This
diaphragm change did yield some taming higher up, between 25 and 60 hz, at the
client couch toward the back of the room.

Any suggestions for a measurement system that might give me better data below 20 hz?

- Analog filters, passive or active, are non-starters for many reasons.

- Fortunately, the EQ required is not extreme nor with peaks and dips all over the
place. For my faux room EQ inside a PT session I've used three filters in two EQ
plug-ins. First, -3.5 dB shelved so that the transition to flat starts at 200 hz and
gets there by 300 hz. Then, one narrow -4 dB notch at 31 hz and one very narrow +4
dB peak at 300.

This measures closer to flat though not all the way there by about 1.5 dB. As one
poster replied (and I agree) making it absolutely flat might not be the practical
ideal nor sound the best.

(I was just listening to a mastering proof of a CD I finished mixing for a Sony
Records affiliate. This was on my cheapo shop boombox and things in the low end
translated very well.)

- I have indeed done many physical measurements trying to relate distances to
frequencies. (Years ago I wrote a handy little app that given wavelength, period, or
frequency, calculated the other two.)

But the relationship just is not there directly -- that's been one of the nutty
things. I suspect the problem might be one of those diagonal interactions I've read
about -- less obvious than the typical standing wave between untreated parallel
surfaces, but still a potential issue.

- it also might indeed be room gain; this room is slightly more solid than the old
room, at least in the flooring (old room was also double sheetrock, wall and
ceiling), so perhaps not as much is getting passed through the floor here.

- Problem with the DEQX or Trinnov is the expense (US$5K for the DEQX and the
Trinnov might be even more). I only need the digital portion of what they do -- so
that's a lot of money in those boxes that would be unused -- assuming they've given
the care they claim to the analog outs. (I already have a pristine and somewhat
expensive analog path from the monitor controller out to the crossover and
amplifiers that would be hard to match; don't want to lose that.)

- The main diff in this new room compared to the old room is the size and shape of
the soffits. These are toe'd in at less of an angle, are not down-angled, and are
somewhat deeper with a little more sheetrock thickness than in the old room. They
each also have way more sand in them than the old ones. It's conceivable the soffits
in the old room were performing some sort of trapping action.

- I shudder at putting a Behringer-anything in that room, but it might be a good
experimental item. It appears to have 10 programmable filter sets (I only need 3 per
channel), so that would be fine -- if I can stay away from their typically cheesy
analog signal paths. The only question is whether I can trust how well it handles a
digital signal in terms of jitter and doing the correct arithmetic...

Is there a similar product made by someone else with better build quality? (I
wonder which company/product was copied by Behringer for this product?)

Sorry to be so long-winded, but a lot of interesting questions and comments were
posted.

Thanks again for any add-on comments/suggestions,
Frank
Mobile Audio
--

Peter Larsen[_3_]
April 10th 15, 06:09 AM
"Frank Stearns" > skrev i en meddelelse
...

> Thanks again for any add-on comments/suggestions,

I reiterate: Behringer DEQ2496, it is not costly, you may or may not want to
later use another processor but then you'll find some other use for it and
getting one will allow you to research what processing can do for that room
at a modest outlay, I paid about USD 120 second hand for one with ECM8000
included.

> Frank
> Mobile Audio

Kind regards

Peter Larsen

geoff
April 10th 15, 07:56 AM
On 10/04/2015 5:09 p.m., Peter Larsen wrote:
> "Frank Stearns" > skrev i en meddelelse
> ...
>
>> Thanks again for any add-on comments/suggestions,
>
> I reiterate: Behringer DEQ2496, it is not costly, you may or may not want to
> later use another processor but then you'll find some other use for it and
> getting one will allow you to research what processing can do for that room
> at a modest outlay, I paid about USD 120 second hand for one with ECM8000
> included.
>
>> Frank
>> Mobile Audio
>

Or as previously mentioned DEQX HDP Series ( http://www.deqx.com/# )

A DSP unit with modules for speaker eq (including individual driver
time-alignment), room EQ, and general EQ.

I got the HDP-Express (ex-demo ?) to do a more accurate speaker EQ job
than the KEF KUBE on my Reference 107s and got the other functions as a
bonus , but not sure I would pay the current RRP - in fact I'm sure I
couldn't/wouldn't !

FWIW I've also got a Beh DEQ2946, but for a different purpose (PA rather
than mastering and pleasure-listening).

geoff

April 10th 15, 12:41 PM
- Fortunately, the EQ required is not extreme nor with peaks and dips all over the
place. For my faux room EQ inside a PT session I've used three filters in two EQ
plug-ins. First, -3.5 dB shelved so that the transition to flat starts at 200 hz and
gets there by 300 hz. Then, one narrow -4 dB notch at 31 hz and one very narrow +4
dB peak ....
--
. I would like to suggest that you listen to what it sounds like with the very narrow 300 Hz peak
Eq turned off. Very narrow peaks or nulls create ringing and ringing at 300 Hz cant be good.
If turning off the 300 Hz peak improves the sound, you may want to figure out what causes the
300 Hz dip and address it directly instead of with Eq.

Mark

Frank Stearns
April 10th 15, 01:46 PM
writes:
>- Fortunately, the EQ required is not extreme nor with peaks and dips all over the
>place. For my faux room EQ inside a PT session I've used three filters in two EQ
>plug-ins. First, -3.5 dB shelved so that the transition to flat starts at 200 hz and
>gets there by 300 hz. Then, one narrow -4 dB notch at 31 hz and one very narrow +4
>dB peak ....
>--

> . I would like to suggest that you listen to what it sounds like with the very
narrow 300 Hz peak
>Eq turned off. Very narrow peaks or nulls create ringing and ringing at 300 Hz cant be good. >If turning off the 300 Hz peak improves the sound,
you may want to figure out what causes the >300 Hz dip and address it directly
instead of with Eq.

True (and thanks for pointing that out), but this is a fairly small nudge. Plus,
this small dip appears in the waterfalls all the way through to 600 ms. As
mentioned, I've chased these various things and cannot find the direct dimensional
correlations. And again, they've been consistent through the waterfalls; the RT30
below 500 hz has been excellent; so this room seems to be the exception where EQ
could actually be helpful.

And I have used this room for 6+ months with no eq. It's doable, but uncomfortable.
I'd spend too much time second-guessing. This lastest mix project, using this
quick'n'dirty room EQ, felt much more "normal". I mixed based on what I heard,
rather than translating what I was hearing. Clients were happy; project translated
well. Seems like I'm on the right track.

And indeed, as you rightly point out, 300 hz is important; that's why I do not want
to have any misrepresentations in that range. In terms of hearing it corrected
versus uncorrected, that's going to depend on program material, given the small size
of the dip. I might be able to take that in and out all day and not hear a diff
until something was actually landing at that point.

I'm a little too busy right now to be listening like that. So, minor corrections
based on measurements, supported by good performance in most areas, seem most
cost-effective both in money and time. It also reassures proper performance when
program material does appear at the frequency.

The frustration now is finding an appropriate equalizer with digital i/o
(3 full parametrics per side, not just a graphic (yarg) or a channel strip). Much to
my consternation, the Behringer appears to be the only box that will do that, even
as one approaches the $5K mark. And I can't do $10,000 for a Weiss or
equivalent. Still searching, though.

Frank
Mobile Audio


--

Scott Dorsey
April 10th 15, 03:14 PM
Frank Stearns > wrote:
>True (and thanks for pointing that out), but this is a fairly small nudge. Plus,
>this small dip appears in the waterfalls all the way through to 600 ms. As
>mentioned, I've chased these various things and cannot find the direct dimensional
>correlations. And again, they've been consistent through the waterfalls; the RT30
>below 500 hz has been excellent; so this room seems to be the exception where EQ
>could actually be helpful.

If it's not the speaker... and it's not a transverse mode or a longitudinal
mode, and it's not something in the room flexing (like a steel cabinet full
of Scully parts that forms a diaphragm resonator when the doors are closed
but not when the doors are open).... then it's some wacky high order oblique
or tangential mode. Don't give up!

>And I have used this room for 6+ months with no eq. It's doable, but uncomfortable.
>I'd spend too much time second-guessing. This lastest mix project, using this
>quick'n'dirty room EQ, felt much more "normal". I mixed based on what I heard,
>rather than translating what I was hearing. Clients were happy; project translated
>well. Seems like I'm on the right track.

Try the Behringer! But don't give up!

And consider also a Helmholz resonator added to the room... generally I am not
a fan of the things but it's easy enough to knock one together from 1/4"
plywood and 2x4s for testing. It'll tell you if the problem really is down
lower than you think or not.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

hank alrich
April 17th 15, 06:51 PM
Frank Stearns > wrote:

> I shudder at putting a Behringer-anything in that room, but it might be a good
> experimental item.

I understand that sentiment, but the DEQ 2496 and the DCX 2496 are very
good pieces of kit. I don't have time to repeat my experiences with
them, but fear not.

Build quality is quite good, excepting ocassional power supply failures,
which in new units seem to occur quickly. I have installed several of
the DEQ's that have been running 10 to 12 or more hours/day 363
days/year since early 2008.

--
shut up and play your guitar * HankAlrich.Com
HankandShaidriMusic.Com
YouTube.Com/WalkinayMusic

Frank Stearns
April 17th 15, 11:22 PM
(hank alrich) writes:

>Frank Stearns > wrote:

>> I shudder at putting a Behringer-anything in that room, but it might be a good
>> experimental item.

>I understand that sentiment, but the DEQ 2496 and the DCX 2496 are very
>good pieces of kit. I don't have time to repeat my experiences with
>them, but fear not.

>Build quality is quite good, excepting ocassional power supply failures,
>which in new units seem to occur quickly. I have installed several of
>the DEQ's that have been running 10 to 12 or more hours/day 363
>days/year since early 2008.

Thanks, Hank. That's an unexpected but solid testimonial.

I did remember that my old TC Electronic M1XL reverb unit had some "spare" DSP for
EQ and dynamics, along with digital I/O. I got it back from a long-term loan
situation and have put it inline, but am not so sure I like the sound of the EQ.

Assuming the Behringer units are 10+ years newer DSP engineering than the TC
Electronic (which was a great reverb at that time), I'm hoping those Behringer units
might be a little more transparent.

Any thoughts about the sound (or hopefully lack of a sound) of the Behringers?

Thanks again,

Frank
Mobile Audio
--

hank alrich
April 19th 15, 09:53 PM
Frank Stearns > wrote:

> (hank alrich) writes:
>
> >Frank Stearns > wrote:
>
> >> I shudder at putting a Behringer-anything in that room, but it might be
> >> a good experimental item.
>
> >I understand that sentiment, but the DEQ 2496 and the DCX 2496 are very
> >good pieces of kit. I don't have time to repeat my experiences with
> >them, but fear not.
>
> >Build quality is quite good, excepting ocassional power supply failures,
> >which in new units seem to occur quickly. I have installed several of
> >the DEQ's that have been running 10 to 12 or more hours/day 363
> >days/year since early 2008.
>
> Thanks, Hank. That's an unexpected but solid testimonial.
>
> I did remember that my old TC Electronic M1XL reverb unit had some "spare"
> DSP for EQ and dynamics, along with digital I/O. I got it back from a
> long-term loan situation and have put it inline, but am not so sure I like
> the sound of the EQ.
>
> Assuming the Behringer units are 10+ years newer DSP engineering than the
> TC Electronic (which was a great reverb at that time), I'm hoping those
> Behringer units might be a little more transparent.
>
> Any thoughts about the sound (or hopefully lack of a sound) of the
> Behringers?

Unobtrusive. Enough so that in every context in which I have used them
in good-quality live sound applications, there has always been something
way more important to worry about in pursuit of "transparency".

Offers an array of extremely useful functions simultaneously.

--
shut up and play your guitar * HankAlrich.Com
HankandShaidriMusic.Com
YouTube.Com/WalkinayMusic