[woodsyi]

Theoretically with my room being 16'x22'x7'(8' discounting drop ceiling), my axial modes are  35Hz, 25Hz and 70Hz.  I have been doing some late night readings!   Do these manifest equally or one stronger then others?  In other words, does room interaction with all the tangential stuff leave us with one primary mode of resonance to deal with?  R-DES offers 4 channels of digital EQ with presumably different slopes at $400.  Parts Express plate amp with 1Kwatt RMS offering one channel of EQ with 24 dB slope is also $400.  If only one channel of EQ is sufficient to tame most bass troubles in a normal room, getting the amp, too, at the same price makes sense. But if more than one channel of EQ is needed for most rooms, then R-DES makes more sense.  Any thoughts on this?

[Marbles]

If you need a sub amp already, the amp route looks like a good first step, if you need more EQ than that, you can buy ABOUT ANY multi-band parametric or 31 band graphic EQ to tame those modes.

Parametric are usually a bit more precise than graphic, but at those low frequencies, a graphic can be very good as the bands are quite small.

If you don't need the amp, then go for a better EQ.

IMO IF you have those noodes, and you probably have those and a whole lot more, more than one band will be needed.

[ctviggen]

This also depends on how you tame the modes.  I'm thinking of purchasing the R-DES, but I have a problem in how my system is hooked up.  I split the signal from my preamp and run my RM40s full range, then use stereo subs to supplement the low end of the bass.  In my case, the R-DES would be used only on the subs, but these are currently crossed over at about 50Hz.  So, I'll only be able to get rid of the modes that are 50HZ or lower (although I might be able to deal with integration -- between the subs and the RM40s -- issues, too).  Tonight, I plan on taking measurements with ETF (a program that shows you low frequency response, among many other things).  The other option would be to use two R-DES modules, one for the bass of the RM40's and one for the subs (I'd have to biamp the RM40s to do this, but I have the cables for biamping).  That seems pretty complex and costly, especially when I could move up to the DEQ-X.

[ted_b]

Woodsyi,
You're putting the cart before the horse.  All rooms are different, even rooms with similar sizes.  The room nodes, the speaker placement, the quantity and quality of the foundation/walls/ceiling, the carpeting/windows/wall coverings, the adjacent rooms.  This doesn't even begin to consider your tolerances (i.e does a 37 hz trough bother you more than a 58hz peak).  I'd not spend any $$ on EQ or amp/EQ before spending a few dollars on measurement devices (this should probably be in the acoustic circle).  ETF is $150 or so and will not only measure your freq response at seating postion, but will also measure ringing and amount of time a bass note decays (through waterfall plots).  Just hook it up to a laptop or whatever pc you can get near the room.

I had the advantage of first using the DEQX to measure my room (I didn't keep the DEQX but that was because of my highly-analog system, not the fault of the DEQX's DSP).  My biggest issue is that I sit in the middle of my room in relation to the side walls, and sit in a 37hz null.  I will verify with ETF.  

Although the VMPS/Parts Express sub amp is an interesting idea (one EQ point at 37hz, bump that bad boy up 6 db), a true null is like a hole; it doesn't really get filled up with EQ boost, most just goes down the hole.

Soo, if I were you I'd measure the room, probably find a few troughs and a few peaks, then first treat the room with some traps (bass), blankets, Sonex (first reflections) whatever.  Then, only then, would I throw electronics at it.  Anything prior to these steps is shooting in the dark, IMO.

Ted

[JoshK]

This also depends on how you tame the modes.  I'm thinking of purchasing the R-DES, but I have a problem in how my system is hooked up.  I split the signal from my preamp and run my RM40s full range, then use stereo subs to supplement the low end of the bass.  In my case, the R-DES would be used only on the subs, but these are currently crossed over at about 50Hz.  So, I'll only be able to get rid of the modes that are 50HZ or lower (although I might be able to deal with integration -- between the subs and  ...

Unless your room is huge, there is a really good chance you don't have any room gain below 50hz.  This is easily calculateable based on the longest room dimension.  Now modes and reinforcement are different things and modes are not well dealt with with EQ of any sorts.  For modes you need to move your subs or treat your room.   Most all room gain is in the upper bass range which is why lots of people prefer dipole subs, that is another issue all together, but in this case it is your RM40s that need the EQ not your sub (presuming your sub is using a steep slope, otherwise its upper harmonics may cause problems).

[ekovalsky]

You will probably find three major nodes at 50, 70, and 140hz.  Probably with suck outs in between.  You will have smaller up beyond 300hz.   Biggest node will probably be around 70hz since the one first harmonic and one second harmonic will be near this frequency.

I agree 100% with the need for a measuring tool.  It is just about impossible to properly set up the VMPS without one -- it will let you get the L-pads set right and find the best possible positions for your speakers and listening position.  Moving the speaker and listening position just a few inches can increase or decrease the amplitude of the nulls and suck outs quite a bit, although they usually will not go away or change frequency.

I don't think a one band ParaEQ will be all that useful, but a four band will probably take care of the major peaks.  Beyond that, a TacT or DEQX (or Meridian) basically acts like an automatic infinite channel ParaEQ and will correct output to your target curve -- usually flat with a little rise in the lower-mid bass).

Fixixng suckouts is a lot harder than peaks.  With the TacT you can drop the target curve to the lowest level of the peaks, so output at other frequencies will be attenuated to the level of the peak.  By doing this you will achiee amazingly flat headphone-like frequency response, at the expense of maximum SPL.  With sensitive speakers like VMPS there is plenty of SPL to spare so this works quite well.

[ctviggen]

Well, I have some Real Traps for the room treatment, although I plan on more room treatment.  As for what my frequency response looks like, I'll see tonight when I use ETF.  If I do need to apply the R-DES to my RM40s only, then I might have to axe that idea.  While I have another two channels of amplification and have cables for biamping (which I did when I first received my RM40s), I don't think that this would work because I'd have to split the output of my preamp three ways -- one to the mids/upper part of the RM40s, one to the bass section of the RM40s, and one to my subs.  The only way I could do this would be to split the signal from my preamp two ways -- one would go to the NHT X2 active crossover.  From there, the high pass would go to the RM40s; the low pass would go to the subs.  The second signal from my preamp would go to the R-DES, then to the amps for the bass part of the RM40s.  That could work.  

Would it be worth it though?

[dwk]

Unless your room is huge, there is a really good chance you don't have any room gain below 50hz.  .

Uh, you have it backwards Josh.  Room gain (or cabin gain) begins to occur BELOW the lowest mode in the room.  In typical rooms this is in the 30-45Hz range, but in large rooms it is lower. Since the room at this point is operating as a single pressurized volume, only straight linear eq is needed to deal with this. In a perfect world, the natural rolloff of your speakers would match the room cain boost, and no EQ in this area is necessary at all. (This is one of the reasons the NHT1259 sub has such a great reputation - the 30-35Hz corner freq and 12dB rolloff below that actually is a very good match for many rooms).

In the modal zone - stretching from the ~30-45 Hz point of the lowest mode up to the point where the modes become dense enough that they effectively merge (the Schroeder freq)- is where room problems manifest themselves.  This is why *BIG* rooms are great - the Schroeder freq is at or below 20Hz, so there are no modal problems to deal with. No room gain either, though, so you need buckets of excursion and power.

In answer to the original poster, the EQ necessary for any particular room is so specific that generalizations are not worth much.  If your room dimensions are close to equal, you may get something that looks a lot like a single dominant mode, but more likely you'll have several to deal with. Parametric eq is one aspect of the solution, but don't expect it to magically solve everything.  If your modes are very hi-Q, it can do a decent job as the notch won't have a broad effect. However, if you have low-Q modes, parametric by itself will probably not be enough.

I do agree with Josh, though - EQ of any sort is the LAST think you do. Make it as good as you can with placement and passive treatments, and only then bring in the EQ.

[woodsyi]

Ted,

I know I have to do the measurements -- I got the software and hardware for this but got no time.   I hope we get a blizzard that will trap me in the house for a couple of days.    Just last weekend, I did put up 2 mondo traps on front wall-wall corner, 3 mini traps on LCR wall-ceiling corners and 5 micro traps at LCR wall and ceiling 1st reflection points as well as 80 sq. ft. of  12" insulation batt in the corner cavities above the ceiling tiles.  I definitely hear difference in tighter bass and coherency in mid/high frequencies from my listening position.  Everynight, I plan to start the measurment process, but I end up just listening and enjoying the music instead!   I will try again tonight!  I also just got my Hurricanes back from Bill and they make seductive music!

[eico1]

I think notch filters are best suited for room correction. Graphics definitely do not provide this function and only some parametrics do.

Here is Meridians take on room correction:


http://www.meridian-audio.com/w_paper/Room_Correction_scr.pdf

steve

[nathanm]

If your room dimensions are close to equal, you may get something that looks a lot like a single dominant mode, but more likely you'll have several to deal with.

I can concur with that.  In my room which is a few feet shy of square I only need one band of EQ (parametric) to tame the 40Hz mega spike.  But that is not to say that I get amazingly flat and tight bass, just that there's no 40Hz mega spike.  It's a 'lesser evil' tweak, not a true 'good' tweak.  Boosting the nulls has been a hit or miss affair so usually I leave off the other three bands.

[JoshK]

dwk,

You are absolutely right, I did have it backwards.  That makes sense though why dipoles would work above that freq better then because they excite less modes (supposedly).

[Russtafarian]

I just re-EQ’d my room this past weekend and ended up using just two parametric bands.  Read on if you’d like to be bored with the details.

First, an overview of my speaker system:  The mains are Gallo Nucleus Reference driven by 60 watt tube amps that give me reasonably decent output to about 40Hz.  Rear channel speakers are Magnepan MMG-Ws driven by two channels of a DIY six channel Tripath TK-2150 amp.  No center channel.

The sub system consists of stereo and multi-channel outputs into an Outlaw ICBM crossover.  Stereo ICBM subwoofer outputs into a Behringer DSP1124P stereo parametric EQ.  Stereo EQ output to the Tripath amp.  I use the four remaining Tripath channels (bridging each pair to mono) to drive two dipole subs using Adire Tempest 15” drivers.  The two dipoles are placed half way along the left and right side walls with the baffle openings facing the front and back walls.  The dipoles are run in stereo with the LFE and rear signals from multi-channel sources mixed to them by the ICBM.

I’ve had the ICBM since Christmas and have been playing with crossover frequencies, slopes, etc.  I first tried running the main L & R signal through the ICBM but found that the crossover’s high pass signal that feed the tube amps & Gallos was unacceptably colored.  So I now run my mains full range and send a parallel stereo signal to the ICBM and cross it over at 40Hz.  

When I measure the room (with both subs and mains playing) using 1/6 octave tones between 16Hz and 160Hz and an RS SPL meter I see peeks at 36, 71, and 125Hz.   I turned the sub output of the ICBM up all the way to get as much gain at 25Hz as possible.  20Hz is only 6dB down at this level (measured by the RS, which is a few dB off at these frequencies).  Using a deep but narrow 36Hz parametric notch gives me a nice +3db peak at 36Hz.  I could get it flat, but I like a little extra meat at that frequency.  Everything else measures within +- 4dB between 25 and 160Hz except those nasty peaks at 71 an 125Hz.

Since the ICBM crossover and Behringer EQ only affect the subs and not the mains, the challenge is how to tame the +12 dB 71 and 125Hz peaks coming through the mains.  I don’t perceive that the 125Hz peak is dramatically coloring the music so I’m just letting it go for now.  Maybe room treatment can help, but I’ll save that project for later.  

The 71Hz peak is another story.  It’s really irritating and severely bloats the mid bass presentation of everything I play.  However, by switching the polarity of one of the subs I can create a null between the subs and mains at 71Hz and a +6dB peak at 63Hz.  By notching the EQ just 3dB at 63 Hz, I get a gentle 4dB peak in the mains at 63 and 72Hz.  Much better!This phase interaction between the mains and subs is fascinating!

I’m now measuring an in-room response of +-4db from 25Hz up to just below 125Hz.  Mission accomplished.  And it sounds good too!

[woodsyi]

Russtafarian,

Not boring at all.  I just did as much room treatment as I can with the time, money and WAF.  I can hear tighter bass, but it is time that I learn the software (trueRTA level 4) and do some measurements.  I also have overlap between the mains and sub between 50 and 80Hz.  I may need to order bunch more frequency modules for my Marchand X-over and another sub before I am done with all this.  Thanks for your information.

[John Casler]

This also depends on how you tame the modes.  I'm thinking of purchasing the R-DES, but I have a problem in how my system is hooked up.  I split the signal from my preamp and run my RM40s full range, then use stereo subs to supplement the low end of the bass.  In my case, the R-DES would be used only on the subs, but these are currently crossed over at about 50Hz.  So, I'll only be able to get rid of the modes that are 50HZ or lower (although I might be able to deal with integration -- between the subs and  ...

Hi Bob,

As I understand it, the R-DES will create an equalization curve for whatever purpose you feed it.

If you take readings they should likely be for the "whole" system, which will then show the peaks and valleys of the system, not just the subs.

Then when you upload the corrections, they will correct using equilization to the subs, but they will (or theoretically should) correct the system as a whole, within the subs response range.

[ctviggen]

John, that's interesting.  I think I'm going to get the R-DES anyway, just to play around.  If I have to choose a fancier way of doing things, I can always do that.

[ryno]

If four bands just isn't enough, and you have a sub with a built in xover, http://www.partsexpress.com/pe/showdetl.cfm?&DID=7&Partnumber=248-656
A 12 band per channel parametric for $100, true RTA software for $100, a mic, mic preamp, and new sound card for ~$130. About the same as R-DES with more bands and everything needed for testing.
Ryan

[thayerg]

ryno

I like your logic. Especially after reading the meridian paper, which suggests that room eq is effective up to 250 hz. The R-DES only goes to 200.  Basic flaw.

[ryno]

If you have the EQ in line with the sub xover, I don't think the difference between 200 and 250hz will be noticed. Like ctviggen was saying, his mains played down to 40 or 50 hz so any eqing would be below that because of the subs xover. If your talking about putting the eq in line with the whole system, I don't know if the eq I mentioned would be a good choice.  You would go through a DA conversion with your cdp or DAC,  AD and DA again in the eq, some, most would find this objectionable. I use a different eq for this, a behringer DEQ2496 that has digital in and out, between a transport and a DAC. 10 band parametric, 31 band graphic, and dynamic eq. I haven't tried the dynamic yet, I think it lets you turn off/on eq bands based on the volume level.
I agree with the other posters who say room treatment is the best , first way to handle frequency response, but my rooms are not dedicated to listening so I can't treat to the extreme. I find eq a good compromise if used correct.
Ryan

[ctviggen]

I've decided that I'd have to buy two R-DESs if I want to tame all the peaks I have.  Using ETF, I've found that I have a nasty peak at about 28Hz (approximately the calculated frequency based on room dimensions).  This is correctable by using an R-DES (or other EQ) on my subs.  However, the other 3-4 peaks, which are above 10db each, are above the operating point of the subs.  (Note that I did a quick test with ETF software and did not apply the Radio Shack compensating data, so these might be a bit off.)  This means that I'd have to buy another R-DES, which would be applied to the bass input on my RM40s.  I'd have to biamp my RM40s, which I can do and have done in the past.  However, that's $800 (MSRP) in two R-DESs, plus a bunch of money in interconnects.  Should I do this or just wait to try to find a used DEQ-X or TACT?  I wouldn't need any new interconnects and would have something that included its own calibrated mic and software and applies to the whole spectrum, not just the first few octaves.