Bfd Room Repair -my Bfd Story (1 Viewer)

[DanRA]

Hi All,
Below are Links to 3 graphs (I hope). I Tried to put the graphs below but if they don't show up you can use the links at the bottom.
First , I will describe my setup. I use a Lexicon DC-1 V4.0 for my Pre/Pro. I have Polk Signature SDS SRS 1.2 mains. These are very big full range mains. These are setup Bi-amped using a Mark Levinson 23.5 for the LF's and a Mark Levinson 29 for the tweets. The mains are setup on slides with custom locking clamps so I can position them in HS position or HT position in a few seconds. It is quite a feat because they weigh 185 #'s each and you can slide them back and forth with one hand. The mains also have heavily modified crossovers, new binding posts and all soldered connections inside. I am using a Behringer Ultracurve for the Mains but only the LF's. I looked at the FBD but the Ultracurve had slightly better specs. Using the Ultracurve on my mains I wanted as good as possible. My HS/HT is in the basement and has some nasty room mode that makes music with a fair amount of bass fatiguing and annoying. Most here use the FBD for their subs but my mains go very low and I like it that way. Don't try to tell me to cross them higher I have tried it and don't like it. The mid-bass punch and LF from the 1.2's will knock you off your chair. I use two Snell 550 subs with updated drivers that are driven by a Carver 35x THX amp. I have the Snells crossed at 40hz to help only with very LF's. All of this can be seen on my extensive web page links below.
I graphed my right main first with and W/O the subs and got a pretty good first pass correction I was happy to make such good progress (First graph labeled Right main). The Dark Blue curve shows the Right main & Right sub W/NO EQ. The Magenta curve shows the Right Main and Right sub with the peaks reduced @ about 67 & 94hz. It has a couple of dips that need some work but the peaks were smoothed out nicely and sounded much better than the peaks. I played with boost but didn't like the results. I will also note the anal side of me made me modify the excel worksheet I downloaded to add 1 hz increments and I also downloaded all the sine wave files (much thanks Sonnie and other contributors!!!).
I graphed my Left main W/my Left sub and had a big surprise. I had this huge dip from about 26-43Hz range I couldn't figure out (Second graph Dark Blue curve). If you look at the Dark blue curve it is also shows the peaks above about 50hz (NO EQ). I removed these peaks in all the other graphs using the Parametric filters so all the other graphs show much better results above about 50hz. It is simply amazing the improvement after eliminating the peaks at about 67 and 92hz. I would never go back and finally can listen to music with LF's. When I complete the LF issues I will fine tune a little more but it looks pretty good for now. This 26-43Hz section in the LF's was so big I thought something was wrong so I checked each driver. Well after playing for a while, I found out the sub on the Left side has almost no effect!!!! You can see this if you compare the Dark Blue curve vs the Red Curve, There is a pretty serious room effect in the Left corner. I turned on the right sub turned off the left sub and boom it looks much better, not perfect but much better (Second graph baby Blue curve). It shows very similar dips in all the curves and as the right channel @ about29 and 39 but raised the area about 10db over the left sub reading. The Magenta graph the left shows the left main and left sub with boost. The curve doesn't look so bad but it sounded like someone puked on a microphone (bad option). The Red graph shows the mains with no sub. Again the red graphs show the dips @ 30ish & 40ish that show in the right channel but that 26-43Hz area about 10 db lower.
Now I know I was on the right track but still didn't like the 2 dips @ 30ish & 40ish even though it sounds pretty good. I moved the Left sub around played with phase but it didn't help those two dips. Well I decided to try stacking the 2 subs in the right corner to see what would happen. I had heard this was a good thing to try so what the hell. The 3rd graph shows the original Dark Blue curve with No EQ and Left sub for comparison from the start. The Magenta curve shows the Left main & 2 Snell subs stacked in the Right corner of the room. This still didn't help the 2 dips @ 30ish & 40ish but the level is up considerably higher (near 85 db @ 10hz!). Now for my take on this and I am looking for comments suggestions and if people agree on what I am thinking. I am thinking I will leave the 2 subs stacked in the right corner but add a BFD to the SUB's and cut the 23 & 35hz peaks (see graph 3). I currently have no EQ on the subs. I figure this will effectively raise the 2 dips without any EQ boost. I figure I can attain a very good curve in the LF's and have zero boost. What do you think?????? The Magenta curve in the 3rd graph shows no boost at all, only cuts.
SIDE NOTE:
I don't think most of you would have picked up one this but my first readings used one sub for each channel measurement even though the sub output from the Prepro is single (not related to right and left). This means LF's routed to the sub would have been to both subs in actual use. I probably wouldn't have noticed this because the placement of that Left sub had almost no effect anyway but I figured I should note my original measurements were not done correctly using each sub with each channel. I realized this when I figured out what was happening with the room effect. Also, in hindsight it was a good mistake because if I had taken measurements with both subs on I might not have found the left room effect relating to the sub as quickly. In other words the left sub would not have been helping and I wouldn't have known it with out testing them separately. All future work will be done with both subs and checking individually to be sure. Interestingly the Right side of the room is cement foundation corner 10" thk. The Left corner is a studded side wall. I assume my different results with the two sides are due to the corner differences. Any Acoustical geniuses want to chime in here about if it is worth doing anything about this. I was thinking I could stiffen up the wall but not sure if it would be worth the effort because it is only very LF that seems to be effected.
GRAPH 1 - RIGHT MAIN WITH RIGHT SUB:
Link Removed
GRAPH 2 - LEFT MAIN 4 CURVES:
Link Removed
GRAPH 3 - LEFT MAIN USING 2 SUBS STACKED IN THE RIGHT CORNER:
Link Removed
LINKS TO GRAPHS & WEB PAGE LINKS INCLUDED ALSO:
GRAPH 1 - RIGHT MAIN WITH RIGHT SUB:
http://photos.yahoo.com/bc/dhummm/vw...iew=t&.hires=t
GRAPH 2 - LEFT MAIN 4 CURVES:
http://photos.yahoo.com/bc/dhummm/vw...iew=t&.hires=t
GRAPH 3 - LEFT MAIN USING 2 SUBS STACKED IN THE RIGHT CORNER:
http://photos.yahoo.com/bc/dhummm/vw...iew=t&.hires=t
HOME PAGE (7 pages of GEEKLAND):
http://www.geocities.com/dhummm/index.html
FRONT VIEW HT AND EQUIPMENT:
http://www.geocities.com/dhummm/page3.html
Sorry for the long post but the geeks will appreciate it (I think)
Ralph-

 

[Ned]

The nice hump in the lower range is nice, I'd want to smooth out that big peak at 35hz though. Looks great otherwise.

 

[Sonnie Parker]

Ralph,

I'm getting dizzy trying to read your widescreen post. Could you shrink your graphs a little?

I noticed your graphs are on 1/12 octave. What type of mic did you use and were mic corrections needed?

 

[DanRA]

Hi Sonnie,
I downloaded all the sine wave files from troycable thru your link. I made up a CD that has frequiencies 10-99 match the track numbers which made it a snap to keep track of which freq I was on. I just put the higher freq on track 1-9 so 10-99 would line up.
http://www.troycable.net/~snap/sinewaves/
I used a radio shack meter with correction factors (from this forum forget link). The Ultracurve has a sine generator but It is only hooked up to the LF mains so the CD allows me to blend everything togeather better from the same source.
I modified the excel worksheet from troycable also thru your link.
http://www.troycable.net/~snap/bfdin...btestmodel.xls
I guess I should also thank Troycable for putting up this info.
Sorry about the wide pics. I tried to make the graph as small as possible and still be able to read it. I will try to make it smaller next time I post. I have high screen resolution and can see the entire graph width full view which is easy for me. I also have a 21" monitor which helps the higher resolution. Also, thanks again for putting all that info and links on your web page. It made life easy for me!!!! Thanks fo rthe comments.
Also general note,
The first graph above shows only one sub. When I re-graphed with two subs stacked the right looks very similar to the left in the 10-45hz range and that's why I think adding the BFD to the sub makes sense in fixing up the very LF. I wasn't sure if the difference in the two graphs might be confusing but was to lazy to re post.
Ralph-

 

[Sonnie Parker]

Ralph, (sorry, where did I get Bob from.... lol)
I'll thank TroyCable for ya!
The reason why I say this about the width is a lot of folks will open up this thread and then shut it down due to the widescreen mode. Some are truly offended. No biggie to me but some it will be.
Many a person will use 800 X 600 screen resolution which will make it scroll left and right forever and ever. I use 1024 X 768 which I believe is fairly normal and excepted for viewing websites. At least the major sites build their pages so that it fits with the latter resolution. You must have some blank space on each side of most pages you view.
As far as the graph and measurements. If you used a Radio Shack Meter, where did you get the correction values for 1/12 octave....... I know I've seen 1/6 (I have those) but I have not seen 1/12. Did you estimate these?
If you will take out the 1/12 and just show 1/6 and post your SPL readings as well, you'll be on the same page as most of us here and might get a little more response on this thread. (Especially if you'll resize those graphs above.....it would be easy to do.) You could use the graph on the excel workbook and just plug in your numbers. It's very readable in a smaller size.

 

[brucek]

Ralph J,
I have to strongly agree with Sonnie's suggestion. While these graphs may look good on your 21" monitor, the fact is, most of the readers here will have 17" monitors and it makes it extremely difficult to read this thread. You'll get far more activity on your thread if you cut it down.
I do have a few comments that may or may not be helpful.
I'd like to talk about your scaling on the graphs. The method you've chosen tends to exaggerate the horizontal axis. It's not representative of the way we hear sounds.
A graph of course can have any scale you like to present data in many different ways. It doesn't change the data, it changes the appearance. You can have linear divisions with non-linear or logarithmic scaling or you can have logarithmic divisions with linear scaling or linear divisions with linear scaling etc, etc. You have chosen linear divisions with linear 1Hz scaling. Not really representative of octaves or the energy in the lower frequencies.
When a graph is made of the results of octave-band sound pressure level measurements, rather than individual linearly increasing frequencies as you've done, the frequency axis is commonly divided into equal intervals of octave based readings. This would be the case in a graph where each successive measurement is a sixth octave higher (or any octave multiple you like). This compresses the graph properly as the frequency increases along the lower axis. Alternately if you had measured instead at 10, 20, 30, 40, 50, 60, 70, 80 etc, then you could have chosen a logarithmic horizontal division.
Sixth octave response graphing is certainly suffice. Further resolution may help somewhat in choosing more accurate filter values, but it isn't necessary. Anyway, just a suggestion.
You of course know your own equipment better than anyone, but I wonder about the tradeoff you're accepting with the BUC (Ultra Curve) in line with your mains, even though it's only on the low frequency drivers of your bi-amp. I don't know the crossover point of these LF speakers, but to subject this chain to another A/D and D/A conversion seems unwise, given the high quality of your equipment. I support BFD and BUC devices in sub chains, because they are usually dealing in the under 80Hz range and so the sampling frequency is suffice and low level input distortion figures are very acceptable. But to sample the signal between the DC-1 and those beautiful ML amps and 1.2 speakers makes me wonder if the compromise is too great.
Anytime you insert an A to D sampler into an analog chain, there's going to have to be level matching. Theoretically to take full advantage of every bit, you want your loudest signal to approach the maximum input level of the sampler (without enabling the clipping compressor), since the number of bits used defines your dynamic range and as a result your S/N ratio.
If your loudest level for example, was just above the noise floor, then you certainly wouldn't be using all the bits available to define the voltage levels of the input samples. So, from the softest input level to the loudest input level, less bits would be used to define the samples and after conversion back to analog your resultant signal wouldn't represent the input accurately.
Every digital machine like this that you use will have the same problems. It's not a big deal at very low frequencies because subs will generally have far higher distortion than what the A/D & D/A will introduce. I suspect your LF drivers chain crosses quite high and may suffer if taken through another digital conversion.
Certainly, when you use 24 bit ADC's, as in the UltraCurve, you will perceive very little quantization error (noise), but there's always going to be quintile errors, but the more bits used in the conversion process, the less it's noticed.
I suspect you've backed yourself into the use of an equalizer for your mains because you've decided to run them full range and so they are at the mercy of some nasty room modes at the very low frequencies. I was in the same situation when I first purchased my $11K ProAcs. There was no way I was going to bass manage those speakers since they were so capable down low. But there's two problem with that thinking. The low frequencies from your mains will be subject to room modes just as a sub is, and so will require equalization, which adds another device in your pristine main speaker chain. You've partially solved that by bi-amping. The other problem is when you run your mains full range and also use a sub (as you do) there is a real risk of signal cancellation because of common out of phase signals. For example, if you produce a 50Hz signal with your mains through an EQ that adds in 180degrees of phase shift, then you also play that same 50Hz signal through your sub without any phase shift, guess what the resultant signal will be - zero. There's a lot of problems associated with running your mains full range and also using, not just one, but two subs. especially if the two subs are not co-located. Yikes.......you have your work cut out for you.
Here's my lame suggestion that you said not to mention. Your red graph line, which is your mains only, shows that your mains would enjoy being crossed at around 50 or 60Hz. If you bass manage crossed your Lex at this frequency and removed the BUC from the mains, then use the BUC or a BFD on your two stacked subs in the right corner, equalized with a nice house curve up to their bass managed 50 or 60Hz - what would it sound like? This would give you a smooth equalized response from the subs up to 50 or 60Hz, and then the mains would take over, unfettered by any further digitizing.
brucek

 

[DanRA]

Hi Guys,
Sonnie first,
I got the corrections from this link:
http://www.hometheaterforum.com/htfo...threadid=53333
Thanks to Kerry Hackney (about a 1/3 into the thread)
Kerry says "I used MS Excel to do a non-linear progression for the in-between values... "
I compared thes to some of the other correction values around and they seemed close. Let me know what you think of them?
I will use smaller graphs in the future! I get the point guys!!!! If I have time I will edit with smaller graphs.
Hi brucek
I mostly used 1 hz increments to fine tune the filters as you say. I have looked at my raw data and I must say I did notice the 1/6 still would have shown the peaks and dips. This surprised me and I assume it is because of the reasons you described. I just figured more data is better and wanted to prove to myself if it is worth while (the engineer in me “trust but verify”). I may or may not use 1/12th data points but will only post 1/6th readings in the future (I promise). I also will say I don't understand the scaling of the graphs (yet) so I will study your comments and research more to understand (thanks for the info).
Now the tough one. I resisted using an EQ on my mains for a long time but found I just didn't listen to music at all anymore and couldn't justify a separate music room. My old music room in my old room sounded great and I miss it. As you guessed the bi-amping is to eliminate the EQ from the HF's. The EQ was noticeable with the HF's. The EQ is noticeable in the LF's as long as I don't add any boost. If I add any boost at all you can hear the wind blowing effect I have heard others talk about. I don't use any boost at all for that reason. Right now I can't tell the difference with the EQ in the chain which is amazing to me (as long as I don’t hit the room peaks). With more time I may get more discriminating and notice something I don't like but for now it is heaven. I have never crossed my mains at 50 and I suppose I could and should try it. I have tried it at 70 and 80 and just lost all mid bass punch. This is the strongest point of the 1.2's and I wouldn't give this up for anything. I listen to a lot of speakers and this is where I leave saying to my self "I wouldn’t give up my 1.2's for those if you gave them to me". I don't want to start a debate hear but I have yet to hear a speaker system crossed high that gives the same effect and it may be totally opinion but that's the way I like it. Crossing at 50 might help deal with phase effects and the peak/dips down in the LF's but it won't help the peaks above that which is the only reason I added the EQ to the mains to begin with. I definitely will try crossing a little higher to se what happens. I currently am crossed at 40 and only recently had the capability to see what’s happening in the room. If you look at the dark blue curves in my graphs the peaks at about 67 & 93ish are really bad Even If I crossed at 80 I wouldn't be fixing the 93hz. I may be more sensitive than others, but when music hits those freq the room sings and I cringe! Like I said I just couldn't listen to music in that room even with different speakers it's a room effect not a speaker effect. Although speakers with poor mid base may sound better. I have had about 15 people in the room recently and bypassed the EQ from the computer and every single person no matter what their listening capability winces when I bypass the eq after listening with the EQ for even a short length of time.
As far as level matching, I am aware of the issues but again, I am still studying and learning. I know am not peaking but I am not sure how close am to optimizing the level. I will probably post separately when I get to that point.
There is no question you don't need double blind testing to hear the difference. I guess the improvement far outweighs the room problems is what it comes down to. But I hear you it still pains me to have it in there. Also, I have had to move the subs around a lot to eliminate canceling effects. I wouldn’t be surprised if I play around a lot more. I am learning, still keep an open mind and may find a way around this but until then I will have an EQ. Like you are commenting I am truly amazed that adding this into my holy chain is so transparent. Someday I will build another room but it will be much different after my experience with this room. Thanks for the suggestions and comments guys.
Keep em coming, I am learning!!
Ralph-

 

[brucek]

Ralph J,

I also will say I don't understand the scaling of the graphs (yet) so I will study your comments and research more to understand (thanks for the info).

Your confusion with regard to my graph scale comments is no doubt due to the rather poor way I explained it. Maybe I can be clearer on that point.
Generally, when decibels are used as the units on the vertical scale when graphing sound pressure, we would use logarithmic divisions on the horizontal axis. For example, if you looked at a logarithmic graph of 10Hz to 100Hz, you would see the scaling units on the horizontal was 10, 20, 30, 40, 50, 60, 70, 80, 90, 100 (linear). The logarithmic division spacing between 10 to 20 would be large and the divisions would get logarithmically smaller until you reach 100. This logarithmic nature is representative of human hearing.
Since an octave is a doubling of a frequency, if you then marked the octaves starting at 10Hz on this logarithmic graph I mentioned above, you would see that they were exactly "evenly" spaced along the horizontal axis. The space between 10Hz and 20Hz would be exactly the same spacing on the graph as 20Hz to 40Hz and so on with every doubling of the frequency. This even octave spacing would follow all the way up to 20KHz on a logarithmic graph.
So this is very convenient when you use a graphing tool like Excel. You can evenly space your horizontal axis lines and use octave scaling on the horizontal axis units and the resultant graph will be logarithmic, which matches human hearing.
In the case of your graph, you used linear rather than octave axis units combined with linear spacing. It makes the division from 20Hz to 40Hz twice as wide as the division between 10Hz and 20Hz. Look how far it is on your graph from 40Hz to 80Hz. This is only one octave and should be only as wide as the distance from 10Hz to 20Hz. It isn't really representative of how we hear and it also makes your graphs really wide. Check out the graph on Sonnie's site and you'll see what I mean......
brucek

 

[Wayne A. Pflughaupt]

Brucek,

Great point about possible phase issues in separately EQing subs and full-range speakers. Never had thought of that.

Ralph,

If you want to EQ the mains full-range, you might consider ditching the Ultracurve in favor of high-quality analog 1/3-octave equalizers. I recommend either AudioControl or Ashly as reasonably-priced alternatives. The AudioControl C-131 has stellar specs that rival the most expensive professional EQs on the market. The late-model Ashly G-131 or 231 models are also excellent, even if they don’t spec quite as well as the C-131. Both these models are discontinued (although Ashly has an updated version), but you can occasionally find the C-131 on e-bay for $250-300 each (you’ll need two of them), and the two-channel G-321 for $325-375.

Regards,

Wayne A. Pflughaupt

 

[Rick Radford]

>Great point about possible phase issues in separately EQing subs and full-range speakers<

Hmm. So from this I assume that mains with self-contained woofs (powered or passive?) would be tough to EQ too?

Say.. PhaseTechs or the ever popular DefTechs?

 

[brucek]

Rick,
You may have misinterpreted. The idea is that when one speaker produces a tone at a specific frequency and another speaker is producing the same tone at equal amplitude and 180 degrees out of phase there will be cancellation and varying degrees from that if you alter the amplitude and shift.
This fact, is one of the problems associated with using capable main speakers on "large" and also using an equalized sub - and worse if using two separated subs. You can easily be producing the same signals completely out of phase. This problem is greatly lessened by using bass management, giving the sub all frequencies below a crossover point and the mains above the crossover point. The two mains will of course be in phase (assuming you've hooked them up correctly). Then the only problems will occur at the crossover frequency where there are common frequencies shared by the mains and sub. The processor manufacturer has gone to some trouble to ensure that phase anaomalies are kept to a minimum within their crossovers. Speakers with self-contained woofs won't be any different in this regard. The point is, that when speakers are run full range, they will "share" a large range of the low frequency spectrum with a sub and the possibility of cancellation is exacerbated.
Now once you introduce an equalizer into the mix you're messing with the phase again because equalizers filters tend to introduce non-linear shifts across the spectrum. You could plot this shift of phase vs. frequency just like you do a SPL level vs. frequency graph. The graph would change with every filter added. This is one of the many reasons when you are equalizing your sub to do it first alone to ensure an equalized response without interaction from any other speakers. Then if you have two subs, equalize the second one alone and then turn on both subs and recheck the interaction between the two. Then add in the mains to check for interaction between the subs and mains, specifically at common frequencies about the crossover area.
While we're talking about it, I should add that a digital equalizer also introduces more than phase shifting. They also introduce a processing delay that must be compensated for using the distance setting on your processor. The DSP inside the equalizer requires time to crunch its numbers. This delay is about 1msec from the time the signal enters the equalizer to the time it exits. With a fixed delay there is a linear increase in shift with frequency and so can be compensated for with a fixed time delay from your processor itself. So, since the speed of sound is roughly 331m/sec or around 1000ft/sec, then the 1 msec DSP processing delay in the equalizer would account for about a foot in distance. If you add a foot to the distance you tell your receiver that your sub is away from your ears, then it will advance the sound the 1 msec required for that sub.
brucek

 

[Rick Radford]

>You may have misinterpreted.<

Heheh. It won't be the first time. nor the last, I fear!

>This problem is greatly lessened by using bass management, giving the sub all frequencies below a crossover point and the mains above the crossover point.<

I had not considered that before, but it makes sense. Excellent point. And overall, just excellent info.

>then the 1 msec DSP processing delay in the equalizer would account for about a foot in distance.<

I hadn't considered that either. Another excellent point!

I see that speaker placement and equalizers can create phase shift. Are room dynamics another source of phase shifting?

 

[brucek]

That's a good question Rick. It's unfortunately outside my field. Maybe someone who knows acoustics like Wayne P or Vince could tell us about that....

brucek

 

[DanRA]

Hi Bruce,

First of all I have to say, thanks so much for your help (& others), it is much appreciated. I have been driving blind for the past few months. I have been trying learn about EQ's, measurements, graphing, understanding room modes, computer hook up, MIDI, and more for the last few months on my own and I hadn't really gone to this sub forum because I was mostly working on fixing the higher freq above my sub settings. I hadn't realized this is where I should be!!!! I have not found any other place on the web discussing this sort of thing. Thanks HTF and SVS for your support.

It wasn't that I didn't understand your first post as much as I need to read things a bunch of times to get them thru my thick skull and hadn't had time to do that. Although I will say your second post did a better job of making it thru.

Now let me see if I have this straight. It sounds like you saying if I cross my mains at the same point as my subs that I may be better off with the interactions between them. Before your last post I was thinking you wanted me to cross higher just to EQ LF's instead of my mains which I agree if my room weren't so bad would be much better. Now I have the ultra curve hooked up to a computer and the Ultracurve software shows the EQ curve and the phase curves for each channel (very slick). I talked to Behringer and basically the tech guy said as long as you're less than 70 degrees your doing very well. Now, not understanding this stuff and I sure he knew it, I think he simplified the detail for me. Actually I sent him my files and he thought it looked pretty good. My phase curves are under 60 degrees and I noticed when I boosted I also shifted the phase above 70. Actually one place is even near 60 degrees. I would love to understand this a little better as well. I am definitely going to play with my crossover points. I hadn't considered the delay thing, I saw the option but didn't see it's application whit respect to the sub and mains position. I didn't see it as being a problem with the positions being so close but I will give it some additional thought. I thought it was only used in situations where speakers my be in different positions like a stage in a professional setting which I'm sure is the same except more exaggerated with greater distances between speakers. Also, I have another question, I have my two subs stacked now in the same plane with the same input (split from the Lex). Same amp cables etc and the driver are even mass matched directly from Snell (they are local and I went there and got a great tour). Wouldn't the room see these as one bigger sub??? Should I consider one bigger single sub or do you think as long as they are matched so such a degree they are OK. The current position and stacking is considerably better than the other positions I have tried when I compared interaction between the mains and subs.

Wayne

I have been looking for better EQ's and will say I will only be interested in Parametric. I found the 1/3 octave EQ's to wipe out a big section of freq's to be very noticeable. I prefer the surgical nature of the parametric filters. I have a Audio Control Richter Scale which is nice but again hitting a specific peak or even worse, two near each other with the 1/3 octave EQ is like cutting off your arm to fix a hangnail. The 1/3 octave EQ's would lower the peaks but would also knock down everything around them. IMHO the 1/3 octave is not a good option at all. I also had a DBX 2231 and found it to be nice but again the cutting off my arm thing. I was considering the Parametric BSS 926 or the BSS 920 if I could get the computer hookup under control I would have picked the 920 because they can be had cheap but you must have it hooked up to a PC. The BSS PEQ's are very expensive and my understanding is they are very nice. Does anyone here have experience with the BSS Parametric EQ's???? What Parametric EQ's would be better for my mains??? I picked the Ultra curve because of the AD/DAC's, It's spec's (a little better than the BFD) and it's cheap. I knew If I didn't like it I could sell it for about the same price as I bought it for. Again, before I found this place, I was driving blind.

Thanks again all, I'm learning,

Ralph-

 

[brucek]

Ralph J,
Yep, you've got it right. Again, there's nothing particularly wrong with some phase shift, it's debatable whether you can hear it or not from a single mono source.
Phase shift becomes a problem when two sources are mixed, whether electrically on a wire or in the air as sound pressure. If you have two signals 180 degrees (worse case) out of phase, you will have complete cancellation at that point. This is the reason it's fairly important to ensure your speaker connections are "in-phase", because there's a very good chance that they'll be producing the same signal, and if it's 180 degrees out of phase, there will be cancellation problems. This was my point about setting your mains to small and let the bass management do its thing. That way your sub and mains won't be producing the same signals.
They'll each have their own part of the spectrum, lessening the chance for phase cancellation in case the equalizer added too much phase shift at certain frequencies. Yes, there's an issue near crossover, but at least you won't have the problems that can occur when running "large" mains combined with a sub.
When I played with my BFD and enabled the filters that I was using at the time when I was doing the testing, I had some fairly wild phase shift at a few frequencies, and yes, the more I increased the gain, the more it changed the phase shift at that frequency. I wasn't using the software that you're using. I had the unit apart in my lab at work and was passing signals through it and monitoring the shift. Same result though - lots of shift. In the end result though, if you're able to put the equalizer in your system and get a final "all speakers on" response that you're pleased with, then that's the goal......
With regard to the sub question. You're correct, that's why it's recommended to match models and position when adding a second sub. When they're co-located and they both possess the same response, they act as "one" with an increased output capability. Start introducing different models and different positions and problems can occur. Stacked is great.
Another parametric equalizer I see talked about is the DSP-30 from QSCaudio. It also has shelf filtering and subsonic filters and a few other goodies. It's in the $500 range. Here's the site - click on DSP-30. Maybe Wayne knows about this one - comments?
http://www.qscaudio.com/
brucek

 

[brucek]

Ralph,

You seem like a very discriminating music listener. Have you ever brought an analog preamp home and listened to music in two channel analog - no sub?

brucek

 

[Wayne A. Pflughaupt]

I found the 1/3 octave EQ's to wipe out a big section of freq's to be very noticeable. I prefer the surgical nature of the parametric filters.

Sorry Ralph, I only recommended 1/3’s because you’re already using one. I can talk parametric.
Since you’re looking at BSS, you might take consider the Klark Teknik DN-410, two-channel 5-band parametric. Reportedly one of the best, used in many studios. If five bands aren’t enough it can be switched to a single ten-band mono unit. You can get them on e-bay from time to time for $550-650, less than a third of what a new one costs.
You also might look at the Ashly PQX-572. I don’t know if it would be as suitable for high-end use as the Klark, but Ashly has a reputation for making very clean equipment. It would certainly be miles above your Ultracurve. The 572 is somewhat more flexible than the Klark in that it has variable high and low shelving filters.
I’m less familiar with it, but the Symetrix 552 might also be an option.
Regards,
Wayne A. Pflughaupt

 

[DanRA]

Hi Again Bruce & Wayne,

Bruce,

Yes, I really do enjoy music listening, I like nothing more than sitting down and listening to music. The HT is something everyone gets to enjoy which helps justify the system. The kids would throw me out if I got rid of the HT. The plan is to add a “music only” portion of the system into the picture at some point but I can’t afford it for now. For now I go to my neighbor’s house to drool. As with all of us it’s a work in progress. Also, I have the sub turned down in 2 channel mode. I find massive bass for HT great but a bit much for music. I just put volume controls in for my bass shakers so I can shut down the bass shakers when listening to music. It’s really odd having the chair shake with music! The kids were watching Tomb Raider yesterday and the entire house was shaking. I had to go back in a sit down thru the HD action scenes. I noticed the bass shakers scaring the S*$% out of a couple of the kids when the serious bass kicks you in the butt. You know when it’s quite and all of the sudden!!!! You just can’t help but jump!! Well getting off track here.

Wayne,

Great, lets talk parametric!!! Like I said I have had no source of good EQ info especially from a home music listening point of view. The Ultracurve has 6 parametric filters in addition to the 1/3’s. I only am using a couple of the 1/3’s with very little cut, just to smooth things out. I think I am only using about a total of about 5 of the 1/3 sliders with no more than a few db cut. I use the parametric filters for the big cuts that I want to be more surgical with. I will check out the Klark & Ashly units. For the kind of $ your talking it’s an easy decision. The way I see it is I pick one up used and if I don’t like it I just sell it for about what I paid. I bought the Ultracurve used also. I know I could sell it probably for more than I paid because I updated the EPROM with the latest software from Behringer. The newer Behringer software is really slick. I couldn’t agree more, I would also prefer not to have additional A/D/A conversions and am looking forward to trying out some of your suggestions. Also, I heard the Ultracurve could be used further back with the AES/EBU in/out. I am not familiar with this type of setup but I heard it’s pretty good.

Could you explain the shelving filters to me? My understanding is shelving filters are to have different slopes on each side of a peak or dip. I haven’t figured them out in the Ultracurve. It seems to me if this were the case, they would need to be used individually on each filter depending on that particular slope you trying to correct. I haven’t figured out how to apply them individually on a parametric or the 1/3’s. If you set the shelving slope globally they wouldn’t seem to be much good unless you were only using one filter. Am I way off base on my understanding shelving filters???

Still Learning!

Ralph-

 

[Wayne A. Pflughaupt]

Ralph,

If you’re only using five filters, you are a suitable candidate for a parametric EQ. I also prefer to buy equipment used when possible, for the same reason you mentioned: considerably less loss should you decide to sell. I just picked up an Ashly 572 on e-bay for $300. It was a good deal - $375 is more common. Considering list price is well over $500, it was a no-brainer.

I used to know a audiophile-type who thought Ashly EQs were clean enough for high-end use. I considered that to be impressive – he was pretty picky about his gear. We were discussing their 1/3-octave EQs, although I’m confident Ashly’s parametrics should also “make the grade” if their 1/3’s do.

If you find the 572’s funky green color off-putting, you may want to look for the older PQ-16 (mono) or PQ-26 (stereo) models, which are a more traditional black. They can be found used for $50-100 less than the current offerings.

However, they are not quite as flexible. Whereas the 571 and 572 filters can each adjust full-range, the 16 and 26 filters have limited range. But this is usually only a problem if you’re looking for a dedicated sub EQ, where you might need 4-5 filters all below 80Hz. Typically the limited range filters are not a problem for full-range EQing you want to do, but still it’s something to keep in mind. The filters do have enough overlap to address most situations. For instance I have a PQ-16 in my bass guitar rig, and I had no problems getting what I needed, using four of the five filters under 500Hz.

The Symetrix EQs should be similar in quality and price to the Ashly models, but they have high and low pass filters (which is more common in parametric EQs) not the shelving filters. They also lack on and off switches for each filter (which I prefer – keeps unused filters totally out of the signal loop). I’ve never used them, but by company reputation I expect the Symetrix EQs to be of better quality than the Rane PE-15 or -17.

Shelving filters are filters that boost or cut everything beyond a designated turnover frequency. The bass and treble controls on receivers are examples of shelving filters.

I’m not sure how the turnover frequency relates to the actual action of the filter, however. A response graph in my DSP-A3090 manual has a picture graph of response of a 1kHz high shelving filter, and turnover frequency seems to be about roughly 1/3 from the filter’s upper extremity of the total affected range.

I’m sure that’s very confusing, but for instance, the graph for 9dB boost or cut shows that the filter’s action begins at about 100Hz. Response at 200Hz is plus or minus 1dB (depending of course on whether you are using the filter to boost or cut). At 1kHz response is at ±7dB. The filter finally reaches its ±9dB “destination” at 5kHz. Response is flat from 5kHz out to 20kHz.

Both high and low shelving filters on the new- and older-model Ashly parametrics have adjustable turnover frequencies.

As for what they’re good for: The high filter could be useful if your speakers started rolling off at 8-10kHz, for instance. The low filter could be very useful in adding a house curve to a sub with flat response.

Regards,

Wayne A. Pflughaupt

 

[DanRA]

Hi Wayne,

Sorry for not getting back sooner but you have been keeping me busy learning. I am really looking forward to trying out an analog EQ and so far the Ashly 572 looks like an easy choice. I like the idea of individual filter bypass switches, full range adjustment of each filter and the summing amplifier design sound good to me. I have done a little searching and also downloaded the manual and specs. I am not too concerned about the color and I will probably make a cover so no one turns any knobs. One of the main reasons I didn’t get an analog EQ was I was wondering how fine the adjustments are. I would think you move that knob a little and the freq etc changes a lot. Do you find it easy and accurate to find say a specific location for a filter etc...

I bought a FBD for my subs and I set it up last night and made a first pass at filters size and location. It was too late to test it. The wife would have gone ballistic if I started playing late. It took me a little while to figure it out but with Sonnie’s info I have it down. Without his info it would not have been figured out last night. It is not nearly as easy as the Ultracurve but it’s fine. I figure I will try it and possible sell the Ultracurve or the FBD depending on which one I like better when I get a analog EQ in for the mains and assuming I like the analog better (I assume I will). I heard there is software available for the BFD but it’s not on the Behringer web page. How can I get the FBD software? I have the Ultracurve hooked up thru a PC and the FBD would be real easy if the software is nearly as good as the Ultracurve software.

Shelving filters:

I still don’t understand the Shelving filters. You see my plots how would you use shelving filter in my setup? Maybe that would help me understand with practical use.

Thanks again for all the help,

Ralph-