Bass EQ'ing

[nzlowie]

Or should that be EQ'ing for bass..... Anyway read somewhere that you should use boost when EQ'ing bass, should only use cut the peaks. Is that true? I've tried that approach but the bass seams over the top even though the response is flat. Will post plots later on. Was trying my DTQWT speaker the by themselves without any eq and the bass sounded far more natural. I'm using these speakers for the bass section of my horns.

 

So.... What's the correct way to eq bass to match the room?

 

Also read on a post here that moving 1 speaker forward in relation to the other can help level out drop outs and peaks, sort of makes sense for bass but surely it will screw up the timing of the higher frequencies?

 

Love to hear your thoughts.

 

Dave

[aechmea]

Hi Dave,

 

I have had a DEQX for 10 years or more.  I too have read that the overall slope should be "downhill" from bass to treble; tried that from time to time.  However I find that to be far too heavy on the bass and I just run the whole spectrum flat.  In my room with excessive treatment the treble dies off naturally without EQ.

 

Troughs in the bass are more than often the result of cancellation of the direct and reflected waves being 180º out of phase.  The peaks are more than often summation of in-phase waves.  If you try to EQ troughs caused be cancellation it won't work because you are just pouring more energy into a void.  Lopping the tops off peaks does work, though.

 

Lots of people say that using EQ may fix the freq response but only does so where you measure and makes it worse elsewhere.  Well that is true, but I don't sit elsewhere! so what happens there is of no concern to me in my chair where I did measure.  I have a single "chair" environment.

 

I set the mains and the chair for best imaging (=mid and high freqs).  Then I solved my room's bass response (every room everywhere has issues) by judicious placement of my 2 subs.  I run the mains full range and then place the subs where they naturally fill in the troughs, finally lopping off the tops of peaks with EQ.  I don't EQ above 200Hz.  My subs are not there for bass reinforcement but to flatten the curve (now there's a novel phrase); in fact they are barely "on" volume-wise.

 

Moving the mains by small amounts has little to no effect on the bass.  Bass wavelengths are best thought of in metres and moving stuff a few centimetres cannot have any effect other than on high freqs where the movement is more comparable to wavelengths.  As you thought, I suspect that moving one speaker will not solve the bass problem but will more than likely stuff-up the mids and highs.

 

[Related but not directly so ...

Room modes (standing waves) are often reported as being the main bass problem.  Well, they weren't for me.  My problems stemmed from the first reflection off the back wall cancelling the direct bass coming from the speakers in the front of the room.  It's called SBIR.  Having a sub in the back of the room solved this to a large extent, as did huge quantities of thick heavy fibreglass sheets and tube traps covering the back wall to kill that first reflection.  It's not pretty but it worked for me.  Of course, I have planar speakers so the way that they interact with the room is different to traditional cone speakers.  I have spent more time taking the room out of the equation rather than manipulating the speakers' freq response.]

[doogie44]

After tweaking and being miserable --especially regarding the bass--I found this 'invaluable' in my system involving 2 REL subs (and the price is right):

 

http://www.dspeaker.com/en/products/anti-mode-8033.shtml

 

Review:

http://www.theabsolutesound.com/articles/dspeaker-anti-mode-8033-dsp-subwoofer-equalizer-tas-204/

 

Just my 2c worth

[nzlowie]

Thanks guys, no subs here. These main speakers have usable response to 30-40 hz so not really thinking about subs. Maybe I should think about a sub to balance things...

 

I also have it reasonably flat but need to dial down certain frequencies,you know the ones where the live crowd clapping cuts through you. 

 

[RoHo]

My approach is to use trial and error combined with my own personal preference.  I tried using a predicted set of eq filters from measurements but was not effective.  So I used the measurements to identify the main problem frequencies and set just a few eq filters to address those frequencies.  Went through a few iterations of "measure, adjust, listen" to optimise things as much as I could.  Much better and a big improvement on the before eq situation.  

The resulting frequency response does slope down from 80Hz to 30Hz but that just sounds best to me.  It could be due to the fact that I have a big room mode at 30Hz which I want to avoid.

The theory says  trying to boost a frequency dip that corresponds to a room null will be ineffective.  I found a little boost did improve things at one frequency.

[davewantsmoore]

12 hours ago, nzlowie said:

Love to hear your thoughts.

Where you measure and how you filter/interpet the data is very important.... otherwise you will be corecting the wrong thing (but won't know).

 

Use 1/6 octave smoothing (or mor) for your data.... this will ensure you're not trying to correct sharp dips......    and use PEQ filters to correct it which have a low-ish Q (eg. <1.5) ... make sure the frequencies below about 150 to 200 are louder than the ones above.

[Nigel]

Bass is a bastard. Very difficult to get a perfect flattish response.

Room treatments struggle due to the long wave lengths they are trying to tame, but they will reduce both peaks and troughs somewhat.  

A good sub with a steep LP filter is a very good solution. A lot can be done using its phase function to blend in with the mains, can even help to reduce a peak a bit. That and its volume knob gives you what an EQ device would do without driving to boost the shyte out of a null (= heat).

 Lastly your listening position... if you are sitting in a group of room modes (ie middle of the 3 dimensions) those cancellations are not far apart and together cause a big hole from 50-75Hz, IIRC. Walk around the room and notice the bass changes. You may want to rearrange your room.

As the phantom agents said "EQ is a last resort".

[nzlowie]

16 hours ago, davewantsmoore said:

Use 1/6 octave smoothing (or mor) for your data.... this will ensure you're not trying to correct sharp dips......    and use PEQ filters to correct it which have a low-ish Q (eg. <1.5) ... make sure the frequencies below about 150 to 200 are louder than the ones above.

Cheers Dave, I've been thinking that peaks were the issue that needed to be tamed..... My thinking was that the speakers are pretty flat near field so its the room doing the damage, creating sharp peaks and dips. Its these sharp peaks I was trying to flatten. Might try a smoother approach.

 

Yea I realise EQ is the last resort but it's just me chasing little bits, just can't help myself......

 

On another point, trying to isolate that frequency when a crowd claps in unison during a live performance. Finding that if i flatten off the top end to tame this frequency range, it sounds to flat overall.   

[davewantsmoore]

4 hours ago, nzlowie said:

Cheers Dave, I've been thinking that peaks were the issue that needed to be tamed..... My thinking was that the speakers are pretty flat near field so its the room doing the damage, creating sharp peaks and dips. Its these sharp peaks I was trying to flatten. Might try a smoother approach.

Forcing "sharp" peaks and dips flat to be flat will often sound quite unnatural, because the measurement you corrected doesn't correspond with what you are hearing.... so what you've done is ADD peaks and dips to the sound (whcih sounds unnatural/bad/exciting/different/etc.)

 

4 hours ago, nzlowie said:

Yea I realise EQ is the last resort

You see people saying this a lot.    It's more wrong that right.

 

At the frequencies you are talking about.... there is very little other than EQ, which will change the sound.    Other than placement of bass sources.... but if your bass sources are your L and R speakers, then you are limited in what you can do as they need to be placed appropriately for "imaging".

[marten]

5 hours ago, nzlowie said:

On another point, trying to isolate that frequency when a crowd claps in unison during a live performance. Finding that if i flatten off the top end to tame this frequency range, it sounds to flat overall.  

If you have an analogue Parm EQ available (or a digital one with a "virtual" freq adjust knob), you could set up a narrow-band boost (sharp peak parametric filter) and, while the music plays, sweep the filter's centre freq from 5k through 16kHz.  When the sweep hits the annoying frequency, your ears should have an "Ah-hah" moment. 

 

[almikel]

On 16/05/2020 at 7:55 AM, nzlowie said:

Anyway read somewhere that you should use boost when EQ'ing bass

typo check...suspect "shouldn't", instead of "should" - but we get what you meant

On 16/05/2020 at 2:04 PM, nzlowie said:

Maybe I should think about a sub to balance things...

definitely, or maybe more than 1

Just because your mains get down to 30-40Hz doesn't mean you don't need subs - subs are also about smoothing the "in room" bass response under 100Hz as well as extending the bottom end...

...but achieving good integration between mains and sub/s isn't trivial.

 

5 hours ago, nzlowie said:

On another point, trying to isolate that frequency when a crowd claps in unison during a live performance. Finding that if i flatten off the top end to tame this frequency range, it sounds to flat overall. 

give up now - just turn the volume down - you would wreck the sound pulling this freq range out.

I know it's annoying - I have several live albums, where the mixer didn't pull the applause down enough - I wouldn't think of pulling that down with EQ

 

5 hours ago, nzlowie said:

Yea I realise EQ is the last resort but it's just me chasing little bits

31 minutes ago, davewantsmoore said:

You see people saying this a lot.    It's more wrong that right.

agreed - it's about knowing when to apply EQ, and more importantly when not to apply EQ.

• at a technical level, EQ works when applying it to "minimum phase" issues, as typical parametric EQ is "minimum phase" - a peak or dip in the room response that's "minimum phase" can be fixed by the appropriate opposite cut or boost of EQ 
• a reasonable (not great) explanation of "minimum phase" is here https://www.roomeqwizard.com/help/help_en-GB/html/minimumphase.html
• broad EQ cut generally works (refer @davewantsmoore's post above about 1/6 octave smoothing) as broad peaks are often "minimum phase" so respond to EQ cut
• sharp dips are never "minimum phase", so won't respond to EQ boost as per @aechmea's post - SBIR is never minimum phase - never try to boost dips, or cut peaks that are SBIR related.
• it's OK to boost dips with EQ you know are minimum phase - they'll never be steep, and applying broad EQ boost is fine
• you can measure where your room is minimum phase - see the REW guide above

I use boost and cut below about 200Hz or so, based on where REW measurements show the response is minimum phase.

 

Mike

 

 

 

[nzlowie]

10 hours ago, almikel said:

give up now - just turn the volume down - you would wreck the sound pulling this freq range out.

I know it's annoying - I have several live albums, where the mixer didn't pull the applause down enough - I wouldn't think of pulling that down with EQ

Yes you're right! But sometimes its good to crank up the volume and "be there". 

 

 

[nzlowie]

On 17/05/2020 at 9:22 PM, almikel said:

On 17/05/2020 at 3:05 PM, nzlowie said:

On another point, trying to isolate that frequency when a crowd claps in unison during a live performance. Finding that if i flatten off the top end to tame this frequency range, it sounds to flat overall. 

give up now - just turn the volume down - you would wreck the sound pulling this freq range out.

I know it's annoying - I have several live albums, where the mixer didn't pull the applause down enough - I wouldn't think of pulling that down with EQ

Wasn't completely an FR issue after all...... I'd made another set of interconnects to try, they had plenty of detail so I'd left then in place, bad move. After the taking in the advise here I thought it can't just be the FR as this is great, must be something else and thought about other changes I'd made. Cables.... Went back to one of my other IC's and a lot better, those ear bleeding frequencies are tamed a lot better. Still loud but they don't cause pain like they were with the other IC.

 

Out of interest it was one I made up using cat 5 cable, wasn't the solid core but multi stranded.

[andyr]

3 hours ago, nzlowie said:

 

Out of interest it was one I made up using cat 5 cable, wasn't the solid core but multi stranded.

 

 

And presumably unshielded, too?

 

Andy

 

[nzlowie]

1 hour ago, andyr said:

And presumably unshielded, too?

Yea, I thought the twisted pair didn't need shielding...... Wrong again!

[pwstereo]

5 minutes ago, nzlowie said:

Yea, I thought the twisted pair didn't need shielding...... Wrong again!

It doesn't when used for it's intended (designed) use.

[andyr]

6 hours ago, nzlowie said:

 

Out of interest it was one I made up using cat 5 cable, wasn't the solid core but multi stranded.

 

 

If you're interested, I could send you over a length of the Cat5 cable that I have ... so you can see whether the 'ear-bleeding frequencies ' problem disappears when you use solid-core Cat5, instead of stranded.  Belden 1585a is 'Plenum-rated' - and so is not sold in Australia ... I suspect neither is it in NZ.  (Our wiring regs don't allow us to run network cabling in aircon ducts - plenums - so industry doesn't use it.)

 

I imported a 300m reel as it is teflon insulated - not PVC insulated - as well as being solid-core.  I find it makes wonderful spkr cable - although it's very labour-intensive! 

 

But of course, it is still not shielded - although it would be pretty simple to feed a tp through a braided shield and connect this to the RCA barrels at the source end.

 

Andy

 

[:) Go Away (:]

20 minutes ago, andyr said:

 

If you're interested, I could send you over a length of the Cat5 cable that I have ... so you can see whether the 'ear-bleeding frequencies ' problem disappears when you use solid-core Cat5, instead of stranded.  Belden 1585a is 'Plenum-rated' - and so is not sold in Australia ... I suspect neither is it in NZ.  (Our wiring regs don't allow us to run network cabling in aircon ducts - plenums - so industry doesn't use it.)

 

I imported a 300m reel as it is teflon insulated - not PVC insulated - as well as being solid-core.  I find it makes wonderful spkr cable - although it's very labour-intensive! 

 

But of course, it is still not shielded - although it would be pretty simple to feed a tp through a braided shield and connect this to the RCA barrels at the source end.

 

Andy

 

Or go with Cat8 cable while you at it. They are cheap as chips.

https://www.fs.com/au/c/cat8-patch-cables-3371

 

Big roll of Cat7.

https://www.fs.com/au/c/cat7-bulk-cables-633

[andyr]

3 minutes ago, N_R said:

Or go with Cat8 cable while you at it. They are cheap as chips.

https://www.fs.com/au/c/cat8-patch-cables-3371

 

Big roll of Cat7.

https://www.fs.com/au/c/cat7-bulk-cables-633

 

Do these cables use solid-core wires ... or stranded?

 

And is the wire insulation PVC ... or teflon?

 

Andy

 

[:) Go Away (:]

2 minutes ago, andyr said:

 

Do these cables use solid-core wires ... or stranded?

 

And is the wire insulation PVC ... or teflon?

 

Andy

 

Both Cat7 and Cat8 are 4-pair 8-core twisted pair OFC with PVC jacket. These are normally used in Datacentres and have top notch quality.

[:) Go Away (:]

48 minutes ago, andyr said:

 

If you're interested, I could send you over a length of the Cat5 cable that I have ... so you can see whether the 'ear-bleeding frequencies ' problem disappears when you use solid-core Cat5, instead of stranded.  Belden 1585a is 'Plenum-rated' - and so is not sold in Australia ... I suspect neither is it in NZ.  (Our wiring regs don't allow us to run network cabling in aircon ducts - plenums - so industry doesn't use it.)

 

I imported a 300m reel as it is teflon insulated - not PVC insulated - as well as being solid-core.  I find it makes wonderful spkr cable - although it's very labour-intensive! 

 

But of course, it is still not shielded - although it would be pretty simple to feed a tp through a braided shield and connect this to the RCA barrels at the source end.

 

Andy

 

Ok, I just re-read what you said. Your point was to use solid core vs stranded. Ignore what I said. 

[pwstereo]

21 minutes ago, andyr said:

Do these cables use solid-core wires ... or stranded?

And is the wire insulation PVC ... or teflon?

The bulk roll is solid core conductors, that is what is used for  fixed runs from data points to patch panels.

It's not suitable for cables that will be handled at all often. As shown on the page linked to, it has PVC insulation.

 

The patch leads or fly leads will be stranded conductors, that is what is used for flexible cables used to patch from patch panels to switches, etc, and to connect devices such as computers to field points. It is suitable for cables that will be frequently handled.

 

Another consideration is termination (I don't think you were intending to use normal RJ series connectors though), the solid core punches down correctly into patch panel and field point connectors, the stranded doesn't. The stranded core crimps correctly into the RJ series line plugs, the solid core doesn't.

 

 

[andyr]

10 minutes ago, pwstereo said:

The bulk roll is solid core conductors, that is what is used for  fixed runs from data points to patch panels.

It's not suitable for cables that will be handled at all often. As shown on the page linked to, it has PVC insulation.

 

The patch leads or fly leads will be stranded conductors, that is what is used for flexible cables used to patch from patch panels to switches, etc, and to connect devices such as computers to field points. It is suitable for cables that will be frequently handled.

 

Thanks for the explanation, Peter.    (I've always wondered why some Cat5 was solid-core ... and others were stranded.)

 

10 minutes ago, pwstereo said:

Another consideration is termination (I don't think you were intending to use normal RJ series connectors though), the solid core punches down correctly into patch panel and field point connectors, the stranded doesn't. The stranded core crimps correctly into the RJ series line plugs, the solid core doesn't.

 

Indeed not - when I use 1585a for spkr cable, I strip the ends of each strand and then solder on the required connector.

 

Andy

 

[:) Go Away (:]

7 minutes ago, andyr said:

Indeed not - when I use 1585a for spkr cable, I strip the ends of each strand and then solder on the required connector.

How does it work? Say, I were to remove the end connector of above mentioned Cat8 Cable and strip them and use 4 individual cables for Positive and 4 for negative in single ethernet cable, will they be better as speaker cables vs the standard speaker cables?

 

I would think, given these cables have 8 individually shielded and twisted cables, they would be better as speaker cable.

[pwstereo]

45 minutes ago, N_R said:

How does it work? Say, I were to remove the end connector of above mentioned Cat8 Cable and strip them and use 4 individual cables for Positive and 4 for negative in single ethernet cable, will they be better as speaker cables vs the standard speaker cables?

 

I would think, given these cables have 8 individually shielded and twisted cables, they would be better as speaker cable.

I don't think they are "better".

It seems like a misuse of a very specifically designed cable to me.

They pairs aren't shielded, they are twisted, so you have multiple TX and RX pairs.  <-- This is wrong for CAT 8, see my later post.

This makes no sense for a speaker cable. The individual conductors are thin, not capable of high current. Using multiple pairs, split, or together as pairs seems like a pure affectation for use as an audio speaker cable.

 

There are Shielded Twisted Pair (STP) cables available, but they are intended for use in a hostile (EMI/RFI) environment and need connectors with the metallic shield outer and patch panels to match. They aren't in common use. Normal situations are dealt with by Un-shielded Twisted Pair (UTP) cables. The pairs are not separately shielded, the shield is an outer foil/braid, similar to that in coaxial cables.

Edited May 22, 2020 by pwstereo
Spelling + correcting mis-information Re: CAT 8