😱 So you will have to fight with nulls, comb filters that do not exist nor can be eqed...or apply some arbitrary smoothing😛
Averaging positions doesn't have to be just about suiting more than one listener, but can be about understanding the affect of modes.
I don't know what you're trying to say, but when you get a massive dip at 60hz or 45hz or 105hz... THAT is a null. Sound waves cancelling other sound waves.
Again: that is position-dependant. Your mic or your ears will sense the dip changing when you move inside the room. Can be as massive of a change to go from a -15db dip to a +5db bump at 1cm distance from the walls near corner/floor.
As a rule of thumb, you get more bass -generally speaking- when the listener's position is close to a wall and corners. On the opposite, if the listener's position is in the middle of a large room (with the usual reflective walls) you will most probably end up with massive dip(s) in the 45-130hz region.
Last edited:
Yes, but you have to be able to act about it.
Software exists that can simulate and predict all that. Nothing new.
Doing something about it, that's the key here...
DSP as ''room-correction'' is mostly a marketing gimmick, even though that can help.
My point is: architectural acoustics is the sina que non condition for optimal results.
Basically, you cannot DSP-change mechanical/electrical limits of a driver no more than you can DSP-change acoustical properties of a given room.
Architectural acoustics - Wikipedia
You start with a carefully designed room/speakers combo, then you get the most from it using DSP.
Not the other way around.
My point is: architectural acoustics is the sina que non condition for optimal results.
Basically, you cannot DSP-change mechanical/electrical limits of a driver no more than you can DSP-change acoustical properties of a given room.
Architectural acoustics - Wikipedia
You start with a carefully designed room/speakers combo, then you get the most from it using DSP.
Not the other way around.
Perhaps I should have said nulls caused by room modes at low frequencies, which is what I was thinking of when I made my extreme statement. In my room I have a large peak at 40Hz which is almost everywhere in the room but I can easily attenuate it. I have a major null at 50Hz which I can boost a little but is so close to the attenuated 40Hz it makes a right mess around the room! However I can sit in a spot where the null almost disappears.
First, I avoid obvious ringing problems, then correct the rest by DSP. Hint: neighbours don't mind pink noise. I don't play any music outside. For audio tests I have recorded rain & helicopter. 🙂
EQ corrections are the easiest and most effective between the 4th and 9th octave, where the limits of the drivers are not so present and where the bass region nulls are more manageable.
Between 160hz and about 800hz, you can EQ pretty much anything if you're within the driver's limits.
Between about 800hz and 8khz, that's a walk in the park, EQ-wise and driver's limits-wise.
Between 8khz and 20khz, you'll hit the driver's limits again; beaming for large Sd's and Mms for the upper region. Not much that can be done with an EQ other than trying to get the most from less-than-optimal drivers...
What's left in the EQ manageable dept is the first octave. Room dependant but a little less; given drivers that can handle it, you can EQ-boost the 25-35hz region quite massively and expect great results. Sealed enclosures are the best to do so obviously.
Between 160hz and about 800hz, you can EQ pretty much anything if you're within the driver's limits.
Between about 800hz and 8khz, that's a walk in the park, EQ-wise and driver's limits-wise.
Between 8khz and 20khz, you'll hit the driver's limits again; beaming for large Sd's and Mms for the upper region. Not much that can be done with an EQ other than trying to get the most from less-than-optimal drivers...
What's left in the EQ manageable dept is the first octave. Room dependant but a little less; given drivers that can handle it, you can EQ-boost the 25-35hz region quite massively and expect great results. Sealed enclosures are the best to do so obviously.
High Q are no fun either 🙁
you might want to find a material that absorbs well in the 40hz as precisely as possible
that might be passive rads tuned at 40hz but that could end up quite expensive though...
In my experience, you cannot boost a null. You have to place a sound source, somewhere else, that does not have a null at that specific spot in your room. I try to "copy" what Geddes does, with 4 subs and have a pretty neat response below 100hz now - which I think, sound alot closer to what you hear with, lets say Beolab90. I have gain and delay, but also remove a peak with the DSP. I know that Geddes writes, that you can get a perfect response, with his method - but I'm only learning here and know that it takes time to perfects things 😱
When we choose to use new tecnology, I think we need to know and understand the pros and cons of that specific tecnology - just like the pros and cons of the human perception.
DSP can do many great things, that is simply not possible with passive solutions - but there is no free lunch 🙂
I will never build a passive design and have great results with my DSP. Building the cabinet properly, choosing the right drivers, measuring/filtering and getting good acoustics - is enough of a challenge to me 🙂
With all the famous hifi names being bought out and becoming part of a group of brands, plus a hifi revival built around convenient high quality streaming systems, I think the move to active is starting to happen already.
Passive or active.........
Why not a mix?
In my next (unfinished) project I will run the mid/high frequencies through a passive crossover while the bass has electronic correction.
I don't like additional and cheap amps in the mid/high freq as I can hear differences between different pre-amps as well.
Why not a mix?
In my next (unfinished) project I will run the mid/high frequencies through a passive crossover while the bass has electronic correction.
I don't like additional and cheap amps in the mid/high freq as I can hear differences between different pre-amps as well.
I agree. My professors also thought room correction could be that magic wand. I developed an advanced room correction system. It works well.
As for my 2c on this subject, I believe DSP is fine, although I'm not for it. I discovered that sometimes my room correction system would decide +20dB was required (bad room, worse speakers). No problem, except that if you're averaging 20 W, you're asking for 2000W (I think - correct me if my maths is wrong). In other words, DSP can give fairly accurate response, and can take room (or should I say system) phase correction into account as well (controversial subject, but in this area it makes sense), but it can't get the room really rumbling, except if the actual speakers have good response, and the enclosures are designed well. In my opinion, if the speakers have good response, and the enclosures are well designed, then a passive crossover should do, along with some room treatment, for extremely good in-room response. I also believe room response changes according to SPL?
I could very much be mistaken in everything I've said.
As for me, I'm passionate about music, and I firmly believe music is best reproduced in 2.0, and that DSP can supplement well-designed passively crossed over speakers.
The question is, how many people agree with me? As for a commercial future, I don't think high-end whatever audiophile has much of a future. The future is in sound-bars, mini BT speakers, mono sound columns, etc., and that audiophile (or proper high-end) has only got a future in DIY.
Agree- in that the audiophile mid market will shrink even more for sure. When I was a kid, all my friends dads had a posh hifi system of some sort. Our generation not so
Yep, I think so too.
And in pro-sound, the move to active is nearly universal. Things that folks like EAW and Martin are doing with dsp on their line-arrays are awesome. And just about every manufacturer of smaller traditional trap boxes is touting their products with a bit of FIR processing. I know, marketing...but still...
And really, I'd have to say FIR is the reason I think passive will all but disappear. IMHO, we simply leave too much sound quality on the table without it...
For a number of reasons that I can think of....
The first is any number of minimum-phase (IIR) eq's can be embedded into a FIR file. This allows corrections at the driver level, before system integration, corrections which actually do work very favorably without any negative consequence.
The same use of embedded IIR filters, can be used to unwrap or level phase, at the ends of each drivers' passband, allowing easier, smoother summation through crossover.
Then, FIR allows linear-phase crossovers to be used keeping phase flat without introducing extra group delay. This takes advantage of all the drivers' IIR mag and phase smoothing previously accomplished.
The xovers can easily be moved up or down in frequency within the bounds of driver overlap, with good summation and without phase issues.
This allows finding xover points for the best directivity, power response, without major xover redo.
It should also be noted, xover slope can be as gentle or steep as desired......and even good ole IIR xovers can be used for stubborns😀
And there is the ability to set delay's.
..... down to 1 sample.
Take a BMS coax CD crossing at 6300Hz...1 sample, 0.02ms, at 48K processing, represents 47 degrees of phase. The tonality of pink noise changes a lot when only 2 to 3 samples off from being completely aligned. Music is obviously harder to tell, but hey, aligned is aligned...
AFAICT, the only 2 downsides to FIR (from an SQ point of view) are latency and concerns with "preringing".
For playback, I can't see latency is a problem unless video sync is an issue. Maybe preringing exists, I dunno, haven't heard it yet..
So, my long winded synopsis is that FIR is really going to help push us into an ever more active world .🙂
Last edited:
I'm still designing passive xovers, but will be moving to DSP eventually. Here's a quote from a man I respect and admire.
"Crossovers may be implemented either as passive RLC networks, as active filters with operational amplifier circuits or with DSP engines and software. The only excuse for passive crossovers is their low cost. Their behavior changes with the signal level dependent dynamics of the drivers. They block the power amplifier from taking maximum control over the voice coil motion. They are a waste of time, if accuracy of reproduction is the goal." SL - October 2009
"Crossovers may be implemented either as passive RLC networks, as active filters with operational amplifier circuits or with DSP engines and software. The only excuse for passive crossovers is their low cost. Their behavior changes with the signal level dependent dynamics of the drivers. They block the power amplifier from taking maximum control over the voice coil motion. They are a waste of time, if accuracy of reproduction is the goal." SL - October 2009
- Status
- Not open for further replies.