DSP and not DSP

[rowuk]

and that a simple single buffer will flip phase.

Not true, a cathode|emitter|source follower will not alter phase.

I stand corrected!
DSP below 100 Hz could work well if the subwoofer was n milliseconds closer to the listening seat. For instance right behind the listening seat and the stereo pair n milliseconds away. That could also be beneficial concerning room modes!

[Wolfgang]

I was wondering if anyone has tried in their audio system ‘spatial equalization’ instead of DSP? DSP effectively corrects some minor room related lower frequency problems but always at the expense of "reality" of sound.

[brig001]

I was wondering if anyone has tried in their audio system ‘spatial equalization’ instead of DSP? DSP effectively corrects some minor room related lower frequency problems but always at the expense of "reality" of sound.

My living room has a 20dB swing in the range 50Hz to 100Hz (and more if you want to get really low), so I wouldn't call that minor as that requires 100 x the power from bottom to top
You could argue that it would be better to fix the room, then the speakers, and I would agree, but practicality comes into play at some point and I can get nearly there with EQ at a fraction of the cost and grief from my wife

What do you mean by "reality"? Sounds more real with corrections than without to me - could you enjoy listening to any bass instrument where the response was varying by 20dB?

Brian

[Wolfgang]

20dB in this important frequency range (bass, kick drum and much more)is not easy to be fixed with EQ/DSP without losing at least some SQ. With high Q filters for small frequency bands it might be possible to cover the filtering effect a little but not for this range (and even lower down as you wrote). 20dB looks to me like your listening position is in an amplitude peak of more than just one frequency and the only way to fix it is to move your listening position and/or speakers in a range of let’s say 1-2m at least and ideally with some room treatment (which is out of the question as you wrote).

What I mean with “reality” is the following. The audio engineer tries to make for example bass and kick drum tight and puts some attack and impact in it. If you listen to this song with bad room acoustics that make it sound too loud and muddy and you add DSP you don’t hit the bad room acoustics but the amplitude of the perfectly fine bass part of this song. In order to tame the bad room acoustics you reduce the amplitude of a certain frequency range with DSP which will sound later exactly like what you have done to it: Very little original impact and attack at typical listening volume levels but still with muddy room acoustics artifacts but now at a lower level so that it is more level with the rest of the song which didn't need to be treated. Like this you will lose some reality of the original song in this bass frequency range (and the original idea/concept of the song) and keep the bad room acoustics, only at more bearable levels.

[brig001]

Hi Wolfgang, I understand what you mean now, but the results are surprisingly good. Bass drum, bass guitar and cello are easy to separate even in the same song - was watching Jules Holland last night. I would argue that I have improved the sound quality since without EQ it sounds awful, but I accept that it’s not ideal and probably never will be here - open plan living, dining and kitchen is about 10m long

I’ll do some measurements at the weekend

Not much room to move anything anywhere really as you can see from the photo

IMG_5775.jpeg (98.53 KiB) Viewed 1692 times

Brian

[Wolfgang]

Hi Brian,

as long as it sounds good to your ears it's really ok. That's in the end all that matters. And indeed there is not much you can do if you want to keep this arrangement. Nice looking living room and optimized listening room are in most cases hard to combine.
Asymmetrical arrangements of speakers and listening position can be helpful in order to get some grip on those amplitude peaks if room treatment is not possible but it looks kind of weird .

In my experience stereo OB subs also take off some pressure in the bass while giving a full satisfying bass experience. The best thing I have found so far is what I call "spatial equalization" but it needs a different preamp.

What got my attention was that you would even consider splitting the audio - using a FR driver! - so that you don’t get digital artifacts in the rest of your analog signal ( I assume) while using DSP for the bass. “Spatial equalization” might be at least interesting as an alternative route. It stays 100% analog, doesn’t add any EQ but uses only what’s already in the audio signal of the song.

This is the shortest explanation I can come up with for what I mean:
DSP works by adjusting the amplitude of frequency bands with the goal to get a more even distribution of sound energy between 20-20kHz in a listening room. It uses normally pink noise for the measurements with equal amount of energy per octave for calculating amount of EQ and filter settings.

IMO music cannot be treated like this if it’s supposed to sound realistic. Many people complain of a flat or uninvolving sound experience after having flattened their room response with DSP. “A little less is more” is the answer then. But it doesn’t solve the underlying problem.
Spatial equalization is a term that belongs to the process of stereo shuffling where stereo is considered as L and R signal but also as M, middle or sum signal of L+R, and as difference signal S (L-R). M contains information about the middle of a stereo signal, S only about the sides.
The stereo recording technique (spaced crossed cardioid mics) leads to phase shift between sum and difference signal. This phase shift can be used to widen the stereo image.

Stereo images at bass frequencies reproduced more narrowly than at treble frequencies for a given intensity ratio in the two speakers. Normally stereo shuffling is being used in the recording studio and different techniques to widen or minimize little differences between microphones etc are used to improve things. There had been consumer amps (in the 1970ies as far as I remember) which used some badly implemented shuffling circuits for increasing general stereo width but that’s a different story, not what I did.

I transferred this technique into a home audio reproduction situation, leaving certain parts out (the adding EQ and adjusting volume in separate channels after widening the image) and using it mostly for getting rid of the overly intense bass in the middle between the speakers by using the effect of natural phase cancelling (M + S information)from within the music signal. Naturally every recording will sound different – differently improved - with the same settings which can be easily adjusted even between songs if necessary like a volume control.

I have listened with different version/amps/speakers for more than a year and can only hypothesize about what going on. But the practical results are overwhelmingly positive. One of my hypotheses is that I reduce the impact of two acoustical environments (acoustics which were used for the final mixdown in the studio and actual listening room acoustics) on each other when brought together in my listening room. Because my listening room also produces phase cancellation/adding by dimensions of my listening position.

I am not adjusting the amplitude by adding/reducing energy with an EQ but use the actual phase of the music signal for doing this,which differs between L + R and is more focused in the middle for the lower mids and bass , in order to avoid overloading the room acoustics with the bass which is a bigger problem with digital music compared to analog music reproduction. Interestingly enough one wouldn’t expect that this could reduce some nasty booming bass or that feeling of “overloading” the room with certain lower frequencies but it does.

[brig001]

Hi Wolfgang,

I have measured my speakers at about as good I can get using DSP alone:
Driver anechoic

IMG_5797.jpeg (151.2 KiB) Viewed 1573 times

SPL at listening position, both speakers

IMG_5799.jpeg (167.5 KiB) Viewed 1573 times

Distortion (mostly below the noise floor)

IMG_5800.jpeg (251.96 KiB) Viewed 1573 times

I agree about flattening the speakers completely - getting a good power response is also important
I personally prefer slightly more bass and a falling treble - does't interfere with my tinnitus as much - they are mine, so I choose

The drivers have a peak at 6kHz, and I'm looking at https://purifi-audio.com/blog/app-notes ... x04-naa-11 as a way to reduce this and improve distortion further in the upper mid range. This will also reduce my reliance on DSP above 200Hz

After that, I'm not sure where to go, hence my original post, and you are right it's about reducing any DSP artefacts in higher frequencies
Bob Carver had a patent in the '70s (I think) for widening the stereo image, and there are a number of threads on DIY Audio for ambiphonics and matrix methods for LCR which sound similar to what you are doing. I am interested in this as an approach, but hadn't considered it as a method of fixing the bass. Looks like I have some reading to do

Brian

[Nick]

Distortion (mostly below the noise floor)

But that's only it seems 40dB or so, so I would say generally providing no information. If you actually had 0.2% 7th Harmonic in a amplifier you would certainly notice it.

Just saying...

[brig001]

Good point Nick, but it does tell you two things:

• Distortion is lower than you might expect for a 55mm driver going to 50Hz

• If you want a quiet life don’t live in a flat near a road

Diet, according to Dayton they are 3 1/2”