Understanding the state of the DSP market

[dualazmak]

....
eg with subs you just don't need to be super precise so the fact room is messy is not a big issue.

Yes, I believe 1 msec precision would be enough for subwoofer to woofer "timing/phase" alignment.

We should note, however, it is critically important to optimize, objectively and subjectively, selection of XO Fq, filter-type(s) and filter-slopes, as well as selection of phase/polarity (inversion or not); the optimization would be greatly dependent on our own individual (specs, behaviors, etc.) subwoofer(s) and woofers, in our own room acoustic environments (ref. here #495 in my case).

 

[levimax]

Those will (or should) be repeatable in your room (if not, seems like a measurement error) but it depends what unreliable means and also what you were trying to do with that info, eg with subs you just don't need to be super precise so the fact room is messy is not a big issue.

You can read up thread that even sub timing in-room measurements and techniques are not so straightforward. It is not so hard to get the timing close enough that the FR looks OK but to get it accurate enough for a clean impulse response / accurate transient response is harder. Audibility of LF timing issues is not settled science but for drum hits and the like having all the energy arriving at the same time rather than smeared over time is a good goal.

 

[3ll3d00d]

Audibility of LF timing issues is not settled science but for drum hits and the like having all the energy arriving at the same time rather than smeared over time is a good goal.

Sure but you have you quantify this to make a meaningful statement

what ms range are you talking about? Also translate that into a phase angle range around your crossover, what is it?

 

[dualazmak]

If I may go back to the fundamentals of this thread started by @Keith_W's wonderful/amazing top three posts under the title of "Understanding the state of the DSP market",,,

We have so many choices of fully-automated DSP solutions, semi-automated ones, and also almost-manual ones; the internal-DSP procedures would be sometimes a kind of complete-black-box or semi-black-box, but many of us tend to do not care about such black-box-ness. Even I myself do not fully understand (which is far beyond my mathematical and technical knowledge/skills) what my beloved DSP software "EKIO" (IIR filters in cascade of 2nd-order stransposed direct form II biquad sections done using 64-bit floating points, ref. here) is actually doing internally.

I believe, therefore, each of us using DSP software/hardware, should have our/your own fully understandable and "validated" objective methods/procedures, in outside of DSP software, for the effects of the DSP implementations; otherwise, we may easily fall into endless spirals of DSP pit-holes/dark-tunnels.

At least in my case, as I repeatedly shared on this thread and other places, primitive "validated" analysis of recorded tone-burst (wave-shape) air-sound, as well as objective analysis of recorded air-sound's "sound energy distributions", both to be performed outside of DSP software, are critically important.

We can easily have so many theoretical and/or in-brain/on-desk discussions under the title of this wonderful thread, but many of such discussions would possibly "sound hollow" in case you/we have no out-of-DSP independent objective validation methods/procedures...

 

[3ll3d00d]

I didn't understand any of that, it was a rather abstract post saying that we should have concrete things without defining any of those things

 

[levimax]

Sure but you have you quantify this to make a meaningful statement

what ms range are you talking about? Also translate that into a phase angle range around your crossover, what is it?

I tried several of the methods given up thread to set the timing of a new co-located sub (which should be relatively easy). The values ranged from .25 ms to 14 ms which is too big a variance and I was not confident in the measurements at all. I can't prove it but I am comfortable with sub timing +/- 1 ms. . I then used a method I have used in the past and @Keith_W also uses which is to delay one driver take a sweep and compared the arrival time of the 2 impulse responses and then subtract out the added delay https://www.minidsp.com/applications/rew/measuring-time-delay. Even this was difficult as it was very hard to tell what was the second impulse response and what was a reflection but by varying the delay I could track "what moved" between different measurements and I finally got some sensical and repeatable measurements and used them to set the mains delay and it sounds OK to me. Upthread there was also what looked to me to be the "best" method, linked by John from REW https://www.prosoundtraining.com/2025/03/21/mains-sub-alignment-a-different-take/ , where you use REW wavelet signal generator and the REW Oscilloscope to adjust polarity, phase, and delay in "real time". Unfortunately I don't currently have the software required for this method but I plan to get it as I think this is one of the only ways you can really be confident you are getting the subs aligned correctly in all 3 related but separate ways i.e. polarity, phase, and delay. Most methods kind of "lump" these all together so even if the timing looks OK at the crossover point it may be misleading.

My point is that in-room acoustic measurements for timing are hard even for experienced people with good equipment. IMO many are being incorrectly led to believe that it is "easy" by the marketing efforts of the automatic room corrections software companies.

 

[boxerfan88]

Upthread there was also what looked to me to be the "best" method, linked by John from REW https://www.prosoundtraining.com/2025/03/21/mains-sub-alignment-a-different-take/ , where you use REW wavelet signal generator and the REW Oscilloscope to adjust polarity, phase, and delay in "real time". Unfortunately I don't currently have the software required for this method but I plan to get it as I think this is one of the only ways you can really be confident you are getting the subs aligned correctly in all 3 related but separate ways i.e. polarity, phase, and delay.

Wavelet method can be challenging in a domestic “smallish room” environment.

https://audiosciencereview.com/forum/index.php?threads/time-alignment-question.62803/

 

[ShadowBoxer]

the improvements I got with MathAudio RoomEQ are night and day.
Plus, as it's free, the only cost

It might be worth repeating that MathAudio is only "free" for the foobar2000 version; but, not the system wide VST version. This has certain implications for HT or streaming usage (to my knowledge, please correct me if I'm wrong).

Note: I do use the MathAudio VST version for 2-channel use. And side-note it is possible to limit the frequency range below a desired threshold. ie below Schroeder.

 

[levimax]

Wavelet method can be challenging in a domestic “smallish room” environment.

https://audiosciencereview.com/forum/index.php?threads/time-alignment-question.62803/

Shoot another hope dashed Thanks for posting this and saving me the trouble and expense of buying new software.

 

[Robbo99999]

If you average multiple measurements or use MMM you can get useful FR information fairly easily and reliably. IMO it is not reliable above Schroeder but it is OK for below Schroeder where the room really messes things up and allows you to make useful adjustments. What I find unreliable is trying to get useful "timing / phase" measurements in-room. Even vector averaging multiple gated measurements yields limited reliable information when it come to "timing / phase". There is just too much going on with sound bouncing around the room in most cases.

Yes, frequency information for the bass, that's what I said. I haven't really looked into phase information. I think people are a bit obsessed by phase in this thread, I'm not convinced it's particularly important. I think I can see why it could be important for DIY speaker designers when designing crossovers for their speakers, but for in room DSP I'm thinking it's less important. I think there's quite a bit talk in this thread where people can't see the wood for the trees, going down the garden path, or other phrases.

 

[LionIT]

Yes, frequency information for the bass, that's what I said. I haven't really looked into phase information. I think people are a bit obsessed by phase in this thread, I'm not convinced it's particularly important. I think I can see why it could be important for DIY speaker designers when designing crossovers for their speakers, but for in room DSP I'm thinking it's less important. I think there's quite a bit talk in this thread where people can't see the wood for the trees, going down the garden path, or other phrases.

I believe that you cannot simplify the phase to a single concept.
1) There is the relative phase between drivers of the same channel.
2) The relative phase between channels (eg left and right).
3) The full band phase response of a channel with respect to a linear system.
4) The absolute phase, or polarity, which indicates whether the energy peak is positive or negative.

For points 1 and 2 it's pretty important, because two perfectly aligned pulses at opposite polarity cancel each other.

 

[Keith_W]

Yes, frequency information for the bass, that's what I said. I haven't really looked into phase information. I think people are a bit obsessed by phase in this thread, I'm not convinced it's particularly important. I think I can see why it could be important for DIY speaker designers when designing crossovers for their speakers, but for in room DSP I'm thinking it's less important.

This is incorrect. All those peaks and dips you see in your frequency response are not errors in the frequency domain. They are actually time domain issues manifesting as superposition phenomena in the frequency domain - a sound wave interacting with a reflected (and thus time-shifted) version of itself, or two drivers producing the same frequency but not time/phase aligned due to group delay, non-coincident positioning, etc. This is why correction of a dip by simply filling it will not work, you will have more success if you adjust the phase. It shifts the dip somewhere else in the room. Because the root cause is a time domain issue, the correction should also be in the time domain. You can try to EQ it out, e.g. you can have some success simply cutting the peaks. But it is not the best way.

 

[LionIT]

This is why correction of a dip by simply filling it will not work, you will have more success if you adjust the phase. It shifts the dip somewhere else in the room.

Sorry, how you do it?
As far as I know you need at least an approximate mathematical model of room reflections, otherwise you have to go to trial and error. But in practice you cannot effectively monitor/control multiple and complex relationships as a deep learning algorithm or something similar would. Too much variables... maybe you can model few specific frequencies. Note that mathematical model of reflections is not the tail of captured IR.
In fact actually the only software that seems to do it, through machine learning, is Dirac. But I'm not even sure about it... nor on its effectiveness.
Then there is MSO but I don't know exactly how it works... I suppose it uses crossed acoustic interference matrices.

 

[Purité Audio]

Reducing the room gain peaks works extremely well.
Keith

 

[Chr1]

Apologies if this is a stupid question but does anyone know if is there a cheap(ish)/free VST which would enable me to adjust phase prior to MathAudio RoomEQ VST, output via Foobar2000?

Ideally I would like to see if I can make improvements through phase adjustments, without a full on deep dive into a complex piece of software I may or may not need all of.
(If that makes sense?)

 

[LionIT]

Apologies if this is a stupid question but does anyone know if is there a cheap(ish)/free VST which would enable me to adjust phase prior to MathAudio RoomEQ VST, output via Foobar2000?

I sometimes use this.

 

[Sokel]

Apologies if this is a stupid question but does anyone know if is there a cheap(ish)/free VST which would enable me to adjust phase prior to MathAudio RoomEQ VST, output via Foobar2000?

Ideally I would like to see if I can make improvements through phase adjustments, without a full run on deep dive into a complex piece of software I may or may not need all of.
(If that makes sense?)

Make the measurement with REW, then export this measurement to rePhase,
in txt format, create the correction and generate the impulse in .wav 32 bts.

Convolve as usual with some Impulse Response Convolver from the ones foobar has.

 

[Chr1]

Splendid!
Thank you... I will look into these options.

Much appreciated.

Follow up question: Has anyone here using MathAudio RoomEQ with Foobar2000 attempted to tweak phase in this way as a means to improve impulse response etc (below Schroeder).
I am curious if this is a viable option.

 

[LionIT]

Make the measurement with REW, then export this measurement to rePhase,
in txt format, create the correction and generate the impulse in .wav 32 bts.

Convolve as usual with some Impulse Response Convolver from the ones foobar has.

You need to start from a Dirac pulse in order to manipulate the phase before MathAudio, not from measured IR.

 

[Sokel]

You need to start from a Dirac pulse in order to manipulate the phase before MathAudio, not from measured IR.

That's what REW measurement is for.
Dirac delta function can be derived from it, that's the info you pass to rePhase after all if I have gotten it right.
Asking REW for minimum phase helps as well.