I wouldn't know, but I consider it worthwhile to study the works of people that are much smarter than I am.
You and many other members of this forum are among those people.
I guess so.
Good sub/infra bass relative to the rest of the spectrum isn't self-evident, especially in club sound.
Last edited:
Camplo, model out the 15h woofer (sub?) to see how big a box you require to get as low as you want. I am normally a high efficiency kind of guy, so I was looking at another set of refrigerator sized boxes for high efficiency 18's but I must admit, I really like Rythmik's direct servo controlled subs in a much smaller package. They do have kits... Rythmik Audio • Subwoofer enclosure plans
Thanks man, it does. diyAudio member DualTriode has a thread looking at the M2 waveguide with different CD's on it - I am waiting patiently to see his results as I like the potential of the wider directivity M2 waveguide compared to the 2384...
Hi plasnu, yes, I have tried both and perhaps it comes down to preference. However maybe we are talking two different things. I use linear phase crossovers because they sum perfectly in amplitude and phase, but use a mixed phase target which is mostly minimum phase with a bit of linear phase down low.
Camplo, I know that I'm on an island when it comes to my driver selections, but I'm just saying that the woofer I settled on does 114db at 30hz if you go vented, although I prefer to give up some output below 40 and go sealed. I know as a 12", excursion will be greater and resulting im distortion (if you believe it's relevant at those insane output levels).. I guess after reading that was your target I had to make one more plug for it as I seem to be the only one who's tried it crossed that high and I found its voice to match so well to a compression driver, as surprising as that is. Dayton 12“ hf. Ok, I've said my peace.
Modern algorithms deal with this pretty well and have reduced the pre-ringing issue to inaudible levels. A mixed phase target like Mitch is using takes it even a step further. Nonetheless, our sensitivity to phase is ambiguous and there is no clear cut preference among listeners for linear or minimum phase.
The delay is a bummer however and renders heavy FIR processing pretty much unusable for home theater, studio and live sound.
If one doesn't filter unneccessarily steep then the delay caused by FIR filtering shouldn't be too large (as long as it is restricted to midrange and HF crossovers) and also the off-axis ringing wouldn't be too severe .
A sub crossover and especially linear phase equalisation in the lower cutoff area however would ask for large delays and a lot of processing power.
If the filtering doesn't have to be extremely steep the subtractive delay approach could be used with DSP. The problem with that is that it can be basically done with every DSP but the usual GUIs wouldn't support it.
Regards
Charles
A sub crossover and especially linear phase equalisation in the lower cutoff area however would ask for large delays and a lot of processing power.
If the filtering doesn't have to be extremely steep the subtractive delay approach could be used with DSP. The problem with that is that it can be basically done with every DSP but the usual GUIs wouldn't support it.
Regards
Charles
I am into computer audio, so using a PC for FIR filter design and generation of FIR filters with 65,636 taps to control the low end plus reflections. Includes a 3 way digital XO (45 Hz to subs, 630 Hz from bass/mids to compression driver), low frequency control, along with tilting eq up top, time alignment of drivers and some excess phase correction. I posted the frequency response earlier, here is the measured timing (i.e. step) response that goes with it using REW measured at the LP:
Modern DSP like Acourate or Audiolense has preringing control, so as one can see there is no preringing, even off axis across my 3 seat couch is inaudible. The small dip before the step is the linear phase part of the mixed phase filter.
InOtIn I have set minphase bass % to 80 and minphase top % to 100 in the target designer.
I use JRiver Media Center on PC for all music and movie content. For movies, JRiver knows the latency of the FIR filter and compensates for it and one gets perfect lip sync. However, content outside of JRiver can still be played through JRiver's convolution engine, but can't account for latency. So a minimum phase filter with a lower number of taps is required if watching Netflix for example.
Back to the regularly scheduled programming, Camplo, have you modelled out your cabs yet?
Cheers,
Mitch
Modern DSP like Acourate or Audiolense has preringing control, so as one can see there is no preringing, even off axis across my 3 seat couch is inaudible. The small dip before the step is the linear phase part of the mixed phase filter.
InOtIn I have set minphase bass % to 80 and minphase top % to 100 in the target designer.
I use JRiver Media Center on PC for all music and movie content. For movies, JRiver knows the latency of the FIR filter and compensates for it and one gets perfect lip sync. However, content outside of JRiver can still be played through JRiver's convolution engine, but can't account for latency. So a minimum phase filter with a lower number of taps is required if watching Netflix for example.
Back to the regularly scheduled programming, Camplo, have you modelled out your cabs yet?
Cheers,
Mitch
Attachments
Off topic, but if someone is looking for an active crossover platform on the computer, I would strongly recommend Reaper. I own Protools, Cubase, Reaper, and have compared them with Sonic Studio (Mastering software). Reaper is the fastest (least processing power needed), least expensive ($50 or so), more versatile than Sonic, and the sound quality is on par with Sonic Studio, a ubiquitous mastering software used by professional mastering engineers all over the world. Actually, Reaper is also used by some mastering engineers because of its superior sound quality. People seems to think all softwares sound the same, but it's wrong. Don't use cheap software. Gain structure, dithering, ..., a lot of things make a big difference. BTW, Protools is the worst sounding DAW.
Bernt R (Audiolense) is working on a convolver for Audiolense, but it is not ready yet.
Maybe the end of 2019 I guess.
Until then, we have to stick with JRiver.
Doing DIY with Audiolense, PC, and pro sound interfaces with many output channels is very fast and fun.
Just swap drivers, run sinus sweeps and set the filter.
I am playing around with good drivers, so its easy to get very good sound with Audiolense.
GPA's, TAD's, and soon I'll get my new Celestion AXi2050.
Will soon order some horns from Auto Tech.
Regards Bergfinn.
Last edited:
I have evaluated several PC audio DSP packages and Audiolense and Acourate are the top in my eval based on requiring digital XO. Acourate is a full on digital DSP audio toolbox and Audiolense is easier to use, so the results posted are from Audiolense. I wrote a couple articles on it here (active 2 way) and here (active 3 way with dual subs).
The DSP software does the measurement, analysis, filter design and generation. JRiver is simply used to host the FIR correction filters in its convolution engine. I also use a DAW for recording/mixing, etc. and can host a convolution VST plug-in. There are several out there, this free one should work: ConvolverVST
Edit - I see Bergfinn beat me to it - cheers!
The DSP software does the measurement, analysis, filter design and generation. JRiver is simply used to host the FIR correction filters in its convolution engine. I also use a DAW for recording/mixing, etc. and can host a convolution VST plug-in. There are several out there, this free one should work: ConvolverVST
Edit - I see Bergfinn beat me to it - cheers!
I've never been a fan of pro tools, it's also superfluous for my purpose.
Running REAPER 5.978 here.
I also like iZotope plugins ao.
Mitch, thanks for the tip.
It appears, I checked the convolver plugin some years ago, but found it was last updated around 2008.
Last edited:
Interesting stuff, when I want to run linear filters, I take measurements with room EQ Wizard, finish off the filter in rephase, and load it into WavesIR reverb, inside of Ableton live.
Like stated heavy filters of the linear kind results in delay unacceptable for movies and video games. I save it for most critical listening, the eq on the input modules of the CTS amps will suffice for everything else before that, I was most concerned about band count and they have 16 per channel.
I have modeled woofers in winISD....I haven’t did any 3D modeling of cabinets, is there some suggested software? Or are you talking about some other type of speaker design modeling?
I’m still on the fence about creating a 2 way with 2 AE15m’s, that result in the woofers having less excursion than 1 2216nd (M2 woofer) which suggest that IMD is under control at that point vs a 3way with an AE 15m and 15H.....which increases lowest woofer Xmax though excursion is higher for the bass woofer the mid would have less excursion than the 2 way....
I have no idea who wins that fight, but the 2 way is better for intentions of closer listening and has one less crossover....while the 3 way reduces IMD in the mid, though it may or may not be warranted, the bass region would have higher IMD for the smaller range it’s handling vs the 2way, both designs being ported.
I haven’t had one person chime in on potential pros and cons ����*♂️
Like stated heavy filters of the linear kind results in delay unacceptable for movies and video games. I save it for most critical listening, the eq on the input modules of the CTS amps will suffice for everything else before that, I was most concerned about band count and they have 16 per channel.
I have modeled woofers in winISD....I haven’t did any 3D modeling of cabinets, is there some suggested software? Or are you talking about some other type of speaker design modeling?
I’m still on the fence about creating a 2 way with 2 AE15m’s, that result in the woofers having less excursion than 1 2216nd (M2 woofer) which suggest that IMD is under control at that point vs a 3way with an AE 15m and 15H.....which increases lowest woofer Xmax though excursion is higher for the bass woofer the mid would have less excursion than the 2 way....
I have no idea who wins that fight, but the 2 way is better for intentions of closer listening and has one less crossover....while the 3 way reduces IMD in the mid, though it may or may not be warranted, the bass region would have higher IMD for the smaller range it’s handling vs the 2way, both designs being ported.
I haven’t had one person chime in on potential pros and cons ����*♂️
Last edited:
You sum up the pros and cons of both designs pretty nicely.
One crossover less is an asset not to be ignored, at least in my opinion.
It remains to be seen whether IMD is really an (audible) issue, considering the output levels we're talking about.
You may want to check lukeamdman's dual 18H + SEOS 24 + BMS4594ND, beause he also likes it "loud 'n clear":
- General info
- Build thread.
One crossover less is an asset not to be ignored, at least in my opinion.
It remains to be seen whether IMD is really an (audible) issue, considering the output levels we're talking about.
You may want to check lukeamdman's dual 18H + SEOS 24 + BMS4594ND, beause he also likes it "loud 'n clear":
- General info
- Build thread.
Last edited:
2 way pros:
1, potential for a more coherent sound
2, vastly easier xo
3, less amp and dsp channels
4, you won't conclude a thread entitled:
Is it possible to cover the whole spectrum, high spl, low distortion with a 2-way?
By building a three way!
3 way pros : dynamic range (that you wont use anyway listening at 2 meters)
1, potential for a more coherent sound
2, vastly easier xo
3, less amp and dsp channels
4, you won't conclude a thread entitled:
Is it possible to cover the whole spectrum, high spl, low distortion with a 2-way?
By building a three way!
3 way pros : dynamic range (that you wont use anyway listening at 2 meters)
I thought ya'll already hashed through 2 Vs 2.5 way........ for 'closer' listening distances, the 2.5 wins due to closest c-t-c spacing and if you're really close, then going with smaller horizontally opposed mids is preferred: augspurger studio monitors - Google Search
GM
GM
Last edited:
Adding an additional 'way' has been reported to better the IM distortion performance by 10dB. I don't recall which paper this was in, unfortunately I can't cite the source for you.
Depending on your intended crossover frequency, the dip present in the TD15M may be an issue.
Look at this post: https://www.diyaudio.com/forums/att...l-low-distortion-2-a-td15m-gdahl-fr-holmi-gif
I have measured 2 different TD15M's with different hardware and computers. The dip I found to be between 400-600 hz. I don't have the measurements anymore, or I would post them.
Here's a short thread about Lambda/AE drivers, including comments by John Janowitz.
Most images are gone, but it's still worth a read.
Most images are gone, but it's still worth a read.