Visit to Audiophile Bill to hear his horns project

[Don Reid]

Another more tuneful option would be a Bill Fitzmaurice sub (such as Tuba) depending on space.

I use Bill Fitzmaurice designed HT Tuba horn woofers with my Oris 150/AER BD3 and Fostex t900a horns. I like them very much. In my DEQX DSP controlled rig their output at twenty-five Hz is identical to the reference 1kHz output. The reason I opted to go the DSP route is because I wanted a fully horn.loaded system and this didn't seem possible without delaying the mids and highs to compensate for the distance/time difference between the bass and upper frequencies. The horn path within the bass enclosures is 11.4 ft. Also I have the bass horns in the corners of the room and the Oris and Fostex horns well out in the room where they image better. The total path between the acoustic centers of the woofers and that of the AER BD3s is over sixteen ft., but the DEQX compensates for this to make them sound as though they are within 3 mm of each others.
I know many on this forum frown on full range DSP, but I find it improves the SQ much more than it takes away.

 

[Solypsa]

It certainly allows arrangements that would not be possible otherwise!

 

[Audiophile Bill]

I use Bill Fitzmaurice designed HT Tuba horn woofers with my Oris 150/AER BD3 and Fostex t900a horns. I like them very much. In my DEQX DSP controlled rig their output at twenty-five Hz is identical to the reference 1kHz output. The reason I opted to go the DSP route is because I wanted a fully horn.loaded system and this didn't seem possible without delaying the mids and highs to compensate for the distance/time difference between the bass and upper frequencies. The horn path within the bass enclosures is 11.4 ft. Also I have the bass horns in the corners of the room and the Oris and Fostex horns well out in the room where they image better. The total path between the acoustic centers of the woofers and that of the AER BD3s is over sixteen ft., but the DEQX compensates for this to make them sound as though they are within 3 mm of each others.
I know many on this forum frown on full range DSP, but I find it improves the SQ much more than it takes away.

Hi Don,

Nice - sounds like you have a very serious setup!

In terms of the whole physics conundrum. At 150hz, the wavelength is 2.3 metres. The distance between the speaker voicecoils should be less than half the wavelength for interoperability. As such your critical distance is just over a metre physically to separate the Tuba from the Oris without “issue.” Of course dsp can correct room anomalies etc too.

Leif is having enormous success with his powersoft parametric EQ function for his bass and mid bass. Fwiw - I have tried to use dsp in my system and always get some losses that fail the upside / downside test.

Best.

 

[iansr]

It seems to me that most people lump all DSP software and hardware solutions into the one basket. You don’t have to dwell on that thought very long to realise that it is on a par with asserting that all DACs sound the same. So once you‘ve got past that you naturally ask yourself, so what are the best DSP solutions? I’m not in a position to tell you, but I would strongly recommend that you seek out the various articles written by Mitch Barnett on the subject. He is a former studio engineer and, based on all I’ve read and seen of his output, highly credible in my estimation. Spoiler alert; he considers the DSP software solutions from Acourate and AudioLense to be in a different league to every other DSP solution.

 

[Audiophile Bill]

It seems to me that most people lump all DSP software and hardware solutions into the basket. You don’t have to dwell on that thought very long to realise that it is on a par with asserting that all DACs sound the same. So once you‘ve got past that you naturally ask yourself, so what are the best DSP solutions? I’m not in a position to tell you, but I would strongly recommend that you seek out the various articles written by Mitch Barnett on the subject. He is a former studio engineer and, based on all I’ve read and seen of his output, highly credible in my estimation. Spoiler alert; he considers the DSP software solutions from Acourate and AudioLense to be in a different league to every other DSP solution.

Dear Ian,

Thanks for your comments alas your case doesn’t apply to me. I was a very early adopter of Acourate and followed all Mitch’s superb articles (and frankly mini manuals) on Computer Audiophile all those years back when he wrote them. Ironically Acourate is such a clunky program that his articles really made the software much easier to use and fwiw I head an economic and predictive modelling team globally for my career so quite au fait with stats software in many guises.

I implemented the full suite of functionality as outlined in the great articles by Mitch and even managed to get my massive Focals sounding reasonable and without a tweeter that was like an ice pick to the skull. That included experiments over the course of 2 years including dsp crossovers, room correction, driver linearisation etc. Correction filter in 64bit floating point by memory.

Anyway - I abandoned Acourate largely because a sense of realism was always lost in the ADDA process. Bass was lovely and tight and my target curves suited my ear well (B&K curve was very much to my preference). Incidentally I couldn’t stand Mitch’s own preferred target curve function - didn’t sound like natural music at all to me. Sounded very artificially detailed. Anyway that is a point of preference and possibly purpose since his end goal was mastering.

Since Acourate I have gone on to try 3 other solutions (including one in the last month) and each one has the same substantial drawback. I suppose I could concede to any of the solutions I have tried if (and only if) I was 100% digital. Alas I am an odd case of 100% analogue. I do believe that better ADDA converters off board can help as you say but I haven’t heard any that can pass the fail test.

So I am quite sure your post wasn’t meant to be condescending in any way but I can assure you that I personally have, and continue even in the last month, to properly investigate dsp rather than making any sweeping statements. If I had to live with dsp implemented below 100hz I could do. Dsp in the 300-20khz range - not for me right now thanks.

Best wishes.

 

[iansr]

Bill, my post wasn’t actually aimed at you specifically - it was more a general comment on the seemingly widely held view that DSP is DSP, and if there was even a whiff of condescension then I apologise because that was certainly not my intent.
It’s very interesting to hear that you have experimented with Acourate but found the negatives outweighed the positives. In the interest of balance it should be said that there are many ecstatic users of Acourate and AudioLense. On the subject of ADDA conversion, Mitch has commented about using his own TT as a source and he thinks its totally transparent and preserves the analogue magic.

So it seems that advanced level DSP suits some people but not all and the only way to find out which camp you are in is to try it for yourself, which is precisely what I intend to do. Once our house re-modelling is complete I will be buying AudioLense and implementing it on a speaker system that will be based around Joseph Crowe’s ES290 horns. Mitch’s evangelism on the subject has drawn me in and I’m fascinated to find out what I will think of it. One thing is certain; it will be an active system but it remains to be seen whether ultimately it will use analogue active crossovers (Nelson Pass designed) or AudioLense DSP.

 

[Audiophile Bill]

Bill, my post wasn’t actually aimed at you specifically - it was more a general comment on the seemingly widely held view that DSP is DSP, and if there was even a whiff of condescension then I apologise because that was certainly not my intent.
It’s very interesting to hear that you have experimented with Acourate but found the negatives outweighed the positives. In the interest of balance it should be said that there are many ecstatic users of Acourate and AudioLense. On the subject of ADDA conversion, Mitch has commented about using his own TT as a source and he thinks its totally transparent and preserves the analogue magic.

So it seems that advanced level DSP suits some people but not all and the only way to find out which camp you are in is to try it for yourself, which is precisely what I intend to do. Once our house re-modelling is complete I will be buying AudioLense and implementing it on a speaker system that will be based around Joseph Crowe’s ES290 horns. Mitch’s evangelism on the subject has drawn me in and I’m fascinated to find out what I will think of it. One thing is certain; it will be an active system but it remains to be seen whether ultimately it will use analogue active crossovers (Nelson Pass designed) or AudioLense DSP.

Hi Ian,

I thoroughly recommend to investigate both approaches. On the analogue side, Joseph recently measured and evaluated the Xkitz active crossovers. Since I am a mad man, I ordered a pair of those boards here - they are on my “to do” list. They were very good value fwiw and measured miles better than the compared mini dsp crossover.

I can often sound inadvertently prejudice or evangelical for a particular philosophy but truly I believe we all need to find our own optimal solutions and there are many paths viable. Important thing is learning and the journey imho.

Very best.

 

[iansr]

Bill, I will definitely be trying both approaches. I already own a First Watt B4 active crossover and I have the Pass designed 6-24 crossover PCBs from the DIYA store so that I can increase the number of active channels beyond the 4 provided by the B4.
Best
Ian

 

[Audiophile Bill]

Bill, I will definitely be trying both approaches. I already own a First Watt B4 active crossover and I have the Pass designed 6-24 crossover PCBs from the DIYA store so that I can increase the number of active channels beyond the 4 provided by the B4.
Best
Ian

Ah right - nice! Sounds like you are well equipped.

 

[pjwd]

Dear Ian,

Thanks for your comments alas your case doesn’t apply to me. I was a very early adopter of Acourate and followed all Mitch’s superb articles (and frankly mini manuals) on Computer Audiophile all those years back when he wrote them. Ironically Acourate is such a clunky program that his articles really made the software much easier to use and fwiw I head an economic and predictive modelling team globally for my career so quite au fait with stats software in many guises.

I implemented the full suite of functionality as outlined in the great articles by Mitch and even managed to get my massive Focals sounding reasonable and without a tweeter that was like an ice pick to the skull. That included experiments over the course of 2 years including dsp crossovers, room correction, driver linearisation etc. Correction filter in 64bit floating point by memory.

Anyway - I abandoned Acourate largely because a sense of realism was always lost in the ADDA process. Bass was lovely and tight and my target curves suited my ear well (B&K curve was very much to my preference). Incidentally I couldn’t stand Mitch’s own preferred target curve function - didn’t sound like natural music at all to me. Sounded very artificially detailed. Anyway that is a point of preference and possibly purpose since his end goal was mastering.

Since Acourate I have gone on to try 3 other solutions (including one in the last month) and each one has the same substantial drawback. I suppose I could concede to any of the solutions I have tried if (and only if) I was 100% digital. Alas I am an odd case of 100% analogue. I do believe that better ADDA converters off board can help as you say but I haven’t heard any that can pass the fail test.

So I am quite sure your post wasn’t meant to be condescending in any way but I can assure you that I personally have, and continue even in the last month, to properly investigate dsp rather than making any sweeping statements. If I had to live with dsp implemented below 100hz I could do. Dsp in the 300-20khz range - not for me right now thanks.

Best wishes.

Wow Bill - you have dug deep into this area - I agree re ADDA - dsp is best for digital systems , convolution in the computer and multichannel DAC ... this is what I aim to do in the next speaker. However I am suspicious of the " black box" in room measurements that acourate and audio lens base the calculations on and will use outdoor fr plots and lspcad to design the transfer function. Technically convolution filters make passive crossovers look sad but how they sound is another matter

Friends tell me that audiolense is a simpler product to use than acourate although Uri has a forum now and apparently that is very helpful - what are your thoughts ( or Ian)

It is pretty remarkable how good minidsp sounds below 170 Hz or so ... even with all the conversion and latency issues but my current project is an active crossover to replace a minidsp so that will be interesting - unfortunately there are not many off the shelf boards that have notch and shelving filters ( linkwitz is the only one I believe and thats 4 drivers ) so it is a ground up for me - will take a while

Like everyone I am vicariously enjoying your journey

Cheers

Phil

 

[spiritofmusic]

One of the most dramatic set of demos ever in this hobby, was re GP Monaco 1.5 TT into Cessaro Liszts horns. Same room, same positioning.
First demo, excluding one factor, was absolutely sublime, real hairs up on back of neck stuff.
Second demo, now utilising Illusonic AD-DA conversion, was as boring as watching paint dry.
Caveat, amps changed from Bakoon 17W Class AB to Mola Mola 700W Class D, and untreated walls to walls slathered in Molex acoustic panels.
But my hunch is that it was the AD-DA process that was the culprit.

 

[morricab]

Dear Ian,

Thanks for your comments alas your case doesn’t apply to me. I was a very early adopter of Acourate and followed all Mitch’s superb articles (and frankly mini manuals) on Computer Audiophile all those years back when he wrote them. Ironically Acourate is such a clunky program that his articles really made the software much easier to use and fwiw I head an economic and predictive modelling team globally for my career so quite au fait with stats software in many guises.

I implemented the full suite of functionality as outlined in the great articles by Mitch and even managed to get my massive Focals sounding reasonable and without a tweeter that was like an ice pick to the skull. That included experiments over the course of 2 years including dsp crossovers, room correction, driver linearisation etc. Correction filter in 64bit floating point by memory.

Anyway - I abandoned Acourate largely because a sense of realism was always lost in the ADDA process. Bass was lovely and tight and my target curves suited my ear well (B&K curve was very much to my preference). Incidentally I couldn’t stand Mitch’s own preferred target curve function - didn’t sound like natural music at all to me. Sounded very artificially detailed. Anyway that is a point of preference and possibly purpose since his end goal was mastering.

Since Acourate I have gone on to try 3 other solutions (including one in the last month) and each one has the same substantial drawback. I suppose I could concede to any of the solutions I have tried if (and only if) I was 100% digital. Alas I am an odd case of 100% analogue. I do believe that better ADDA converters off board can help as you say but I haven’t heard any that can pass the fail test.

So I am quite sure your post wasn’t meant to be condescending in any way but I can assure you that I personally have, and continue even in the last month, to properly investigate dsp rather than making any sweeping statements. If I had to live with dsp implemented below 100hz I could do. Dsp in the 300-20khz range - not for me right now thanks.

Best wishes.

Have you tried a good studio AD?

 

[morricab]

One of the most dramatic set of demos ever in this hobby, was re GP Monaco 1.5 TT into Cessaro Liszts horns. Same room, same positioning.
First demo, excluding one factor, was absolutely sublime, real hairs up on back of neck stuff.
Second demo, now utilising Illusonic AD-DA conversion, was as boring as watching paint dry.
Caveat, amps changed from Bakoon 17W Class AB to Mola Mola 700W Class D, and untreated walls to walls slathered in Molex acoustic panels.
But my hunch is that it was the AD-DA process that was the culprit.

My hunch is the amp switch…

 

[morricab]

Have you tried a good studio AD?

I ask be because I am considering something like this…rather than an analog active.

 

[Audiophile Bill]

Have you tried a good studio AD?

I have a Tascam DA3000. I also have a Focusrite Scarlett.
So both pro but can do better. Not tried the Grimm etc.

 

[Audiophile Bill]

I ask be because I am considering something like this…rather than an analog active.

Try one of the Xkitz boards in your 2 way. So cheap. You can build it a really nice dc power supply if you wish.

 

[microstrip]

(...) Caveat, amps changed from Bakoon 17W Class AB to Mola Mola 700W Class D, and untreated walls to walls slathered in Molex acoustic panels.
But my hunch is that it was the AD-DA process that was the culprit.

With such changes and in such conditions how can you be sure of anything?

 

[spiritofmusic]

With such changes and in such conditions how can you be sure of anything?

Because the sound changed nature SO comprehensively. I also heard the same spkrs in a much bigger room (better suited to them), with a different vinyl front end (TW AC3), and different amps (Tron 211), and the sound though different again, had more familial similarities to the first demo. This DSP based presentation was brutally unengaging. They didn't sound like Liszts, they didn't even sound like horns. I had to terminate the demo early, so poor was it.

 

[iansr]

Wow Bill - you have dug deep into this area - I agree re ADDA - dsp is best for digital systems , convolution in the computer and multichannel DAC ... this is what I aim to do in the next speaker. However I am suspicious of the " black box" in room measurements that acourate and audio lens base the calculations on and will use outdoor fr plots and lspcad to design the transfer function. Technically convolution filters make passive crossovers look sad but how they sound is another matter

Friends tell me that audiolense is a simpler product to use than acourate although Uri has a forum now and apparently that is very helpful - what are your thoughts ( or Ian)

It is pretty remarkable how good minidsp sounds below 170 Hz or so ... even with all the conversion and latency issues but my current project is an active crossover to replace a minidsp so that will be interesting - unfortunately there are not many off the shelf boards that have notch and shelving filters ( linkwitz is the only one I believe and thats 4 drivers ) so it is a ground up for me - will take a while

Like everyone I am vicariously enjoying your journey

Cheers

Phil

According to Mitch, AudioLense is indeed more user friendly than Acourate and that is precisely why I will be choosing it.
Re your comment about using outdoor measurements to design the transfer functions, I think you need to read more about what Acourate and AudioLense do and how they work. The whole point is to deliver optimal sound quality at the listening position.

 

[pjwd]

According to Mitch, AudioLense is indeed more user friendly than Acourate and that is precisely why I will be choosing it.
Re your comment about using outdoor measurements to design the transfer functions, I think you need to read more about what Acourate and AudioLense do and how they work. The whole point is to deliver optimal sound quality at the listening position.

Thanks for that confirmation.

I do understand the room correction component of those programs however my goal is to achieve the best response ( B&K I believe - in agreement with Bill) from the speakers, provide the optimum acoustic environment in the room and then, if needed , room correction. I am sure the less you do of the latter the better - and you need some clever averaging of response from multiple locations to get a good overall room response

I am sure there are some very clever algorithms in the software that address these issues but I don't see how it can be as good as a staged approach.
When I have finished I can always test the theory out

Cheers
Phil