The 18sound tetracoil subs would be my choice for 'ultimate sub driver':
Eighteen Sound - Professional loudspeakers
Eighteen Sound - Professional loudspeakers
They are very deep though so probably not suitable for DBA. With sub drivers quantity should be maximized before quality of each individual driver (until you get very low on quality). This is because a doubling in excursion gets you +6dB and also a doubling in the number of drivers +6dB. After ~10mm of one way excursion even super drivers are going to be struggling by 20mm. E.G I would expect 2*18tbx100 to have more output than 1*18NTLW5000 (and at less cost). What the expensive driver gives you is maximum output per unit volume.
Eighteen Sound - Professional loudspeakers
Eighteen Sound - Professional loudspeakers
They are very deep though so probably not suitable for DBA. With sub drivers quantity should be maximized before quality of each individual driver (until you get very low on quality). This is because a doubling in excursion gets you +6dB and also a doubling in the number of drivers +6dB. After ~10mm of one way excursion even super drivers are going to be struggling by 20mm. E.G I would expect 2*18tbx100 to have more output than 1*18NTLW5000 (and at less cost). What the expensive driver gives you is maximum output per unit volume.
No in general. You can use both vented and closed subs in the same room. Geddes states that the biggest sub should be the one behind the main speakers.
I'm a big fan of listening 😀 of the room and speakers as a total system and using room boundaries.
Phobic's room seems *BA compatible in general, but *BA performance may be affected by room irregularities, primarily in upper range of *BA. Accordingly, the speakers must have a suitable low frequency capabilities.
For example, my current 9*4.2*3 m room, with a lot 😀 of doors and windows everywhere, DBA 2+2 real (no cross, no smoothing, NO EQ) and theoretical:
Attachments
Last edited:
If so - second width mode will start to appear. In 3.3 m room you can get good results up to ~80 Hz.
REW RoomSim is very simple tool to see it.
Can you find and listen DBA and multisubs somewhere?
For DBA you need at least 4 drivers and 2 amp/processing channels.
Last edited:
I started down a similar path several years ago, and yes, it is taking me years and have been learning a ton and having fun along the way.
Not sure why you are considering a computer as a low quality source. A dedicated, "audiophile" optimized computer can be pretty amazing and reliable. It's just another box on the rack, no screen, using the Roon phone app to manage playback. Using a computer and a "soundcard", as was mentioned before, in my book is a really powerful, high performance and cost effective solution. Mind you, soundcard does not necessarily mean a card you insert into your computer. Professional multichannel DAC/ADC units are commonly referred to as soundcards, but they are stand alone boxes and some are quite refined from an audio standpoint. Since you have a nice budget, check out Merging Technologies Hapi with ADA8P card, as an example.
The central concept is a multichannel ADC/DAC connected to a computer that runs a software to derive convolution filters, and amps connected to the analog outputs from the DAC. I use Acourate software, which is sophisticated and super flexible, at the cost of steep learning curve. You can do as simple as 2-channel digital room correction, all the way to active 5-way with digital crossovers, time aligned, room corrected filters. And of course you need a measurement mic, which is why you need the ADC capability. Modern software volume controls are pretty good - a non-issue these days.
FWIW, my learning journey: started with 3-way towers (B&W 804S) plus two subs with analog xo (also tried the 3-ways full range and overlapping with subs). To test the water, I purchased a Lynx Hilo DAC/ADC and implemented active xo between subs and mains, learned to measure and create filters, play music with them, etc. Then purchased a couple Hypex amps, removed the woofer passive xo and replaced with active (now an active 3-way, where the mids-to-treble was still passive), and so forth. Used the manufacturers xo points and slopes initially to reduce the number of variables. Then went into DIY speakers...a rollercoaster! Currently running active 4-ways, experimenting with dipoles.
The computer never failed. The nice CD player and Lamm pre were sold, never looked back. While my speaker is not final, my current expectation is that I will in fact achieve much better sound than I started with, at a lower system cost. As I'm ignorant on passive crossover design, it saved me a huge learning curve there, and fully active systems allow for use of a wider range of amps for mids and treble if you so desire.
I hope you explore it and decide if it's good for your needs/preferences. I don't think it's a matter of audio quality, but rather the pros and cons of each approach. You can get a used multichannel ADC/DAC and try out for little to no loss if you decide it's not for you and sell it.
Enjoy the process! 🙂
Not sure why you are considering a computer as a low quality source. A dedicated, "audiophile" optimized computer can be pretty amazing and reliable. It's just another box on the rack, no screen, using the Roon phone app to manage playback. Using a computer and a "soundcard", as was mentioned before, in my book is a really powerful, high performance and cost effective solution. Mind you, soundcard does not necessarily mean a card you insert into your computer. Professional multichannel DAC/ADC units are commonly referred to as soundcards, but they are stand alone boxes and some are quite refined from an audio standpoint. Since you have a nice budget, check out Merging Technologies Hapi with ADA8P card, as an example.
The central concept is a multichannel ADC/DAC connected to a computer that runs a software to derive convolution filters, and amps connected to the analog outputs from the DAC. I use Acourate software, which is sophisticated and super flexible, at the cost of steep learning curve. You can do as simple as 2-channel digital room correction, all the way to active 5-way with digital crossovers, time aligned, room corrected filters. And of course you need a measurement mic, which is why you need the ADC capability. Modern software volume controls are pretty good - a non-issue these days.
FWIW, my learning journey: started with 3-way towers (B&W 804S) plus two subs with analog xo (also tried the 3-ways full range and overlapping with subs). To test the water, I purchased a Lynx Hilo DAC/ADC and implemented active xo between subs and mains, learned to measure and create filters, play music with them, etc. Then purchased a couple Hypex amps, removed the woofer passive xo and replaced with active (now an active 3-way, where the mids-to-treble was still passive), and so forth. Used the manufacturers xo points and slopes initially to reduce the number of variables. Then went into DIY speakers...a rollercoaster! Currently running active 4-ways, experimenting with dipoles.
The computer never failed. The nice CD player and Lamm pre were sold, never looked back. While my speaker is not final, my current expectation is that I will in fact achieve much better sound than I started with, at a lower system cost. As I'm ignorant on passive crossover design, it saved me a huge learning curve there, and fully active systems allow for use of a wider range of amps for mids and treble if you so desire.
I hope you explore it and decide if it's good for your needs/preferences. I don't think it's a matter of audio quality, but rather the pros and cons of each approach. You can get a used multichannel ADC/DAC and try out for little to no loss if you decide it's not for you and sell it.
Enjoy the process! 🙂
Very nice 😀 discussion about DBA vs MS https://www.diyaudio.com/forums/multi-way/280249-controlled-directivity-diys-4.html#post4465503
these drivers look interesting thanks.
that makes sense on maximising quantity of drivers.
I still need to read more about Geddes, I guess given I've got a fireplace that I could put the biggest sub in the corner behind the speakers? little worried it will be too close though.
I was hoping that I'd not need to put any sub behind the main speakers ideally, if that's not an option I might instead need to consider a wall mounted sub and building out the alcove so it can be fitted there.
yes the french doors are a concern for DBA, but then again they will be for the mains as well - I've got to figure out what I'm going to do about them though that's a problem for another day
I don't know anyone with a DBA setup and given we're in lockdown it's going to be hard to visit someone anyway.
in all honesty I'm game for just giving it a go, I just need to make my mind up on the right approach. lots more reading to do!
I've done a lot more reading about PC based solutions and I actually think it's going to give me the best answer.
I think my existing main computer running Roon Core, HQplayer (server) with DSD encoding and a DSP program (most likely Acourate for conv filters) , connected over ethernet to a small headless NUC (client PC) running the HQplayer client which is outputting over USB to a multichannel DAC with provide the best quality from what I've read.
I just need to figure out if I can build things this way. I'm not 100% sure that the low processing power NUC outputting over USB works, I think it does, but need a bit more research.
I can then make/buy a very high spec silent NUC with good PSU.
thanks for the link, I'll give it a read
Multichannel Dac for PC crossover active speakers.
Okto Research
Reasonable price, little less them 1.000 €
Okto Research
Reasonable price, little less them 1.000 €
Here's an active speaker that is pretty nearly perfect, the JBL 708P
708P | JBL Professional Loudspeakers
I'd even argue the smaller 705P is a better choice given your room size.
705P | JBL Professional Loudspeakers
Just my opinion, but I don't think you should DIY. The reason is you said you are thinking long term. In the long term streaming audio services are going to offer multi-channel music. The DIY open baffle stuff is intended to make two channel stereo sound better by combining a type of directivity in conjunction with room reflections. You don't need that with multi-channel. I doubt you'd even want that with good streaming service multi-channel source material.
The other thing that's going to happen long term is patents are going to expire. Companies like Sonos will build inexpensive multi-channel CBT arrays (they hired the guy who developed JBL's CBT arrays).
From what you've said I think you'd get limited satisfaction from building DIY. First, your DIY speakers won't be as good as the JBLs listed above. Second, it no longer makes sense to think in terms of "forever" speakers because music source technology is right at the start of big changes.
A lot of DIY effort is expended trying to make two channel stereo sound better. That's where the open baffle stuff came from. They combined a type of directivity with a type of room reflection in an attempt to make two channel stereo sound better. You aren't going to need that in the long term.
Hi,
re - PC based solution, post lockdown, you're welcome to come and have a listen, but I'm happy with ~ 50 Hz bass. Details here - System Pictures & Description
re - PC based solution, post lockdown, you're welcome to come and have a listen, but I'm happy with ~ 50 Hz bass. Details here - System Pictures & Description
Attachments
coincidently I was looking at the non-pro version earlier today
Review and Measurements of Okto DAC8 8Ch DAC & Amp | Audio Science Review (ASR) Forum
I've got a very large 2 channel collection so I'll still be listening to it for a long time to come, though it's an interesting point to consider.
I think though that I should consider the pros and cons of build versus buy for the subs, something to think through.
why won't DIY speakers sound better than these JBLs? I'm not familiar with them but I'm aiming for a high performance system
very kind offer thanks, I've just looked up some of the components, think that's a similar approach to what I was talking about only I'd have the core PC elsewhere.
Those horns look amazing! 🙂
Note you need the ADC section to measure and implement convolution filters, such as with Acourate. BTW, Uli (the man behind Acourate) recommends using the same unit for ADC and DAC to avoid "clock drifting" - basically using different clocks for ADC and DAC.
exaSound has a good 8 channel DAC too, but no ADC either.
Some multichannel ADC/DAC: I mentioned the Hapi, also Prism Titan, Lynx Aurora and Hilo, several Focusrite.
Glad you are looking into it.
FWIW, my chain is similar: a general purpose PC runs Roon Core where I upsample to 176 or 192 and convolve the filters derived on Acourate. Roon outputs 8-channels (stereo) thru ethernet to an "audiophile" PC running Roon Bridge on WS2019 with Audiophile Optimizer that outputs 8 channels thru USB into an UpTone ISO Regen and into the Lynx Hilo.
Before Roon could do convolution I used to output Roon thru ethernet into HQPlayer in the audiophile PC where convolution took place and multichannel was sent thru USB to the Hilo.
FWIW, my audiophile PC is not so powerful by today's stds: 7 years old Xeon 1256v3 processor on an Intel 1200 server motherboard, 8GB RAM, headless. All very reliable. Of course linear power supplies, no fans, no HDD, no motors of any type, etc. Acourate is installed in this machine for measurement and filter creation. Running 192kHz/24bit material in 8 channel the CPU load is under 10%. Cruising. DSD is more demanding, of course.
Roon has been very reliable on convolution and streaming to the audioPC, so I simplified and took HQPlayer out since my DAC can't take advantage of it. If/when I move to Hapi with native DSD capability...that might change. BTW, Hapi is a ethernet attached device, so you don't need an audiophile PC to feed it through USB, which ends up streamlining the system and potentially lowering overall investment.
For your consideration: before buying a NUC try using any computer/laptop you have available (with the rest of the setup as you are planning to have) and see how demanding it is on resources, then ask at computeraudiophile and knowledgeable folks will likely be able to directionally extrapolate from your results in that PC to expected performance in a NUC.
cheers!
A NUC in a silent case works really well as my main system, I've got everything on an SSD with NVME for system drive. I had to get a USB wireless dongle as the inbuilt NUC Antenna didn't connect very well to the case.
I use Jriver and run a boatload of DSP and convolution. Most Audio processing barely causes it to blink. HQ Player is another story but streaming the audio from a server and using the NUC as a client is no problem at all. I run the Jriver DSP studio on my server and have any processing I want stream directly to any DLNA device with all the heavy lifting done on the server. It's my way of R128 volume levelling music to a Sonos system.
Controlling the whole thing from an Android tablet is good enough for as far a future audio is concerned 🙂
Attachments
Phobic,
I find it interesting that discussion seems to have avoided measurement.
If you plan on building a system, especially an ambitious one, then I think you ought to start by being able to quantify how it performs. If you plan on spending 10's of K on the system, you really owe it to yourself to do better than by ear.
Several people have noted that there are no "secrets" to designing stuff, but there are an awful lot of things you will learn by doing. In my mind:
- Learn to measure
- Design
- Measure what you have designed (and I promise you will be surprised)
- Iterate until the design looks good
- Now listen to them...
The advice to start with this at a modest scale is sound. Try a 2 way, hell try a sub.
Try different drivers and listen. You will learn an awful lot. One thing that you will observe is for two speakers with ostentiably flat responses there can be massive differences between how they sound.
Now get your head around dropping a stack of cash on a system without the benefit of knowing what to expect.
I find it interesting that discussion seems to have avoided measurement.
If you plan on building a system, especially an ambitious one, then I think you ought to start by being able to quantify how it performs. If you plan on spending 10's of K on the system, you really owe it to yourself to do better than by ear.
Several people have noted that there are no "secrets" to designing stuff, but there are an awful lot of things you will learn by doing. In my mind:
- Learn to measure
- Design
- Measure what you have designed (and I promise you will be surprised)
- Iterate until the design looks good
- Now listen to them...
The advice to start with this at a modest scale is sound. Try a 2 way, hell try a sub.
Try different drivers and listen. You will learn an awful lot. One thing that you will observe is for two speakers with ostentiably flat responses there can be massive differences between how they sound.
Now get your head around dropping a stack of cash on a system without the benefit of knowing what to expect.
thanks for mentioning this, it's not something I had considered.
I'd assumed I'd just plug in my miniDSP USB mic and use Acourate to measure like I do with REW right now.
I've never used Acourate so need to dig into why it needs to measure this way. I'm not sure I understand why, assume it's trying to track the impact of the DAC and signal chain on the filter changes?
Great that gives me a lot more confidence I'm onto the right track.
I might just start off by trying Roon + REW convo filter to prove it all works, then add in Acourate, then later HQ Player.
Windows, I could use Linux but would rather not if I can avoid it, while I can use it I find it too much of a PITA unless there's a very good reason.
sounds great.
I like the look of your case, I'm likely to buy a pretty chunky 6-8mm aluminium case to make sure there's not EMI/RFI leakage.
did you do anything special with the PSU?
you're 100% right measurement are key, think that's kind of implicit in the discussion, Acourate includes measurement capability and I'm already using REW to measure the room. I will have to measure any sub/speaker that I build, as well as being able to measure drivers electrically as well.
key part of the whole process.
You will get cleaner measurements with the UMIK if you set REW to use the acoustic timing reference. That will also allow you to take the time locked measurements you need to set the relative time delays between drivers etc.
Here is an image that shows the difference between measuring out of one DAC with an independent clock to the ADC. Look at the start of the impulse on the top graph it rolls in slowly before the peak.
When I measure with my preferred DAC what I do to avoid this is to set the audio interface as the output from REW and then take a digital output from the interface to the DAC. The interface is then controlling the clock on both and you avoid the drift. Then it looks like the bottom impulse.
It is a solid chunk of aluminium, I have no issues with EMI/RFI. The PSU is just the wall wart it came with. From some of your posts I get the impression that you like tweaking components etc. I am now cured of that affliction so I can't help there 🙂
They have a whole range now so maybe one of these suits. I chose mine to be small and thin to sit under my TV. The only tweak I made was to put some Kapton tape over the front LED to stop it from blinding me 😀
Akasa Thermal Solution
there is also this method if you don't have an electrical timing reference:
Using REW to Determine Time Delays Between Drivers - Technical/Modifications - The Klipsch Audio Community
Using REW to Determine Time Delays Between Drivers - Technical/Modifications - The Klipsch Audio Community