High-end full active DSP 2.1

This is very good idea and I would start here. The Linn speakers are not rubbish by any means and could be improved with the proper measurement and integration. Getting a multi channel interface and computer based DSP would be a really good way to learn. When you know what you like and have experimented with that and have done all the research needed to determine what you need to do better, that is the time to buy drivers and start designing 🙂

Then don't spending some money on an interface mic and software is money well spent in this hobby that will never be wasted. If you want to start building then build some sub boxes as they are much less critical and I think you would be surprised how much better your existing speakers sound when you get the bass right.

That's very much the 1st step, in fact I've already tried it with REW + Rephase with limited success versus Linn's own DSP implementation.

to make more progress I need to finish sorting my room out in the next few months, then integrate subs, so I think it make sense to do that with DSP at the same time. Not sure this is too big a step to make to be honest, especially considering that I have an existing sub I can use as well.



Phobic,
Please apologize to make your thread derail and brings 'bad vibes' to it. The overall ambience of this troubled time makes me overreacting about 'attitude'.

This is reccurent to me and i try to work on it but difficult to change my own nature.

Once again please accept my excuse.

No need to apologize, It's all very interesting conversation, It's good to get different views, ultimately it's for each individual to take in the opinions, look at the facts & evidence, and make a decision for themselves based on their own values.

There's no right or wrong answer here, just a lot of complex decisions!
 
I'd rather avoid going passive so I can bypass having to build crossovers.

This is exactly what I was trying to say. I have no idea how to build passive xo.

Lojzek said it best: first try to improve the sound of Akubarik using dsp crossovers.

I guess I was referring to first implement a digital crossover between your subs and mains (keeping the passive xo in the Akubarik). Then you can remove one passive xo at a time in the Akubarik and replace with a digital xo. Of course, starting with Linn's xo points and slopes is a natural starting point, which makes it all easier than starting from scratch.

Be sure to use a protection cap on your tweeter to avoid destruction from a routing mistake, amp turn on pop-up, etc.

Starting with subwoofers, a measurement mic and REW is a worthwhile experimentation route. Mono subs in distributed bass array, stereo subs? Different schools of thought and they all have good arguments. Read Earl Geddes paper on distributed subs. use the mic and REW to try it out in your room.

There is a lot to learn around DSP and step by step allows focusing on a narrow problem and disregarding a host of variables that have been tackled by Linn in the design of Akubarik. Even placing the speakers in different places and applying digital room correction and then listening to see if you like it better.

Anyway, I'm not even close to being among the knowledgeable around here. Just happened to have started on your same path a couple of years earlier so just talking from my limited experience.

Have fun!
 
I've looked at Dirac before, don't think it's going to be powerful enough for what I want though the automation is very good.
Dirac is designed as a room correction product for use with an existing set of speakers where the crossovers are already fixed.

Acourate and Audiolense are more powerful in that they offer crossover generation and individual driver correction as part of the overall target.


I don't really have a view on what sampling rate & number of taps I need, but unless I'm missing something the PC route seems more powerful and actually pretty cost effective when you compare it to high end DSP.

I've used Rephase quite a lot already, it takes far too long, I find it a bit painful!
Most good audio playback software can change sampling rate to suit the source and convolution files at the different sampling rates can be created and loaded as needed.

The other way to do it is to pick the highest rate you need and upsample everything to that rate or downsample to a specified rate which is what I do mostly.

The number of taps sets the frequency resolution that the filter can resolve and the latency that will be introduced. The resolution is needed at low frequencies for detailed corrections. For example a 65,536 tap filter has a resolution of about 1Hz but it will introduce a few hundred milliseconds of latency. For audio playback at home this is no real issue it takes a little while for sound to appear after pressing play 🙂 Issues can occur with video sync and in live sound situations where the latency is unacceptable.

You can run as many of those convolutions as you like in a modern PC, there is very little hardware than can come close to that without being extremely expensive.

Whether it is necessary to have that resolution depends but with a PC if you can deal with the latency there is no real downside.

As you increase the sampling frequency the number of taps needed goes up but because the sampling period is smaller the latency doesn't have to increase. There is no hardware that can do very high sample rates and high resolution filters.

rephase is quite easy to use when you get the hang of it so I'm not sure what you find difficult or painful about it but I would work that out before you spend money on commercial software.


I'd been advised elsewhere about the benefit of always converting to DSD to reduce the noise generating in the DAC. If PCM is going to sound better then I'm happy to stick with that, I'm simply just driven by the solution with the best quality.

I also don't have any preference on SD v R2R, think it's more a case of just picking a well measuring DAC with the right features I need and then figuring out what's the best way to drive it.
I don't understand the noise comment as DSD generates huge amounts of noise but it is usually shaped to be placed above the audible band.

Sounds better is relative and can be personal preference but in a multichannel active situation DSD is impractical. DS DAC's almost always measure better which is why they became popular in the first place, that and ease of manufacture.

Trying to match resistors in the ladder at the resolution needed to get low noise and distortion is no easy task.
 
There is a lot of silliness around digital audio (MQA, DSD as an exchange format, High sample rates as an exchange/internal DSP format). We don't even have to do experiments to prove all of this is a waste of time as the maths is well developed. Stick with conventional PCM systems and select DAC/ADCs based in their performance rather the particular technology.
 
I'm starting to make a bit of progress, I now have a dedicated listening room....it still needs a partition wall building but that will come once I've managed to find a builder, seems to be a bit of a struggle right now due to covid.

first step is to do some basic DSP with the single Sub that I already have as a way to learn the ropes and get a grip on DSP.

I've not yet decided which of the PC based DSPs solution I'm going to use however my current pre-amp only supports HDMI, SPDIF and TosLink. the DAC is built into the active mains so no way to use USB.

I don't think I have any choice but to buy an external multi channel DAC to connect into my PC to allow me to connect to the pre-amp AND to the sub via RCA?

Due to the architecture of the Linn setup I can't connect the sub to the pre-amp without buying a different unit which is a bit silly as I'll only need it short term.

Maybe my 1st step should just be to try the PC DSP software instead - I wonder if I can do this over HDMI....
 
Last edited:
Buy for example minidsp with coax input and 4xcoax output ie. 8 channelse. Then buy a ‘coax to RCA’, that go into the sub.

Keep in mind this introduces a non known delay.
A delay is also introduced when you connect the Linn speakers. Don’t know how much it is. Should be mentioned in the specs.

Important to keep in mind, that the external DSP has enough ms to delay the difference plus or minus the actual required delay when implementing the sub.
 
Last edited:
I'm starting to make a bit of progress, I now have a dedicated listening room....it still needs a partition wall building but that will come once I've managed to find a builder, seems to be a bit of a struggle right now due to covid.

first step is to do some basic DSP with the single Sub that I already have as a way to learn the ropes and get a grip on DSP.

I've not yet decided which of the PC based DSPs solution I'm going to use however my current pre-amp only supports HDMI, SPDIF and TosLink. the DAC is built into the active mains so no way to use USB.

I don't think I have any choice but to buy an external multi channel DAC to connect into my PC to allow me to connect to the pre-amp AND to the sub via RCA?

Due to the architecture of the Linn setup I can't connect the sub to the pre-amp without buying a different unit which is a bit silly as I'll only need it short term.

Maybe my 1st step should just be to try the PC DSP software instead - I wonder if I can do this over HDMI....

One of the the BIG leaps of faith for me was letting go of my beloved Lamm pre amp. The multichannel DAC needs to drive the amps directly. In your case the sub late amps and the amp driving the Akubarik.

How are you planning to get the mic connected into the computer? Regardless of which DSP software you use. You could start trying out REW for free.
 
I've not yet decided which of the PC based DSPs solution I'm going to use however my current pre-amp only supports HDMI, SPDIF and TosLink. the DAC is built into the active mains so no way to use USB.

I don't think I have any choice but to buy an external multi channel DAC to connect into my PC to allow me to connect to the pre-amp AND to the sub via RCA?

Maybe my 1st step should just be to try the PC DSP software instead - I wonder if I can do this over HDMI....
Have you listed the preamp somewhere, without knowing what it is makes it hard to check whether it could work.

HDMI could work if the preamp has the right EDID to expose all of the channels HMDI has available. If it is a home theatre type of preamp then it should.

Easy to check by using Jriver or some other multichannel audio player/DAW to see if you can address the HMDI channels separately.

In my experience balanced PC multichannel should be used with balanced amplifiers.
Mixing balanced and unbalanced, might give you sound, but also give you hissing issues.
Most have both TS and TRS outputs, I've never had a problem doing this.
 
Whilst building my horns, I started with a Behringer DCX2496 (3 way) unit with the idea that I would build a passive crossover using the crossover points & slopes that I had derived from measuring using REW. I quickly realised that 1st or 2nd order crossovers were not going to work for me so went LR4. This meant a passive crossover was no longer an option. I then started looking for a 4 way DAC, etc that would replace Behringer. Options included previously mentioned Oktodac8 or Xilica XP-4080 units. Fortunately, I came across the PC + Lynx AES16 + Apogee Rosetta 800 2nd hand and was given help to set it up. I think there are a few on here that use Xilica units.
 
Busy weekend for me with DIY and sorting out my new living room, and a little work on my new listening room

One of the the BIG leaps of faith for me was letting go of my beloved Lamm pre amp. The multichannel DAC needs to drive the amps directly. In your case the sub late amps and the amp driving the Akubarik.

How are you planning to get the mic connected into the computer? Regardless of which DSP software you use. You could start trying out REW for free.

The Linn gear all needs to go, I think it's just a case of when!

I was hoping I'd be able to do sub integration 1st in a sensible way so I wouldn't need to replace any component twice.

for the Mic I have a UMIK-1 and and only analogue mic, so for now I can use USB but in the future I'll look at a new analogue mic and new sound card/ADC options.

I already have REW and have been using it, it's certainly an option short term.

In my experience balanced PC multichannel should be used with balanced amplifiers.
Mixing balanced and unbalanced, might give you sound, but also give you hissing issues.

good point & on the ground loop thanks. I think this just makes it a bit harder to choose the right DAC, seems I'm quite limited on options

Have you listed the preamp somewhere, without knowing what it is makes it hard to check whether it could work.

HDMI could work if the preamp has the right EDID to expose all of the channels HMDI has available. If it is a home theatre type of preamp then it should.

Easy to check by using Jriver or some other multichannel audio player/DAW to see if you can address the HMDI channels separately.

Most have both TS and TRS outputs, I've never had a problem doing this.

The Pre-amp/streamer is a Linn Akurate System Hub, Speakers are Linn Akubariks, they have on board power amp and DAC, the system hub is the DAC master clock and connection between the 2 devices is proprietary so I can't replace just the pre-amp.


Akurate System Hub - LinnDocs
Akubarik Aktiv - LinnDocs

Pre-amp supports HDMI 2.0 and there is an upgraded surround module available so I'd hope EDID should be find but can't find any real documentation on things.

Multichannel preamplifier.

PreAmplifier v2.0

Pay attention to product page. Among others items, balanced receiver, 8 channel preamplifier, ......

PCB-Kits

Regards

this looks quite interesting, 1 for me to consider more when start looking at amps
 
good point & on the ground loop thanks. I think this just makes it a bit harder to choose the right DAC, seems I'm quite limited on options
The options are not limited and the previous warning is not at all generally applicable. Ground loops are always a possibility in a multichannel setup especially when USB and a computer are being used. Using an amplifier with a balanced receiver input is a good idea and goes a long way to avoid the issue presenting. Connecting the single ended output to the balanced input amp with a pseudo balanced cable gets you some common mode rejection and help to avoid completely tying the main grounds of the devices together in the wrong place. I did exactly this with a Najda connected via USB with single ended RCA's to Hypex balanced amplifiers and it was completely immune to any noise or ground issues. The lowest noise combination of equipment I have used.
 
I would like to thank the contributors to this thead I have learned a lot.



to the OP from a layman newby POV I was in you situation without the budget. I had to sell my 802s and decided to try diy so bought used components for the seas Thor and with a lot of help from the people on this site built the small Thors. Loved them they were a noticeable improvement from the 802Ms . So building a known design for the first build is highly recommended.



If you are going to design your own speakers. I would recommend you build a kit first.

The other thing I would recommend is that you figure out what you like and do not like in a speaker.



Type - MT. TM, MTM. OB, Dipole. Line array, planer do you like the way the wide baffle strats sound or do you prefer the wilson individual adjustable boxes. there a so many choices

Material. metal drivers, Paper. kevlar planer so many choices



Each of these choices can help you narrow your preferences and compromises.

In the correct design each can sound great. but you will likely have a preference.



As far as moving to building your own design for your first build I would not recommend that unless you are willing to live with what you build or to make changes after you think you are done so in other words make it a hobby. my story - The magnesium drivers in the thors corroded. and because of my environment I replaced them with nextel drivers thinking I could change the drivers and maybe change few components in the crossover and have Thors again. After pestering and frustrating many people on this site. buying a unic mic and leaning new software (Holms) I got close and they sound good but I still need to build new cabinets and buy more crossover components or go DSP. After many months of reading and building a second set of full range speakers I feel I can attempt to build a good sounding speaker with the thor components plus additional components and I think I will have some degree of success. But again I am working with a crossover designed by a Master speaker builder and only changing a few components. I am not designing a new speaker. But now it is a hobby not really a build and forget sort of thing.



I have come to the conclusion that speaker design is advanced physics along with a good degree of Art. It is far more complex than it seems.



It is a great hobby.