View Full Version : Digital Active Speakers - My experience to date
Hello to All
I am new to this forum and as requested I put a welcome note . One of your fellow AOS members suggested it might be worth putting a thread on detailing some of things I have recently completed turning some passive units in to Active speakers. Not sure if too many will be interested but will detail a few things here ad see what the response is. If nothing else it will give a starting point for a little discussion on this area.
Up to June or so last year I had always been a passive buyer of Hi Fi equipment and like many did a little bit of box swapping to get a different perspective from time to time. Having had a room at the Wam Show last year I found that some of the systems in other rooms that I liked to sound produced were all active speakers often with a full construction of cabinets and crossovers.
Sadly my practical and technical skills are fairly limited but having go involved with setting up my own Music Server Network at home based on LMS , Raspberry Pi and various players I have found I am quite adept at following instructions and getting things in place .
A further chat (admittedly while very drunk) with some of those at the show who had already made active speakers I cam away with an itch that I needed to scratch . Looking at the problem as a whole I tried to get a plan together that would be as simple as possible and also that would cost as little as I could get away with .
So my first project was defined as a Two Way Speaker , With Bi Wire terminals and sealed baffle design as a personal preference . I already knew that I was going to buy a MiniDSP HD 24 as this would take most of the heavy lifting out of the crossover part of the build . So scoured E-bay and found a pair of cabinet distressed but driver good condition Monitor Audio MA11 Monitors from around the 1986 era. Bought them for £15 and the project was now a reality . Almost by chance a MiniDSP HD24 cam on e-bay and joined the speakers .
A UMIK Microphone soon followed and everything was in place.
Next with the help of two Wammers I was shown how to set up a speaker to measure the units and establish the crossover points for the drive units using REW software . The speakers were then taken apart and the cabinets got a little refurbishment and the passive crossover disconnected and the drive units wired directly to the Speaker binding posts . Running the MiniDSP software I was able to put the crossover points in and switch between different slopes and types of crossover .
I then used two power amplifiers that I had in store (Trio Basic 1 (treble ) Quad 405 (Bass) ) and played a little music through them for the first time . The results were very good much better as a first trial than I had expected and the active speed of sound (how the notes stopped and started ) was very evident . Best of all the speakers were now sounding so much better than they did as passive units .
Flushed with my brilliance (yeah right) I started to study the REW software a little more and found that using the original sound measurements I had done on the treble and bass drivers separately it was possible to generate an automatic filter that made the out put come closer to the flat ideal that is supposed to be the goal . After many days and nights playing with these different filters I what I would call finessed the speakers the basic sound was the same but the bass treble and mid all became a single sound without any jumps between them the sound also started to Image really well.
Again I was happy with what I had done and the sound was for me still much better than I had expected . Then on reading through some of the threads from other MiniDSP users I was introduced to Dirac Live which is available as a bolt on software download and changes my unit to a MiniDSP DDRC 24 . Now this was a whole new world . Not only would my MiniDSP take care of the crossover but I now had a really powerful DSP tool that would calibrate and measure the actual sound from the listeners position and offer a very comprehensive set of filters that would produce a sophisticated change to the sound.
The first time I heard these speakers with a full DIRAC treatment I knew I was never going to go back to using the REW filters . It did everything they did and so much more . Best of all for the results the process was very simple and one that anyone could complete without any technical knowledge . Full instructions are given including on screen step by step help . The results are a real syrprise with very extended and musical bass and real tangable imaging .
I have since given these units away and moved on to my second project but will see how this thread is received before I add that .
Comments , suggestions , questions all welcome .
sounds great. youve done very well, and obviously got to grips with the problems. look forward to how the new set progress
How did you deal with the difference in output levels between the power amplifiers? Does the mini-DSP handle that?
Hi @Macca , with the standard MiniDSP HD24 unit the differences in output levels are indicated in the REW measurements and you can then adjust the individual units up or down to compensate within the MiniDSP software controls . However if you use the MiniDSP DDRC 24 with DIRAC Live then this software does this automatically for you .
I guess that makes life easier and you can use whatever amps you have lying around instead of having to buy matching units.
Hi Martin yes it does make things easier but my own paranoia still made me use two Quad 405 power amps and I had both sensitivity matched because if I did not then I would always have the nagging feeling that the digital adjustment would be too much or that it was stressing the unit . Silly I know but it amkes me breathe a little easier to know that the basic matching has already been done before I apply any treatment .
Hi Andrew
It was me that invited you to post about your DSP experiments. Not surprised to read that you found a big improvement to the sound, especially the speed of the sound. It will be interesting to see what next steps you take on this journey.
I've also played around with active crossovers but took a slightly different path to you (there is a thread somewhere I did where I got a lot of help from people here).
Essentially, as an experiment, I used an 7.1 AV amp (to access multi channels) and a 7.1 dac (again for multi channels) and use Roon to provide the active crossovers (via their parametric filters) driving a 3 way speaker (removed passive xovers and wired direct). So 6 (2*3) channels in total.
My next step is to measure the in-room response (via REW) and dial that into the picture (using Roon's convolution filters).
Not sure if the minidsp or the dirac software allows you to tweak the horizontal and/or vertical alignments, but if so you can increase quite dramatically the 'focus' of the sound.
All good fun .... :cool:
Well with my first project having gone well with a few setbacks but nothing major I now felt that I could tackle the real project I had in mind . The MA Speakers were unknown to me and were a cheap option to use and cut my teeth on making a passive speaker active .
I have owned for about eight years a pair of 1973/4 KEF 103 Reference Speakers . I first heard these at a friends bake off and was astonsihed at how good they sounded . I managed to find a pair on e-bay in fair condition at the huge sum of £63.42 , had to collect them from Lincoln but that was just a pleasant day trip for myself and my wife .
These have been my front pair in my AV set up , my speakers in my spare room and the main speakers in my music system . In all these uses they have continued to amaze me how great they sound considering how long ago they were designed and made .
So given that these sound so good in passive mode once I was confident of what I could do practically as it was important I did not destroy these speakers in the process as I do love them this was my second project .
Firstly the plus points , This is a straight Two Way Speaker design so the MiniDSP DDRC 24 was perfect . Secondly they are sealed baffle unit my preference as in my view this type of loading means that units are able to be used hard or close to a wall and that bass output is tight and controlled . The crossover in these speakers is easily accessed with the front mount removed and is bolted in to place so the full board just needs desoldering and then removing along with the mounting legs etc (these are stored away in case i ever decide to go pack to passive ) .Thirdly KEF very kindly published full specifications for both drivers including the measured and correct crossover point so no need for any REW measurement sessions as I already have enough to put the correct values straigh in to the DDRC 24
Drawbacks these are very old drivers they are in superb condition but they are likes hens teeth to buy in paticular the tweeter as it is a version that was only really used in this speaker . Secondly these are single wired units and need to be converted to Bi Wire . The speaker connections are in a hard plastic exstrusion on the rear with a scooped out shape which is not condusive to adding a second pair of 4mm sockets.
So a very intense day of deconstruction and bagging of all bits removed . Careful removal of the foam padding inside the unit and a quick note of how it came out so it can be replaced the same way . I had considered looking at the internal padding and possibly looking changing the amount (reducing the amount) or even switching to some thing like long wool but I very quickly came to the conclusion that KEF knew what they were doing far more than I di so keep things as they were . Then it was heat up a suitable skewer and burn the new holes for the additional 4mm banana sockets . Would have preferred to drill this but the shape of plastic extrusion made this very difficult so quick burn was the best option. Burn a smaller hole then you need then use a srewdriver (flathead) to slowly gouge the whole to the right size testing the fitting of the sockets till you are happy that tey can go in tightly but easily . Another essential with a sealed baffle of this age is to source and buy some sealing foam gasket material . Help from KJF Audio here with the neoprene tape sealant he sells which has a very good ability to spring back to shape after sealing so in case you need to open the front again after construction the material will still provide a good sels .
THen it was a case of wiring the individual drivers to the sockets for me keep with the norm Bass to the bottom two sockets treble to the top two . Fit the gasket sealing strip then replace the front panel and tighten it up carefully to ensure a good all round seal . With all the mechanical work now complete it is just a plug in to the DDRC 24 start a DIRAC Live session take your time and measure out the ten measurement points . Then as shown on screen optomise the filter and apply to the speakers .
Final result is honestly one of the best sounds I have heard at home . All the good aspects of the KEF 103 units are maintained but then you add in the speed and clarity of active and you somehting very special . I know anyone reading this will think , well you would say that after all the work you put in and I am sure there is an element of that but I have been listening to music on good sytems for over fifty years and I know what a good sound is.
Rather than go on about it I woud like to make anyone interested in hearing this sytem an offer. I am going to have a room at the Hi Fi Wigwam Show which is held at the Kegworth Hotel in Kegworth in the Midlands . The show runs from Saturday 16th (Set up day) with the main show on Sunday the 17th . Anyone who wants to can come along and this will be my main system . Come along listen and see if you agree that this is now a very capable system indeed and worth the work that has been put in to achieve it.
If I review the costs involved they are not that cheap but many of the additional costs are only at my choice the whole thing could be done for well under £1000 which given the quality of sound you get for me is money well spent as this will last you for many years of happy music enjoyment .
Only draw back which I freely admit is real is that once you start down this road you do have a tendancy to keep fiddling , just one more meaurment session with really accurate positions , or slight higher , shlightly lower mic placements just to see if it changes . Various changes to the House curve you constructed a little tweak here a little tweak there . Use the Power Amps one left one right or one for treble and one for Bass . This can go on forever if you let it . I think I am over most of that now and just leave things alone and love the music .
Thank you to anyone who has bothered to read through this I know it is a bit long but it has helped me to get it out and written down . If you are thinking about trying this then I am always happy to help either with what little knowledge I have , practical help to set up or even just support. I think active and DSP is a very positive way forward with music in a domestic setting as it gives you the real chance to change the weak link in our music chain which is the listening room which for most of us were never designed or built to be listening rooms and are always a compromise.
Interesting read Andrew. Thanks for that. I plan on going to to Kegworth and will look you up. I see you are also doing the suitcase challenge.
Great write up. Speaks sound great
Thanks for sharing that.
I've made my 3-way Heybrook Sextets active using 2 miniDSP 2x4HDs. I started out with what amps I had available, initially Marantz MA22 monobloc power amps, Quad 405 and the power amp section of a vintage Marantz 2285B receiver, the control the miniDSP has on these elements is quite amazing and dealt perfectly with the various sensitivity values of both the amps and the speaker drivers.
The first part of the conversion to active (after removing the crossovers) was to work out the crossover points of the units from the crossover and then replicate this using the miniDSP which was very easy with the crossover types that are available to you. This enables you to pretty much reproduce the sound of the speaker, that's when you can really start to play. Using REW (Room Equalisation Wizard) and the UMIK-1 microphone you can get a response curve to baseline your work so you can see the effects of phasing, the cone responses, the tail off of the drivers etc and then you start to move the crossover points, taking advantages of the crossovers types.
I changed the amp set up once I became happy with the results, so now I have two Linn AV5105 (two LK240 in a single box) providing the bass and the highs, with my MA22s covering the midrange.
I'm going to make the leap to Dirac Live, what Andrew said about the REW filters etc really has helped make that decision. Whilst I enjoyed the listening, the monitoring and then using REW to model the response, it became quite tedious and you always felt the need to fiddle (well I did!) however the results are amazing, just what I was looking for and breathed further life into my Sextets.
Interesting read Andrew. Thanks for that. I plan on going to to Kegworth and will look you up. I see you are also doing the suitcase challenge.
Yes I will have very minimal system which I must say I am surprised at how good it sounds. Look forward to meeting you .
252052520625207
I am putting this here because it concerns the MiniDSP DDRC 24 but it is also connected to Quad ESL57 speakers which I realise should be in Past Masters apologies if I have got this wrong . As well as the above Active system which I use in my spare room I also have a MiniDSP DDRC 24 in my main system which is SBT Clone - Toslink out , MiniDSP DDRC 24 then two analogue outs to my Krell KSA50 connected to a pair of refurbished Quad ESL57 speakers.
Now I use the MiniDSP in this instance in full bypass mode for Crossover and Equalisation but use the DIRAC Live filter to adjust the Quads in the room . However I have had a very frustrating time recently since the newly refurbished Quads have come back with issues over distortion when the volume is raised and when there is heavy bass content . I have changed the routing of my system to the MiniDSP from analogue via a Concordant Valve Pre and Metrum Quad DAC in to the analogue inputs on the MiniDSP to a direct digital in but this dod not solve the problem .
After help and advice from others that use the MiniDSP and based on the graphs I have copied in below I was made aware that the target curve that was generated (Graph 2) was asking toom much to drag the Quads dip to the required level and this was causing the over load and distortion in the MiniDSP . So instead I have set the revised target in Garap 3 and I am now able to play the Quads to a very decent level with no distortion . I have lost a little power in the Bass but what is produced is at least clean and easy to follow. I am sure the settings can be tweaked and I can get a little mor lower bass out of this without setting off the MiniDSP but at the moment I am just going to listen to the musc for a few weeks and let my ears grow used to the current settings before I try any changes.
This just shows the very different roles you can use the MiniDSP unit for and how it can make a real difference in almost any system .
Discopants
02-02-2019, 10:14
Very interesting read Andrew thanks for sharing. Could the bass response you are noting after the quad service , just be changes in room response ?
andy335touring
02-02-2019, 21:02
Thanks for sharing, i've been contemplating going active for ages but with new drive units/design rather than converting an existing speaker.
A side from not having a lot of spare cash at the moment i'm constantly changing my mind on which sort of design i want to go for but one or two things are fixed is that like you i'm going for a sealed design and it needs to be a fairly compact floor standing speaker.
If i'm not working i hope to see you at the Kegworth show :wave:
Very interesting read Andrew thanks for sharing. Could the bass response you are noting after the quad service , just be changes in room response ?
@Discopants Sorry i might have misled you the problem with the distortion was there before the Quads were refurbished and in no way would I make any suggestion that the work done has been anything but excellent . Thye are miles away as plain speakers from when I bought them . All issues are with the DIRAC filter and my room and the fact that DIRAC will try to correct a problem to try and achieve the ideal house curve response even if this is nt really possible with the speaker you are using. Like lots of modern equipment and software they make much of what they do simple and easy to use we forget that they need you to still work and try and understand what they are doing . They can still get things wrong you need to be able to spot the areas and address them .
Andrew (bencat), yes your experience matches what I found as well (even though I have not used minidsp/Dirac). Always best to concentrate on EQing the peaks and not try to EQ the bass dips too much. So setting the target curve appropriately will be the way forward. Or just remove the big +ve bass EQ points manually (not sure if minidsp/Dirac allows that).
Generally large bass dips (assuming the amp is flat to such levels) is caused by poor room acoustics which suck out bass. Trying to EQ these dips is asking the amp to do far to much (=distortion). Best to cure the room response passively with say bass traps or diffusers as the case may be. Passive room treatment is still on my todo list so lots to learn there.
Andrew (bencat), yes your experience matches what I found as well (even though I have not used minidsp/Dirac). Always best to concentrate on EQing the peaks and not try to EQ the bass dips too much. So setting the target curve appropriately will be the way forward. Or just remove the big +ve bass EQ points manually (not sure if minidsp/Dirac allows that).
Generally large bass dips (assuming the amp is flat to such levels) is caused by poor room acoustics which suck out bass. Trying to EQ these dips is asking the amp to do far to much (=distortion). Best to cure the room response passively with say bass traps or diffusers as the case may be. Passive room treatment is still on my todo list so lots to learn there.
This is my experience too. Never try to bring up a dip of more than 3dB always lower the peaks and plateaus.
Right now back to an update on my active set up . I was planning this system to be a Raspberry Pi with Picore and Justboom Digihat digital out to a MiniDSP DDRC 24 then from the MiniDSP out to two Quad 405 amps upgraded and matched by Fidele Audio . Problem was when I got the Quads back one developed a problem which was complete silence . So back to Fidele which diagnosed an Speaker Protection Baord that was wrongly configured so not fast enough to turn off before damge to the boards were done (nothing to do with Fidele these were Net Audio mods) .
So while the Quad was away being repaired I used one Quad for the Bass on my KEF 103's and a Trio Basic 1 for the treble . A DIRAC Live mesaurement session and this worked very well indeed and sounded excellent .
Wednesday the repaired Quad arrived back and as I only had the evening to mess about I just did a straight switch of the Quad for the Trio still looking after the treble for both speakers. Another DIRAC measurement session and put some music on . Firstly and best result music came out and it sounded good , just good not really as good as the Quad / Trio combination which was a bit of set back . Did some adjustments with music to the filter curve and this improved things but still not better than the Quad / Trio .
Just listened to the system through Wednesday night and Thursday until this morning as i do not work on Fridays it is a great day for getting stuck in to a project . So first made up a set of longer speaker cables for both speakers as this has been just a spare rooom system in a small room and only needed short cables. In five weeks I will be taking this to Kegworth so needed a new set of longer cables in order to be able to get a wider spread for the spekers in a bigger room . While I was doing a full refit of the amplifier s seemed the ideal time. Then I set the Quad amplifiers up as a single stereo amplifier for each speaker right and left with each amp doing Bass and Treble (Right for bass Left for treble) . Once completed turn on and sigh of relief nothing , no pops no buzzes no hums everything as it should be .
The out with the tape measure and the coloured notice board pins and measure out and complete a complete new DIRAC Live session . This went as expected having done quite a number of these over the last months i now seem to be able to get everthing in sync and completed very easily . The suggested House Curve shown at the end of the session looks pretty much like all of the previous ones I have done . However I know from experience that it needs a little tweaking here and there to suit my ears and my system . applied this and then press the music play button . Smile as wide as the Mersey as msuic comes out and the sound is not just good bu excellent . Switch back to the filter page in DIRAC and i am able to make some very small adjustments on the fly and hear the improvements (or were i ahve gone too far you know how it is) .
To me supreme delight this new set up is now not only better than the Quad / Quad for Bass and Treble set up but also much better than the Qaud / Trio set up . Main areas on improvement is in the texture of instruments they sound more real with little elements like the sound of astring now being much easier to hear. Then imaging this has just snapped in to place and is so much better than the passive speakers and all my other active ones as well . These are sealed baffle design and I use them hard against a back wall which reduces the imaging ability well this now brings lots of that back but i retain the speed of notes and bass reinforcement I was gettin from placement .
I am really pleased with my finak results and will be very interested to hear any comments (good or bad) at Kegeworth . I have suddenly realised that this is a very personal system made up for my own personal use in a single space and I like what it does . If that can be tranferred to a different room and the sound I get will be enjoyed by others is a different thing. My Wife Linda has just come and said that the last track (Ladysmith Black Mambazo) sounded very good so maybe I have got on the right track at least. I am sure there are little tweaks and changes that will imrpove this but I am satisfied with how it sounds now and I think I will leave the DIRAC button alone and just enjoy the music .
Andrew.
Excellent write up and seems like you intuitively know what to do when adjustments are required.
I'm confused how you set up the Quad amps. Are you saying that you use one Quad amp to drive the LH speaker and the other Quad amp to drive the RH speaker splitting the stereo channel on each to drive the bass and tweeter respectively?
Do you have a picture of the house curve you use showing the full spectrum? How many EQ points is Dirac coming up with to match the target curve?
§
Hi Edward
Will try and be clear as it is vital that we are talking about the same set up . I use one of the Quad 405 amplifiers for the Left Speaker with the Right channel connected to the Bass input and the Left channel connected to the treble input . These use MiniDSP outputs 3 and 4 which are powered by Input 2 . I then use the second Quad 405 Amplifier connected to the right hand speaker again the right channel connected to the Bass inputs and the Left channel connected to the Treble inputs . This is connected to the MiniDSP via outputs 1 and 2 which are driven by input one . The DDRC units is set up using the DDRC software to be the digital crossover for the speakers in the case of the Kef 103 units this is set with Bass from 10HZ - 3000HZ and Treble from 3000 HZ upto 20000 kHZ I used a L&R 24 DB slope CR .
I have added a couple of screen prints from the DIRAC porject which I hope are what you want. One of the things I find most useful about the DIRAC is that once you have finished all the measurements and the Filter is generated with the project still open you can play music through the system and then from your normal listening position you can make changes to the curve there and then and they apply straight away so you can hear what the change is and make little (or large) adjustments as you want . If you mess up then no bid deal just hit the button and it will take it back to the proposed house curve the system gives you which I am sure for many will be good enough .
Thanks Andrew. That is clear how you set up the Quads. However I don't understand why that config sounds better than just the 'normal' way of one amp driving the bass and the other the tweeters - all four channels of the Quads ought to be the same (give or take). And if one amp is mismatched (relative to the other) then in your setup you would get a mismatch between L and R (correctable of course via Dirac).
Your screenshots have come out too small to be easily legible - sorry.
Do you have a valve amp you could use to drive the tweeters? Could add some 'magic' to the top.
Oh sorry, another q if I may? Any reason why your are crossing at 3kz? Can you cross outside (either below or above) the voice frequency of say 250hz to 3.5kz
The main reason I have for explaining why this set up configuration sounds better is that there is zero cross talk . Both speakers and amplifiers are completely electronically and physically seperated. My two Quad amps are matched both have the identical same upgrades ,they have been matched for input sensitivity and both have the current limiting removed. The only physical difference they have is two different Mains smoothing Capacitors which should and I think does make no difference at all . While I enjoy measurments and the things they can show I also think we have to use the best measuring instruments we have which is the ears on the side of our heads and our brains . If you listen and it sounds good then it is good . Someone telling you that this or that is off or wrong is not really helpful or useful if you like the music being produced. Often it is stated as if it is fact that a perfectly flat response from 10 - 20000 is the ultimate goal . Well I have heard perfectly flat and frankly it is dull grey and boring and if I had to listen to music like that I would stop listening .
The crossover points are set by KEF who designed and made the drivers in my speakers. I checked using REW and guess what the 3000 kHZ crossover points were spot on they are the natural mechanical roll off of the bass drivers were they start to degrade and the strong rising point of the curve to the treble units . KEF as a compnay had some very gifted and high quality sound engineers working for them and they had much greater technical and engineering skill than I ever had or will have. If it was right for them I certainly can confirm that it is right for me . Testing , using information supplied by the makers and listening all has to be balanced and then you need to have conviction in your own ability to decide what you like and hear .
Yes I have tried nudging the cr point down a little and up a little but in each different case the sound was worse . The cr points I use are the right ones for this speaker with these drivers , sometimes we have to not try and invent the wheel but trust what more gifted individuals have already proved. That does not mean you should not try and see but do not expect that when others have done the same work you do your result will be different . In many cases if you do the exacr same measurment test the chances are it will be exactly the same . it should not stop you doing that work if you feel you want to but often all you do is confirm others results.
With regard to the attachements did you right click on them ? This gives you a bigger image which I checked and while not huge it was perfectly readable .
Powered by vBulletin® Version 4.2.3 Copyright © 2026 vBulletin Solutions, Inc. All rights reserved.