[versus rider]

Yes thanks John, read and reread those, lots.
Oddly, the part that worries me most is the fact I have only an iMac, no MacBook or laptop, and not sure what gear I will need for measurements.
I have an old AudioControl active bass crossover device (this proved singularly disappointing) which came with a microphone so perhaps the mike could be used for in-room measurements?

ears are the best tools for measuring unless you plan to market your speakers to the general public. speakers that measure well tend to sound dull by comparison, which is the reason why I tend not to use the room eq function on the DEQX

[scorpion]

I have been using the SRC2496 and DCX2496 combination for a lot of years now and been quite satisfied all the time. I use steep crossover slopes were I for instance have found Butterworth filter better performing sometimes than L-R type. For mid and tweeter I have used small Trends 10.1 T-amps with excellent results. However what makes the greatest difference in sound is the quallity of the speaker units. I am very pleased with the last (?) version of the Blindstone. A B&G Neo10 might better it though.

/Erling

[studiotech]

I have been using the SRC2496 and DCX2496 combination for a lot of years now and been quite satisfied all the time. I use steep crossover slopes were I for instance have found Butterworth filter better performing sometimes than L-R type. For mid and tweeter I have used small Trends 10.1 T-amps with excellent results. However what makes the greatest difference in sound is the quallity of the speaker units. I am very pleased with the last (?) version of the Blindstone. A B&G Neo10 might better it though.

/Erling

Are you handling volume control pre or post SRC/DCX combo?

[studiotech]

ears are the best tools for measuring unless you plan to market your speakers to the general public. speakers that measure well tend to sound dull by comparison, which is the reason why I tend not to use the room eq function on the DEQX

Sorry, this is bad advice IMHO.  This is the kind of thinking in the hifi industry that leads to bad decisions and poor systems that we accept as OK.  I'm not saying that you HAVE to have flat response, but to NOT know what your system is doing, especially as a DIYer is bad form.  It all just becomes a big guessing game that you may or may not get lucky with and stumble upon some good sound.

Greg

[scorpion]

Greg,

Nowadays pre or in the SRC as I play through computer with Itunes, Airfoil and Airport Express with fiber access. Some analogue with the SRC volume control, vinyl and tuner. SRC to DCX digital 24/96 via AES interface.

/Erling

[JohnR]

I'd have to agree with Greg. "Measures well but sounds bad" suggests to me that there's a gap in understanding of the "measurements" and the correlation with subjective performance - please note that is not a criticism, I have and am still very much going through this learning process. I realize it's a completely different type of system than the DEQX but I've found manual setting of filters in the miniDSP to be much preferable than REW's automatic mode, which takes (in my experience) a lot of mucking with the algorithm parameters to produce a reasonable result. VR - you must surely be using (gated?) measurements on the drivers at least?

[Luigi]

Finally pulled the pin myself and ordered a miniDSP box to begin playing with X-over for DE10s in waveguides and B200s, with Alphas now powered by the Gallo 3 Ref amp.

Only problem with this set up, as I see it, is sorting power amp for the tweeters. But given these are hi sens I will probably try to find something with about two valves and one watt per channel. Like a Decware or some such.

Cannot wait to rid myself of soldering resistors, caps and inductors ad infinitum.

[versus rider]

I'd have to agree with Greg. "Measures well but sounds bad" suggests to me that there's a gap in understanding of the "measurements" and the correlation with subjective performance - please note that is not a criticism, I have and am still very much going through this learning process. I realize it's a completely different type of system than the DEQX but I've found manual setting of filters in the miniDSP to be much preferable than REW's automatic mode, which takes (in my experience) a lot of mucking with the algorithm parameters to produce a reasonable result. VR - you must surely be using (gated?) measurements on the drivers at least?

if this is in reply to my comments, firstly I said dull not bad. What is REW? Gated? again I have no idea what that is. I let the DEQX do the measuring of the drivers and carry out any error corrections. I choose the crossover points and filter slopes and then listen to the three options I have chosen , changing from one to another on the fly, sometimes changing one crossover point leads to me needing to change the other, i.e. I run a 4 way with the horns from 1500Hz being on what would be called the tweeter output from the DEQX through 2A3 mono's with a cap to low pass the 1000Hz horn, the 400Hz horn runs from 1500Hz until it runs out of steam.  I run the alpha's up to 200Hz and an Audax AP170MO FROM 200Hz - 1500Hz. This is the best combination I have found so far. The system as a whole is a good as anything I have heard and so I have decided it is time to disassemble the baffles for painting.

[matevana]

Wouldnt mind hearing experiences of any miniDSP users here.

Unfortunately my experience with MiniDSP was not so good. I really wanted to like the product, due to the promise of simplicity with almost no negative returns. I put many hours into it and ended up selling it after about a month. My comments were:

(1) Rev B offered enough input gain but the output level was not sufficient to drive any of my amps to full volume. The new balanced 2x4 unit may be a bit better, but its more complex and cant be USB powered. 
(2) The bit depth is questionable.
(3) It seems to boost by cutting. I was only able to address satisfactory dipole roll-off compensation by cutting critical mids and highs.
(4) The onboard DAC is fair at best. Most of my nicer DACs lost SQ being passed through the MiniDSP's. 

For now Im sticking with my DBX 2231 EQ my Ashley x/o despite the pure analog signal path. In my experience, they sound quite a bit better.

[JohnR]

Um, Rev B has lower gain... people do seem to have problems with the gain structure issue with DSPs.

How are you doing your dipole compensation now?

[Redefy Audio]

i would like to know also about the gains structure of minidsp.

right now im using F5 clone and SymaSym diy. im using digital volume on foobar with WASAPI.

i have 3 diff dacs to play with.
TwistedPEar ESS9018, POPPULSE PCM1796 & AST DAC PCM2793.

twistedpear and poppulse have similar gain.

AST pcm2793 with output voltage 2.75V was almost 2 times louder than other dacs.

but for me, its all very good performance from minidsp point of view. I cannot imagine setting up the system without the flexibility of the minidsp.

cheers
henry

[Redefy Audio]

hi johnR,

how to do the dipole compensation on miniDSP?

with ALPHA15 + BETA15 per side right now, i used +6db on 30hz, thats all.

cheers
henry

[JohnR]

hi johnR,

how to do the dipole compensation on miniDSP?

Hi Henry, I use a "HIGH-SHELF" shelving filter. You can see my curves in Figure 8 and Figure 10 here:

http://www.hifizine.com/2011/03/refining-a-4-way-open-baffle-speaker-minidsp-2x4/

It will vary with the woofer and the frequency range used.

[matevana]

Um, Rev B has lower gain... people do seem to have problems with the gain structure issue with DSPs.

Well, it's a balancing act between both stages. Rev B's higher sensitivity of 2V may be offset by a lower net gain in output, compared to Rev A's 0.9V input stage. My board allowed me to switch from A to B and neither drove the amps to full volume w/o clipping the input. Many people seem to have had the same problem. The balanced 2x4 has slightly more output, 2Vrms if I remember correctly, so maybe one day I wil try that.

In the meanwhile I continue to use my DBX 2231 which is an analog prosound dual 31 band EQ with type III noise reduction. There is zero noise plus all the clean output one could ever need, both balanced and unbalanced. It's signal/noise ratio is noticably higher than MiniDSP, there is no bit depth issue, and no degradation in SQ by introducing an extra A/D/A conversion step. 

[JohnR]

Interesting to hear, thanks. The digital card would avoid the "extra conversion" issue. The 2x8 card has higher output levels, although I've not actually measured it - not needed in my system.

i would like to know also about the gains structure of minidsp.

Basically the 0.9V output level (on the regular miniDSP) is a hard limit. That has to be high enough to drive your amps to full (or adequate) output levels. With the "Rev B" option the input sensitivity is decreased, purely for the reason of not clipping the A/D convertor with sources that generate higher levels. However you decrease the overall gain of the system by about 7 dB in that case.

[panomaniac]

Good topic.  Gotta say I like both.  Right now I run active 3-way with a modified DCX but some of the best systems I've ever heard were passive - 2, 3 and even 4 way.  True world class stuff.

So I'm working  my way back toward a passive crossover.  It just makes it all so much easier.

[studiotech]

As more manufacturers like Digmoda, Powersoft and Hypex make integrated solutions, the "easy" factor goes up for active.  As simple as changing virtual passive components in simulation, but with  real results and no tangles of wires from processors to amps and from amps to drivers.  Doesn't DEQX even have a solution with 4 channels of amplification built in now?  Anyone out there try the Hypex yet?  I will jump ship from Digmoda for my next studio monitor build and go with Powersoft.  Smaller footprint, cheaper prices.  Better for a midfield or nearfield application.

Greg