Welcome, Guest
Username: Password: Remember me

TOPIC: Re: Active crossover listening test

Active crossover listening test 9 years 9 months ago #1511

  • paulspencer
  • paulspencer's Avatar
  • Offline
  • Senior Boarder
  • Posts: 40
  • Thank you received: 1
  • Karma: -19
Not sure if this is the right part of the forum, but I thought you guys might be interested in a recent series of listening tests. We compared 4 different active crossovers:

MiniDSP
DCX
DCX (with about $1k of mods)
DEQX

We tested in two systems on two different days. Each event took about 5 - 6 hours for setup, measuring, listening and lunch in the middle. On the second day we compared also the digital conversions. On the first, we had a time delay in-between change over. On the second, we had a custom made switch and level matching box. It meant we could instant switch to better reveal differences. If there was any shift in imaging, any chance in the sound we would hear it right away.

5 were present.

In the first event, we had a compatibility issue. It seemed the valve pre wasn't suited to driving it. It was a very high end system and in a blind test, the owner could pick MiniDSP over DCX 3 times out of 3 easily. He thought that MiniDSP was better.

On the second event, the system was solid state with Unity horns. We did pick some differences in both tests, but they were quite subtle. Again, MiniDSP was a slight improvement over a modified DCX. That's quite an endorsement since MiniDSP costs $100, while the DCX here is over $500 and it had $1000 worth of upgrades.

We didn't get to test DEQX since it was skewing the response and something clearly wasn't right.

The full write up, along with links to the original discussion threads, are here:

redspade-audio.blogspot.com/2010/12/acti...listening-tests.html
The administrator has disabled public write access.

Re: Active crossover listening test 9 years 9 months ago #1519

  • devteam
  • devteam's Avatar
  • Offline
  • Administrator
  • Posts: 8757
  • Thank you received: 1140
  • Karma: 163
Nice writing Paul!

Glad to hear that you guys are having fun down under with your listening sessions.. must be quite some entertaining time! :-)

Thanks again for sharing with the community your findings,
miniDSP, building a DSP community one board at a time.
The administrator has disabled public write access.

Re: Active crossover listening test 8 years 11 months ago #3617

  • Eftegari
  • Eftegari's Avatar
  • Offline
  • Fresh Boarder
  • Posts: 10
  • Karma: 0
@paulspencer, Impressive results! This review instantly persuaded us to buy two miniDSPs!


One question, what power supply did you exactly use to power the miniDSP's?
The administrator has disabled public write access.

Re: Active crossover listening test 8 years 11 months ago #3620

  • paulspencer
  • paulspencer's Avatar
  • Offline
  • Senior Boarder
  • Posts: 40
  • Thank you received: 1
  • Karma: -19
You will find the answer here:
gainphile.blogspot.com/

I don't recall, but I think Andi made up his own power supply on a breadboard.
The administrator has disabled public write access.

Re: Active crossover listening test 8 years 10 months ago #4008

  • klc
  • klc's Avatar
  • Offline
  • Fresh Boarder
  • Posts: 3
  • Karma: 0
Hi' all

I've just done some listening tests.

Squeezebox > Linn Numerik > Linn Kairn > Linn Klout > Linn Ninka

Squeezebox > Linn Numerik > Linn Kairn > miniDSP > Linn Klout > Linn Ninka

Play'n FLAC files.

All settings in the miniDSP set to bypass so I guess it is the quality of the ADC - DAC I hear.
The Nink's are passive.

This was a test of the miniDSP vs. no miniDSP and I have to say that the miniDSP is no where near a straight wire. I didn't expect it to be but I was quite surprised how bad it really was.

Like the noise floor was raised 30 dB, lack of low level details and dynamics, music just became lifeless, flat and boring.

As I said it was miniDSP vs. straight wire and I have no idea how it performs compared to other DSP's. Given the test on this forum it doesn't look good for other DSP's.

That said I'm still impressed by the concept and usability and software etc. It just works. It's quite fun to play with.

Maybe I'm expecting to much.

Anyone else done that comparison?

Before the above I run one of my Nink's active through the miniDSP and one Nink's passive through the miniDSP (all settings bypassed).
The active setup to roughly same response of the original Linn active modules

The result proved the advantages of going active.

I've not tested (AB) active Nink(minidsp) vs a passive Nink(no miniDSP) but but from memory I'm quite convinced that the passive configuration wins by a fair margin.

I'm still a miniDSP fan.
Last Edit: 8 years 10 months ago by klc.
The administrator has disabled public write access.

Re: Active crossover listening test 8 years 10 months ago #4009

  • curryman
  • curryman's Avatar
  • Offline
  • Platinum Boarder
  • Posts: 792
  • Thank you received: 181
  • Karma: 101
Very interesting test! price of miniDSP is quite low compared to the rest ;-)

Did you measure raised noise floor (30dB) or is it just a guess?
Guess you know, that the setup (Kairn Pre - miniDSP - Klout Amp) is not the optimum for a DSP based setup. What was the output level of the pre when doing the listening tests? Optimizing the gain structure can improve things significantly!

I am using volume control after the DSP and don't have any noise issues.
The administrator has disabled public write access.

Re: Active crossover listening test 8 years 10 months ago #4010

  • klc
  • klc's Avatar
  • Offline
  • Fresh Boarder
  • Posts: 3
  • Karma: 0
Noise floor is estimated :)
It just seems so since low level detail lacks.
It's not that I hear noise it's just that finer details blur some way. Hard to describe but your'e welcome to have a listen :)

The level thing cross'd my mind but the output was about 50 on my Kairn and 80 is 0dB (straight through) so the level should be around 3-400 mV.

Btw. I have both USB and a 12V SMPS powersupply connected. Maybe I should carry out some power supply experiments.
The administrator has disabled public write access.

Re: Active crossover listening test 8 years 9 months ago #4011

  • Sean M. ONeill
  • Sean M. ONeill's Avatar
  • Offline
  • Fresh Boarder
  • Posts: 10
  • Thank you received: 3
  • Karma: 1
if the Kairn measures in db, then 30db down would lead to a max input voltage at the minidsp of 63mv (2v input)

--> that is not sufficient input level - its not shocking you would see reduced detail.

It would be much better to do the squeezbox spdif into the minidsp and then output to the Kairn.
Last Edit: 8 years 9 months ago by Sean M. ONeill.
The administrator has disabled public write access.

Re: Active crossover listening test 8 years 9 months ago #4012

  • paulspencer
  • paulspencer's Avatar
  • Offline
  • Senior Boarder
  • Posts: 40
  • Thank you received: 1
  • Karma: -19
It's not easy to comment on tests done when you weren't there, but in our test we found MiniDSP performed very well, equal in sonics to the much more expensive DEQX (4.5k compared to 0.1k!) Forty five times the price!

Since this test I have been involved in another DAC comparison where we heard DACs costing about 3 - 12k. Guess what? The differences between them were greater, not hard to pick. For all that extra money I'm not convinced they are better than MiniDSP in terms of their DACs.
The administrator has disabled public write access.

Re: Active crossover listening test 8 years 9 months ago #4015

  • devteam
  • devteam's Avatar
  • Offline
  • Administrator
  • Posts: 8757
  • Thank you received: 1140
  • Karma: 163
@ KLC,

Hard to know how an "estimated" measure (i.e. subjective most of the times) relates in dB.. 30dB sounds a lot by any means so please use measurement gear to get more accurate statements.

I think that most likely it's a gain structure problem as already highlighted by a couple of members here. In a DSP (or any audio system for that matter), matching the sensitivity of input and output levels matters. What's the SNR of your gear you have in line? You can read the measured DR of the miniDSP (circa 96dB). Loosing 30dB would mean running a system at 60dB SNR.. I think that most likely, you're feeding too small of a signal to the unit (i.e. maybe need tweeking on the input jumpers) and trying to amplify that small signal after that through your amp. So yes, it wouldn't sound good.. :-)

So maybe before looking at new PSU, check your gain structure (i.e. good RMS gain at input of DSP, good RMS gain at output). Might bring more improvements.

Hoping this helps

DevTeam
miniDSP, building a DSP community one board at a time.
The administrator has disabled public write access.

power supply 8 years 9 months ago #4033

  • Eftegari
  • Eftegari's Avatar
  • Offline
  • Fresh Boarder
  • Posts: 10
  • Karma: 0
KLC & others,

How about this powersupply?

www.minidsp.com/forum/6-diy-hifi-project...ly-ever-for-minidsp-
The administrator has disabled public write access.

Re: power supply 8 years 9 months ago #4034

  • klc
  • klc's Avatar
  • Offline
  • Fresh Boarder
  • Posts: 3
  • Karma: 0
I have no idea and I haven't had the time to investigate power supply improvements yet but I will.
I'm thinking about making some recordings with and without miniDSP in the chain. Just need to figure out where to store them. Then we can see if I'm nuts :)
The administrator has disabled public write access.

Re: Active crossover listening test 8 years 9 months ago #4035

  • hendrix
  • hendrix's Avatar
  • Offline
  • Fresh Boarder
  • Posts: 7
  • Karma: -1
hello curryman
i have a simular noise problem,

I think the output voltage on the 2x8 board is to high for my amps or you can say the input sensitivty of my amps (1.9 volts unbalanced 3.8 volts balanced) is to high ( mcintosh mc352 350 watts per channel) so I think this is a gain problem.
This causes an audible hiss in my high frequenzy drivers (tad 2001, sensitivity 109db)
I tuned the board at the lowest possible output with the dipswitches down and using unbalanced outputs.
The VU meters on my amps indicate also a raised noise level of about 30db unbalanced about 35db with balanced connections i think i cannot use my input jumpers on the board to lower the input sensitivity to 0.9Vrms since my pre amp (mcintosh c42) has a rated output of 2.5 Vrms and a maximum output of 10Vrms can you please explane to me your:whistle: :whistle: :whistle: volume controls after mindsp and how these volume controls effect the SNR.
The administrator has disabled public write access.

Re: Active crossover listening test 8 years 9 months ago #4040

  • curryman
  • curryman's Avatar
  • Offline
  • Platinum Boarder
  • Posts: 792
  • Thank you received: 181
  • Karma: 101
My current system setup:
CD Player digital out -> miniDIGI -> miniDSP (digital volume control, RCA out) -> Naim NAP 160 + NAP 110 -> Audax HM170 + B&G Neo3

With respect to the gain structure the important hint is, that I do not attenuate the signal before the DSP (actually using digital input without digital attenuation ;-)).

I am just using the build in volume control of the miniDSP 2x8 (digital volume control after DSP controlled by build in IR receiver) which works really good! I think the loss of resolution of the digital volume control can be neglegted and the noise of the DACs is very low. In another setup with miniDSP 2x4 I used an external PGA2310 based volume control between miniDSP analog outputs and poweramps, however for 2x8 this is not necessary!

What makes it quite easy with respect to noise (hiss) in my setup is that the speakers are not that sensitive (arround [email protected],8V and 1m);-)

With speakers having 109dB even the noise of the power amp (input shortened) might be an issue?

Have you ever tried to use the digital volume control of the miniDSP? using the build in IR receiver and a suitable remote control this is quite easy. When synchronized with the plugin software you can watch the actual attenuation in the software. Just try setting it to -40dB or -60dB, then you can raise the level in your preamp and go into the miniDSP analog input with a higher signal. In this case you will not loose that much in the A/D and the DSP;-)

Looking forward to hear your findings!
The administrator has disabled public write access.
Moderators: devteam