• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Minidsp Flex Review (Audio DSP)

Rate this product:

  • 1. Poor (headless panther)

    Votes: 4 0.7%
  • 2. Not terrible (postman panther)

    Votes: 16 2.8%
  • 3. Fine (happy panther)

    Votes: 117 20.6%
  • 4. Great (golfing panther)

    Votes: 431 75.9%

  • Total voters
    568
Joined
Mar 16, 2022
Messages
57
Likes
95
Location
Earth
As I mentioned in my previous post the Flex does Dirac on the input, therefore you will not be able to apply independent correction to each output. This is clear from the user manual which shows Dirac on the 2 channel input upstream of the routing matrix / xover / PEQ.

View attachment 204478

Michael
That's unfortunate - thank you for the information.
 

levimax

Major Contributor
Joined
Dec 28, 2018
Messages
2,348
Likes
3,462
Location
San Diego
The internal DSP of the MiniDSP Flex allows digital processing, including precise EQ and 2-way crossover functionality. However, the almost completely unique feature is its ability to implement FIR based phase coherent crossovers between woofer and tweeter.

It is a fact that any normal analog crossover, and any normal IIR digital crossover that emulates an analog crossover, that has a slope over 6 db/octave, has uncorrectable phase shift. While the woofer and tweeter outputs can be in phase (or 180 degrees out of phase, correctable with a polarity reversal of the tweeter) at the crossover frequency, there is a net phase shift across the audio spectrum. For example, a 12 db/octave crossover has a net 180 degree phase shift from lows to highs across the audio spectrum. A 24db/octave crossover has a net 360 degree phase shift from lows to highs.

This wreaks havoc with phase coherency. It is thus impossible for a speaker system consisting of a separate woofer and tweeter with anything greater than a crossover with a 6 db/octave slope to reproduce a recognizable square wave. To reproduce a square wave correctly, both the amplitude AND phase of each harmonic must be reproduced accurately.

Here is a snapshot of what a 4th order 24 db/octave LR crossover does to a square wave:
View attachment 204538

The only practical way around the inherent phase shift is to use an FIR (Finite Impulse Response) based digital crossover. This MiniDSP product is one of the very few devices on the market that has FIR based crossover filter capability. It allows a woofer/tweeter based speaker to actually reproduce a recognizable square wave.

Here is one cycle of a square wave from one of my homemade line source speakers using a digital FIR based 100db/octave crossover, in room, with the microphone about 5 feet from the speaker. Note the room reflections starting immediately to the right, and even in the lower half of the square wave.
View attachment 204549

The question is, is being able to line up the fundamental and harmonics of a square wave accurately audible? Most people can't answer that because they have never heard fundamentals and harmonics properly lined up with near zero phase shift across the audio spectrum from a speaker, and thus can't say if it's audible or not.

Having read of studies on the audibility of phase coherence, and then having experimented with it myself with the MiniDSP 2X4HD, phase coherence (or lack of it) is indeed audible. Here's a link to a study on the audibility of midrange phase shift:
https://www.audioholics.com/room-acoustics/human-hearing-phase-distortion-audibility-part-2

The main audible effect of lining up all harmonics with the fundamental is a less 'speakerish' and more realistic sounding midrange, especially with percussive musical instruments with lots of harmonics. It sounds more like live music and less like speakers reproducing music. 'Midrange transients' sound more realistic. Well recorded applause sounds more like applause. A well recorded solo singer sounds more realistic and 'present'.

I am not financially associated with MiniDSP in any way. I have just experimented with phase coherent crossovers using the older Mini DSP 2X4, and have become strongly convince that phase coherence (or lack of it) is indeed audible. Phase coherency is indeed a step forward in audio reproduction. I am looking forward to getting this new and improved product.

From the study you linked to:
  • There exists in this study no statistically significant evidence supporting the audibility of phase distortion in the musical samples provided, using the all-pass filter implementations chosen by the researchers.
Are you saying that you can hear a difference with music? If so this is an extraordinary claim that contradicts the study and it would be helpful if you could provide some evidence for this claim such as a level matched blind ABX test.
 

Sluraad

Member
Joined
May 14, 2021
Messages
51
Likes
50
That's unfortunate - thank you for the information.
While it would be sweet if the MiniDSP Flex was a 2.1 device, you can still use the PEQ on your sub, tune the delay with your L/R, apply a crossover and THEN run Dirac on the combined crossed over signal. It works pretty fantastic.
 

Dan B

Member
Joined
May 5, 2022
Messages
10
Likes
5
From the study you linked to:
  • There exists in this study no statistically significant evidence supporting the audibility of phase distortion in the musical samples provided, using the all-pass filter implementations chosen by the researchers.
Are you saying that you can hear a difference with music? If so this is an extraordinary claim that contradicts the study and it would be helpful if you could provide some evidence for this claim such as a level matched blind ABX test.
Yes, I can hear a difference. Perhaps some of these studies did not use phase coherent speakers in their testing.

If you have a MiniDSP Flex, take it out of Dirac mode and implement a phase coherent FIR crossover yourself. I use the program rePhase.exe to generate the coefficients. It's all explained at the MiniDSP site.

To quote from the same website you did:
" Although not in large numbers, previous research in investigation of the audibility of phase distortion has proven that it is an audible phenomenon. Lipshitz et al. has shown that on suitably chosen signals, even small midrange phase distortion can be clearly audible. Mathes and Miller and Craig and Jeffres showed that a simple two-component tone, consisting of a fundamental and second harmonic, changed in timbre as the phase of the second harmonic was varied relative to the fundamental. The above experiment was replicated by Lipshitz et al., with summed 200 and 400 Hz frequencies, presented double blind via loudspeakers resulting in a 100% accuracy score.

That means that 100% of everyone who listened could hear a difference when the phase of the second harmonic was varied relative to the fundamental of a simple tone. Based on the 400Hz frequency of the harmonic and the statement that a 'small' amount of phase distortion was 100% audible, and that a 'large' amount would logically be 180 degrees, or 1.25 ms, we might be able to surmise that less 60 microsecond phase shift would probably be easily 100% audible.


John Dunlavey used to make highly respected phase coherent loudspeakers. To quote him from an old Stereophile article

"Dunlavy: We participate regularly in the recording of our symphony orchestra here in Colorado Springs. We've also recorded instruments like violins and cellos and timpani in our big anechoic chamber using instrumentation-quality microphones and equipment. And we find that in order to reproduce those sounds with a level of accuracy such that you can not literally hear any difference between the live and the recorded sound, you have to have a speaker that exhibits almost perfect impulse and step responses. The only way to do that is to time-align the drivers very, very accurately, usually within a matter of a few microseconds, then use a minimum-phase, first-order crossover network and get everything right. And you have to have an on-axis response of better—well better—than ±2dB."

Now, the same accuracy in step and impulse responses can be achieved with steep FIR filters, reducing out of band problems with the drivers. Inter driver time-alignment differences can easily be dialed out. Besides the phase coherent crossover itself, phase and frequency response anomalies of the drivers themselves can also be compensated for.

You may not be able to hear any differences with phase coherency yourself, but you can't tell without listening. You might even be more sensitive to correct phase than I am.

Phase shift is actually a differential time delay vs frequency. If the delays were long enough, they would be extremely obvious. Everyone could tell if the higher harmonic delays were smeared over 10 seconds in arrival after the arrival of the fundamental. The only question is how small of a delay is audible -- and what it sounds like.

If you can, try it!
 

Geronimo.USMC

Active Member
Joined
Jun 24, 2021
Messages
102
Likes
59
The balanced FLEX is not in stock. I ordered April 25 and still have not received a shipping email. They emailed me the wait is about 3 weeks to late May. Most likely the orders after May 3 will ship June.

The Balanced to RCA cable output will simply drop to 2V going into your sub. It's not a concern for most users.
Using an RCA source to balanced input on FLEX can drop SINAD almost 20 points. However, that may not be drastically detectable to the ear.

As far as cables, shielded 1/4" TRS to RCA should work.

I think you loose about 15dB of Sinad with the RCA to balanced input. Flex Sinad is about 115, 2x4hd isn't given as Sinad as far as I can see, but second and third harmonics alone are above -100, so yes, you'll be at least as good as the 2x4 HD. In any case, the SINAD of your subs will probably be much worse.

On the input side, you can use standard RCA to TS connectors. (DO NOT use these on the output). I'm using these:

If you want to make your own, pin of RCA to tip of TS, Ring of RCS to sleeve of TS.


I'm trying to buy and spend the minimum amount of cables/adapters, since RCA is temporary for me.

IMG_0188.jpg


I rummaged around my audio interconnects bin, and found this brand new Radioshack 1/4 TS mono plug to RCA adapter above. I should be able to use this with a nice shielded RCA to RCA cable for the FLEX's input from my AVR, correct?

Now for the Flex's 4 outputs, I already have so called "proper" XLR output to RCA adapters from Emotiva. I was thinking of buying 4 Balanced TRS to male XLR cables to connect to the adapters I have below.


Emotiva says this of their adapters:

The Proper Way To Connect a Balanced Output To an Unbalanced Input

When connecting a balanced output to an unbalanced input on modern audio gear, the proper way to do so is to connect the non-inverted output on the balanced output to the unbalanced input, and leave the inverted output unconnected. Many adapter cables, designed primarily for use with microphones, short the unused output pin to ground. While most equipment, including ours, will tolerate this without damage, it is not recommended, and may unnecessarily stress the output circuitry on your processor or preamp. This adapter was specifically designed to work optimally on modern balanced audio gear - and leaves the unused output line unconnected.

 
Last edited:

antcollinet

Master Contributor
Joined
Sep 4, 2021
Messages
7,408
Likes
12,291
Location
UK/Cheshire
I'm trying to buy and spend the minimum amount of cables/adapters, since RCA is temporary for me.

View attachment 204588

I rummaged around my audio interconnects bin, and found this brand new Radioshack 1/4 TS mono plug to RCA adapter above. I should be able to use this with a nice shielded RCA to RCA cable for the FLEX's input from my AVR, correct?

Now for the Flex's 4 outputs, I already have so called "proper" XLR output to RCA adapters from Emotiva. I was thinking of buying 4 Balanced TRS to male XLR cables to connect to the adapters I have below.


Emotiva says this of their adapters:

The Proper Way To Connect a Balanced Output To an Unbalanced Input

When connecting a balanced output to an unbalanced input on modern audio gear, the proper way to do so is to connect the non-inverted output on the balanced output to the unbalanced input, and leave the inverted output unconnected. Many adapter cables, designed primarily for use with microphones, short the unused output pin to ground. While most equipment, including ours, will tolerate this without damage, it is not recommended, and may unnecessarily stress the output circuitry on your processor or preamp. This adapter was specifically designed to work optimally on modern balanced audio gear - and leaves the unused output line unconnected.

That all sounds fine. As you say RCA to TS adapter on input only. The description from Emotiva is exactly correct about leaving the cold output un-connected. The suggestion from Mini DSP for outputs is to use a two conductor shielded connector, as shown here:

Since I made mine by butchering an existing RCA Cable, mine only have a single conductor (presumably same as the Emotiva) and it works fine.
 

MCH

Major Contributor
Joined
Apr 10, 2021
Messages
2,581
Likes
2,197
Yes, I agree. And I mentioned this in my first post to @SubsonicLurker. :)



Michael
I think many folks (including me when i discovered this product) were hoping for a newbie friendly solution that would allow you to integrate one or two subs pressing a couple of buttons and measuring a few sweeps. Without having to worry about phases, delays, crossover stuff, db per octave and all those things. (specially phase, i hate phase. When i read the word phase i know the instructions are not for me :D)

I guess the confusion results from being a 2 channel in 4 out. The flex would be the perfect cheap small candidate for this.

Most of you already have the knowledge and experience and it is perfectly fine as it is, but some others will continue dreaming of that small magic box.
 

Sluraad

Member
Joined
May 14, 2021
Messages
51
Likes
50
I think many folks (including me when i discovered this product) were hoping for a newbie friendly solution that would allow you to integrate one or two subs pressing a couple of buttons and measuring a few sweeps. Without having to worry about phases, delays, crossover stuff, db per octave and all those things. (specially phase, i hate phase. When i read the word phase i know the instructions are not for me :D)

I guess the confusion results from being a 2 channel in 4 out. The flex would be the perfect cheap small candidate for this.

Most of you already have the knowledge and experience and it is perfectly fine as it is, but some others will continue dreaming of that small magic box.
I would see these things as a fun learning point! It's not particularly challenging and with a tool like a minidsp flex + dirac you can get end game level performance out of both your subs and mains.
 
  • Like
Reactions: MCH

asrUser

Senior Member
Joined
Apr 11, 2022
Messages
354
Likes
216
I'm quite interested in a miniDSP Flex TRS balanced. There you got 4 TRS outputs numbered from 1 to 4. Is it possible to connect those outputs to two headphone amps simultaneously? So jacket 1 to left, 2 to right, 3 to left, 4 to right?

Flex%20Balanced9.jpg
 

antcollinet

Master Contributor
Joined
Sep 4, 2021
Messages
7,408
Likes
12,291
Location
UK/Cheshire
I'm quite interested in a miniDSP Flex TRS balanced. There you got 4 TRS outputs numbered from 1 to 4. Is it possible to connect those outputs to two headphone amps simultaneously? So jacket 1 to left, 2 to right, 3 to left, 4 to right?

Flex%20Balanced9.jpg
Yes - But you could also use a splitter cable to drive two amps from any analogue output.
 

Sluraad

Member
Joined
May 14, 2021
Messages
51
Likes
50
I'm quite interested in a miniDSP Flex TRS balanced. There you got 4 TRS outputs numbered from 1 to 4. Is it possible to connect those outputs to two headphone amps simultaneously? So jacket 1 to left, 2 to right, 3 to left, 4 to right?

Flex%20Balanced9.jpg
Yeah, each channel is mono and with the mixer you can select whichever output receive whichever input. Capabilities on the mixer side are the same in non-Dirac mode.
1651781888730.png
 

asrUser

Senior Member
Joined
Apr 11, 2022
Messages
354
Likes
216
Yes - But you could also use a splitter cable to drive two amps from any analogue output.
Oh right. I haven't thought about that. I could buy a cheaper DAC, but then it's annoying to use an Equalizer on every device/computer.

You can save up to 4 Eq profiles on the Minidsp, right? Is it like PEQ with 10 bands? Then if I use Dirac Live on some computer, does it need to run 24/7?
 

Sluraad

Member
Joined
May 14, 2021
Messages
51
Likes
50
Oh right. I haven't thought about that. I could buy a cheaper DAC, but then it's annoying to use an Equalizer on every device/computer.

You can save up to 4 Eq profiles on the Minidsp, right? Is it like PEQ with 10 bands? Then if I use Dirac Live on some computer, does it need to run 24/7?
Correct, 4 different profiles of completely separate configurations. Thats matrix assignments, dirac calibrations, per output EQ/delay/crossover/etc settings, the whole thing.

In Dirac mode you have 10 band PEQ bank per output, along with gain, delay, crossover, etc
In regular mode you have an additional 10 band PEQ bank per input (2 more) along with per output FIR.

Dirac has two components, there is the measurement & filter construction that is done off of the PC, and then the processor that runs on the device. You run the Dirac component on your pc during setup and then push the filters to the device and then you don't need anything on the PC anymore.
You can buy a software Dirac Processor for your desktop, which does need to run all the time but the processor part is pretty low overhead. It also has a free 2 week trial which can be a fun way for you to dip your toes into the world of Dirac.
 

antcollinet

Master Contributor
Joined
Sep 4, 2021
Messages
7,408
Likes
12,291
Location
UK/Cheshire
Yes, if you want separate eq profiles for the different devices you are connecting, then the flex is good for that.
 

poxymoron

Senior Member
Joined
May 7, 2021
Messages
459
Likes
588
Flex Digital arrived yesterday, haven't had a chance to use it yet. Bit disappointed to see one of the 4 rubber feet is missing.
 

aikofan

Member
Joined
Apr 13, 2022
Messages
87
Likes
88
Location
US
Maybe I’m dense, but the plugin UI (standard, non-Dirac) really needs to label the box where you enter the delay. The label could be off to the left of the columns. And the units (ms) should be displayed inside the box too. These are just basic principles of design and usability.
 

Geronimo.USMC

Active Member
Joined
Jun 24, 2021
Messages
102
Likes
59
Flex Digital arrived yesterday, haven't had a chance to use it yet. Bit disappointed to see one of the 4 rubber feet is missing.
Pretty sure they could send you one free of charge, but yeah I definitely feel you.
 

Sluraad

Member
Joined
May 14, 2021
Messages
51
Likes
50

antcollinet

Master Contributor
Joined
Sep 4, 2021
Messages
7,408
Likes
12,291
Location
UK/Cheshire
Flex Digital arrived yesterday, haven't had a chance to use it yet. Bit disappointed to see one of the 4 rubber feet is missing.
Have a look in the packaging - especially the plastic bags. I found two of mine stuck to bits of plastic packaging.

Yes though : disappointing they seem to fall off in transit.
 
Top Bottom