Minidsp Flex Review (Audio DSP)

[staticV3]

Whats the DAC chip used in MiniDSP Flex?
AKM, ESS.. ?

Flex: 2x TI PCM1795
Flex Eight: 2x ESS ES9080Q

 

[MCH]

Flex Eight: 2x ESS ES9080Q

how could i have missed it, now i have a benchmark for my diy project. They bested the datasheet SINAD by 1 dB, damm it! i need a teardown asap

btw, once they use 2x they could have made it balanced...

 

[mglobe]

DSP processor is stated, but DAC chip isnt.
Hopefully, no high pitch noise from unit while operating idle.
Thx

There's non noise when in standby on my Flex

 

[ItSoundsNatural01]

Does it make sense to use this for a Hi-Res streaming solution, because its bandwidth is limited to an Internal sample rate of 96kHz, whereas a Topping E50 like unit is capable of sample rates like 44.1 KHz- 768 KHz? Any idea about this limitation?

 

[TheBatsEar]

Does it make sense to use this for a Hi-Res streaming solution, because its bandwidth is limited to an Internal sample rate of 96kHz, whereas a Topping E50 like unit is capable of sample rates like 44.1 KHz- 768 KHz? Any idea about this limitation?

Higher sampling rates than on CD (44.1kHz) are great in the studio, if you want to mix instruments to a track and make sure artifacts are beyond hearing. At home, as a consumer, it makes no difference. They probably have chosen 96kHz because the DSP is powerful enough to do it. Once you install Dirac, it drops to 48kHz, likely because Dirac needs more computing power. Higher incoming sampling rates get downsampled.

This is a high level overview about how 44.1kHz and 48kHz came to be:

44,100 Hz - Wikipedia

en.wikipedia.org

And this is why, for consumption of audio, we don't need more:

Nyquist–Shannon sampling theorem - Wikipedia

en.wikipedia.org

As for buying or streaming songs that have far more than 44.1kHz or 48kHz, it's just a higher number and higher numbers are better for sales, but there is no benefit in terms of fidelity.

 

[mike70]

Higher sampling rates than on CD (44.1kHz) are great in the studio, if you want to mix instruments to a track and make sure artifacts are beyond hearing. At home, as a consumer, it makes no difference. They probably have chosen 96kHz because the DSP is powerful enough to do it. Once you install Dirac, it drops to 48kHz, likely because Dirac needs more computing power. Higher incoming sampling rates get downsampled.

This is a high level overview about how 44.1kHz and 48kHz came to be:

44,100 Hz - Wikipedia

en.wikipedia.org

And this is why, for consumption of audio, we don't need more:

Nyquist–Shannon sampling theorem - Wikipedia

en.wikipedia.org

As for buying or streaming songs that have far more than 44.1kHz or 48kHz, it's just a higher number and higher numbers are better for sales, but there is no benefit in terms of fidelity.

That's very simple to confirm ... Take a 192 khz recording and make 96khz and 44.1 versions.
Then hear it with an abx software and decent phones / phone amplifier.

If 192 is really better, you'll identify that track more than 7 of 10 times.

Done.

(I failed with total success )

 

[MCH]

And this is why, for consumption of audio, we don't need more:

Nyquist–Shannon sampling theorem - Wikipedia

en.wikipedia.org

The way I understand it (that could be wrong, I am no expert), what it tells you is the minimum you need, does not tell anything about if or why you need more or not.
(It is not that I think that more is better, but just to clarify)
edit: well yes, from the encoding point of view it also tells you that it is enough.

 

[Purité Audio]

There is an interesting article somewhere, wherein Mark Waldrep , AIX records and a huge proponent of ‘hi-res’ conducted some unsighted tests of his own material and he could not detect any difference in sound quality.
If it’s any consolation many years ago when I bought my first hi-res reference recording I expected ( because I was technically ignorant) that there would be greater resolution, sadly not!
Of course the actual quality of the recording/mastering plays a huge part.
Keith

 

[mike70]

The way I understand it (that could be wrong, I am no expert), what it tells you is the minimum you need, does not tell anything about if or why you need more or not.
(It is not that I think that more is better, but just to clarify)

But ... If that minimum reconstructs the original signal, then, using more ... what it will achieve? more signal? It doesn't work in that way.
It's only inaudible information.

(Or other scam to take your money)

 

[onlyoneme]

- [All Models] Noise performance improvement on filter response for IIR filters
- [All Models] Volume control moved before ASRC in DSP chain to help against potential distortion caused by intersample peaks

Device Console ChangeLog

The following software changeLog summarizes the changes made to the miniDSP Device Console (DC). As a reminder, please note that XMOS and DSP firmware are packaged together inside the miniDSP DC and that you may be requested to upgrade your firmware in order to use the latest version of miniDC...

www.minidsp.com

Finally

 

[Ron Texas]

There is an interesting article somewhere, wherein Mark Waldrep , AIX records and a huge proponent of ‘hi-res’ conducted some unsighted tests of his own material and he could detect any difference in sound quality.
If it’s any consolation many years ago when I bought my first hi-res reference recording I expected ( because I was technically ignorant) that there would be greater resolution, sadly not!
Of course the actual quality of the recording/mastering plays a huge part.
Keith

Does this need an edit?

 

[esl 63]

Want to see distortion measurements when MiniDSP flex is running some processing, phase correction, EQ. What is the difference with the current bypass measurement? Is there a difference?
When will MiniDSP FLEX Eight be tested?
And is SHD a better performer than FLEX Eight? During processing.... SHD measures better than flex in the multitone testing even though it is much older.

 

[Purité Audio]

Does this need an edit?

Yes and done.
Best
Keith

 

[mike70]

- [All Models] Noise performance improvement on filter response for IIR filters
- [All Models] Volume control moved before ASRC in DSP chain to help against potential distortion caused by intersample peaks

Device Console ChangeLog

The following software changeLog summarizes the changes made to the miniDSP Device Console (DC). As a reminder, please note that XMOS and DSP firmware are packaged together inside the miniDSP DC and that you may be requested to upgrade your firmware in order to use the latest version of miniDC...

www.minidsp.com

Finally

who's the first brave to make the upgrade?

 

[AaronJ]

After having lived with the Flex for some time I am pretty much convinced it is the biggest bargain in all of hifi, my most important component, and I don’t even use any of the digital inputs. I have it in the pre-out/main-in loop of a Sansui AU-5500, Sansui volume pot at 12 o’clock to prevent the Flex from clipping its analog input, an inexpensive volume pot between the Flex and main-in to protect the amp and speakers in case something goes haywire. And I use the Flex as the volume control. The 2.2 system is integrated perfectly and I have never been happier. I’d recommend it highly even at 5x the price.

 

[Robocop]

who's the first brave to make the upgrade?

You don't have to be brave it happens when you next open the console.

 

[Robocop]

I updated my minidsp flex balanced and all works great out of the box. It is slightly faster moving between Preset settings and connecting to PC console. No PEQ settings were changed or Presets in this process. It is two separate update processes which I conducted one after the other. The flex was rebooted twice in this process. I did turn off all amplifiers down stream before the update.

I thoroughly recommend updating your minidsp devices.

 

[antcollinet]

I'll wait a month to see what happens in the forums.

 

[Robocop]

I guess ant you will need Amir to put this on the test bench before trying. Bearing in mind we will not make a call on what it sounds like if anything like this makes a difference.

 

[onlyoneme]

I updated my minidsp flex balanced and all works great out of the box. It is slightly faster moving between Preset settings and connecting to PC console. No PEQ settings were changed or Presets in this process. It is two separate update processes which I conducted one after the other. The flex was rebooted twice in this process. I did turn off all amplifiers down stream before the update.

I thoroughly recommend updating your minidsp devices.

No issue on my side either. Two separate updates for XMOS and DSP. And Flex volume level affects the headroom level before ASRC as announced, no more intersample peak clipping if the volume is properly attenuated.