TOPIC:

Hi,

I just bought a Flex Eight for a DIY project with linear phase/brickwall crossovers, but while I was reading the documentation after receiving it I sadly realized that the FIR capabilities of the unit are only available on the inputs...
Maybe the description of the product can be modified because it is misleading:
"As always, a full suite of miniDSP’s tried and proven audio processing functionality is included: parametric EQ, IIR & FIR filters, crossovers, advanced biquad programming and delay on each output channel."

The Flex Eight is sadly useless for me without FIR on output channels.

And miniDSP refuses all returns... I need to replace it with two 2x4HD but the Flex Eight consumed all of the DSP budget.

If FIR can only be one place, it's preferable on the inputs.

Linear-phase corrections (which is what I assume you're interested in) is not necessary on the output channels since those can be equalized to minimum-phase using the tradition PEQ correction capability.  The overall result can then be adjusted to linear-phase using the global (input) path.

This topic (FIR location) has been discussed on the miniDSP forum numerous times already.

Dave.

Hi,

You missed the point: the product description is wrong, so it concerns the Flex Eight.

BTW, I am not interested in linear FIR correction, but in brickwall linear phase FIR filters for crossovers. That is not the same thing, as linearizing phase after the crossover do not correct interferences between the drivers. Brickwall filters do.

I purchased the Flex Eight for its given output FIR capabilities (according to the description). I would really love to find a solution.
 

A simple look at the user manual would have shown you the pertinent configuration.

docs.minidsp.com/product-manuals/flex-ei...gnal-flow/index.html

Dave.

Yes, thank you.

but the product description is wrong and led me to buy it.

Also the documentation is not listed on the product page, and I didn’t find it at first.

Is there any way to solve this ? I am stuck with a unit that I can’t use, and it cost me a lot of money. I don’t want to be rude, but since the product description is wrong, it is clear to me that this is the fault of miniDSP.

Thank you for your understanding.
 

All you had to do was scroll down. 
www.minidsp.com/products/minidsp-in-a-box/minidsp-flex-eight

How much $$$ do you want for it????

Dave.

Why are you that condescending and dishonest ?

The description is clearly wrong. My father, who is a sound engineer was also misled by it and assumed that it worked the same way as his 2x4HD.

All I ask is an exchange of products or a return and refund.

aldem wrote:

Why are you that condescending and dishonest ?

The description is clearly wrong. My father, who is a sound engineer was also misled by it and assumed that it worked the same way as his 2x4HD.

All I ask is an exchange of products or a return and refund.

You need to contact miniDSP support if you want to make requests like that. This is a community forum as noted in bold at the top of the page.

It's helpful to other potential customers that you've highlighted this possible source of confusion though.

Besides offering sharper 'brick wall' phase coherent crossover filters, FIR filters on the outputs also allow for independent phase compensation of possible phase shifts in the low and high frequency divers themselves, most likely near and at the crossover frequency.  Since the 8 channel Flex device appears to use the same DSP processor as the 4 output channel Flex models, it would require the same processor to handle twice as many FIR filters if they were on the 8 outputs, which may not be practical in the current design.

MiniDSP's next generation products may incorporate a more powerful DSP that can handle more output FIR filters.  On the other hand, the consensus seems to be that phase shift is totally inaudible, rendering FIR filters irrelevant (a contention I don't share).  Let's see what features the next generation products have.  Hopefully they will retain and expand FIR capability.

If you are experimenting with brick wall FIR filters using rePhase, I suggest you try selecting 100 db/octave LR type filters.  There seems to be a slight difference in the sound, possibly due to round off errors in the FIR coefficients of either the brick or 100db/octave filters, or both.  I find that the 100 db/octave filters sound better, and the successive approximation math that rePhase does to calculate them is faster.

dreite wrote:

If FIR can only be one place, it's preferable on the inputs.

Linear-phase corrections (which is what I assume you're interested in) is not necessary on the output channels since those can be equalized to minimum-phase using the tradition PEQ correction capability.  The overall result can then be adjusted to linear-phase using the global (input) path.

This topic (FIR location) has been discussed on the miniDSP forum numerous times already.

Dave.

true , except the outputs gives a level of accuracy far more superior then doing it in the sums….


I would much rather use a pair of flex 4ch over a single flex8 

the issue with doing it in the sums is you have very little control over the speakers phase itself, so many many measurements and carefully tuning the crossovers so each driver is in proper phase (not polarity) maybe by stacking a few out of band APFs in the subsonics to align them….

with driver fir , none of that is needed 

 

And brick wall filters are not at all very good sounding unless used for a very specific purpose (beam forming) 

low phase angle crossovers almost always sound better in fir and iir. 

if it’s fast attenuation for a sub perhaps, an LR8 will do just about everything (with some weird off axis stuff goin on) 

Been there done that!!!  LR4 is as steep as I would ever go.  Whenever I hear FIR filters with brick wall in the same paragraph (besides this one of course 🤪) then I know it is almost always no experience fir users.


the flex in 2x4hd mode is amazing 

And brick wall filters are not at all very good sounding unless used for a very specific purpose (beam forming) 

low phase angle crossovers almost always sound better in fir and iir. 

if it’s fast attenuation for a sub perhaps, an LR8 will do just about everything (with some weird off axis stuff goin on) 

Been there done that!!!  LR4 is as steep as I would ever go.  Whenever I hear FIR filters with brick wall in the same paragraph (besides this one of course 🤪) then I know it is almost always no experience fir users.


the flex in 2x4hd mode is amazing 

Quote denver8me4dinner :
Been there done that!!!  LR4 is as steep as I would ever go.  Whenever I hear FIR filters with brick wall in the same paragraph (besides this one of course 🤪) then I know it is almost always no experience fir users.

Interesting to speculate why that may be the case.  Maybe math errors in the calculation of steeper slope FIR filters, possibly including the exact matching of crossover frequencies in the high and low pass FIR filters.  Or maybe some kind of audible effect of a rapid transition between drivers over a very narrow frequency range.

I'll experiment with FIR linear phase LR4 slope, even though my 5 foot tall ribbon tweeters will get more power and more motion below the crossover frequency.

It’s about the dispersion of the larger cone to smaller cone/horn tweet, or whatever.


that rapid of a change to a arguably completely different characteristic is quite obvious! 

there’s no point in crossing any speaker to try to get to the bleeding edge of power handing and flat power responses… besides , in fir on each output you can simply flatten the phase on all of the driver , so it becomes a non issue with crossover summation. It’s virtually perfect summing and even dispersion with lower order filters.

 

That is another discussion fellas.  (And it's been hashed around on this forum previously many times.)

What we have here is (apparently) a buyer who didn't perform even the most basic due diligence to determine if the product would meet his objective.
That said, I have my doubts whether he's even an actual owner, or if this is yet another anonymous troll operating on this forum.

Dave.