Yes, it seems that one need to do 2 or 3 iterations to really get it right. It means that the last adjustments is now your "system," and you go in there, measure it, and create a new set of filters which you add to the previous system, I.e. you add more filters.
//
//
I think i am beginning to understand where i don't understand, i have only 10 bi quads available, these are all used in the first iteration so i dont believe i can simply add another layer, can I?
Aha 1p PEQs... - I understand. I think your speakers deserve a much more refined DSP platsform than that.
//
//
Maybe someone out there knows how to encourage the miniDSP kit to iterate?
I do have PEQ on both input and output channels but at the moment i don't wish to use the input ones as i will use the second set of outputs to do the house streaming, but i might not!
I do have PEQ on both input and output channels but at the moment i don't wish to use the input ones as i will use the second set of outputs to do the house streaming, but i might not!
Assuming the FLEX allows real time EQ adjustments in normal mode, an alternative method to REW auto-EQ and the bi-quad batch generation process,
is to dial in EQs in real time.
Using real time dual channel measurement software, like donate-ware Open Sound Meter https://opensoundmeter.com/en/
you can dial in one EQ at a time, as pink noise is played continuously (and doesn't have to be loud at all).
I typically start with what needs to be done most, EQ that, and go to the next pressing need, and the next..........etc
I remember my minidsp opendrcs or their icepower amps, having real time capability (not sure which).
Apologize if the FLEX doesn't and this post of no value....couldn't tell from a look at the manual.
is to dial in EQs in real time.
Using real time dual channel measurement software, like donate-ware Open Sound Meter https://opensoundmeter.com/en/
you can dial in one EQ at a time, as pink noise is played continuously (and doesn't have to be loud at all).
I typically start with what needs to be done most, EQ that, and go to the next pressing need, and the next..........etc
I remember my minidsp opendrcs or their icepower amps, having real time capability (not sure which).
Apologize if the FLEX doesn't and this post of no value....couldn't tell from a look at the manual.
Does anyone know please if what Mark suggests above is possible with the Flex?
Another question, does anyone have thoughts about the optimum aim to the seating positions?
I have tried crossing just in front of what could be regarded as the "main" seat. One advantage i feel these speakers have is they behave quite well for lots of different positions, i read that they would and mine do that.
However the path length from each speaker is significantly different, any knowledgeable feedback on that subject would be interesting!
Another question, does anyone have thoughts about the optimum aim to the seating positions?
I have tried crossing just in front of what could be regarded as the "main" seat. One advantage i feel these speakers have is they behave quite well for lots of different positions, i read that they would and mine do that.
However the path length from each speaker is significantly different, any knowledgeable feedback on that subject would be interesting!
I have no clue about anything miniDSP. But crossing slightly in front of the listener would have advantages for the seating next to it on both sides.
I haven't done it myself because I didn't quite warm up to that look, these arrays are rather imposing in our room and it just looked odd and out of place.
I cross them behind the listener instead, making the exact sweet spot about 10 degree off axis of both speakers. That's what I used in the simulations too.
I tried just about anything and ended up with this angle as the most pleasant sounding to my ears. Which no doubt had something to do with off axis reflections too.
In other words: Experiment with it and find your preferred setup.
I still think or believe crossing at a similar (about 10 degree) angle in front of the listener could be the best solution, but I'd have to cover the baffle to make it look a bit better or even normal. Crossing in front of the listener should give the positions next to the central sweet spot a better intensity trading to preserve Stereo effects. Then again, it could also bring the nearest speaker too far off axis, as these arrays are full range speakers that simply don't have the directivity of a tweeter. My compromise works well for us, even in movies without a center channel.
I haven't done it myself because I didn't quite warm up to that look, these arrays are rather imposing in our room and it just looked odd and out of place.
I cross them behind the listener instead, making the exact sweet spot about 10 degree off axis of both speakers. That's what I used in the simulations too.
I tried just about anything and ended up with this angle as the most pleasant sounding to my ears. Which no doubt had something to do with off axis reflections too.
In other words: Experiment with it and find your preferred setup.
I still think or believe crossing at a similar (about 10 degree) angle in front of the listener could be the best solution, but I'd have to cover the baffle to make it look a bit better or even normal. Crossing in front of the listener should give the positions next to the central sweet spot a better intensity trading to preserve Stereo effects. Then again, it could also bring the nearest speaker too far off axis, as these arrays are full range speakers that simply don't have the directivity of a tweeter. My compromise works well for us, even in movies without a center channel.
Last edited:
I remain in the learning period for sure, many thanks for the various feedbacks.
I will continue to try to understand!
Meanwhile I learnt to input a couple of shelf filters by hand to bring the high and low frequencies near the target curve before asking REW to match the target line and got the attached measured channel match after that. They match the target curve well.
I will continue to try to understand!
Meanwhile I learnt to input a couple of shelf filters by hand to bring the high and low frequencies near the target curve before asking REW to match the target line and got the attached measured channel match after that. They match the target curve well.
Attachments
Thanks Karsten, i hear it being not harsh/tiring.
One learns with use, and i have a long way to go!
Inevitably other things change at the same time because to achieve that filter input using REW and Biquads and there are only 10 one has to free up some resource, so i find it very difficult to make and listen to and measure only one change at a time. one way to free resource is to ask for a looser match to the target (2Db instead of 1)
That may be a limitation of such user friendly approaches.
So far i have found it very convenient to input the filters this way and the Flex also provides good input selection and controls.
Now, it also allows/provides FIR filtering, see a page top of the manual attached, although i must admit i have never seen it yet, that may be because i already used PEQ and that one can only have one or the other, not both?
Is this a more powerful/refined approach that TNT suggested earlier? - a whole new learning curve perhaps, but we have storms this weekend!
By the way, concerning bass, back in the discussions before build various respondents have questioned the ability to produce bass, in my opinion they produce lots of bass, including as measured down at 20Hz although i don't know what recordings include that low, but areas like kick drums are there and impactful/dynamic.
One learns with use, and i have a long way to go!
Inevitably other things change at the same time because to achieve that filter input using REW and Biquads and there are only 10 one has to free up some resource, so i find it very difficult to make and listen to and measure only one change at a time. one way to free resource is to ask for a looser match to the target (2Db instead of 1)
That may be a limitation of such user friendly approaches.
So far i have found it very convenient to input the filters this way and the Flex also provides good input selection and controls.
Now, it also allows/provides FIR filtering, see a page top of the manual attached, although i must admit i have never seen it yet, that may be because i already used PEQ and that one can only have one or the other, not both?
Is this a more powerful/refined approach that TNT suggested earlier? - a whole new learning curve perhaps, but we have storms this weekend!
By the way, concerning bass, back in the discussions before build various respondents have questioned the ability to produce bass, in my opinion they produce lots of bass, including as measured down at 20Hz although i don't know what recordings include that low, but areas like kick drums are there and impactful/dynamic.
Attachments
I have now spent a bit of time listening to various sources and music after getting them basically set up.
1. In my opinion these are remarkably competent and a well worthwhile project
2. They sit within 25 cm (10") of the corners as required
3. They reproduce clear detail, separating instruments and voices clearly, and i have heard some details and differences/separation better than on other (good) speakers i have elsewhere
4. They reproduce bass all the way down to 20Hz if you set it up that way, and that is clear bass with no boominess which every other speaker in a corner seems to produce and the bass has good tunefulness. I shall experiment further with how low i set the DSP to go, maybe 30Hz, as there isn't much material, if any down there.
5. It is noticeable that the whole experience is tolerant of where you are in the room, no one-seat only effect.
6. The set up with the Flex for source selection, volume control and DSP is very user friendly
7. I haven't yet completed the front baffle corner round off and sock which will imho make them less visually noticeable, and may yet take some time
8. I don't notice big differences with where they are aimed/cross paths in relation to the listening position, I have tried aiming at the prime seat and in front of it
9. I am sure also there is lots to come as i learn about DSP, i would like to learn more about this now, with frequency correction and possibly phase, with different methods if i can work those out
10. Its been great to have the support and input of folks on this thread and those who built previously and put their ideas out there, and i look forward to continuing
M
1. In my opinion these are remarkably competent and a well worthwhile project
2. They sit within 25 cm (10") of the corners as required
3. They reproduce clear detail, separating instruments and voices clearly, and i have heard some details and differences/separation better than on other (good) speakers i have elsewhere
4. They reproduce bass all the way down to 20Hz if you set it up that way, and that is clear bass with no boominess which every other speaker in a corner seems to produce and the bass has good tunefulness. I shall experiment further with how low i set the DSP to go, maybe 30Hz, as there isn't much material, if any down there.
5. It is noticeable that the whole experience is tolerant of where you are in the room, no one-seat only effect.
6. The set up with the Flex for source selection, volume control and DSP is very user friendly
7. I haven't yet completed the front baffle corner round off and sock which will imho make them less visually noticeable, and may yet take some time
8. I don't notice big differences with where they are aimed/cross paths in relation to the listening position, I have tried aiming at the prime seat and in front of it
9. I am sure also there is lots to come as i learn about DSP, i would like to learn more about this now, with frequency correction and possibly phase, with different methods if i can work those out
10. Its been great to have the support and input of folks on this thread and those who built previously and put their ideas out there, and i look forward to continuing
M
I'm glad you've got these promising results! Take it slow, but most of all, enjoy the journey. 🙂
I just learnt something else. Earlier i wrote about and folk answered about iterative filtering and it was questioned about what the flex can do.
The Flex is not just DSP but can be a crossover etc and to do that is set up to have inputs and outputs, and one chooses a route in between.
I am basically using just two in and two out, being left and right channels.
Previously i was only using the Output filter capability. i now find that if i use the input side as well i can use more filter power/capability.
The reason i did that was that i wish to send another pair of outputs to other rooms which would need the unfiltered signals. now i realsie i can send the unfiltered signal even if i have processed it on the way in by inverting that on the way out. Consequently i can use both sets on the arrays
I believe Flex only allows use of FIR on the output side, so currently my thought is to use IIR on the inputs and FIR on the output.
Separately i will look at FIR and phase.
M
The Flex is not just DSP but can be a crossover etc and to do that is set up to have inputs and outputs, and one chooses a route in between.
I am basically using just two in and two out, being left and right channels.
Previously i was only using the Output filter capability. i now find that if i use the input side as well i can use more filter power/capability.
The reason i did that was that i wish to send another pair of outputs to other rooms which would need the unfiltered signals. now i realsie i can send the unfiltered signal even if i have processed it on the way in by inverting that on the way out. Consequently i can use both sets on the arrays
I believe Flex only allows use of FIR on the output side, so currently my thought is to use IIR on the inputs and FIR on the output.
Separately i will look at FIR and phase.
M
Hi Folks, i find i have mistakenly been posting some material that belongs on here on the Two Towers thread, sorry about that, is there an easy way to copy post my last few posts on there over here as well please?
M
M