Trying the rePhase thread for some ideas. Its strange that my Hypex DLCP setting, replicated in rePhase just using 4 low Q filters sound so different. Beyond different - something is not right.
I will try the biquad/IIR solution with the DAM instead. Same problem with scaling of coefficients and I get "owerflow" error in MKROM. So I need to scale them somehow but don't know if I can do simple multiplication over all biquads... Lacking signal theory knowledge...
And now I realise that a and b isn't even necessary used in the same way :-/
The 4 biquads used in DLCP:
1: -14,4dB at 210 hz Q 0,65
2: 14,4dB at 23 hz Q 0,55
3: -3,8 dB at 3000 hz Q 0,5
4: 2,2dB at 17000 hz Q 0,5
B1:
a0 = 0.9172257802529049
a1 = -1.792384354953564
a2 = 0.8759611472790664
b1 = -1.792384354953564
b2 = 0.7931869275319711
B2:
a0 = 1.0143276105115997
a1 = -1.9937904939813273
a2 = 0.9794745396614566
b1 = -1.9937904939813273
b2 = 0.9938021501730564
B3:
a0 = 0.8921442650005743
a1 = -1.147917583936177
a2 = 0.3692549599892966
b1 = -1.147917583936177
b2 = 0.26139922498987106
B4:
a0 = 1.1028532631739936
a1 = 0.9067249064166488
a2 = 0.10268425080507655
b1 = 0.9067249064166488
b2 = 0.20553751397907014
MROM says overflow on e.g. B1 a0. Shuould not be more than 0,125... but wait there is a scaling parameter in the DAM filter defintion file...
//
I will try the biquad/IIR solution with the DAM instead. Same problem with scaling of coefficients and I get "owerflow" error in MKROM. So I need to scale them somehow but don't know if I can do simple multiplication over all biquads... Lacking signal theory knowledge...
And now I realise that a and b isn't even necessary used in the same way :-/
The 4 biquads used in DLCP:
1: -14,4dB at 210 hz Q 0,65
2: 14,4dB at 23 hz Q 0,55
3: -3,8 dB at 3000 hz Q 0,5
4: 2,2dB at 17000 hz Q 0,5
B1:
a0 = 0.9172257802529049
a1 = -1.792384354953564
a2 = 0.8759611472790664
b1 = -1.792384354953564
b2 = 0.7931869275319711
B2:
a0 = 1.0143276105115997
a1 = -1.9937904939813273
a2 = 0.9794745396614566
b1 = -1.9937904939813273
b2 = 0.9938021501730564
B3:
a0 = 0.8921442650005743
a1 = -1.147917583936177
a2 = 0.3692549599892966
b1 = -1.147917583936177
b2 = 0.26139922498987106
B4:
a0 = 1.1028532631739936
a1 = 0.9067249064166488
a2 = 0.10268425080507655
b1 = 0.9067249064166488
b2 = 0.20553751397907014
MROM says overflow on e.g. B1 a0. Shuould not be more than 0,125... but wait there is a scaling parameter in the DAM filter defintion file...
//
Last edited:
Some success!
After swopping the a and b in the biquads the system now sound at least like music
-- IIR for FR10 Line Source
dam1021,44100,8,16,5,0.8
16 IIR biquad 1
b0 = 0.9172257802529049
b1 = -1.792384354953564
b2 = 0.8759611472790664
a1 = -1.792384354953564
a2 = 0.7931869275319711
dam1021,44100,8,17,5,0.8
17 IIR biquad 2
b0 = 1.0143276105115997
b1 = -1.9937904939813273
b2 = 0.9794745396614566
a1 = -1.9937904939813273
a2 = 0.9938021501730564
dam1021,44100,8,18,5,0.8
18 IIR biquad 3
b0 = 0.8921442650005743
b1 = -1.147917583936177
b2 = 0.3692549599892966
a1 = -1.147917583936177
a2 = 0.26139922498987106
dam1021,44100,8,19,5,0.8
19 IIR biquad 4
b0 = 1.1028532631739936
b1 = 0.9067249064166488
b2 = 0.10268425080507655
a1 = 0.9067249064166488
a2 = 0.20553751397907014
The biquads was calculated from this page:
Biquad calculator v2 | EarLevel Engineering
SQ? Hmm, but still, at least in the ballpark!
MROM overflow fixed by setting the last term here to "0.8".
dam1021,44100,8,16,5,0.8
//
After swopping the a and b in the biquads the system now sound at least like music
-- IIR for FR10 Line Source
dam1021,44100,8,16,5,0.8
16 IIR biquad 1
b0 = 0.9172257802529049
b1 = -1.792384354953564
b2 = 0.8759611472790664
a1 = -1.792384354953564
a2 = 0.7931869275319711
dam1021,44100,8,17,5,0.8
17 IIR biquad 2
b0 = 1.0143276105115997
b1 = -1.9937904939813273
b2 = 0.9794745396614566
a1 = -1.9937904939813273
a2 = 0.9938021501730564
dam1021,44100,8,18,5,0.8
18 IIR biquad 3
b0 = 0.8921442650005743
b1 = -1.147917583936177
b2 = 0.3692549599892966
a1 = -1.147917583936177
a2 = 0.26139922498987106
dam1021,44100,8,19,5,0.8
19 IIR biquad 4
b0 = 1.1028532631739936
b1 = 0.9067249064166488
b2 = 0.10268425080507655
a1 = 0.9067249064166488
a2 = 0.20553751397907014
The biquads was calculated from this page:
Biquad calculator v2 | EarLevel Engineering
SQ? Hmm, but still, at least in the ballpark!
MROM overflow fixed by setting the last term here to "0.8".
dam1021,44100,8,16,5,0.8
//
Success :-D
I succeeded to implement a compensation filter for the line arrays along with a DA brickwall filter, placed in the first FIR stage on a DAM DAC using rePhase to generate the filter. Hurray And it sound really promising I think but still a few pitfalls there I'm sure. Due to lack of taps the desired EQ can not be acomplished but it seems one can fiddle with to approach what is desired.
//
I succeeded to implement a compensation filter for the line arrays along with a DA brickwall filter, placed in the first FIR stage on a DAM DAC using rePhase to generate the filter. Hurray
And it sound really promising I think but still a few pitfalls there I'm sure. Due to lack of taps the desired EQ can not be acomplished but it seems one can fiddle with to approach what is desired.//
Attachments
Yes I'm quite happy with it. This is the first rev1 1021, with a few "std" mods. I also run 2 pcs 1121 which I prefer as digi-mono-power-blocks with 5m opto to each and very short analog path. The 1121 seem to lag wrt. FW upgrade which enable 4k taps. Hopefully Sören delivers eventually.
Dead silent. The 1021 had pop problems but was dealt with I think in latr revs. The 1121 I flick the switch to and there is nothing. This hook up to Hypex nCore/SMPS600.
No promises
I think dare recommend it as long as you are not deep into tubes and 2:nd order dist :-D
//
Dead silent. The 1021 had pop problems but was dealt with I think in latr revs. The 1121 I flick the switch to and there is nothing. This hook up to Hypex nCore/SMPS600.
No promises
I think dare recommend it as long as you are not deep into tubes and 2:nd order dist :-D
//
I choose the 1121 due to the fact that it does re-clocking on the backside of the FPGA. The clocking scheme is apparently why the new FW lingers - it cant be reused straight off the other products.
Compared to my 1021I felt that had distinct level of "easyness", yet clarity, over it that my 1021 didn't have. Now, again they are the first revision and I haven't gone bananas modding them. Dont know about the current crop 1021. Now ther are new boards coming up to watch for.
//
Member pos wrote:
"4000 taps at 352.8k rate is equivalent to 500 taps at 44.1k. That is more than enough for an antialiasing filter, but probably not enough for real world EQ. In comparison an openDRC does 6144 at 48kHz."
I think I have to compare letting the DAM do 4 IIR filters instead.
//
Last edited:
2 things that hold it back, first: there is a maximum number for boost settings (yours is 9 dB, the gap is bigger) and second: in general it usually does not boost after roll of at low frequencies. Not all users of REW will have multiple drivers that can take the abuse in the low end.
I'd advise to bring the target a bit lower (half way between the max amount of boost and cut needed) and use a few manual PEQ's to bring up the bottom end. Don't go completely wild on the boost part. It's quite easy to do better with manual EQ once you get the hang of it. You can use the split window to see the impulse or waterfall in the lower window to see what you're doing isn't creating new problems.
I'd advise to bring the target a bit lower (half way between the max amount of boost and cut needed) and use a few manual PEQ's to bring up the bottom end. Don't go completely wild on the boost part. It's quite easy to do better with manual EQ once you get the hang of it. You can use the split window to see the impulse or waterfall in the lower window to see what you're doing isn't creating new problems.
Last edited:
I have heard it rumble so I think it will be enough
Some DAC measurement here:
Soekris DAM1021 R2R DAC Measurements | Audio Science Review (ASR) Forum
//
Some DAC measurement here:
Soekris DAM1021 R2R DAC Measurements | Audio Science Review (ASR) Forum
//
Very much so - thanks for asking!
Even more after replacing that bass peak above and introduced an LT instead. That made the bass come home. I was also gentle with that so need som more tweaking. Also it seems that tighter span for REW EQ pays off so I will see what I can fit into 13 BQs. Reinstated one the absortion-worms on the side walls. This widens the soundscape in what seems a natural way but also seem to eliminate some kind of "fluttering" sound - is this loobing?
I have now also been able to move the speakers all the way into the corners. Well, almost, something is obstructing the last say 5mm. Probably this old house is not square enough so this will be it. I reused som of the cuts from the cabinets to fill that little gap on the front side wall - I will paint them eventually
I also now use my B&O ASP1000 - wohooo ;-) the solidity, the power. OMG :-D and still some tuning on EQ to be done.
Project beyond expectations. I will do new measurements and present them and I hope you all have yet the will to share more of your valuable advice and tips.
//
Even more after replacing that bass peak above and introduced an LT instead. That made the bass come home. I was also gentle with that so need som more tweaking. Also it seems that tighter span for REW EQ pays off so I will see what I can fit into 13 BQs. Reinstated one the absortion-worms on the side walls. This widens the soundscape in what seems a natural way but also seem to eliminate some kind of "fluttering" sound - is this loobing?
I have now also been able to move the speakers all the way into the corners. Well, almost, something is obstructing the last say 5mm. Probably this old house is not square enough so this will be it. I reused som of the cuts from the cabinets to fill that little gap on the front side wall - I will paint them eventually
I also now use my B&O ASP1000 - wohooo ;-) the solidity, the power. OMG :-D and still some tuning on EQ to be done.
Project beyond expectations. I will do new measurements and present them and I hope you all have yet the will to share more of your valuable advice and tips.
//
Attachments
Last edited: