Thanks Bengt.
Was that with Bass/Treble enabled?
Is that with Bass/Treble enabled?
What bass clipping are you referring to? Bass/Treble unit?
Ok, the bass controll is now dissabled again. It apperantly went to "on" with the new software.
Cpu usage 32/37.
Edit: Bass clipping was on one of the leds and at the bass/treble monitor light
Edit2: cpu 42/42 with no channels dissabled.
Edit3: cpu 41/41 with sum/diff on input dissabled
30/37 with 1/2 and 9/10 dissabled as well.
Last edited:
Thanks Bengt for these details.
So you've gone from DSP load 42% (Core 0) and 41% (Core 1) down to 30% and 37%.
I'd say you could run this setup at 192 kHz if you wanted that.
Or at the same 96 kHz sampling rate, you've made enough CPU savings for 16 additional biquads (1 biquad is about 1% DSP load at 96 kHz on one core).
Good man! You've just convinced me to make more biquads available per channel.
I'd say 15 biquads is appropriate.
RE your other note: when you update the firmware, you loose all your settings. They are reverted to their defaults, and Bass/Treble is enabled by default. This explains the clipping you have noticed. Also, even if you disable all input channels, the Bass/Treble unit is still fed with the input signals (see the signal flow diagram in the manual).
So you've gone from DSP load 42% (Core 0) and 41% (Core 1) down to 30% and 37%.
I'd say you could run this setup at 192 kHz if you wanted that.
Or at the same 96 kHz sampling rate, you've made enough CPU savings for 16 additional biquads (1 biquad is about 1% DSP load at 96 kHz on one core).
Good man! You've just convinced me to make more biquads available per channel.
I'd say 15 biquads is appropriate.RE your other note: when you update the firmware, you loose all your settings. They are reverted to their defaults, and Bass/Treble is enabled by default. This explains the clipping you have noticed. Also, even if you disable all input channels, the Bass/Treble unit is still fed with the input signals (see the signal flow diagram in the manual).
Thanks Bengt for these details.
So you've gone from DSP load 42% (Core 0) and 41% (Core 1) down to 30% and 37%.
I'd say you could run this setup at 192 kHz if you wanted that.
Or at the same 96 kHz sampling rate, you've made enough CPU savings for 16 additional biquads (1 biquad is about 1% DSP load at 96 kHz on one core).
Good man! You've just convinced me to make more biquads available per channel.
RE your other note: when you update the firmware, you loose all your settings. They are reverted to their defaults, and Bass/Treble is enabled by default. This explains the clipping you have noticed. Also, even if you disable all input channels, the Bass/Treble unit is still fed with the input signals (see the signal flow diagram in the manual).
I have actually come to the conclusion that I like the 96khz sampling more than the 192, go figure
I will se if I can rearrange the channels to get a more even load. There will
be some addition of eq bands in the future so there is no real hurry though.
The linkwitz transform function would be nice
That would actually be more helpful than pairing 1-2..... etc
Some more reflexions..
I am trying the DSP now and there is sound comming out even thow I have removed the opamps... is there a feedback path around the opamp?
If I wanna use the volumecontrol with my own DAC should I remove the small e-lyts behind the opamps?
I am playing now with an external DAC and using the volumecontrol (digital) in my laptop,when I try to mute the channels on the monitorpage they wont work (that is when using only the digital out)
I am trying the DSP now and there is sound comming out even thow I have removed the opamps...
is there a feedback path around the opamp?If I wanna use the volumecontrol with my own DAC should I remove the small e-lyts behind the opamps?
I am playing now with an external DAC and using the volumecontrol (digital) in my laptop,when I try to mute the channels on the monitorpage they wont work (that is when using only the digital out)
I was actually thinking of making this 1-2/3-4 etc pairing available as well. So what, no need for it?
I am trying the DSP now and there is sound comming out even thow I have removed the opamps...
If I wanna use the volumecontrol with my own DAC should I remove the small e-lyts behind the opamps?
I am playing now with an external DAC and using the volumecontrol (digital) in my laptop,when I try to mute the channels on the monitorpage they wont work (that is when using only the digital out)
Yes, it's normal you still get a signal out if you just remove the op-amp. The best way to re-inject the signal of your DAC into the volume control chip is by removing the coupling caps in front of the op-amps. This results in disconnecting the onboard DAC from the volume chip.
If you remove also the op-amp, then you have easy access to the ground plane for the signal return path.
On another subject.
I have a friend that is putting together a 5-way system (dipole)
and he wants to be able to use both step filters,compensation in
almost all registers and a high samplingfrequency. If he would end
up hitting the cpu's maximum capasity before the systems done,
would it be possible to run two "Najda's" in parralel from one usb
reciver over the I2S input using one for left and the other for right channel ?
Next subject.
If there is more then enough cpu capasity available, is there any
possibility to use that for "double presicion mode" ?
I have a friend that is putting together a 5-way system (dipole)
and he wants to be able to use both step filters,compensation in
almost all registers and a high samplingfrequency. If he would end
up hitting the cpu's maximum capasity before the systems done,
would it be possible to run two "Najda's" in parralel from one usb
reciver over the I2S input using one for left and the other for right channel ?
Next subject.
If there is more then enough cpu capasity available, is there any
possibility to use that for "double presicion mode" ?
If he's happy with 96 kHz sample rate, then one single Najda will cope with 2x5 ways and plenty of filters.
If he wants 192 kHz, then a pair of boards will be necessary I suppose, based on your description. In this case, he would need to share all USB I2S signals between both Najdas and just set the routing appropriately on each unit.
It's already full double-precision. Please see the NUC manual fig. 10 p. 13.
I must say I am very pleased with the filter,it brings more out of my spekares than ......nanoDIGI does a more realsistic soundstage.
And I havent finnished tweaking yet..
Isn't there supposed to be a comma after NanoDigi ? Or are you
saying that the NanoDigi is presenting a more belivable stage in your system?
Yes there should be a coma,this is how it should be:
I must say I am very pleased with the filter,it brings more out of my spekares than ......nanoDIGI , a more realsistic soundstage.
And I havent finnished tweaking yet..
Thank you for reporting it.
Since I'm truly satisfied with the holografic presentation I thought there
might be something like that
If all goes well I'm fitting the tweeters tomorrow and if I get hold of some polyfill I'll fit the mids and woofers as well. The Najda is now in the box and is ready for duty. I'm currently testing the SMPS but will test the analog one more later.
Edit: For some reason I cannot attach pics..
Edit: For some reason I cannot attach pics..
Last edited: