Filter brewing for the Soekris R2R

[analog_sa]

Setting equal gain will certainly make sense, especially if some of the filters are unusable at 0db. No idea if this is simple, trivial, very involved, or even possible.

 

[TNT]

There is the parameter "multiplier" one can change for any given filter having access to the text file - even if you didn't generate the filter coefs. It seems that every single filter line will be ehhh multiplied by this before computation.

-- signature, samplerate, interpolationrate, type, numbercoefficients, multiplier

dam1021,44100,8,4,1975,7.8
0
-0.00000000000000016622018796799955
-0.00000000000000030396469096397812
..... 1972 lines more

(so in the above example, 7,8)


So even if you didnt produce the taps you can change the gain of the filter. But it's also possible to set the gain in a filter generation program like rePhase. As I stated, I did -3 dB in rePhase. But if I set the multiplier to >8 I'd be back at zero or more again. This is my current understanding.

If I understood Sören correctly, these kind of clipping can be avoided by turning the digital volume down say 3 dB. Does that work?

//

 

[analog_sa]

Thank you for the explanation.

Yes, turning back the volume works fine. -2db seems enough. Not sure what the situation is if using the hardware volume control with the pot: is it possible to set a max value when the pot is active?

Do you know if the attenuation is proportional to the fraction of 8 on that line?

 

[TNT]

Setting volume by command or by potentiometer/encoder connected to header control pins should be no difference - it manipulates the same thing inside the DAM.

I don't think you can set a level guard value when volume adjustment is active.

I haven't measured on how the multiplier works.

//

 

[spikestabber]

Setting volume by command or by potentiometer/encoder connected to header control pins should be no difference - it manipulates the same thing inside the DAM.

I don't think you can set a level guard value when volume adjustment is active.

I haven't measured on how the multiplier works.

//

You can't, people have requested it before but the reality is you need to workaround it by adding a 2nd trim pot for your 0dB+ guard. II personally don't use the analog pot for the volume control, however I thought it was once possible to go to up to +10dB with the vol pot, even greater need for a guard band if that is the case.

 

[TNT]

Yes, I think the command: "V+10" is accepted.

//

 

[analog_sa]

Thanks Spikestabber for refreshing my memory about the pot situation. So, it is pretty much unusable, yet it seems easy to fix. Something like the dysfunctional i2s port on the 1941

 

[bmxmen]

Hello did any one have problem with i2s connection after change 1.05 to 0.99 with nos party pack? I have soekris v4 please help i need to have 0.99 with party nos pack

 

[analog_sa]

Not me. I did roll back to 0.99 for a while and recall no issues.

 

[bmxmen]

What filters did You use? I have an two same soekris, v2 and v4 onn Full batery Power suply . V2 with 99 nad party nos filters. Outstanding Sound quality and need to do same with v4.

 

[bmxmen]

For now i have 1.19 Full pack and Sound Good, near same like an 0.99 party pack nos

 

[spikestabber]

Me and TNT both posted some updated packs that work fine on 1.19/1.21, TNT replaced the Soekris 4K with his, attached is the one with Soekris 4K as filter F4 with the text.

04 Linear Phase Soekris, 44.1 Ksps, 21.3 Khz - 1 db, 21.9 Khz -130 db
04 Linear Phase Soekris, 48 Ksps, 23.2 Khz - 1 db, 23.8 Khz -130 db
05 EQHQ_lpbr_b4, 44.1 Khz, 0-20.45Khz +-0.00000005dB, 22.50Khz -174dB
05 EQHQ_lpbr_b4, 48 Khz, 0-22.26Khz +-0.00000005dB, 24.49Khz -174dB
06 C128dp Linear Phase NYQUIST, Interpolation 16 SAMPLES, 44.1Khz.
06 C128dp Linear Phase NYQUIST, Interpolation 16 SAMPLES, 48 Khz.
07 NOS FIR1, 44.1 Khz.
07 NOS FIR1, 48 Khz.
04 Linear Phase Soekris, 88.2 Ksps, 38.7 Khz - 1 db, 39.8 Khz -130 db
04 Linear Phase Soekris, 96 Ksps, 42.1 Khz - 1 db, 42.2 Khz -130 db
05 EQHQ_lpbr_b4, 88.2 Khz, 0-40,9Khz +-0.00000005dB, 45Khz -174dB
05 EQHQ_lpbr_b4, 96.0 Khz, 0-44,5Khz +-0.00000005dB, 48.9Khz -174dB
06 C128dp Linear Phase NYQUIST FIR1, Interpolation 16 SAMPLES, 88.2Khz
06 C128dp Linear Phase NYQUIST FIR1, Interpolation 16 SAMPLES, 96Khz
07 NOS FIR1, 88.2 Khz.
07 NOS FIR1, 96 Khz.
04 Linear Phase Soekris, 176.4 Ksps, 77 Khz -1 db, 80 Khz -130 db
04 Linear Phase Soekris, 192 Ksps, 84 Khz -1 db, 87 Khz -130 db
05 EQHQ_lpbr_b4, 176.4 Khz, 0-163,6Khz +-0.00000005dB, 180Khz -174dB
05 EQHQ_lpbr_b4, 192.0 Khz, 0-178Khz +-0.00000005dB, 196Khz -174dB
06 C128dp Linear Phase NYQUIST FIR1, Interpolation 16 SAMPLES, 176.4Khz
06 C128dp Linear Phase NYQUIST FIR1, Interpolation 16 SAMPLES, 192Khz
07 NOS FIR1, 176.4 Khz.
07 NOS FIR1, 192 Khz.
04 Bypass FIR1, 352.8 Ksps
04 Bypass FIR1, 384 Ksps
07 NOS FIR1, 352.8 Khz.
07 NOS FIR1, 384 Khz.
04 Linear Phase DSD64 2.8 Msps, 30 Khz -3db
05 Mixed Phase DSD64 2.8 Msps, 30 Khz -3db
06 Minimum Phase DSD64 2.8 Msps, 30 Khz -3db
04 Linear Phase DSD128 5.6 Msps, 60 Khz -3db
05 Mixed Phase DSD128 5.6 Msps, 60 Khz -3db
06 Minimum Phase DSD128 5.6 Msps, 60 Khz -3db
04 Linear Phase DSD256 11.2 Msps, 70 Khz -3db
05 Mixed Phase DSD256 11.2 Msps, 70 Khz -3db
06 Minimum Phase DSD256 11.2 Msps, 70 Khz -3db
08 Linear Phase Soekris FIR2, 352.8 Ksps, 102 Khz -1 db, 177 Khz -130 db
08 Linear Phase Soekris FIR2, 384 Ksps, 111 Khz -1 db, 192 Khz -130 db
09 EQHQ_lpbr_b0, 352.8 Khz, 0-30.00Khz +-0.00000004dB, 264.60Khz -176dB
09 EQHQ_lpbr_b0, 384.0 Khz, 0-32.65Khz +-0.00000004dB, 268.40Khz -176dB
10 C128dp Linear Phase NYQUIST FIR2, Interpolation 16 SAMPLES
10 C128dp Linear Phase NYQUIST FIR2, Interpolation 16 SAMPLES
11 NOS FIR2, 352.8 Khz.
11 NOS FIR2, 384 Khz.
30 Deemphasis IIR, 352.8 Khz Samplerate, hiself f=5600 Hz, Q=0.485, gain=-10.1
30 Deemphasis IIR, 384 Khz Samplerate, hiself f=5600 Hz, Q=0.485, gain=-10.1 dB

 

[TNT]

Would that be the latest release from Soekris or is this a custom pack of some sort?

//

 

[spikestabber]

Would that be the latest release from Soekris or is this a custom pack of some sort?

//

It's the same custom pack you edited, just with the Soekris 4K filter. I like this one because it is good to compare the Soekris 4K with NOS and 2 of Paul's best work. IMO anyone not using 3rd party alternative filters is doing themselves a disservice with this DAC and should only review the unit based on disclaimer they are using the stock (IMO rather bad) filters.

 

[TNT]

Aha - thanks!

//

 

[analog_sa]

attached is the one with Soekris 4K as filter F4 with the text.

Thank you!

Are volume levels of F5,6,7 still excessive?

 

[bmxmen]

For now i hve 1.19 pack and it is plain well. Something beter is pasible?

 

[spikestabber]

Thank you!

Are volume levels of F5,6,7 still excessive?

Try this one. Tweaked the multipliers, tested a bunch of various values, ended up at 6.4 which matches the Soekris one and so on, amplitude seems spot on.

 

[spikestabber]

That pack is tested good for anyone still on the fence, all samplerates work also. Tested using firmware 1.21. A mod can kindly delete the file from my previous post @1692 with the varying gains.

 

[TNT]

And the above with the same as before TNT44/48.

//