Filter brewing for the Soekris R2R

Disabled Account
Joined 2005
No need bitchin' around here...


If you'd read my posts you'd notice my system is Danish speakers and DAC and the amp made in Netherlands.

The track in question was written by one of my favorite old school Scandinavian trance acts, and the compilation was released on Dutch label.

I thought it was quite appropriate. :)
 
Last edited:
Snow Patrol - no clipping. Somehow not sounding so 'boomy' this morning. Etta James is not right however. Audio looking good on scope but sounds awful. Can't believe she's blowing into the mic, but same track upsampled to 384k sounds exactly the same - not the filters then?

New Psu for main PC arrived, play time....
 
Well I've had this dac for a month or so, tried all the filters , and generally over the long term unimpressed.
If I was to give my Ian FIFO/isolater/double clock/PCM-i2s/AYA2014 100 points then the dam dac would get about 70 points. It's ordinary sounding , maybe better than the typical commercial offering using opamp outputs but one can do a lot better if diy'ing.
I tried it with stats, horns, moving coil speakers, tube and solid state very competent systems, but it never floated my boat.
I expect the power supply guys to tell me I need to to fix the power supply, but I simply don't think it's worth the effort .
This ones going in the reject box.
 
supra,

I don't know if you prefer tubes but after I switched from balanced output (nothing special) to raw + 6n30p SRPP I must say this is something that is better that anything I heard before even ESS Sabre with Ian's Isolator + Dual XO board with Crysteks 957.
I must admit that currently I'm playing from Foobar with Sox resampler (with sox mod plugin) set to 384kHz through Amanero interface.
I can't say if it is a merit of raw output or output stage (for sure tube grid make almost no load to DAC output). Anyone tried discrete transistor output stage?

I did try a few (not all) filters including NOS but till now none of theme can match with Sox results no matter if I set slider to linear, intermediate or minimal.

spzzzzkt, do you have any idea what is the diffrernce between your filters and those used by Sox?
 
C128x110v8 attached.

After having a listen to David Sylvian's "River Man" vs the Izotope up sampled version, I think this is probably closer in terms of tonality and impact of percussion bass attack. Without going to minimum/intermediate phase it's going to be hard to get closer in these aspects I think.

Yesterday, I spent some time listening to this filter after I moved my speakers some centimeters apart. I caught my self just listening to the music and singing for an hour, without spending effort to find inaccuracies or problems. And this was with 192kb mp3 files also (missing details, air or whatever in comparison to an flac but the music was there, in a natural way). I am not sure if it was the filter, the small change in the speakers placement, or the combination of these two, but I was very satisfied with what I was listening. So, thank you Paul!
 
Well I've had this dac for a month or so, tried all the filters , and generally over the long term unimpressed.
If I was to give my Ian FIFO/isolater/double clock/PCM-i2s/AYA2014 100 points then the dam dac would get about 70 points. It's ordinary sounding , maybe better than the typical commercial offering using opamp outputs but one can do a lot better if diy'ing.
I tried it with stats, horns, moving coil speakers, tube and solid state very competent systems, but it never floated my boat.
I expect the power supply guys to tell me I need to to fix the power supply, but I simply don't think it's worth the effort .
This ones going in the reject box.

Suit yourself but I wouldn't be so hasty. This thing took over a month to really break in. The output stage/filters/PS all make a noticeable difference and can be tailored to your system and taste. How can it be "ordinary" with so many user configurable options? One thing I can agree with here- it requires a lot of effort. In my experience it is more than worth it.
 
If you'd read my posts you'd notice my system is Danish speakers and DAC and the amp made in Netherlands.
The track in question was written by one of my favorite old school Scandinavian trance acts, and the compilation was released on Dutch label.
I thought it was quite appropriate.

Paul,
my response surely was not directed at you...:)
I'd never question your choice of music while you are evaluating filters...
In fact I really appreciate your ongoing effort to understand digital filters while
providing files to the community by the dozen... soon enough. ;)

I have a very good collection of music by scandinavian artists myself...
 
I've fitted my dac board into a Tube Technology DAC64 box - the original is still there although the digital and front panels are disconnected. I modified the TT tube output stage a while ago, which has a balanced input by the way, fitting Black Gate, Audionote paper in oil, and Mundorf Supremes where appropriate. I find little to choose between the dac op-amp balanced output (running unbalanced) and the TT analogue stage fed from the resistor divider, although perhaps the TT offers slightly more body and space. Both are blessed with pace, transparency, and dynamics, if you like that sort of thing. Be interesting to put it up against something like a Chord Hugo, dCS, and other high-end competition...

Looking forward to the next round of filter development. Will have a go myself when I fix my main pc, cooler leak caused motherboard failure!
 
Snow Patrol - no clipping. Somehow not sounding so 'boomy' this morning. Etta James is not right however. Audio looking good on scope but sounds awful. Can't believe she's blowing into the mic, but same track upsampled to 384k sounds exactly the same - not the filters then?

New Psu for main PC arrived, play time....

What Etta James recording was this? I briefly listened to a compilation of hers the other day (on headphones) and remember thinking it sounded pretty bad.
 
Etta James: I Only Want To Make Love To You


Comparing the Tag McLaren DVD32 driving the dac digital input (Bambadoo's filter & C128_100v8) with the Tag AV192 processor, I find them idential more or less. There is perhaps the tiniest fraction more focus (less phase error?) with the Tag, but not that you'd notice unless you pinner your ears back. The Tag DVD cost about 5000 pounds new, the 192 around 8000-12000 depending on which options were fitted, and was (still is, I'd say) highly regarded. Surely then, the sound quality from our dac speaks for itself? Thanks to Mr eBay for my Tag kit, could never afford to buy new.
 
Etta James: I Only Want To Make Love To You


Ah, OK, the only version of this I have (I Just ? Want To Make Love To You) is by Muddy Waters. so can't confirm your findings...

A bit OT - have you tried Xivero MusicScope yet? It can be enlightening as well as a bit scary too, lots of clipping reported on my 24bit/96kHz version of 'Rumours' compared to the 16bit/44.1kHz version, I will need to investigate further.
 
Disabled Account
Joined 2005
supra,

I don't know if you prefer tubes but after I switched from balanced output (nothing special) to raw + 6n30p SRPP I must say this is something that is better that anything I heard before even ESS Sabre with Ian's Isolator + Dual XO board with Crysteks 957.
I must admit that currently I'm playing from Foobar with Sox resampler (with sox mod plugin) set to 384kHz through Amanero interface.
I can't say if it is a merit of raw output or output stage (for sure tube grid make almost no load to DAC output). Anyone tried discrete transistor output stage?

I did try a few (not all) filters including NOS but till now none of theme can match with Sox results no matter if I set slider to linear, intermediate or minimal.

spzzzzkt, do you have any idea what is the diffrernce between your filters and those used by Sox?


Given that sox can be configured in countless ways it's a bit difficult to answer. The IPv4 filters were made with SoX but the SoX upsampling is not necessarily done in the same way the DAM does it, so there may be something in that aspect. If you can give me one specific setting you use in SoX I can probably have a go at replicating to see if it's the limitation of my filter creation skills (such as they are) or something more intrinsic.

sox upsampling enforces a stopband that starts at nyquist by default. Unless you are overriding that the SoX will be using very steep filters. If it's a third party gui all bets are off.
 
Last edited:
Disabled Account
Joined 2005
I'm loosing track of what's been posted in terms of filters :scratch:

Anyway after a bit of oversample switching Audirvana+ decided it was all too difficult and decided if I was going to swap that often it wasn't going to play anything until I quit the app and iTunes, then had the sulks every time I toggled rates. :dead:

So flipped back to Pure Music for a check against it's oversampling, which is pretty good. From what I can tell C128x90v8 does a better job of retaining the "air" and over-all balance vs PM compared with the other C128 filters.

I was having a listen to a C128x120v8 and it was really upfront, and lacking a bit of bottom end impact cf the other filters.

I've tried the C128x90v8 against a selection tracks, none "audiophile"...

Black Data I & II from Phantom City's Shiva Recoil recorded live at Tampere Jazz Festival fwiw.
https://www.youtube.com/watch?v=TgyrfOJ-dOY

Against the Upper House from Bill Laswell's Means of Deliverance - which is beautifully recorded acoustic bass.

I also revisited a couple of the problematic tracks and didn't have major issues either with this or the other C128's, so it seems it could have been Audirvana+ having a "moment".

ADD: I should probably mention that the XXX in C128xXXXv8 is the amount of side lobe attenuation on the filter. My take is that the lower levels of sidelobe attenuation (which means more ringing in the impulse) sounds more natural in this series. I think it's mainly case of the higher levels of attenuation result in a wider transition band which starts biting into the audio frequencies.
 

Attachments

  • 1021filtNQ_C128x90v8.skr.zip
    1.7 KB · Views: 72
  • 1021filtNQ_C128x90v8.txt
    23.9 KB · Views: 62
Last edited:
supra,

I don't know if you prefer tubes but after I switched from balanced output (nothing special) to raw + 6n30p SRPP I must say this is something that is better that anything I heard before even ESS Sabre with Ian's Isolator + Dual XO board with Crysteks 957.
I must admit that currently I'm playing from Foobar with Sox resampler (with sox mod plugin) set to 384kHz through Amanero interface.
I can't say if it is a merit of raw output or output stage (for sure tube grid make almost no load to DAC output). Anyone tried discrete transistor output stage?

I did try a few (not all) filters including NOS but till now none of theme can match with Sox results no matter if I set slider to linear, intermediate or minimal.

spzzzzkt, do you have any idea what is the diffrernce between your filters and those used by Sox?

The unbuffered output seems to pair extremely well with tube stages. Everything I listen to I keep hearing things I never knew were there on my old DAC. In my case the first preamp tube in my system is a 6SL7.