Why don't you upload a .skr file for us to try?
//
By using external x8 filter we can bypass onboard... It doesn't have to be x8, if you only have 96Khz input I think that can show some differences too. My point is to compare the default brickwall in Resampler-V with a much less aggressive one that sacrifices FR and alias suppression. Just to throw the idea out there if anyone hasn't tried. Should be like a 5-minute setup if you're using Foobar already.
I'm out.
//
Actually you might be right again. Out of curiosity tried an ABX with everything I had in mind, and I'm not doing much better than chance, even though I thought it'd be easy in the last couple of days. I would conclude from my experience that there might be systematic confounding factors that influences our judgment on this. Heck, the live music experiences that we so trust as ground-truth are probably the most heavily biased, since concertgoing is even more social and multi-faceted. It doesn't mean that there isn't a difference, but maybe at some point we'd get more enjoyment out of training our imagination than training our equipment. Even if there's a difference, ABX might not be all powerful as we'd like to believe - there's a lot that happens during test-taking...
Does dam1121 use 1021 filter file as default? the newest pre-4k filter set, 1.05?
the 1021 filters txt shows 1021 signature, maybe it doesnt matter for 1121?
the 1021 filters txt shows 1021 signature, maybe it doesnt matter for 1121?
Last edited:
Thanks, it shouldnt be too hard to compile a select few filters (like your 44.1 linear and the NOS) into one filter file then.
The file im using is called TNT441Lin2noDC.skr and it seemed to be stock with only linear filter changed, not sure where I found it, cant find the post again.
I found the most recent 4k tap 1021 files posted by spikestabber and you on page 170, I suppose your filter and the NOS filters will work on 1121 but some of others need 4K tap firmware.
I found the most recent 4k tap 1021 files posted by spikestabber and you on page 170, I suppose your filter and the NOS filters will work on 1121 but some of others need 4K tap firmware.
Filter brewing for the Soekris R2R
Post #1803 ;-)
I have not postad any 4k filters... yet. Not until 1121 gets 4k support. If ever.
//
Post #1803 ;-)
I have not postad any 4k filters... yet. Not until 1121 gets 4k support. If ever.
//
Aha - thats how you do it 😉 Tuff crawd :-D
This plays music as it can be heard in real life (44,1 (maybe 48?)). I stand by that.
Never loaded it - be careful...
//
This plays music as it can be heard in real life (44,1 (maybe 48?)). I stand by that.
Never loaded it - be careful...
//
Attachments
Last edited:
Ill let you know how it sounds 🙂
btw is the deemphasis filter only truly active with (rare) specially flagged audio files or something like that?
If not you would expect more discussion about its influence on sound/disabling it,
There was one mention I found of someone disabling in an old post .
btw is the deemphasis filter only truly active with (rare) specially flagged audio files or something like that?
If not you would expect more discussion about its influence on sound/disabling it,
There was one mention I found of someone disabling in an old post .
Really like that filter, super smooth compared to stock linear filter and maybe even a touch better detail retrieval at the same time, gain is noticeable lower I wonder if the extra headroom is part of that.
Best way to describe NOS filter sound would be direct but sloppy,
though that's not a surprise, I was more interested use NOS only for FIR2 but appears you cant ''mix and match''.
Best way to describe NOS filter sound would be direct but sloppy,
though that's not a surprise, I was more interested use NOS only for FIR2 but appears you cant ''mix and match''.
I’m interested in getting a 1% tolerance dam1021. I’ve almost designed my own R2R but for the price the only annoying thing is use of a Spartan 6 rather than something a little more modern with a little more resource (Artix7 perhaps with DDR). I’ve programmed FIR filters for Astro image convolution, matching poles to the scope psf on GPUs.
So a see questions (the filter designer doesn’t work on my iPhone)
* is the FIR/IIR filter able to implement a sinc filter/upsampling without introducing too much error? it looks like a Chebyshev filter by default? I was thinking butterworth perhaps with a bessel as options too.
* I take it that it can implement a pipeline but at a reducing pipeline rate as the rate increases?
* has anyone added a final passive RC low pass as a softer filter after the pwm switching?
* is the Firmware is closed source? I see sdk files (encrypted?) and a “makerom” dos exe (I’m on OS X/Linux) so it smells of secure boot using prehistoric unsupported tooling? So in 10 years the firmware filters can’t be updated easily when the resistor tolerances have slackened, much like the owner 😀
* I was going to model the design in python/ngspice to experiment with filters. How close are the filters to matlab/octave? I could model the coefficients in a jupyter notebook if the tooling is aging.
So it’s either go down this route or DIY it (with 2-3x cost).
So a see questions (the filter designer doesn’t work on my iPhone)
* is the FIR/IIR filter able to implement a sinc filter/upsampling without introducing too much error? it looks like a Chebyshev filter by default? I was thinking butterworth perhaps with a bessel as options too.
* I take it that it can implement a pipeline but at a reducing pipeline rate as the rate increases?
* has anyone added a final passive RC low pass as a softer filter after the pwm switching?
* is the Firmware is closed source? I see sdk files (encrypted?) and a “makerom” dos exe (I’m on OS X/Linux) so it smells of secure boot using prehistoric unsupported tooling? So in 10 years the firmware filters can’t be updated easily when the resistor tolerances have slackened, much like the owner 😀
* I was going to model the design in python/ngspice to experiment with filters. How close are the filters to matlab/octave? I could model the coefficients in a jupyter notebook if the tooling is aging.
So it’s either go down this route or DIY it (with 2-3x cost).
I noticed that the FIR2 filters are always running at 384kHz with 44.1kHz input when viewed in teraterm, I just reverted to stock filter file and its still doing this... I thought they were running at 352.8 in the beginning but not sure now.
normal or?
on dam1121 btw
normal or?
on dam1121 btw
- Home
- Source & Line
- Digital Line Level
- Filter brewing for the Soekris R2R