[SlimDevices: Audiophiles] Quick question about DAC "filters"

[ralphpnj]

i just finished reading a really poorly written review of Auralic's
Altair streaming DAC and in the review extensive coverage is given over
to the various filters in the unit.

So my question is:

How are these "filters" different from a simple graphic equalizer or
simple DSP or even, god forbid, tone controls?

Note: the review is the March 2017 issue of Stereophile.



Living Rm: Transporter-SimAudio pre/power amps-Vandersteen 3A Sign. &
sub
Home Theater: Touch-Marantz HTR-Energy Veritas 2.1 & Linn sub
Computer Rm: Touch-Headroom Desktop w/DAC-Aragon amp-Energy Veritas 2.1
& Energy sub
Bedroom: Touch-HR Desktop w/DAC-Audio Refinement amp-Energy Veritas 2.0
Guest Rm: Duet-Sony soundbar
Garage: SB3-JVC compact system
Controls: iPeng; SB Controller; Moose & Muso
Server: LMS 7.9 on dedicated windows 10 computer w/2 Drobos
'Last.fm' (http://www.last.fm/user/jazzfann/)
------------------------------------------------------------------------
ralphpnj's Profile: http://forums.slimdevices.com/member.php?userid=10827
View this thread: http://forums.slimdevices.com/showthread.php?t=106979

[pablolie]

I haven't read it.

But in a perfect world, I'd like this simple feature in a DAC+Preamp, or
in an integrated DAC+Amp unit into active speakers:

* High quality and configurable (30-40-50-60-etc) frequency divider in
the digital domain with both analog and digital *dual* outputs, one for
low and one for high domain. Some audiophiles may prefer more, but
beyond 3 I never would see the point.

* Option for room optimization in both domains.

My current main setup [see sig] provides me with the output options, but
not with the built-in configurability. I drive my speakers full range,
and optimize the sub on the side of it. The system I set up at my SO's
place (NAD D7050 integrated amp) allows for total optimization. You can
configure the cutover freq.



...pablo
Server: Virtual Machine (on VMware Workstation 12) running Ubuntu 16.04
+ LMS 7.9
System: SB Touch --optical->- Benchmark DAC2HGC --AnalysisPlus Oval
Copper XLR->- NAD M22 Power Amp --AnalysisPlus Black Mesh Oval->- Totem
Element Fire
Other Rooms: 2x SB Boom; 1x SB Radio; 1x SB Classic-> NAD D7050 -> Totem
DreamCatcher + Velodyne Minivee Sub
Computer audio: workstation --USB->- audioengine D1 -> Grado
PS500e/Shure 1540
------------------------------------------------------------------------
pablolie's Profile: http://forums.slimdevices.com/member.php?userid=3816
View this thread: http://forums.slimdevices.com/showthread.php?t=106979

[Mnyb]

Post by ralphpnj
i just finished reading a really poorly written review of Auralic's
Altair streaming DAC and in the review extensive coverage is given over
to the various filters in the unit.
How are these "filters" different from a simple graphic equalizer or
simple DSP or even, god forbid, tone controls?
Note: the review is the March 2017 issue of Stereophile.

They are probably talking about the DAC 's reconstruction filters both
the digital and analog part, the measurements side note did it have a
lot of different impulse responses ? They are a necessary part of every
DAC ( except some NOS DAC's but they are broken by design )

These are not really eq filters but the final stage in giving you a nice
analog signal after the DAC process without ultrasonic gunk in the
signal .
The hot topic a couple of year have been these filters impulse response
and steepness and resulting ringing fancy terms like "apodizing" is used
. In some cases like Ayre DAC they even sacrifice the hf response a bit
to get less ringing .

And also how they are implemented in the chips or software is point in
good DAC design .
Not just the chosen filter topology but how you do it , for example the
dreaded intersample overshoots you can get if these filters have no
margins for very high levels in the signal and you are oversampling .
It can be done better with floating point math in software instead of
hard wired digital filters ? Or can it ?

Some extrem setting can be "EQ like" in the highest treble , there is
always some trade offs between frequency response phase response and
ringing .

The ringing in DAC filters has always been assumed benign as the
frequency of it is above human hearing and you have pre and post ringing
in these impulse response test .
Which are just test this kind of signal is not present in music the DAC
normally converts ( due to normal ADC filtering when recording) but a
way to test one property of the filter .

Meridian and Ayre and some others argue that apodizing aka no pre
ringing sounds better even if it sacrifices performance elsewhere

All in all usually very subtle differences if they are at all audible .
In the old days a thing the DAC designer just decided for the design at
hand .
Now for good and bad it's a knob to twiddle on some DAC .

And in software if you oversample with Squeezelite on PI based player
you got the whole range of possibilities to twiddle with for yourself :)
a real handful the SoX lib can any filter you like ? I'm testing it to
remove my MeridianG98DH ovesampling of inputs by letting Squeezelite do
it with the important diff of -3 dB attenuation to handle intersample
oversshots



--------------------------------------------------------------------
Main hifi: Touch + CIA PS +MeridianG68J MeridianHD621 MeridianG98DH 2 x
MeridianDSP5200 MeridianDSP5200HC 2 xMeridianDSP3100 +Rel Stadium 3
sub.
Bedroom/Office: Boom
Kitchen: Touch + powered Fostex PM0.4
Misc use: Radio (with battery)
iPad1 with iPengHD & SqueezePad
(spares Touch, SB3, reciever ,controller )
server HP proliant micro server N36L with ClearOS Linux

http://people.xiph.org/~xiphmont/demo/neil-young.html
------------------------------------------------------------------------
Mnyb's Profile: http://forums.slimdevices.com/member.php?userid=4143
View this thread: http://forums.slimdevices.com/showthread.php?t=106979

[Archimago]

Post by ralphpnj
i just finished reading a really poorly written review of Auralic's
Altair streaming DAC and in the review extensive coverage is given over
to the various filters in the unit.
How are these "filters" different from a simple graphic equalizer or
simple DSP or even, god forbid, tone controls?
Note: the review is the March 2017 issue of Stereophile.

Hey Ralph,
Like Mnyb said, they're just referring to variations of the digital
oversampling antialiasing filter used in the DAC. Subtleties like the
presence of pre-ringing in the impulse response, passband, length of
filter can change the frequency response and determine how well aliasing
distortion is filtered out when the signal is upsampled.

My post this past week on the blog models what is seen with the MQA
filter for example:
http://archimago.blogspot.com/2017/02/musings-discussion-on-mqa-filter-and.html

SoX can do it but if you want a bit more control with the GUI interface,
iZotope RX does a great job with modelling the filter parameters. Have
fun experimenting and seeing if you can hear the difference.

Bottom line - unless you really fool around with the parameters and
strongly affect the frequency response (like with Ayre's filter and the
PonoPlayer), you're not going to hear much difference. If you play a
very loud track with clipping and square waveforms, then you might see
the ringing in the upsampled signal as well as intersample overload
distortions. With properly low-passed signal (as in all frequencies
<22.05kHz in a 16/44.1 file), then there's not going to be any ringing
in the output...



Archimago's Musings: (archimago.blogspot.com) A 'more objective'
audiophile blog.
------------------------------------------------------------------------
Archimago's Profile: http://forums.slimdevices.com/member.php?userid=2207
View this thread: http://forums.slimdevices.com/showthread.php?t=106979

[Wombat]

When people know what filter they play they will tell you all kind of
stories.
When people don't know what filter they play they dig in the dark.
Like Archimago pointed out there can be filters that change parameters
very much so things like frequency response or phase can be audible.
Only the legendary Maridian filter listening test had a tiny positive
pecentage positive result with very steep, strong ringing from
upsampling.
Forget about the nonsense people wirte about every tiny parameter in a
filter changes the sound.



Transporter (modded) -> RG142 -> Avantgarde Acoustic based 500VA
monoblocks -> Sommer SPK240 -> self-made speakers
------------------------------------------------------------------------
Wombat's Profile: http://forums.slimdevices.com/member.php?userid=4113
View this thread: http://forums.slimdevices.com/showthread.php?t=106979

[Archimago]

Post by Wombat
When people know what filter they play they will tell you all kind of
stories.
When people don't know what filter they play they dig in the dark.
Like Archimago pointed out there can be filters that change parameters
very much so things like frequency response or phase can be audible.
Only the legendary Maridian filter listening test had a tiny positive
pecentage positive result with very steep, strong ringing from
upsampling.
Forget about the nonsense people wirte about every tiny parameter in a
filter changes the sound.

Yeah, that epic Meridian listening test that also either truncated or
used rectangular dithering that they admitted were suboptimal.



Archimago's Musings: (archimago.blogspot.com) A 'more objective'
audiophile blog.
------------------------------------------------------------------------
Archimago's Profile: http://forums.slimdevices.com/member.php?userid=2207
View this thread: http://forums.slimdevices.com/showthread.php?t=106979

[Wombat]

Remember the "Ringing Less" filter you tested?
http://archimago.blogspot.de/2016/10/measurements-hifiberry-dac-pro-pcm5122.html
Alone the positive name of the filter had effect on the listener that
reported its 'better' sound it seems.



Transporter (modded) -> RG142 -> Avantgarde Acoustic based 500VA
monoblocks -> Sommer SPK240 -> self-made speakers
------------------------------------------------------------------------
Wombat's Profile: http://forums.slimdevices.com/member.php?userid=4113
View this thread: http://forums.slimdevices.com/showthread.php?t=106979

[Wombat]

Post by Archimago
Yeah, that epic Meridian listening test that also either truncated or
used rectangular dithering that they admitted were suboptimal.

Or simply the used beryllium metal tweeter distorts different with or
without HF content.



Transporter (modded) -> RG142 -> Avantgarde Acoustic based 500VA
monoblocks -> Sommer SPK240 -> self-made speakers
------------------------------------------------------------------------
Wombat's Profile: http://forums.slimdevices.com/member.php?userid=4113
View this thread: http://forums.slimdevices.com/showthread.php?t=106979

[arnyk]

Post by Archimago
Yeah, that epic Meridian listening test that also either truncated or
used rectangular dithering that they admitted were suboptimal.

Also significant was the fact that they used ultra-sharp, narrow
bandwidth filters that generated a lot more ringing than just about any
real world DAC.


------------------------------------------------------------------------
arnyk's Profile: http://forums.slimdevices.com/member.php?userid=64365
View this thread: http://forums.slimdevices.com/showthread.php?t=106979

[Archimago]

Yup... Good points all.

*Bottom line: *The Industry -needs -us to believe the filters are a "big
deal". After all, how else to differentiate digital audio of which
essentially any decent DAC over the last 20 years has sounded very good
already!? Maintaining this sense of wonder and belief in claims
underpins many of the advertised benefits for companies as disparate as
PS Audio, Ayre / Pono, Meridian, T+A, all the way to MQA and its promise
of "revolutionary" sound quality.

What the audiophile world truly needs to evolve and "revolve" is a
realization of the objective mindset in terms of adjudicating the
engineering around devices and file "formats". Time for the pendulum to
swing back to objectivity... A long time coming but I remain hopeful
that we can all contribute to making an impact.



Archimago's Musings: (archimago.blogspot.com) A 'more objective'
audiophile blog.
------------------------------------------------------------------------
Archimago's Profile: http://forums.slimdevices.com/member.php?userid=2207
View this thread: http://forums.slimdevices.com/showthread.php?t=106979

[Mnyb]

Sadly this kind of product works very well with the audiophile mindset ,
you could also have a twiddly knob that did not do anything at all . (
placebo twiddly knob anyone ? ) this one does subtle things that or may
not be audible , but thanks to sighted testing etc the imagined part is
much bigger than the real part .

The engaged users feels involved now that he can tweak the product +1
and probably do this during the whole "burn in period" aka when you
psychologically adjust to the new thing , ( the only that gets burned in
is you ;) ) and you can probably share your revelations trough a forum
that product xyz with setting abc finally lifted the last veil and got
you blacker places and more air between the instruments ....

I having nothing really agianst this feature but the implications gets
blown out of proportion .



--------------------------------------------------------------------
Main hifi: Touch + CIA PS +MeridianG68J MeridianHD621 MeridianG98DH 2 x
MeridianDSP5200 MeridianDSP5200HC 2 xMeridianDSP3100 +Rel Stadium 3
sub.
Bedroom/Office: Boom
Kitchen: Touch + powered Fostex PM0.4
Misc use: Radio (with battery)
iPad1 with iPengHD & SqueezePad
(spares Touch, SB3, reciever ,controller )
server HP proliant micro server N36L with ClearOS Linux

http://people.xiph.org/~xiphmont/demo/neil-young.html
------------------------------------------------------------------------
Mnyb's Profile: http://forums.slimdevices.com/member.php?userid=4143
View this thread: http://forums.slimdevices.com/showthread.php?t=106979

[arnyk]

Post by Archimago
Yup... Good points all.
*Bottom line: *The Industry -needs -us to believe the filters are a "big
deal". After all, how else to differentiate digital audio of which
essentially any decent DAC over the last 20 years has sounded very good
already!? Maintaining this sense of wonder and belief in claims
underpins many of the advertised benefits for companies as disparate as
PS Audio, Ayre / Pono, Meridian, T+A, Chord all the way to MQA and its
promise of "revolutionary" sound quality.
What the audiophile world truly needs to evolve and "revolve" is a
realization of the objective mindset in terms of adjudicating the
engineering around devices and file "formats". Time for the pendulum to
swing back to objectivity... A long time coming but I remain hopeful
that we can all contribute to making an impact.

Begs the question - how much would audio technology progree if so much
time, capital, and intellect were applied to things that actually
mattered. I found a list of commercially-available DACs and it had about
400 items on it. This is truely amazing to me, partically because I
know from practical experience that upwards of 90% of those products are
not audibly different in any practical way. As I recall the list did
not include other products that incorporated DACs as critical components
such as AVRrs, equalizers, and PCs so the actual number of alternatives
in this field add up to thousand's of redundant products. All of that
redundant engineering, marketing, sales, labor and capital flushed down
the porcelain convenience!


------------------------------------------------------------------------
arnyk's Profile: http://forums.slimdevices.com/member.php?userid=64365
View this thread: http://forums.slimdevices.com/showthread.php?t=106979

[foxesden]

If manufacturers are serious about providing a compelling reason for
buying their products I don't understand why more of them don't
incorporate a usable DSP for example for speaker/room correction into
their products. The difference between a £30 raspi dac and a £2K dac is
pretty marginal in my experience, but sorting out dodgy peaks makes a
massive difference when listening to music. I guess it boils down to
marketing...


------------------------------------------------------------------------
foxesden's Profile: http://forums.slimdevices.com/member.php?userid=56286
View this thread: http://forums.slimdevices.com/showthread.php?t=106979

[arnyk]

Post by foxesden
If manufacturers are serious about providing a compelling reason for
buying their products I don't understand why more of them don't
incorporate a usable DSP for example for speaker/room correction into
their products. The difference between a £30 raspi dac and a £2K dac is
pretty marginal in my experience, but sorting out dodgy peaks makes a
massive difference when listening to music. I guess it boils down to
marketing...

This product comes pretty close to what you seem to be talking about:
https://www.minidsp.com/products/minidsp-in-a-box/minidsp-2x4


------------------------------------------------------------------------
arnyk's Profile: http://forums.slimdevices.com/member.php?userid=64365
View this thread: http://forums.slimdevices.com/showthread.php?t=106979

[drmatt]

I have a mytek DAC with such options. Upsampling on/off, filter
sharp/slow. Can't say, thus far, that I've noted any difference between
them. Perhaps I've not fed it any material that tickles the ultrasonic,
or jitter is not too bad in my system, or my tweeters behave well in the
presence of ultrasonics, or.. I'm just too old to hear anything that
high up..



--
Hardware: 3x Touch, 1x Radio, 2x Receivers, 1 HP Microserver NAS with
Debian+LMS 7.9.0
Music: ~1300 CDs, as 450 GB of 16/44k FLACs. No less than 3x 24/44k
albums..
------------------------------------------------------------------------
drmatt's Profile: http://forums.slimdevices.com/member.php?userid=59498
View this thread: http://forums.slimdevices.com/showthread.php?t=106979

[ralphpnj]

Post by drmatt
I have a mytek DAC with such options. Upsampling on/off, filter
sharp/slow. Can't say, thus far, that I've noted any difference between
them. Perhaps I've not fed it any material that tickles the ultrasonic,
or jitter is not too bad in my system, or my tweeters behave well in the
presence of ultrasonics, or.. I'm just too old to hear anything that
high up..

Or maybe it's just that you happen to be human rather than SUPER-human
or an electronic measuring device and so are physically unable to hear
ultrasonic frequencies and pico second jitter.



Living Rm: Transporter-SimAudio pre/power amps-Vandersteen 3A Sign. &
sub
Home Theater: Touch-Marantz HTR-Energy Veritas 2.1 & Linn sub
Computer Rm: Touch-Headroom Desktop w/DAC-Aragon amp-Energy Veritas 2.1
& Energy sub
Bedroom: Touch-HR Desktop w/DAC-Audio Refinement amp-Energy Veritas 2.0
Guest Rm: Duet-Sony soundbar
Garage: SB3-JVC compact system
Controls: iPeng; SB Controller; Moose & Muso
Server: LMS 7.9 on dedicated windows 10 computer w/2 Drobos
'Last.fm' (http://www.last.fm/user/jazzfann/)
------------------------------------------------------------------------
ralphpnj's Profile: http://forums.slimdevices.com/member.php?userid=10827
View this thread: http://forums.slimdevices.com/showthread.php?t=106979

[drmatt]

Post by ralphpnj
Or maybe it's just that you happen to be human rather than SUPER-human
or an electronic measuring device and so are physically unable to hear
ultrasonic frequencies and pico second jitter.

Nah, can't be that..



--
Hardware: 3x Touch, 1x Radio, 2x Receivers, 1 HP Microserver NAS with
Debian+LMS 7.9.0
Music: ~1300 CDs, as 450 GB of 16/44k FLACs. No less than 3x 24/44k
albums..
------------------------------------------------------------------------
drmatt's Profile: http://forums.slimdevices.com/member.php?userid=59498
View this thread: http://forums.slimdevices.com/showthread.php?t=106979

[darrenyeats]

The reconstruction filters used in up-sampling are anti-imaging
(anti-aliasing filters are used in down-sampling, ADCs etc).

In theory, no filter should make an audible difference if the transition
band (the curved bit of the graph) is above audible frequencies.

In practice, distortion is generated by content above audible
frequencies, and this distortion can leak into the audible band. I
believe this the only route for images, ringing etc to become audible.

See here for an interesting discussion:
http://www.pinkfishmedia.net/forum/showthread.php?p=3040013#post3040013



Check it, add to it! http://www.dr.loudness-war.info/

SB Touch
------------------------------------------------------------------------
darrenyeats's Profile: http://forums.slimdevices.com/member.php?userid=10799
View this thread: http://forums.slimdevices.com/showthread.php?t=106979

[darrenyeats]

Post by ralphpnj
How are these "filters" different from a simple graphic equalizer or
simple DSP or even, god forbid, tone controls?

When the sampling rate is increased, "images" of the original signal are
created at higher frequencies - the up-sampling (anti-imaging) filters
are intended to remove these. Any filter placed after the DAC
up-sampling stage could do a similar job in principle. In fact NOS DACs
rely on the fact that amps, speakers, ears - all after the DAC output -
are natural filters. And even up-sampling DACs use a combination of
digital and analogue anti-imaging filters.

But if the filter is applied (e.g. with equaliser in a computer player)
before up-sampling, it won't be doing the same job.

You can -up-sample and filter- on a computer, and pass the music at the
higher sample rate to the DAC. This could avoid a stage of up-sampling
and filtering in a typical DAC (but with typical DACs there is more than
one stage).



Check it, add to it! http://www.dr.loudness-war.info/

SB Touch
------------------------------------------------------------------------
darrenyeats's Profile: http://forums.slimdevices.com/member.php?userid=10799
View this thread: http://forums.slimdevices.com/showthread.php?t=106979

[arnyk]

Post by darrenyeats
When the sampling rate is increased, "images" of the original signal are
created at higher frequencies - the up-sampling (anti-imaging) filters
are intended to remove these. Any filter placed after the DAC
up-sampling stage could do a similar job in principle. In fact NOS DACs
rely on the fact that amps, speakers, ears - all after the DAC output -
are natural filters. And even up-sampling DACs use a combination of
digital and analogue anti-imaging filters.
But if the filter is applied (e.g. with equaliser in a computer player)
before up-sampling, it won't be doing the same job.
You can -up-sample and filter- on a computer, and pass the music at the
higher sample rate to the DAC. This could avoid a stage of up-sampling
and filtering in a typical DAC (but with typical DACs there is more than
one stage).

This is conditionally and not uncommonly false. If the DAC is operating
with an excessively high sample rate as is common with audiophiles, say
96 KHz and above, then an anti-imaging filter further down the signal
chain operating at a more sane design frequency such as 44 KHz can still
remove any audible imaging.

However, upsampling is generally just superstitious audiophile
wheel-spinning. Once certain kinds of mistakes are made, they cannot be
fixed with subsequent processing. Audiophiles have proven their
inability to hear what they claim as shown by their willing acceptance
and adulation for so-called high resolution recordings (SACD and DVD-A).
As of about 2008 about half of them or more had low resolution
recordings in their provenance, and the inherent audible problems with
resolution and bandpass built into in them cannot be overcome with
post-processing. No audiophile or audiophile reviewer that I know of
reported thins in the usual Golden Ear publications.

If one of the so-caled NOS DACs is followed by a properly-designed
reconstruction filter, the intentional design error that they embody,
which is the elimination of said filter, can be overcome.


------------------------------------------------------------------------
arnyk's Profile: http://forums.slimdevices.com/member.php?userid=64365
View this thread: http://forums.slimdevices.com/showthread.php?t=106979

[darrenyeats]

Post by arnyk
This is conditionally and not uncommonly false.

So a part or all of what I wrote ("this") is sometimes not true? What
exactly did you mean by "this"?

Post by arnyk
If the DAC is operating with an excessively high sample rate as is
common with audiophiles, say 96 KHz and above, then an anti-imaging
filter further down the signal chain operating at a more sane design
frequency such as 44 KHz can still remove any audible imaging.

Assuming by "down the signal chain" you mean "after the DAC output", I
don't see how any of my statements contradict the above - some at the
least imply the above. Please explain further.



Check it, add to it! http://www.dr.loudness-war.info/

SB Touch
------------------------------------------------------------------------
darrenyeats's Profile: http://forums.slimdevices.com/member.php?userid=10799
View this thread: http://forums.slimdevices.com/showthread.php?t=106979