Mnyb
2014-10-29 04:30:06 UTC
This may be very old news to the mastering engineers at this forum .
More about intersample peaks .
The company TC electronics have some of their AES papers freely aviable
. I'm not affiliated with them similar research may be aviable elsewhere
. Please provide me with more reading in the tread if you have some god
reading to suggest .
http://www.tcelectronic.com/tech-library/mastering/
You get four different papers describing some consequences of brick wall
audio at 0dFs especially with very artificial limits .
Yes this is about intersample peaks or intersample overs as they are
called .
The one where they run borked signals trough consumer CD players are
really interesting .
A perverse consequence can be that really old school DAC can behave more
benign when overloaded , such are the times we are living in .
Benchmark audio has taken it into their design of their latest DAC .
This has been discussed before on this forum
http://benchmarkmedia.com/products/benchmark-dac2-hgc-digital-to-analog-audio-converter
It's sales blurb ,but they actually added some headroom for this
phenomena -3.5 dB
Hydrogen audio forum have an ongoing discussion .
I've read treads in this forum where members have had good results by
actually lower the digital gain before the DAC or transporter with SoX
.
I was just surfing the web for good reading about this problem , hence
sharing with you . Appreciate even better reading about it if you help
out .
Reflections so far , my thoughts .
This is an understood and we'll known problem for at least a decade ( if
not more )yet it s consequences has not changed much in mastering or
consumer audio .
Master engineers still use outdated methods to evaluate peak levels (
not considering the true signal level after reconstruction ).
Equipment manufacturers and chipmakers does not seem to care much .
Happily making a bad problem worse .
Modern productions methods can easilly produce signals inside the
"system" that violates the nyquist requirements .
Perceptual coders are also suspectible in their filters mp3 and AAC OGG
and all are not at all happy with this .
(There is a reason for why Apples mastered for iTunes program suggest -1
dB and delivery as 24/96 before their conversion to 256kbs AAC .)
This would not be a problem if all where as carefull as some old school
sound engineers regarding levels .
And finally a Squeezebox thought , if you are using a DAC or in my case
a HT processor hence you untilise the digital output.
Would it be a good idea to just pull down the digital volume a little
bit prior to your DAC when listening to music with these "production
values" . I think this is sufficient ? Or is it not .
Or are more elaborate measures using SoX and custom-convert.conf really
needed ? Configuring SoX in this case can have some potential problems
done wrong you can run into the same problems as discussed in these
papers !
The simple undithered 24 bit volume in a squeezebox simply move the bits
down the "ladder" and they fall off at the end ( truncated ) harmless
with 16 bit , a theoretical issue with 24 bit sources where maybe you
should have used dither .
The Pro people usually always suggest dither when changing bit depth ,
but a small volume change in a squeezebox does not really reduce the
bits .
Any suggestions .
I will ofcourse I'll try at home in the weekend , the first thing a
16/44.1 input sees in my system is a sample rate converter to 24/96 then
probably conversion to float before entering the DSP processes .
--------------------------------------------------------------------
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
(in storage 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=102378
More about intersample peaks .
The company TC electronics have some of their AES papers freely aviable
. I'm not affiliated with them similar research may be aviable elsewhere
. Please provide me with more reading in the tread if you have some god
reading to suggest .
http://www.tcelectronic.com/tech-library/mastering/
You get four different papers describing some consequences of brick wall
audio at 0dFs especially with very artificial limits .
Yes this is about intersample peaks or intersample overs as they are
called .
The one where they run borked signals trough consumer CD players are
really interesting .
A perverse consequence can be that really old school DAC can behave more
benign when overloaded , such are the times we are living in .
Benchmark audio has taken it into their design of their latest DAC .
This has been discussed before on this forum
http://benchmarkmedia.com/products/benchmark-dac2-hgc-digital-to-analog-audio-converter
It's sales blurb ,but they actually added some headroom for this
phenomena -3.5 dB
Hydrogen audio forum have an ongoing discussion .
I've read treads in this forum where members have had good results by
actually lower the digital gain before the DAC or transporter with SoX
.
I was just surfing the web for good reading about this problem , hence
sharing with you . Appreciate even better reading about it if you help
out .
Reflections so far , my thoughts .
This is an understood and we'll known problem for at least a decade ( if
not more )yet it s consequences has not changed much in mastering or
consumer audio .
Master engineers still use outdated methods to evaluate peak levels (
not considering the true signal level after reconstruction ).
Equipment manufacturers and chipmakers does not seem to care much .
Happily making a bad problem worse .
Modern productions methods can easilly produce signals inside the
"system" that violates the nyquist requirements .
Perceptual coders are also suspectible in their filters mp3 and AAC OGG
and all are not at all happy with this .
(There is a reason for why Apples mastered for iTunes program suggest -1
dB and delivery as 24/96 before their conversion to 256kbs AAC .)
This would not be a problem if all where as carefull as some old school
sound engineers regarding levels .
And finally a Squeezebox thought , if you are using a DAC or in my case
a HT processor hence you untilise the digital output.
Would it be a good idea to just pull down the digital volume a little
bit prior to your DAC when listening to music with these "production
values" . I think this is sufficient ? Or is it not .
Or are more elaborate measures using SoX and custom-convert.conf really
needed ? Configuring SoX in this case can have some potential problems
done wrong you can run into the same problems as discussed in these
papers !
The simple undithered 24 bit volume in a squeezebox simply move the bits
down the "ladder" and they fall off at the end ( truncated ) harmless
with 16 bit , a theoretical issue with 24 bit sources where maybe you
should have used dither .
The Pro people usually always suggest dither when changing bit depth ,
but a small volume change in a squeezebox does not really reduce the
bits .
Any suggestions .
I will ofcourse I'll try at home in the weekend , the first thing a
16/44.1 input sees in my system is a sample rate converter to 24/96 then
probably conversion to float before entering the DSP processes .
--------------------------------------------------------------------
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
(in storage 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=102378