>15 kHz? (+Master filtering)

Have an urge to learn, or a calling to teach? Want to share some useful Youtube videos? Do it here!
User avatar
selig
Moderator
Posts: 7998
Joined: 15 Jan 2015

Post 27 Sep 2019

Chizmata wrote:
27 Sep 2019
if you lose your hearing of high frequencies, you can still hear the indirect effects of those frequencies on the stuff you can still hear. just like sounds above 20khz leave their traces in the audible region for everyone because its still physics, no matter if you hear it or not.
I got that from your first post - what I'm unclear on is what you are saying is happening to cause this? Can you give an example, or post a link, that describes the effect you are mentioning here?

For example, what traces are left in the audible region by a sound that is inaudible? I've not found any evidence that sounds above our range of hearing have any effect on sounds we can hear. That was the whole argument for why we needed to have much higher sample rates. If true, then ALL audio gear is messing with what we hear, since ALL audio gear has some limit to how high a frequency they can reproduce.

I'm asking with an open mind, even though in the past I've not found evidence of what you suggest. I just want to be sure I'm not missing something, which happens! ;)

[this is also why I'm asking about the reasoning behind cutting the high frequencies…]
Selig Audio, LLC

User avatar
Chizmata
Posts: 432
Joined: 21 Dec 2015

Post 27 Sep 2019

selig wrote:
27 Sep 2019
Chizmata wrote:
27 Sep 2019
if you lose your hearing of high frequencies, you can still hear the indirect effects of those frequencies on the stuff you can still hear. just like sounds above 20khz leave their traces in the audible region for everyone because its still physics, no matter if you hear it or not.
I got that from your first post - what I'm unclear on is what you are saying is happening to cause this? Can you give an example, or post a link, that describes the effect you are mentioning here?

For example, what traces are left in the audible region by a sound that is inaudible? I've not found any evidence that sounds above our range of hearing have any effect on sounds we can hear. That was the whole argument for why we needed to have much higher sample rates. If true, then ALL audio gear is messing with what we hear, since ALL audio gear has some limit to how high a frequency they can reproduce.

I'm asking with an open mind, even though in the past I've not found evidence of what you suggest. I just want to be sure I'm not missing something, which happens! ;)

[this is also why I'm asking about the reasoning behind cutting the high frequencies…]
quite hard to actually EXPLAIN the physics just from the top of my head, but what ive read is that using 48khz (and higher) sampling rates has something to do with the "nyquist frequency" and preventing aliasing (maybe i was wrong about the phasing in my first post?). From what i understand cutoff above 20khz is done for the same reason, but a higher sampling rate is better, because the cutoff introduces other artifacts.

https://en.wikipedia.org/wiki/Nyquist_frequency

hope that helps.

User avatar
selig
Moderator
Posts: 7998
Joined: 15 Jan 2015

Post 27 Sep 2019

Chizmata wrote:
27 Sep 2019
selig wrote:
27 Sep 2019


I got that from your first post - what I'm unclear on is what you are saying is happening to cause this? Can you give an example, or post a link, that describes the effect you are mentioning here?

For example, what traces are left in the audible region by a sound that is inaudible? I've not found any evidence that sounds above our range of hearing have any effect on sounds we can hear. That was the whole argument for why we needed to have much higher sample rates. If true, then ALL audio gear is messing with what we hear, since ALL audio gear has some limit to how high a frequency they can reproduce.

I'm asking with an open mind, even though in the past I've not found evidence of what you suggest. I just want to be sure I'm not missing something, which happens! ;)

[this is also why I'm asking about the reasoning behind cutting the high frequencies…]
quite hard to actually EXPLAIN the physics just from the top of my head, but what ive read is that using 48khz (and higher) sampling rates has something to do with the "nyquist frequency" and preventing aliasing (maybe i was wrong about the phasing in my first post?). From what i understand cutoff above 20khz is done for the same reason, but a higher sampling rate is better, because the cutoff introduces other artifacts.

https://en.wikipedia.org/wiki/Nyquist_frequency

hope that helps.
OK, that's different - that's not high frequencies affecting lower frequencies. And as a rule, higher sample rates do not ALWAYS sound better than lower rates…

What you're explaining is how a low quality filter can adversely affect the frequencies around nyquist. And what you're talking about has more to do with the A/D which has to use analog filters, rather than the D/A which can use digital filters. So it affects the recording of analog signal into a digital system, rather than the results of working in an all digital workflow as many folks do when using Reason.

What I've heard and experienced over the years of the transition from analog recording to digital, is that with cheep interfaces you'll notice a difference in sample rates. But with higher quality interfaces, the difference goes away. And that data was from the early days of convertor design. These days even the cheapest converters do an amazing job at all sample rates. So what you've heard in the past may or may not apply to the present (always something to consider).

So why do some still claim higher sample rates sound better? Because is SOME cases, the CAN. These are very case specific, and have to do with how a digital processor such as a compressor, or even a synth, handle audio and issue such as aliasing. Simply put, a poor design functions better at a higher rate because there's less work for a filter that has 1-4 octaves to work with than when it has less than an octave to work with, which require steeper slopes. Some of the better designs sound identical (or nearly so) at all sample rates - as always, your melange may vary…

But there's a catch. There's always a catch. The perceived advantage may be so small compared to the "cost" in CPU hit and file size. You certainly don't get twice as much "quality" in most cases when you double sample rate (which doubles CPU hit and file size). And if you're going to deliver at a lower rate in the end, any potential advantages can be wiped out in the end.
Selig Audio, LLC

User avatar
Kalm
Posts: 258
Joined: 03 Jun 2016
Location: Austin

Post 27 Sep 2019

I only high pass in mastering due to DC offset, low pass mastering if a client sends me to much hiss, white noise, and air. Other than that there's no reason to high pass and low pass if its taken care of in the mix. Doing it twice won't do you any good. If your air and sub bands add up they should be taken care of with busses, not influence every sound in the mix + some phase shift across the summed mix
Courtesy of The Brew | Mac Mini Intel i7 Quad-Core | 8 GB RAM | Apple 1 TB & Samsung 850 EVO 250 GB | Reason 10 | Studio One 3 Professional | Ozone 7 | Pro Tools 12 | Adobe CC | Presonus FireStudio Project | M-Audio Axiom | And many plugs

RobC
Posts: 1161
Joined: 10 Mar 2018

Post 29 Sep 2019

Kalm wrote:
27 Sep 2019
I only high pass in mastering due to DC offset, low pass mastering if a client sends me to much hiss, white noise, and air. Other than that there's no reason to high pass and low pass if its taken care of in the mix. Doing it twice won't do you any good. If your air and sub bands add up they should be taken care of with busses, not influence every sound in the mix + some phase shift across the summed mix
Agreed on the 20 Hz cut in case of DC offset. Very useful in case of wave-shaping effects, too - especially rectifying.

I'll try to avoid high cuts as much as I can. Truth be told, I want to cut as little as possible.

If everything goes well, I should be able to be able to render a clean mix as a final one, which needs no mastering. And as long as my music is between -16 LUFS and -23 LUFS, I'm more than cool with it.

User avatar
Kalm
Posts: 258
Joined: 03 Jun 2016
Location: Austin

Post 29 Sep 2019

RobC wrote:
29 Sep 2019
Kalm wrote:
27 Sep 2019
I only high pass in mastering due to DC offset, low pass mastering if a client sends me to much hiss, white noise, and air. Other than that there's no reason to high pass and low pass if its taken care of in the mix. Doing it twice won't do you any good. If your air and sub bands add up they should be taken care of with busses, not influence every sound in the mix + some phase shift across the summed mix
Agreed on the 20 Hz cut in case of DC offset. Very useful in case of wave-shaping effects, too - especially rectifying.

I'll try to avoid high cuts as much as I can. Truth be told, I want to cut as little as possible.

If everything goes well, I should be able to be able to render a clean mix as a final one, which needs no mastering. And as long as my music is between -16 LUFS and -23 LUFS, I'm more than cool with it.
Yup, you'll be surprised though what plugins can do in your chain. I've had instances when simply adding an L2 on a mix caused DC Offset :shock: this one im still trying to test out why.
Courtesy of The Brew | Mac Mini Intel i7 Quad-Core | 8 GB RAM | Apple 1 TB & Samsung 850 EVO 250 GB | Reason 10 | Studio One 3 Professional | Ozone 7 | Pro Tools 12 | Adobe CC | Presonus FireStudio Project | M-Audio Axiom | And many plugs

exxx
Posts: 84
Joined: 12 Sep 2016

Post 30 Sep 2019

RobC wrote:
23 Sep 2019
Be it mixing, or mastering, doesn't matter. General audio engineering.
I think there are some problems with Reason's high frequency processing.

If you're feeling uncomfortable with that part of your job, read this thread.

It may be my own idea, but I think it will help.

viewtopic.php?f=4&t=7514201

RobC
Posts: 1161
Joined: 10 Mar 2018

Post 30 Sep 2019

Kalm wrote:
29 Sep 2019
Yup, you'll be surprised though what plugins can do in your chain. I've had instances when simply adding an L2 on a mix caused DC Offset :shock: this one im still trying to test out why.
I think I experienced something like that with NN-XT, but not sure if the cause was that 'high quality interpolation' was enabled.
exxx wrote:
30 Sep 2019

I think there are some problems with Reason's high frequency processing.

If you're feeling uncomfortable with that part of your job, read this thread.

It may be my own idea, but I think it will help.

viewtopic.php?f=4&t=7514201
I saw that one yesterday ~ what do you use for analyzing?

User avatar
selig
Moderator
Posts: 7998
Joined: 15 Jan 2015

Post 30 Sep 2019

exxx wrote:
30 Sep 2019
RobC wrote:
23 Sep 2019
Be it mixing, or mastering, doesn't matter. General audio engineering.
I think there are some problems with Reason's high frequency processing.

If you're feeling uncomfortable with that part of your job, read this thread.

It may be my own idea, but I think it will help.

viewtopic.php?f=4&t=7514201
I don't think that explains anything about high frequency response. It shows how the automation system affects a pure sine wave. And it shows that the issues are worse on many more popular DAWs, the ones that have made much of the music we enjoy.
Selig Audio, LLC

exxx
Posts: 84
Joined: 12 Sep 2016

Post 01 Oct 2019

selig wrote:
30 Sep 2019
exxx wrote:
30 Sep 2019


I think there are some problems with Reason's high frequency processing.

If you're feeling uncomfortable with that part of your job, read this thread.

It may be my own idea, but I think it will help.

viewtopic.php?f=4&t=7514201
I don't think that explains anything about high frequency response. It shows how the automation system affects a pure sine wave. And it shows that the issues are worse on many more popular DAWs, the ones that have made much of the music we enjoy.
Using another daw. Call up the instrument of reason. Play the above note c5. No automation is required. Play the chords or move the keys. It just sounds different. More clean. You may not hear the difference, but there is a difference.

User avatar
diminished
Posts: 1306
Joined: 15 Dec 2018

Post 01 Oct 2019

exxx wrote:
01 Oct 2019
It just sounds different. More clean. You may not hear the difference,
Yeah, uhm.
:reason: Most recent track: The Test (feat. MrFigg) || Others: on my YouTube channel •ᴗ•

User avatar
selig
Moderator
Posts: 7998
Joined: 15 Jan 2015

Post 01 Oct 2019

exxx wrote:
01 Oct 2019
selig wrote:
30 Sep 2019


I don't think that explains anything about high frequency response. It shows how the automation system affects a pure sine wave. And it shows that the issues are worse on many more popular DAWs, the ones that have made much of the music we enjoy.
Using another daw. Call up the instrument of reason. Play the above note c5. No automation is required. Play the chords or move the keys. It just sounds different. More clean. You may not hear the difference, but there is a difference.
I've done that test more times than I care to remember, as have dozens of other folks. No difference.

Now, what YOU may be talking about is how a synthesizer CAN alias at very high frequencies - the kind of frequencies you are likely never to use in actual music, so any affect you hear in those extreme cases will be far less in most applications. And for one, Subtractor can be forced to exhibit what you're talking about.

But you're mixing issues here. It's not the audio "engine" of Reason causing this, it's only the instrument - switch to another instrument and it will sound different.

Again, I beg to differ about the raw "sound of Reason" being any different than any other DAW. I've done the comparisons countless times over the decades, the last one as recently as a year ago or so (search this forum for the sordid details), comparing not only the sound of one channel in Reason and Logic, but also the summing of 24 channels. Results proved "no difference whatsoever". Others performed the same tests on other DAWs with the same results.

If you want to post the test you have done that prove otherwise, feel free to do so. But that poor horse is a skeleton now… ;)
Selig Audio, LLC

  • Information
  • Who is online

    Users browsing this forum: CommonCrawl [Bot] and 0 guests