Skip to main content ↑ ↓

Notice

Please note that most of the software linked on this forum is likely to be safe to use. If you are unsure, feel free to ask in the relevant topics, or send a private message to an administrator or moderator. To help curb the problems of false positives, or in the event that you do find actual malware, you can contribute through the article linked here.
Topic: Impulse ringing embedded on cd (Read 12029 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

Impulse ringing embedded on cd

Reply #25 –
Quote from: Wombat on
I heard theories that the created ringing already does harm the way the DAC works internal no matter how it filters.

What are those 'theories'?

Ringing is not really created. We don't add anything, we actually "take away".


Btw, iirc they did the test with their own test software implemented in matlab.
"I hear it when I see it."

Impulse ringing embedded on cd

Reply #26 –
Quote from: xnor on
Quote from: Wombat on
I heard theories that the created ringing already does harm the way the DAC works internal no matter how it filters.

What are those 'theories'?

Ringing is not really created. We don't add anything, we actually "take away".


Btw, iirc they did the test with their own test software implemented in matlab.

There are endless reports of people descriping lower ringing with not so steep filters at resample stage create better sound. Even the Weiss resampler does everything slow and i guess they know what they do.
I also ask the veterans here. Wasn't it an early Philips DAC that had a problem with ringing and distortion and this established the whole paranoia?
Is troll-adiposity coming from feederism?
With 24bit music you can listen to silence much louder!

Impulse ringing embedded on cd

Reply #27 –
That's anecdotal at best, not theories.
"I hear it when I see it."

Impulse ringing embedded on cd

Reply #28 –
1) Anybody can theorize, including people who don't have the slightest clue as to what they're talking about.

2) Without getting too far into the other topic, does anyone recall exactly where in the (recording -> production -> distribution -> playback) chain is said "typical filter" supposed to exist?

Impulse ringing embedded on cd

Reply #29 –
Sharp filter at resampling means full scale at the DACs filter frequency, filter at resampling below the DACs filter frequency means less but lower frequency ringing. Theory enough?
Is troll-adiposity coming from feederism?
With 24bit music you can listen to silence much louder!

Impulse ringing embedded on cd

Reply #30 –
Quote from: Wombat on
Sharp filter at resampling means full scale at the DACs filter frequency, filter at resampling below the DACs filter frequency means less but lower frequency ringing. Theory enough?

I don't understand the first part.

I prefer the filter (anti-imaging or resampling) during playback to dominate.

"I hear it when I see it."

Impulse ringing embedded on cd

Reply #31 –
Full scale means full amplitude, content so the DACs ringing is stronger as with a lower already filtered amplitude.
I see all kind of filters used in the musuc i own. I doubt there is the one and only answer.
Is troll-adiposity coming from feederism?
With 24bit music you can listen to silence much louder!

Impulse ringing embedded on cd

Reply #32 –
The "amplitude" at the cutoff frequency is the same for both a steep and less steep filter. Even the cutoff frequency can be the same. Now if you refer to the length of ringing... sure, steeper means longer. But that is no theory regarding audibility or "sounding better".

Also, it completely ignores what is happening during playback.
"I hear it when I see it."

Impulse ringing embedded on cd

Reply #33 –
Quote from: xnor on
The "amplitude" at the cutoff frequency is the same for both a steep and less steep filter. Now if you refer to the length of ringing... sure, steeper means longer. But that is no theory regarding audibility or "sounding better".

Also, it completely ignores what is happening during playback.

Still a misunderstanding. I talk about playback all the time. The amplitude is not the same with a resampled file using a filter at 20kHz as with a resampled file with full content up to 22kHz. When the DACs playback filter is at 21kHz the ringing will be different at 21kHz.

Edit: You suggest for downsampling we should use a very steep filter as near to Nyquist as possible?
Is troll-adiposity coming from feederism?
With 24bit music you can listen to silence much louder!

Impulse ringing embedded on cd

Reply #34 –
You initially talked about creating 44.1 kHz files... filtered such that "the DAC has not so much influence anymore".

The cutoff frequency for these filters is typically defined as -6.02 dB point, so by definition the amplitude is the same at that frequency. A typical filter for 44.1 kHz is one where the cutoff frequency is 22.05 kHz. What's variable is the steepness (and therefore length of the filter's impulse response or ringing).

On creation this will obviously cause aliasing, during playback it will allow some imaging.

For creation I'd suggest rather a steeper filter with higher cutoff frequency than a slow one with lower cutoff that would roll-off high frequencies, because I might use such a filter on playback which would then result in twice the attenuation of high frequencies and ringing which could potentially be audible.
"I hear it when I see it."

Impulse ringing embedded on cd

Reply #35 –
One simple attempt. The above pic is done with SoX 99.7% from 24/96 to 44.1 and the below is done with 92%. To simulate a filter applied at playback i upsampled both with 96%. That is a valid value for a recent DAC i suppose.
The distribution of ringing energy is very different.
Is troll-adiposity coming from feederism?
With 24bit music you can listen to silence much louder!

Impulse ringing embedded on cd

Reply #36 –
Well, of course, you undercut the final filter by several % before.

When doing this experiment with sox the problem is that you do not control the cutoff frequency and steepness. You control some passband % (which is not clearly defined afaict) and the resulting filter is somewhat unpredictable.
"I hear it when I see it."

Impulse ringing embedded on cd

Reply #37 –
At least the pic shows what can happen with embedded ringing from cd at playback.
Now you vote for keeping all content near to Nyquist and it does not matter how steep. Me votes  for above 20kHz doesn't matter and i prefer filtering myself not steep.
I really want to know what others think and why on the famous SRC comparison page many professional solutions prefer no steep, not maximum bandwith solutions? I don't think swiss Weiss engeneering has no clue how to calculate a steep filter.
Don't count some steep solutions offered, the defaults are often much unsteeper.
Is troll-adiposity coming from feederism?
With 24bit music you can listen to silence much louder!

Impulse ringing embedded on cd

Reply #38 –
What is not steep? Most solution I've seen do not roll off before 21 kHz, have a cutoff frequency of 22.05 kHz and therefore allow some aliasing (such that significant energy will not fold down to below 21 kHz).

Now on playback you see all kind of crazy filters. From some people using steep SoX filters with cutoff below 22.05 kHz and reaching the stopband before that (no imaging) to some DACs with min phase filters that already have significant roll off at 20 kHz (allowing lots of imaging).

Both cases would significantly attenuate the ringing on the recording.
"I hear it when I see it."

Impulse ringing embedded on cd

Reply #39 –
The fact alone they allow some imaging means it could have been done steeper.
Anyone else?
Is troll-adiposity coming from feederism?
With 24bit music you can listen to silence much louder!

Impulse ringing embedded on cd

Reply #40 –
I don't see that it makes any difference if it can't be heard and isn't doing any harm.
If you're in a position to appreciate or make use of frequency content at 22kHz in non-placebo way then that's another matter.

Impulse ringing embedded on cd

Reply #41 –
Quote from: Wombat on
The fact alone they allow some imaging means it could have been done steeper.


Given that cutoff frequency (22.05 kHz) there is always some aliasing even if it is a tiny amount unless your filter is infinitely steep.


"I hear it when I see it."