26

Re: Competent adjustment of loudness on a sound card

romanrex
What exactly is not measured? Dynamic range?
In electronics it is accepted to name dynamic range relation RMS of a maks-signal of the form a sine, not subject to a clip to level of stationary noise.  - SNR. And it does not concern numeral resolution. Analog circuits have congenital property to be linear, and to remain such even for signals below noise level. Numeral - demand higher digit capacity and application  at its reduction for saving of the given properties.
Here it (example)... If it is a question the friend the detail information is necessary more THAT it would be desirable to measure and AS.
This example does not show dynamic range (which it is measured much easier), internal resolution of mathematics of the chip - which even for 24 bits lies below schedule lower bound, and demands application of special signals for display (in particular not-60dB, and a range in low bit, and also frequency should be such that on the durable samplings of frequency of products of distortions did not departure, spreading the schedule). Good  show resolution more than 24 bits, even being "24 bit".
About  is that there is unique true algorithm, no them a lot of different under different tasks.
However there are correctness conditions. And "the different task" is only one -  noise a little (16) - bit record not heard. For all other applications modern (24) - TPDF (a range 2LSB) it is unique true and it - the elementary correct. Therefore as questions of audibility of noise, and accordingly noise shaping it is not necessary.
. And strictly speaking - noise shaping is not  in itself, and additional "forming" a spectrum of noise by back coupling and all uses inside the same TPDF.

27

Re: Competent adjustment of loudness on a sound card

igorzep :

Diter not "the theory from a ceiling". It is a pure mathematics in the first, and in the second - a subject for a long time and perfectly researched. Though some designers of chips of "quite known firms", for some reason are guided, as you noted -  nonsense, selecting from  coefficients from calculations.

I thank you, you very competently stated the question theory. And now it would be necessary to go down on the guilty earth and to face practice. And to recall a question .
In what position it is necessary to install a regulator of level of a signal on "the numeral side" taking into account nonoptimal coefficients in  in chips of a digital-to-analog coder and possible oversampling in driver ZK to cut possible artifacts from "numeral " and not to lose a dynamic range part on "the analog side". Times ENVY24 with its automatic machine of frequency of sampling left irrevocably. In modern DSP, applied in , oversampling of an entering flow is fulfilled in the majority of operation modes. And recommendations to play back a flow on its frequency with reference to  are not quite well-grounded.
Adding from 12/19/2016 00:29:
ikm :

About TI  - I like such behavior did not note, in CS4382 such behavior is watched on an one-acoustical signal.
To that I have been madly surprised, watching it on H-Faj Titanium.
It was necessary to mix an ultralow-frequency signal (hertz in 5) with a minimum level (-96 if I am not mistaken less,  is not able) as all became on the place.

If walk on Vegu, there there is an interesting subject . The effect "numeral ", under certain conditions, is possible in any modern chip of a digital-to-analog coder, especially at its usage in modern .

28

Re: Competent adjustment of loudness on a sound card

vitamir
In what position it is necessary to install a regulator of level of a signal on "the numeral side"
So it is answered already - in position implementing the maximum dynamic range. I.e. 0 if it is not restricted by any factors of the subsequent analog part, as, for example, the fixed structure .
Taking into account nonoptimal coefficients in  in digital-to-analog coder chips
They influence only the lower part of dynamic range.  on top - it is not necessary to be afraid. On any __ muses-records those 1-2 LSB "clips" will not be shown, even if they somewhere and exist - to hear them there are no chances.
In modern DSP, applied in 
All superfluous "DSP", is not clear that doing in OFF. Oversampling in itself does not cause a clip. If oversampling qualitative - it is possible to leave, but is better for this purpose to use checked means.

29

Re: Competent adjustment of loudness on a sound card

ikm
To That I have been madly surprised, watching it on H-Faj Titanium.
In was not  Creative X-Fi Xtreme Gamer Fatal1ty Professional Series (the same CS4382) on 0dBFS, and for all modes.
http://reference-audio-analyzer.pro/rep … ode-44.php
Adding from 12/19/2016 00:57:
igorzep
In electronics it is accepted to name Dynamic range relation RMS of a maks-signal of the form a sine, not subject to a clip to level of stationary noise.  - SNR.
This determination suits only amplifiers and is not applicable to a digital-to-analog coder. For a digital-to-analog coder is a relation of a maximum to distortions on  a signal, in particular to a sine with amplitude-60 dBFS.  here generally any side.
This example does not show dynamic range (which it is measured much easier),
It is necessary to study as well as why the dynamic range is measured at-60dBFS and why for a digital-to-analog coder other standard technique worldwide, rather than a technique for amplifiers. The information truth already in English since our GOST chronically lagged behind with "transfer".
Good  show resolution more than 24 bits, even being "24 bit".
I will repeat a question WHAT it would be desirable to measure and HOW? "Digit capacity" or  "resolution"?
And "the different task" is only one -  noise a little (16) - bit record not heard.
The driver sound or a loudness regulator in a digital-to-analog coder already the automatic machine translate a flow in 24 or 32 bits. A store on volume control for 16 bits in 24 bits - 48 . For 16 bits in 32 bits - 96 .
Adding from 12/19/2016 00:59:
vitamir
Times ENVY24 with its automatic machine of frequency of sampling left irrevocably. In modern DSP, applied in , oversampling of an entering flow is fulfilled in the majority of operation modes. And recommendations to play back a flow on its frequency with reference to  are not quite well-grounded.
In the modern OS and ENVY24 does not rescue but if to use WASAPI the automatic machine works. Though at Creative in Z series remained artful SRC in the driver.

30

Re: Competent adjustment of loudness on a sound card

romanrex
This determination suits only amplifiers and is not applicable to a digital-to-analog coder.
It is not necessary to invent "special" determinations for the same term! It always means same.
For a digital-to-analog coder...
Imaginations from illiteracy. Given "determination" is senseless enough. As indefinitely and the result "floats".
Dizering here generally any side.
ALWAYS sideways to numeral signals.
It is necessary to study as well as why the dynamic range is measured at-60dBFS and why for a digital-to-analog coder other standard technique worldwide, rather than a technique for amplifiers.
This "standard" already  lagged behind years life! dithered the numeral system submits to all analog "rules". And if does not submit is  numeral system. Really it is necessary to drag till now that delirium from 80 in each discussion? And-60dBFS - it is far not a "small" signal for 24 bit systems with  more 140 that it however  authentically to test their resolution.
The Store on volume control for 16 bits in 24 bits - 48 .
Only at multiple to levels to a twain. It is the extremely difficult to me to name a regulator with step in 6 "a loudness regulator". However, at calculations with the big digit capacity (32 bits for only volume control - are quite normal) and the further application correct  at digit capacity reduction - problems are not present generally as a class.

31

Re: Competent adjustment of loudness on a sound card

igorzep
Imaginations from illiteracy.
Correctly I understand, what AP does dynamic range measurement incorrectly?
https://www.ap.com/analyzers-accessories/2700-series/
Also what vendors of a digital-to-analog coder the round idiots using AP for obtaining of passport characteristics in ? It is necessary to measure dynamic range one in one as a signal/noise?
And-60dBFS - it is far not a "small" signal for 24 bit systems with  more 140 that it however  authentically to test their resolution.
Try to formulate that such "resolution" that it would not be same fuzzy and ostensibly clear parameter as "". And as it is connected to digit capacity and dynamic range.
For some reason as speech that it is necessary to measure and as, so any answer comes. Only that all measure all incorrectly.:D

32

Re: Competent adjustment of loudness on a sound card

romanrex
Correctly I understand, what AP does dynamic range measurement incorrectly?
I do not know, from what you took that AP does it how you think...
Also that vendors of a digital-to-analog coder round idiots
We follow as an example two  ...
BurrBrown/TI:
http://www.6moons.com/audioreviews/weiss4/pcm1792.pdf
SNR and DR coincide in accuracy.
AKM:
http://www.akm.com/akm/en/product/datas … o=AK4497EQ
And these even in single line wrote down.
Try to formulate that such "resolution"
As well as everywhere define resolution - the relation maximum to the least statistically to a perceptible signal. Both set forth above are quite definitely capable to transmit signals much less, than RMS their own noise. wink
For some reason as speech that it is necessary to measure and as, so any answer comes.
Because you it is banal ignore answers. Doing not wish to see, yes does not see. However recently here it is fashionable.

33

Re: Competent adjustment of loudness on a sound card

igorzep
SNR and DR coincide in accuracy.
They coincide only in case of ideally made , and in a value reality will be different. The data methodically on a miscellaneous turns out.
As well as everywhere define resolution - the relation maximum to the least statistically to a perceptible signal.
It is described as  to a sine, the meander to a meander Is, etc. calculated on peak, to average, amplitudes of separate harmonics or still somehow? How it was specific is measured? From the general dim concepts to a reality.
Or it nevertheless is exceptional subjective  parameter? Any relation to measured values?

34

Re: Competent adjustment of loudness on a sound card

romanrex :
ikm
To That I have been madly surprised, watching it on H-Faj Titanium.
In was not  Creative X-Fi Xtreme Gamer Fatal1ty Professional Series (the same CS4382) on 0dBFS, and for all modes.
http://reference-audio-analyzer.pro/rep … ode-44.php

the Novel, recall as the spectrum  a signal looks smile
[img=620x369, 90.6Kb] http://reference-audio-analyzer.pro/rep … 0Series%20 (Creation%20Mode%2044) _Center_3_No%20load_-_-_150_20-22k_lin_amp_thd.png [/img]
And
http://m.eet.com/media/1064549/sound7_fig1.47.jpg [off] (966x573, 32.2Kb) [/off]
Long tail from odd harmonics.
Reduce an output literally by pair , and all becomes strong .