I am glad that the web page linked by Scott seems to confirm my "ridiculous" idea, and explains it better than I could.
Always good to know that I can arrive at something close to the known truth simply by thinking; it encourages me that I can still sometimes arrive at previously unknown truth by thinking too.
Always good to know that I can arrive at something close to the known truth simply by thinking; it encourages me that I can still sometimes arrive at previously unknown truth by thinking too.
This is quite good about scales. Scales: Just vs Equal Temperament
I wasn't sure what scale cbdb was referring to earlier, but all keys, major or minor, in the equal temperament scale have the same ratio between notes and sound the same.
I wasn't sure what scale cbdb was referring to earlier, but all keys, major or minor, in the equal temperament scale have the same ratio between notes and sound the same.
Master oscillator jitter is critical, same as any other DAC.
It is the nature of the clock jitter that makes (or breaks) the 'sound' of electronic instruments and can be acceptable through to very 'unnatural' sounding.
Tweaking the master clock stage in electronic instruments can bring very nice sounding dividend.
Dan.
If you look at a stringed instrument (such as a piano) in the hands of a skilled tuner you will find the 'best' subjective results occur when the octaves are 'stretched' meaning the treble is tuned progressively sharp and the bass correspondingly slightly flat. If this isn't done then the instrument doesn't sound 'right' even though analysis of the fundamentals shows it might be 'spot on'. The middle octaves are tuned pretty much 'true'. The difference in the extreme treble and bass can be up to 30 or more cents. Given that there are 100 cents per semitone that is quite a deviation from the theoretical values predicated.
The reason this has to be done is because the harmonics of a mechanical vibrating string can be higher that theory predicts.
The reason this has to be done is because the harmonics of a mechanical vibrating string can be higher that theory predicts.
The numbers from the actual physics of the instruments are orders of magnitude more than a typical crystal.
I did not get Dan's input on master clock jitter when we speak about real instruments and their frequency fluctuations. Modern DACs are close to perfect re jitter. Non issue.
I asked a local piano tuner that i know about the "flat/sharp"thing and i found his comments interesting!
according to him the primary reason to go ever so slightly flat with bass notes is it is it damps the harmonics makes it less "bright" and the note fades faster so that with attack(dynamics) the pitch character changes with playing technique, and we can't neglect that there's a different number of strings,with middle notes you have three strings per note and must be tuned more precisely in relation to one another, with higher notes going ever so slightly sharp increases the decay of the note because of dissonace it increases perceived loudness
according to him the primary reason to go ever so slightly flat with bass notes is it is it damps the harmonics makes it less "bright" and the note fades faster so that with attack(dynamics) the pitch character changes with playing technique, and we can't neglect that there's a different number of strings,with middle notes you have three strings per note and must be tuned more precisely in relation to one another, with higher notes going ever so slightly sharp increases the decay of the note because of dissonace it increases perceived loudness
Last edited:
There is probably more than one explanation for why this needs to be done. The ear finds any treble notes that are flat very objectionable, and the opposite occurs in the bass, any sharpness sounds awful. That's probably to simplistic and avoids technical reasons why it is so... just that it is.
I've tuned my piano on more than one occasion using both intervals and 'Tunelab' software. There is an ancient (really ancient) book available on the web for free download called 'Piano Tuning' by J Cree Fisher. Its well over 100 years old now but still absolutely relevant today in both technique and approach.
I found this which you might find interesting:
The Equal Tempered Scale and Peculiarities of Piano Tuning
I've tuned my piano on more than one occasion using both intervals and 'Tunelab' software. There is an ancient (really ancient) book available on the web for free download called 'Piano Tuning' by J Cree Fisher. Its well over 100 years old now but still absolutely relevant today in both technique and approach.
I found this which you might find interesting:
The Equal Tempered Scale and Peculiarities of Piano Tuning
A test "byproduct" - recorder flute playing E5 note. I have processed it by the tested amp distortion profile - there is no chance to hear the difference. In analysis, there is a slight change in amplitude of the 2nd harmonic and even slighter of higher even harmonics. The difference file avg level is -47dB below avg level of the original tone. No chance to hear a difference.
Attachments
if you feel there's no chance, why not make it a test? or is the thought that someone may actually have the ability to reliably hear a difference too frightening?
admittedly it would take a large number of participants to show that the previously accepted limits of thresholds of audibility of distortion are to be questioned.
admittedly it would take a large number of participants to show that the previously accepted limits of thresholds of audibility of distortion are to be questioned.
Oh no, no frightening 🙂. I can post it without any problem. However, it would be just only another sample so no speaking about known thresholds. And the known thresholds are valid for pure sine tones, that are without any doubt most sensitive to distortion perception. I do not think there are any defined thresholds of audible distortion levels or profiles for music. That's why we are testing it, and we have only indications yet and not a single indisputably positive test result, in this current test.
This is the sine test with the same distortion profile as the hybrid amp has.
http://pmacura.cz/sines.zip
Everyone should hear a difference in ABX and get result 16/16 from every run of the test. The only condition is volume level used, it should not be high (ear distortion) and it should not be too low (sensitivity).
So we can see that the disrortion profile of this amp is quite easily audible on a sine tone. Not so easy on a music sample. One should get 16/16, again, to be almost absolutely sure.
http://pmacura.cz/sines.zip
Everyone should hear a difference in ABX and get result 16/16 from every run of the test. The only condition is volume level used, it should not be high (ear distortion) and it should not be too low (sensitivity).
Code:
foo_abx 2.0.2 report
foobar2000 v1.3.7
2018-12-03 08:38:50
File A: sig1.wav
SHA1: 621ecd158928499510d523dc5a85abd019cb1d44
File B: siga.wav
SHA1: 640783db9340fde87db66db7127031862c1d8814
Used DSPs:
Resampler (PPHS)
Output:
WASAPI (event) : OUT (DUO-CAPTURE EX), 24-bit
Crossfading: NO
08:38:50 : Test started.
08:39:08 : 01/01
08:39:12 : 02/02
08:39:16 : 03/03
08:39:19 : 04/04
08:39:23 : 05/05
08:39:26 : 06/06
08:39:29 : 07/07
08:39:32 : 08/08
08:39:35 : 09/09
08:39:38 : 10/10
08:39:41 : 11/11
08:39:44 : 12/12
08:39:47 : 13/13
08:39:51 : 14/14
08:39:54 : 15/15
08:39:59 : 16/16
08:39:59 : Test finished.
----------
Total: 16/16
Probability that you were guessing: 0.0%
-- signature --
69c3c3490031763834ba3e8d57de70945008d943So we can see that the disrortion profile of this amp is quite easily audible on a sine tone. Not so easy on a music sample. One should get 16/16, again, to be almost absolutely sure.
