Hi Andreas,

No I did not change anything between both takes except for the samplig rate. I am monitoring by means of German Maestro GMP400 headphones and cannot hear what you are hearing. My recording of this piece is already on Pianosociety but I am no longer happy with my performance. I would like to improve it. Still a long way to do so...

Hi Didier,
I suppose to be able to hear differences between 44, 96 and 192 Khz-recordings has very much to do with the speakers. All components are "falsifying" a bit the transients, but the speakers do most of it. So, if you have speakers, which are able to reproduce the transients correctly you have more original respective natural sound (see the wave-schemes in the pdf-file on the Myro-page, please). And may be this is a reason more to hear a difference between the samplerates, but I´m not really sure. Most of the headphones and speakers today are not able to reproduce the transients correctly.
Anyway, I´m attending your improved version of that nice piece with curiosity!

Dear Andreas,

you were comparing two 44 kHz converters, one being your 44 converter, the other one being a combination of your 192 converter and sample rate conversion (SRC). Some professional analog-to-digital converter (ADC) work like that and it is very common in digital-to-analog conversion (DAC): most hi-fi CD player have such a converter. Your result prove that your converter works better at 192 than at 44, not that 192 is intrinsically better than 44. So the right test would be to compare 192 ADC listened to by means of 192 DAC with 192 ADC + 192-to-44 SRC + 44-to-192 SRC listened to by means of 192 DAC.

Dear Didier,
thank you for your reply.

Didier wrote:


O.k., I think, I have got this. So I have to compare a recording originally recorded with 44 Khz with one originally recorded with 192 Khz, isn´t it? I also have done that several times and everytime I found out, that the 192 Khz-recordings are of a better quality than the 44 Khz-recordings.




I disagree, my friend: the right test would be to compare 192 ADC listened to by means of 192 ADC with 192 ADC + 192-to-44 SRC + 44-to-192 SRC listened to by means of 192 DAC listened to by means of 192 ADC with 192 ADC + 192-to-44 SRC + 44-to-192 SRC listened to by means of 192 DAC, isn´t it?

No Andreas you did not get my point. If the question of interest is whether 192 kHz sampling performed by an ideal converter is better, in general, than 44 kHz sampling performed also by an ideal converter, then you cannot compare a recording done by means of your ADC at 44 with a recording done by means of your ADC at 192 since your test proved to your ears that your 44 ADC is not as good as the 44 ADC that you get by combining your 192 ADC with 192-->44 SRC.

So your reference 44 ADC should be this latter one. How to compare it with a 192 ADC ? Of course you must use the same DAC performing at the same rate. If you would use your 44 DAC, you would have to convert the 192 kHz digital output of your 192 ADC in a 44 kHz digital signal and would get exactly the digital signal from your reference 44 kHz ADC... That's why I suggested that you should use rather your 192 DAC. But 88 or 96 DAC would be fine also.

Your converter may be far worse at 192 kHz than this one at 44 kHz, which costs more than 7000 € and the designer of which is famous in the pro audio community for his claim that sampling above 96 kHz does not provide any audible benefit, a statement agreed by many, may be most, professionals today.

This issue is not so important. The most important is that you have found the setting from which you can get the best performance from your recorder. I just wanted to stress the point that one cannot derive from such an experience a statement as general as '192 is better than 44', despite that there is no doubt that it is larger. Back to the piano ?

Hi Didier,
thank you so much for your detailed help and explanations. I really appreciate that very much! You are the professional audio engineer here and I´m only an unknowing layman compared with your great knowledge concerning recording-technique. So, I want to try my best here (as you do at the piano with great success, btw)!



O.k., I think, I have got that know, really. To find out in general, if 44 or 192 Khz is the best I should test the two best converters of the world in each samplerate-category, isn´t it?


I have a question here. If you talk about a 44 ADC, 192 ADC and so on, do you mean two different devices or do you mean one device? I have only one device: a sound-card by M-Audio (Audiophile 192), which I can set to different samplerates for recording like 44, 96 and 192 Khz.
In that sense I don´t hear a difference between 192 ADC-->44 SRC and a 44 ADC recording taken with my sound-card. But I hear a difference between a 44 ADC, 96 and 192 ADC recording. The 192 ADC has the best quality, so I suppose that to be the best setting for my device. The 96 is also very good, but has a bit less of the "ambience" and "soul" respective "atmosphere" of the sound.



Nope, I would like to learn as much as possible about these things.

In the land of the blind the one-eyed man is king.
We are a few one-eyed men here. I don't want to be the king.


It would be a way for minimizing as far as possible the impact of the hardware so that to get a result where the impact of the data sampling rate is maximized. But why would you do this ? A question of more practical interest is how to choose between two AD sampling rates available from your ADC. You need obviously two recordings of the same piece made at both AD rates. Then you should determine for each recording what is the best DAC and sampling rate available to you for listening to it. It would be likely the same for both. Then you could decide whether one of both such listenings is better. This latter comparison should be a blind test.

Well, this is a very cumbersome process... I prefer to consider that the fact there are a lot of piano CDs that sound great, much better than my own recordings, is a proof that I don't need more than 44 AD samplig rate. A great room would be much more efficient.

Hmm. Any chance this thread can be continued in a new thread not titled: "Bach: Prelude & Fugue e-minor, WTC II, BWV 879"
It will also make this technical subject much easier to find later when someone (like me) needs it.

Didier wrote:


Of course, you are right, and I think, for my hard ware it´s always 192 Khz. Thank you for your useful tips!

Hi Andreas,

I was quite impressed by your playing of this wonderful Bach P & F. Your articulation was clean and precise, and I loved how when the melody in the fugue shifted into the left hand, you highlighted it so well. Everything in your playing was thoughtful and always most musical. Some of the ornamentation sounds difficult to do, but you executed it all in a natural way that always fit and blended well into the musical line. You show a thorough musicianship in this music. Excellent playing!

David

Hallo Andreas, I enjoyed your performance of this set very much. I think the prelude in the improved version is much better than the previous one on youtube. More beautiful and more fluent! Just I hear a bit weak sounding notes sometimes which are probably your artistic choices, but some of them sound rather unconvincing, of which you could take care of. But it may be just me...
And I especially liked the fugue which seems require great intelligence and concentrations of the interpreter. An excellent job you did.

David wrote:


Hi David,
how nice to read your encouraging lines here, my good friend! I´m so happy to read your feedback here especially concerning my voicing and my execution of the ornamentations! Coming from you it means a valuable affirmation for me. Though my trills are not always the fastest, I really try them to play with feeling, sense for voicing and dynamic. So it makes my really happy you feel them to be natural and fitting into the musical line. Thank you for your wonderful feedback!
I wish you a Merry Christmas!

Andreas

Hye-Jin wrote:


Dear Hye-Jin,
I´m very happy to read your comment here, thank you! But I have to say, I did not re-record the prelude, I just edited a little digital distortion, which was caused by my hardware by accident, I think, and I cut out a very small, nearly inaudible slip. So, it can only be more beautiful, because I have deleted that little "click" noise and that small slip, nothing else is changed.



I have tried to make a quite lyrical interpretation in each first part, during the repetitions I changed the character (and articulation), may be that lyrical character is not your taste? (It´s no problem, if not. Not everyone has to like my way of interpretation.)But tomorrow I will listen again and see, if there are some notes, which are too weak, may be you are right. So in every case thanks for the advice.



Thanks for that, Hye-Jin. Nice to have your helpful and appreciated comments here again!

Hi Andreas,

And a very Merry Christmas to you too!

David