advertisement


Vinyl is not obsolete, Will CD become obsolete ?

Modern ADCs sample at several MHz and then down-convert to required kHz, so the filtering can be done more benignly in the digital domain for any of the usual sample rates.

This came up on Gearslutz recently.
Darren
 
darrenyeats said:
Modern ADCs sample at several MHz and then down-convert to required kHz, so the filtering can be done more benignly in the digital domain for any of the usual selection of sample rates.

This came up on Gearslutz recently.
Darren
Click to expand...
I'm assuming that that is a reference to Delta sigma ADC's- or are there 24/16 bit converters working at that rate. I hope not because if there are, someone's going to insist that you need one in your dac.

Incidentally I'm really intrigued by this Anti alias/decimation filter thing. It seems to be to be the 600lb Schrodinger's gorilla in the room except it's invisible and you're in the room with it and my mixed metaphors fail me. It would be great if we had information about that invisible gorilla with every recording so that we could make at least partially sensible decisions about the potential effects of digital filters.
 
Plutox said:
I might be wrong but I assumed, on a quick read, that he was referring to A to D conversion when recording, not the converse with which we are all far more familiar.

A to D conversion is generally held to be a more fickle process than the later D to A, if only because many A to D devices internally incorporate a partial D to A in a kind of 'driving in a wedge' analogy.

It was certainly the case a few years ago that a better result could usually be achieved by down-sampling a 96kHz master to 44.1kHz than converting to 44.1kHz directly.
Click to expand...

I think I managed to copy a partial link - so you may have read some random , but related, article!

The original was (I hope):

http://trustmeimascientist.com/2013...rates-when-higher-is-better-and-when-it-isnt/

And I'm failry sure he was talking about playback...
 
Plutox and Werner, I stand corrected. Actually that's good news - one less thing to worry about as a consumer.
 
Rodney gold said:
http://www.dspguru.com/dsp/faqs/multirate/decimation
A readable and not too mathematical "explanation" of decimation and downsampling..
some other nice sections on that site , FAQs etcc..
Another quite nice book / site is
http://www.dspguide.com/
Click to expand...

I have looked at dsp guide before and found it helpful. In a fit of enthusiasm I bought the book version for £38 even though it's available free on the internet. However it is well known fact that merely owning books is enough to make one clever and knowledgeable, although there are sometimes additional benefits to actually reading them. The book hasn't arrived yet, but i am feeling more knowledgeable already.
 
adamdea said:
The peak thing is not how it works because the figures are all relative. You Can record the peak whether you have 10-12 -14 -16 - 24 bits .
Click to expand...
True. But if you don't have enough dynamic range given by the bit depth, either the peak will be less peak (everything compacted), or the quietest parts will be lost/mix in the noise. None of the options is ideal.
Now consider that if the orchestra really does reach a 120 dB peak, what was the background/mic noise level? If it was above 30 dB or so, is the quantisation noise going to make a material difference?
Click to expand...
But I see where we getting and you have a point.

The other day I was listening to some Diana Krall FLAC and I was hearing some mechanical sound from the piano. Seemed to me like the retracting sound of the little hammers that strike the chords inside. First I thought, that's incredible, I can even hear the sound of the mechanical parts of the instrument.
...than I thought better. ...This isn't really part of the music, is it?
:)

Still, if it was recorded, I'm glad I was able to hear it.
...and yes, I believe they would be audible at 16/44.1 too.

Michael
 
Mikas said:
True. But if you don't have enough dynamic range given by the bit depth, either the peak will be less peak (everything compacted), or the quietest parts will be lost/mix in the noise. None of the options is ideal.

But I see where we getting and you have a point.

The other day I was listening to some Diana Krall FLAC and I was hearing some mechanical sound from the piano. Seemed to me like the retracting sound of the little hammers that strike the chords inside. First I thought, that's incredible, I can even hear the sound of the mechanical parts of the instrument.
...than I thought better. ...This isn't really part of the music, is it?
:)

Still, if it was recorded, I'm glad I was able to hear it.
...and yes, I believe they would be audible at 16/44.1 too.

Michael
Click to expand...
Almost certainly-if it's swamped by quantisation noise at 16/44 you would be very unlikely to hear it at 24 bits either at sensible volume.
I agree with you though that it can give you a real kick to hear incidental noises like people turning the page on music or shifting in their seat. As long as it isn't Glenn Gould humming.


Frankly (and heretically), though, for ordinary listening it's useful to have a wee bit of dynamic compression if you want to hear the triangle solo and the full orchestra crescendo without having to sound proof your room and play at full volume. uncompressed 16 bits is already more than you need.
I wonder how many orchestral recordings really have absolutely uncompressed dynamics.
 
Vital said:
Maybe "Baton Hero" would work better. Yes, I think I can see that.
Click to expand...

Air Baton is much more plausible. Air conducting is a very real phenomenon: I was once at Covent Garden with someone who was discreetly asked in the interval to stop it.
 
adamdea said:
Frankly (and heretically), though, for ordinary listening it's useful to have a wee bit of dynamic compression if you want to hear the triangle solo and the full orchestra crescendo without having to sound proof your room and play at full volume. uncompressed 16 bits is already more than you need.
I wonder how many orchestral recordings really have absolutely uncompressed dynamics.
Click to expand...

I understand you, because it's difficult, at least for me, to listen to some classical tracks. For instance, this version of Ravel's Bolero. It starts with a single guy playing some kind of flute and ends with everyone playing loud and hard (everyone knows it for sure).
If I set the correct volume for the first accords, I find myself realizing the room's to small for this track.

Michael
 
Mikas said:
I understand you, because it's difficult, at least for me, to listen to some classical tracks. For instance, this version of Ravel's Bolero. It starts with a single guy playing some kind of flute and ends with everyone playing loud and hard (everyone knows it for sure).
If I set the correct volume for the first accords, I find myself realizing the room's to small for this track.

Michael
Click to expand...
Yes, it really can be a problem, and even if the room isn't too small, it just isn't practical for everyday listening. An extreme example is radio 3 in the car- if you set it so that you can hear everything you would deafen yourself and/or blow up the speakers. And AFAIK even radio 3 is usually a little bit compressed.
 
You must log in or register to reply here.


advertisement


Back
Top