![]() |
|
Post Reply ![]() |
Page 123 5> |
Author | ||
Dean ![]() Special Collaborator ![]() ![]() Retired Admin and Amateur Layabout Joined: May 13 2007 Location: Europe Status: Offline Points: 37575 |
![]() Posted: June 06 2008 at 13:02 |
|
Good question.
![]() The dynamic range of a Cd is 96.33dB - usually we reference everything negatively with respect to FS (ie 0dB), so the ideal minimum noise-floor for a 16-bit ADC is expressed as -94.56dBFS
Looking at the maths for Replay Gain they calculate the reference level as being -20dBFS of Pink Noise to produce 83dB(SPL) output in a listening environment (whatever that means
![]() So a quick answer is No.
|
||
What?
|
||
![]() |
||
MikeEnRegalia ![]() Special Collaborator ![]() ![]() Honorary Collaborator Joined: April 22 2005 Location: Sweden Status: Offline Points: 21847 |
![]() |
|
^ Yes, that's what I meant. But I'm still trying to figure out how the Album Gain figure calculated by ReplayGain relates to the figures usually mentioned in articles. ReplayGain calculates the difference between the album gain and a mysterious "reference" (either 83dB or 89dB - sources vary), while articles usually specify the difference between the album gain and 0dB (full scale). Am I right in assuming that "full scale" means 96dB?
Edited by MikeEnRegalia - June 06 2008 at 12:35 |
||
![]() |
||
Dean ![]() Special Collaborator ![]() ![]() Retired Admin and Amateur Layabout Joined: May 13 2007 Location: Europe Status: Offline Points: 37575 |
![]() |
|
^ just add an "Album Gain" field to your database and voilą!
|
||
What?
|
||
![]() |
||
MikeEnRegalia ![]() Special Collaborator ![]() ![]() Honorary Collaborator Joined: April 22 2005 Location: Sweden Status: Offline Points: 21847 |
![]() |
|
I did a bit of research, but I couldn't find any database which you can use to find out which recordings suffer from extreme compression and which don't ... that's too bad. Such a database would be quite interesting ... maybe I'll even add a tag to my website which indicates the "audiophility". ![]() |
||
![]() |
||
MikeEnRegalia ![]() Special Collaborator ![]() ![]() Honorary Collaborator Joined: April 22 2005 Location: Sweden Status: Offline Points: 21847 |
![]() |
|
^^ sorry, I still don't quite understand. But is this all really important today? I mean, even cheap computer sound cards do the digital/analog conversion correctly. I looked at the wikipedia page:
http://en.wikipedia.org/wiki/Sampling_(signal_processing)#Audio_sampling Consider the following statement: "One advantage of higher sampling rates is that they can relax the low-pass filter design requirements for ADCs and DACs, but with modern oversampling sigma-delta converters this advantage is less important" Doesn't this mean that whatever happens during sampling, in the end a low pass filter eliminates this high frequency noise which could affect the original signal? Edited by MikeEnRegalia - June 06 2008 at 11:53 |
||
![]() |
||
Dean ![]() Special Collaborator ![]() ![]() Retired Admin and Amateur Layabout Joined: May 13 2007 Location: Europe Status: Offline Points: 37575 |
![]() |
|
^ I agree too
![]() I am a mixed-signal test engineer by profession so I tend to over-analyse things for a living
![]() |
||
What?
|
||
![]() |
||
Dean ![]() Special Collaborator ![]() ![]() Retired Admin and Amateur Layabout Joined: May 13 2007 Location: Europe Status: Offline Points: 37575 |
![]() |
|
My mistake - it's 22.05kHz ... (Excel rounded down to 1dp and I missed it
![]() When a signal is sampled at 44.1kHz the bit-rate is present in the resulting data as fs/2 [a bit-rate 44.1kHz is one sample every 22.676µS - two samples are needed to reconstitute a signal (Nyquist) so the resultant is 22.05kHz].
This is regardless of the original sampling frequency - keeping the bit-depth and bit-rate high during recording process shifts this quantisation frequency out of the audio-band preventing it from affecting the recorded/mixed data, but it appears once the data is downsampled to 44.1kHz to encode onto the CD - what comes off the CD is still limited to 16bits @44.1kHz and contains all the inherent artifacts of that (low) sampling rate.
Therefore this is not an effect of the sampling ADCs or the reconstituting DACs, but is an inherent trait of the encoded data.
|
||
What?
|
||
![]() |
||
MikeEnRegalia ![]() Special Collaborator ![]() ![]() Honorary Collaborator Joined: April 22 2005 Location: Sweden Status: Offline Points: 21847 |
![]() |
|
^ I agree ... almost. I do like vinyl. And although I think that digital is superior in many ways, it doesn't mean that there can't be a vinyl disc which sounds superior than many digital recordings. It all depends on what the engineers make of the medium. For example, there are many audiophiles which prefer to listen to classical music on CD and/or SACD/DVD-Audio. I don't think it has much to do with the actual music, but with the fact that classical music is rarely played "competitively" on the radio, and thus has been spared from the loudness war.
Edited by MikeEnRegalia - June 06 2008 at 10:39 |
||
![]() |
||
arcer ![]() Prog Reviewer ![]() ![]() Joined: September 01 2004 Location: United Kingdom Status: Offline Points: 1239 |
![]() |
|
And we're off! Again.
I think the bottom line here is that it's horses for courses. I like vinyl and analogue, so does Oliver, Mike's a digi fiend. I find digital flat and hard, he champions its accuracy. The correlation here I suppose is as simple as whether you like a smooth, warm sounding system or something that delivers anlaytical detail and crispness. Personally I want it all - warmth and a tangible midband but with crispness treble and tight bass but I guess I veer towards the warm side of the spectrum. Others prize that "hi-fi" sound of forensic detail (which I find comes at the expense of musicality). In the end it doesn't matter one whit. Both formats have their strengths. If you get something that makes your ears tingle when you slap on a CD or a record then voila - happiness. And that's it. None of us are ever going to agree on this eternal circular debate so best to probably leave it. I still say vinyl rule though ;-) |
||
![]() |
||
MikeEnRegalia ![]() Special Collaborator ![]() ![]() Honorary Collaborator Joined: April 22 2005 Location: Sweden Status: Offline Points: 21847 |
![]() |
|
Some people just can't be taken seriously ... ![]() BTW: At least *I* try to increase my knowledge. ![]() |
||
![]() |
||
oliverstoned ![]() Special Collaborator ![]() ![]() Honorary Collaborator Joined: March 26 2004 Location: France Status: Offline Points: 6308 |
![]() |
|
But nobody does! You see, things are a little more complex than it seems. |
||
![]() |
||
MikeEnRegalia ![]() Special Collaborator ![]() ![]() Honorary Collaborator Joined: April 22 2005 Location: Sweden Status: Offline Points: 21847 |
![]() |
|
^ where does that 22.1kHz signal come from ... why should it be "automatically" present in the digital signal? Or in other words: When/How is it added to the signal?
|
||
![]() |
||
Dean ![]() Special Collaborator ![]() ![]() Retired Admin and Amateur Layabout Joined: May 13 2007 Location: Europe Status: Offline Points: 37575 |
![]() |
|
^ the 20kHz is part of the audio signal, but the 22.1kHz is the bit-rate/2, which is inherent in the digitised signal, so is present in the reconstituted analogue signal.
|
||
What?
|
||
![]() |
||
MikeEnRegalia ![]() Special Collaborator ![]() ![]() Honorary Collaborator Joined: April 22 2005 Location: Sweden Status: Offline Points: 21847 |
![]() |
|
^ I checked the various wikipedia pages about upsampling and digital/analog conversion ... they don't contain much information about the effect you're describing. But: If a signal contains both 20khz and 22.05khz waveforms, the modulation effects are not unwanted at all. Only if by the conversion new signals were introduced, that of course would affect the signal. Are you sure that this happens during digital/analog conversion? I thought that it was primarily a problem when mixing several digital signals into one, which is why professional digital audio in studios usually works with 24bit/192khz.
Don't take me too serious here ... I currently don't have the time to brush up on my knowledge in this domain. |
||
![]() |
||
Dean ![]() Special Collaborator ![]() ![]() Retired Admin and Amateur Layabout Joined: May 13 2007 Location: Europe Status: Offline Points: 37575 |
![]() |
|
True, but if it is present in the audio signal it will beat with other audio signals and the products will be within human hearing range. For example 22.1kHz beating (i.e. modulating) with a 20kHz component will produce unwanted signals at 2.1Khz, 4.2kHz, 6.3kHz, 8.4kHz, 15.8kHz and 17.9kHz (and many others - those are just the 2nd and 3rd order components). Consider then how a complex music signal will beat with 22.1kHz. and the myriad of unwanted modulation products that will result in. This is how quantisation affects A-weighted SNR figures even though the A-weighting filter should 'ignore' the 22.1kHz.
Filtering off 22.05kHz without affecting frequencies below 20kHz is possible, but the filters are complex and expensive - removing 48kHz is easier as simpler, cheaper filters can be used.
|
||
What?
|
||
![]() |
||
MikeEnRegalia ![]() Special Collaborator ![]() ![]() Honorary Collaborator Joined: April 22 2005 Location: Sweden Status: Offline Points: 21847 |
![]() |
|
22khz is already well outside the hearing range ... isn't it? I doubt that there are many adults around who can even hear anything beyond 18khz.
|
||
![]() |
||
Dean ![]() Special Collaborator ![]() ![]() Retired Admin and Amateur Layabout Joined: May 13 2007 Location: Europe Status: Offline Points: 37575 |
![]() |
|
Upsampling CD players upsample the bit-rate and the bit-depth - it would be meaningless to do just one. Upsampling the bit-rate without changing the bit-depth does not alter the original signal, so the quantisation frequency is unchanged. Upsampling bit-rate and bit-depth interpolates extra samples between the original ones in both time and amplitude, which, as you say, smoothes the waveform, but also shifts the quantisation frequency to 48kHz. The whole purpose of upsampling at play-back is to move the quantisation frequency to something well outside the audio band so that a) it can be filtered off without affecting the audio band and b) any modulation products between the quantisztion frequency and the audio signal also fall outside the audio-band.
|
||
What?
|
||
![]() |
||
MikeEnRegalia ![]() Special Collaborator ![]() ![]() Honorary Collaborator Joined: April 22 2005 Location: Sweden Status: Offline Points: 21847 |
![]() |
|
|
||
![]() |
||
oliverstoned ![]() Special Collaborator ![]() ![]() Honorary Collaborator Joined: March 26 2004 Location: France Status: Offline Points: 6308 |
![]() |
|
One more time, i'll repeat myself, and you will too, until your death...
I don't need to listen to your "whisper boxes" to know how crappy it is. You can't judge nothing with that, whereas i had the chance to listen to some of the best digital and analog sources on real big systems in bi amplification. So i know what the best of the best digital source does (imagine a 60 000 dollars Mark Levinson CD setup), i know that it's impressive for some CD, but it's ridiculized by a good turntable of the third of its price. So please don't repeat endlessly what other ignorants and/or jealous say on Internet. I know a little what im talking about, whereas you have no experience in Hifi. You can have a look at my portable system into the heaphones thread, and everybody who listen to it agrees that the good Sony Discman is completly smoked by the Sony cassette Walkman. Much more low, highs, dynamic, detail, image, mateer, precense, smoothness...Now i don't say that vinyl and cassette is simple or convenient, but it's so superior to digital... Now it's not a mater of price, a second hand 100 dollars Nad Cd player will explode any computer or Ipod, and "Imod" (modified version of the Ipod by the american company "Red wine audio") beats the Ipod and so on...well you can listen to music through the worse equipment and still enjoy it...your brain will "fill the gaps" just like with digital... but you're more likely to end up with an headache than when listening to good digital and a tube amp or better to a decent turntable such as a Rega 3, a little good tuner such as a Rega Radio. Edited by oliverstoned - June 05 2008 at 05:13 |
||
![]() |
||
MikeEnRegalia ![]() Special Collaborator ![]() ![]() Honorary Collaborator Joined: April 22 2005 Location: Sweden Status: Offline Points: 21847 |
![]() |
|
^ Funny how people like you completely ignore when people like me say that they *have* heard good analogue equipment.
Have you ever heard my Logitech system? No? Then shut the f*ck up. ![]() Edited by MikeEnRegalia - June 05 2008 at 04:53 |
||
![]() |
Post Reply ![]() |
Page 123 5> |
Forum Jump | Forum Permissions ![]() You cannot post new topics in this forum You cannot reply to topics in this forum You cannot delete your posts in this forum You cannot edit your posts in this forum You cannot create polls in this forum You cannot vote in polls in this forum |