What's wrong?
One of the things about HiFi which is universally ignored is the fact that real instruments generate sounds way above human hearing. For example trumpets or cymbals have harmonics up to 90kHz or so.
Now you can argue, and many do, that since we cannot hear sounds above 15-18kHz, and as we get older down to 8-10kHz, that we do not need to record, store or reproduce higher frequencies.
But this is UNTRUE. If the true output sound of the instrument is not recorded then the timbre will be lost.
It is also true that making digital recordings with low sample rates, e.g. the CD at 44.1kHz, limits the upper frequency which can be stored to about half of that, i.e. 20kHz or so. And it is also true that lossy compression techniques, e.g. MP3 or AAC, although designed to use psycho-acoustic sampling, i.e. throwing away those things that humans cannot hear, allows us to make recordings smaller.
But both of these delivery method prevent the true range of the instrument, the phase relationship and stereo image to be delivered.
More over the limit of digital recording on CD and most MP3/AAC songs, means that the dynamic range of the music cannot be recorded. A full orchestra has an SPL range of 30-130dB, whereas a 16bit CD can only store 35-110dB at best. Consequently engineers turn down the loud bits and turn up the soft bits to compensate.
Through all this we have four problems
- Tibre
- Dynamics
- Integrity
- Delivery
Solution?
The recording chain must be upgraded to capture as high a frequency spectrum as possible, this is why studios to day are often using 192kHz sample rates, giving the possibility to record frequencies up to half this, or 96kHz. There is a strong movement to deliver 192kHz or the lesser 96kHz files over the internet today. Moreover these files are usually 24bit not 16bit, thus improving the dynamic range. Lossy compression is also not used, but lossless FLAC files.
Thus almost all the problems are addressed. So what we need is a HiFi systems capable of handling the new frequency spectrum up from the 96kHz recordings. This means a 24bit DAC for conversion, and amplifier with a response to 50kHz or so, and some very good loudspeakers - these are the critical part of the chain.
Amplifier
A good amplifier can be fairly easily constructed, this on uses the Class D modules UcD180 from Hypex.
Here's a photo of the amplifier, using a toroidal transformer and a power supply module from HK.
No comments:
Post a Comment