FW: arsclist ["David Satz" <DSatz@msn.com>]

For the Arhoolie purposes a well recorded 16-bit transfer would be sufficient, but *well* recorded is the key. There are 2 primary reasons for using 24-bit media.

First one can raise the level of recorded material without losing resolution. A 16 bit recording has 96 dB of dynamic range (6dB per bit 0r 2 to the 16th power) If a recording is recorded with 12 dB of headroom - a possibility when recording a large amount of material of varying volumes, You're only using 14 bits, for a dynamic range of 84 dB. Human hearing has a dynamic range of approximately 98 dB so in this case the system noise would be raised into an area that could be heard during quiet material. Additionally if dither is added, which is usually around - 84 dB (the 2 least significant bits), this noise will also be raised to -72 dB which can easily be heard, sounding a little like tape hiss. If the same material was recorded with 24-bit resolution which has 144 dB of dynamic range, raising the level by 12 dB would still put the system noise and dither below the threshold of hearing. So in short, when recording to 16-bit media, it's necessary to carefully optimize the volume of the recording. With 24-bit this is not as necessary.

Obviously in the case of records the surface noise will be far louder than the system noise or dither. Which brings me to the second point. There are is an advantage to using 24-bit transfers if further processing, such as denoising is to be done. There are 2 reasons for this. First the dither can affect the processing, and if EQ is used it can change the sound of the dither making it more apparent. Second, 24-bit material gives the processing more information to work with. Even though you may not be able to hear it, there is important information below the 16th bit. If this low level information is retained, any subsequent processing will be more accurate. On the other hand if this information is thrown away during the initial transfer it cannot ever be recovered. Is this really a factor with old records with lots of surface noise? Well, yes it is because the surface noise is being processed along with everything else, so it must be processed as accurately as possible. Keep in mind that dithering, rounding or truncation of the digital word will result in varying amounts of distortion at the level of the least significant bit (-96 dB in the case of 16-bit) This distortion can be amplified during further processing. The result can be the "grainy" or "harsh" sound and collapsing of the stereo image that used to be the main complaint about digital sound.

I think that cost is no longer an issue with 16 vs. 24 - bit transfers. And now that the Alesis Masterlink, is available storage isn't a problem either. You can post this to the ARSC list if you'd like.

--
Konrad Strauss
Recording*Editing*Mastering*Production
http://members.surfnetcorp.com/konrad

[Table of Contents]

FW: arsclist ["David Satz" <DSatz@msn.com>]

[Subject index] [Index for current month] [Table of Contents]