This quote is from a web site explaining Adobe Auditon's (CoolEdit 
revisited) operations:

http://www.covingtoninnovations.com/audio/audition/#bitdepth

I think it states the use of upsampling and downsampling in editing 
quite clearly in layman's (or laywoman's) terms.

>
>       Bit depth (16 or 32)
>
> Standard CD-quality audio files have 44,100 samples per second, 16 
> bits per sample. However, many of us now have 24- or 32-bit sound 
> cards and can record 32-bit audio files. These give higher quality and 
> can be downsampled to 16 bits before burning to CD.
>
> Further, Adobe indicates that highest quality is obtained by 
> upsampling your file to 32 bits if it was not already recorded at 32 
> bits, editing it, and then downsampling again to the original format.
>
Rod Stephens

Jeffrey Kane wrote:

>Perhaps my understanding is faulty; I'd always thought processing should be
>done at a higher resolution if possible to minimize adverse effects on the
>signal and preserve as much resolution as possible. Applying certain effects
>to 16 bit audio if editing in 16 bit resolution could result in loss of
>resolution could it not? 
>
>-----Original Message-----
>From: Association for Recorded Sound Discussion List
>[mailto:[log in to unmask]] On Behalf Of Alyssa Ryvers
>Sent: Monday, February 20, 2006 10:42 PM
>To: [log in to unmask]
>Subject: Re: [ARSCLIST] Cassette obsolescence - digitizing standards
>
>OK...so am I understanding correctly that you are transferring through 
>your converter into your computer at 16bit, but then importing it into 
>ProTools for 24bit?
>
>If this is the case, you're not "enhancing" the sound quality, but just 
>using some kind of algorithm(s) in order to get a more complex file - 
>information that was added, by the way, and has nothing to do with the 
>original.
>
>Alyssa.
>:)
>
>"If someone, holding fast to the name of Bodhisattva Perceiver of the 
>World's Sounds should enter a great fire, the fire could not burn 
>him...If one were washed away by a great flood and call upon his name, 
>one would immediately find himself in a shallow place." (The Lotus 
>Sutra)
>On 20-Feb-06, at 10:08 PM, Lou Judson wrote:
>
>  
>
>>Interesting perspective. When I transfer cassettes for clients, I use 
>>16 bit, and if they want it processed in any way, I import it to 24 
>>bit Protools sessions for the added range... Best of both worlds, I 
>>like to think.
>>
>><L>
>>
>>Lou Judson . Intuitive Audio
>>415-883-2689
>>
>>On Feb 20, 2006, at 10:00 AM, Mike Richter wrote:
>>
>>    
>>
>>>Lou Judson wrote:
>>>      
>>>
>>>>What about using 24 bit at 44.1 so that any noise reduction or 
>>>>processing done later is higher definition?
>>>>        
>>>>
>>>Given that the best dynamic range on standard cassettes - assuming 
>>>Dolby  B in proper calibration which is highly questionable - is 
>>>unlikely to exceed 60 db, one might suspect that 16 bits is 
>>>sufficient. Of course, processing could consume several bits and one 
>>>only has half a dozen to spare (~30 db).
>>>
>>>For that potential, infinitesimal advantage, one is likely to spend 
>>>four to ten times as much to make the transfers counting both 
>>>equipment and time. Given infinite resources, a case can be made; 
>>>with a budget less than that of a typical multinational corporation, 
>>>such overkill is hard to justify even on theoretical grounds.
>>>
>>>Mike
>>>-- 
>>>[log in to unmask]
>>>http://www.mrichter.com/
>>>      
>>>
>
>  
>