dithering/resampling...is it worth it?

[ghetto3jon]

i'm mixing from a 16 bit/44100 digital recorder, into an edirol 16 bit/48000 spdif-usb interface, into sound forge 6. i was going to record a 16 bit/48000 file in sound forge (or 24 bit if it would let me) and then kick that file up to a 32 bit float/88200 sound file to do some processing (cutting and pasting, plugin stuff, etc).

then, i was going to dither/resample back down to 16 bit/44100 to burn it.

my question: is all that dithering and resampling worth it? would recording and working with 16 bit/44100 in sound forge be better? i'm concerned that all that dithering and resampling would damage my file more than if i was to just do my processing in 16 bit/44100. is this a valid concern?

jon.

 

[shredfit]

If your initial sound source is 16bit 44.1 Khz, No your not really benifiting from any sample rate conversion or bit depth change!

Shred

 

[Slackmaster2K]

Your audio application will automatically bump it up to a 32bit float, because that's what it, along with VST effects and so on, all use internally. In this case there isn't any benefit to increasing the bit depth of your initial sources.

Now, if you're going to process a file in multiple applications, that's a little different. In that case it might be beneficial to crank it up to 32bit, such that you preserve the detail from one application to the next. Of course that's all in theory. In reality the noise floor of your 16bit converter probably negates any real benefits of bumping up

Upping the sample rate is just asking for trouble.

Slackmaster 2000

 

[ghetto3jon]

thanks!

but what about this:

when i record into sound forge, i can set the sound forge recording rate to 48000, and the bit rate to 24. my source is still 44100/16 (fostex fd-8), but the file i'm creating in sound forge is 48000/24.

so my output from the fd-8 is 44100/16, but soud forge is recording it onto a 48000/24 file. so the file becomes 48000/24, right? now i know that i don't gain any sound quality by this, but maybe processing (editing, pasting and such) a 48000/24 is better than processing a 44100/16 file. is that true? then, after editing, i bump it down to 44100/16.

does that make sense?

thanks for the responses! i just had to reword my question a bit.

 

[M.Brane]

Slack is right. You gain nothing by upping your sample/bit rate. All your doing with that is chopping your 16/44.1 file into smaller pieces. You cannot reclaim lost resolution.

As for dithering, you only need to do that if your downsampling, say 24 to 16 bit. The dither should be the last thing in the chain.