FYI, you don't "dither down" from one sample rate to another. You convert the rate. Dither is for reducing bit depth.
You got me there, I stand corrected. My mind was in many places at once as I quickly wrote the post.
I don't really mean to get into this argument here, but, in practice, you'd want to apply dither after doing any editing on a computer. Wavelab, for example, allows you to use 32 bit temp files. so if I open up a 16/48 file, and resample to 16/44.1, the computer uses more decimal places, and it actually creates a 32 bit, 44.1khz. you can then apply dither to get it down to 16 bit. (or, with WaveLab, you can set it up to do it all in one step). I'm pretty sure this is better than using 16 bit temp files to begin with, because the dither can hide the small digital artifacts left over from the conversion process.
Another thing to consider, is that sooner or later, you'll probably want to burn DVD's to listen to your recordings. I know for me, even with my 16 bit recordings from DATs, I like to burn DVD-Audio discs, because you can ge the whole show on one disc. it's convenient. depending on your DVD player, however, you may want to use audio-DVD-creater to create DVD-Video discs, because not all players can read DVD-Audio discs. While the DVD-Audio format spec includes many possible sample rates (32, 44.1, 48, 88.2, 96, 176.4, 192kHz), the DVD-Video spec only allows for 48 and 96kHz. so if your "masters" are at 44.1kHz, and you want to burn a DVD at some point, you'll either have to use the DVD-Audio format, or resample from 44.1 to 48.
just some other things to think about...
[hijack]
I've got a bunch of Dats that were recorded, and I transferred to my computer, at 48 khz. What is the *best* way to resample these to 44 khz, and lose the least amount of fidelity? If wavelab is the program to do it in, what are the steps if dithering after resample is the preferred method? Also, where is the setting that allows me to make a 32/44 file when I resample, instead of a 16/44?
It would be great if someone could write up a quick little tutorial on how to properly resample and get the highest quality result in wavelab. anyone willing? It seems a lot of people would benefit from it, including myself. 
[/hijack]
Tim - I was just about to send you a PM, but I thought others might like to hear it as well, considering this is a 44.1 vs 48 thread, it's not
that off topic
first off, let me say that I'm still at work, and I don't have WaveLab in front of me. Somewhere in the general "options" or "preferences" there is a choice for "when opening files" and then there is a pull down list, and it lets you choose "16 bit temp files", "24 bit temp files" or "32 bit temp files" it didn't take me that long to find initially, you can probably find it easily. One thing to note is, if you've got a 3 hour 16 bit file that is already close to 2 gigs, you won't be able to take advantage of the additional bits for temp files. WaveLab doesn't like files over 2 gigs, so if the temp file goes over 2 gigs as it's processing, it'll say that it can't complete the task. So you might have to ues 24 bit temp files instead of 32 bit files. if you're 16 bit file is over 2 hours, than even a 24 bit temp file would go over the 2 gig limit. unfortunately, we have to live with this limit in Wavelab.
then you go to "edit sample rate" (offhand, I'm not sure exactly what phrase they use, could be "convert" sample rate, something along those lines.)
so, if you do this and you don't have any dither selected, it'll process the file, and then if you use the bit-meter, you'll see that it's now 24 bit, or 32 bit, or whatever. you can then select the dither scheme you want to use (I always use uv22hr), and hit "render" to apply the dither and save the new file.
however, you can set the dither that you want to use before you convert the sample rate, and then it applies the dither automatically, after resampling the file. so if you've got the dither selected, and then resample, and then play the file with the Bit Meter on, you'll see that it's at 16 bit.
lastly, a third way of doing this is to take your 48kHz file, choose which dither scheme you want to use, and then hit render. you can choose to create a new file, and then you can choose the properties of the new file. go in there, and where it says "sample rate", instead of leaving it as "same as original" you can change it to "44,100 Hz". this method will also resample, and apply the dither scheme afterwards, and create a your new file.
I do not think there is any difference in the way that WaveLab performs the operations, regardless of how you do it. so all three methods in wavelab should be equal to one another.
hopefully this explanation was clear, maybe someone else can verify what I wrote. I probably won't be back tonight, because I'll be at the Trey show. I'll check back in this thread tomorrow
Topic: recording @ 48 khz vs. 44 khz to a jb3 (Read 5824 times)