Notch Filter Oddity

I'm starting on a turd polishing project which involves using the notch filter.  I set all the frequencies in the notch filter on the 44.1 KHz file I pulled of a CD.  I tried upsampling the file to 96 KHz and when I applied the notch filter all the noise was still there.

I opened the filter dialog and all the freq settings had been multiplied by the increase in sampling rate.  That doesn't seem right, does it?

<edit>  the same thing happened on the FFT filter frequency scale & in the NR dialog.  Am I expecting too much? </edit>

Well, apparently the frequencies are stored as frequencies, but as the number of FFT bin, or something like that. When the sampling frequency changes, the frequency "assigned" to a given bin is different. This is not so convenient, true. The program might rescale, and that would be nice. Though rounding-up or interpolation would have to be done in most cases.

Also, imagine that you set something (a notch, an FFT filter point) at 40 kHz in a 96 kHz file. Now you resample to 48 kHz. The 40 kHz setting won't make sense anymore. What should the program do?

Zemlin,
I'd suggest running your notch filtering before you upsample the files.
Think about it... by upsampling a file from 44kHz to 96kHz, you're not actually going to introduce any frequency information above 22050Hz anyway, so filtering before upsampling will still result in the same end result.
On top of which, filtering the 96kHz files will just take 2.17 times as long to achieve the same result!
I'm guessing you're mastering stuff for DVD (?), but anyway you look at it, the files aren't going to be improved by upsampling. Therefore, you may as well do all your "restoration/turd polishing" at 44kHz and only jump to 96kHz when you're ready to compile the master.
Unless I'm missing something...

Sorry for being Green but why would you resample UP to 96kHz for Mastering?

It was an experiment - I had some serious noise reduction to do on the crappy recording and I thought that NR artifacts might be better at higher sampling rates.  I can't say conclusively that it made a difference.

Sorry for being Green but why would you resample UP to 96kHz for Mastering?

I was having a guess that perhaps he was preparing audio for DVD-A.

Sorry for being Green but why would you resample UP to 96kHz for Mastering?

Some processing apparently sounds better at higher sampling rates. Some plug-ins upsample internally too.

Whether you'll hear an improvement depends on the particular settings used and whether you are using decent monitoring. It isn't something that I do routinely.

Cheers

James.

[decent monitoring]

Whether you'll hear an improvement depends on the particular settings used and whether you are using decent monitoring.

Well, that rules me out! but Ohhh! how I'd appreciate a decent set of speakers and cans. Investment time beckons