Changing LE session from 48 to 44.1 sample rate

[ansanity]

I unfortunately had my A/D converter (Mytek ADC 8x96) switched to the sampling rate of 48, and of course now when I bounce to CD my songs sound like they were recorded underwater. Is there any way for me to bounce the songs to CD as I hear them in my ProTools session playing at 48?

Thanks a bunch -
Andy


Digi 001 PT LE 5.1
Mac G4 dual 60G 867 mhz

[ansanity]

A related question...

What is the difference in sound quality between recording at 48 then converting to 44.1 and just recording at 44.1 to begin with and staying there? If you're going to end up at 44.1 in the end, does tracking at 48 even produce a different result?

Also, with respect to my 48 to 44.1 issue, if I were to just take an analog out and run my 48 tracks back as audio into a 44.1 session, does anyone have any suggestions as far as hardware/processes they've used?

(If anyone knows of any other method besides running analog out, I would LOVE to hear it!)

Thanks -

[globaleric]

I've been never forced to test this so far, but did you make sure to select a 44.1 sample rate in the Bounce dialog and set the conversion quality to "Tweak Head" (highest)?
You also might getting better results with the "Convert after Bounce" option.

[mindnoise]

HI,

in addition to the above this thread might hold useful information for you:

44.1 vs 48kHz:
http://duc.digidesign.com/showflat.p...rt=&PHPSESSID=

Personally I find my 001 sounds better with 48kHz but it´s a taste really. much more important
is that you record in 24 bit where you DON´T need to get hot with recording.

[globaleric]

If you're going to end up at 44.1 in the end, does tracking at 48 even produce a different result?

Yes, it's likely to be less accurate at 48khz than if you'd do everything in 44.1. Why?
Imagine the conversion algorithm having to transfer the amplitude (values) from the 48khz "grid" to the slightly wider 44.1khz grid. Assuming the tempo of the material has to be preserved, then the algorithm has to interpolate ("guess") the amplitude of every generated 44.1 frame from between 2 neighbor frames of the 48khz material. At least with simple conversion algorithms, those "guesses" won't always exactly match the actual amplitude of the original analogue signal, and this inaccuracy generates some additional distortion. Of course, some quality and price differences do exist with such algorithms, but don't expect having a state-of-the-art type implemented in PT LE

Downsampling from 88.2 to 44.1 is far less critical as the algorithm will simply skip every second 88.2 frame, but every other frame will always be accurate. (Filling the missing gaps is done later by the digital filter of the DAC)

[mindnoise]

HI,

but don't expect having a state-of-the-art type implemented in PT LE

sorry eric but if don´t bring hard evidence for that I consider this statement slightly BS.


Following the mentiooned thread above concludes for the complexity of downsampling the latest finding was that IF you are downsampling that is average samples it does not matter where you average them.
And if you do it mathematically the greater error must be when averaging every second sample (50%"error" on 88.2->44.1) than when averaging every 9th (9% error on 48->44.1) sample.

there are no frames skipped but averaged with downsampling.


I guess what you really mean is that the Frames averaged remain within an even multiplier of the timestamp and I agree totally with you here.

if this is actually audible, well, just look at the numerous discussions about it.

perhaps this helps a little.

here´s also some light in the issue:
http://www.theprojectstudiohandbook.com/directory.htm


regs

[globaleric]

I don't think it's the right place here to continue those discussions here neither

Can we just agree that the disadavantages from downsampling from 48 to 44.1 outweigh the benefits from the higher 48k sample rate?

[mindnoise]

hi eric,
it should be avoided where uncecessary for sure!

regs

[ansanity]

now that that's settled ( ), back to the original question... does anyone have any suggestions as to the best way for me to go from 48 to 44.1 since I am unfortunately stuck at 48 right now and need to bounce to CD?

and eric, yes, thanks... i've tried setting the bounce settings to all different settings, the problem is that it was all tracked at 48, so it always bounces with that "slow and low" sound.

thanks too mindnoise for that thread - that's very helpful for the future!

thanks -

[globaleric]

Strange - what happens when you bounce your session with 48k, close the session, create a new 44.1k one and try to import the bounced tracks to a stereo track?

Otherwise I think you'll need a DAT machine or similar recording (at 44.1) from the analogue outputs of your Digi Hardware. Then record the tape back to PT via SPDIF in a new 44.1 session.