44 to 48kHz conversion

MetalSir

Member
Nov 17, 2008
177
0
16
hi!
it's a stupid question but I wanna ask you the same.. I have 2 protools sessions, one at 48kHz and one at 44kHz. I need to import on the 48kHz one some files from the 44kHz one. (basically someone gave to me some DI to reamp.. they recorded at 44kHz and I want to reamp at 48kHz, because of low ending matters).

in your opinion te best procedure should be converting (from 44 to 48kHz) when exporting the files (from the 44kHz ses) or when importing it (on the 48kHz ses)?

I honestly think that it's the SAME, you?


really tks a lot for your time! :kickass:
 
hi!
it's a stupid question but I wanna ask you the same.. I have 2 protools sessions, one at 48kHz and one at 44kHz. I need to import on the 48kHz one some files from the 44kHz one. (basically someone gave to me some DI to reamp.. they recorded at 44kHz and I want to reamp at 48kHz, because of low ending matters).

in your opinion te best procedure should be converting (from 44 to 48kHz) when exporting the files (from the 44kHz ses) or when importing it (on the 48kHz ses)?

I honestly think that it's the SAME, you?


really tks a lot for your time! :kickass:

It should be the same. Personally I'd convert my 48k stuff to 44.1 so that you're only doing one set of conversion before you burn your master, instead of going 44.1>48>44.1.
 
Could you please first explain why you need to operate at 48kHz if the end product is for human hearing?
mm.. low endings?


@Trevoire520: basically, since I have all guitars and the bass DIs to reamp, the situation should be very close to recording again everything at a better resolution.. so i was thinking to reamp at 48kHz, and i was wondering if in your opinion should be better resampling from 44 to 48 on the export or on the import..

tks guyZ! :kickass:
 
Hey metal sir I kinda get where you're coming from. Seen as your re amping you want to re record at the higher sample rate, then I'd just import into your 48 k session and get cracking on those tunes.

:cool:
 
Either export and convert down or do file, "save copy in..." then select 44.1 as your sample rate. It will convert your entire session.

Could you please first explain why you need to operate at 48kHz if the end product is for human hearing?

Some people just like 48. Its the standard audio sampling rate used by video gear too.

EDIT: I'd be tempted to reamp at 48 then convert the reamps down
 
I'd convert to what ever your intended final destination format sample rate is. So if it's going to be broadcast 48 and if it's CD 44.1. As to why we use a higher SR than the human audible range, without going on at length it's due to the Nyquist frequency and anti-aliasing in digital.
 
Could you please first explain why you need to operate at 48kHz if the end product is for human hearing?

Converters work better at 48khz compared to 44.1.

I'd go with AEL89's method, seems most reasonable here.
If you want to have a backingtrack for the mixdowns to reamp to then export it at 48khz, for the purpose it has the conversion wont really matter.

Even if that really doesnt answer your original question if it makes any difference if you convert them by inporting or exporting lol
 
Either export and convert down or do file, "save copy in..." then select 44.1 as your sample rate. It will convert your entire session.
you solved my day!! I love ya!!


whats about my idea? because of my DI recordings are at 44100.. so, my starting files are at 44khz.. and I'd try to record the reamps at 48khz.. because of I CAN HEAR a HUGE DIFFERENCE between 44 and 48khz trough my apogee rosetta.. i think that distorted guitars and especially the bass guitar MUST be recorded at 48. so I'd try to convert ALL the project (containing ONLY clean tracks to reamp) from 44 to 48khz and then reamping at 48..

what do you think?
 
It might be worth trying a blind AB or ABX test at a friendly pro mastering studio using their high end convertors (ie Lavry, Cranesong, Forsell, Prism, Mytek...) to see if you still hear a difference. If you don't then any difference between 44.1 and 48 on your Rosetta is more likely to be down to distortion and ringing in its convertors. If you do hear a difference on a high end convertor as well as your Rosetta I still think you should SRC to the intended final source SR. Multiple SRC will almost certainly result in a poorer end resut then recording and mixing and mastering at a single SR.
 
you solved my day!! I love ya!!


whats about my idea? because of my DI recordings are at 44100.. so, my starting files are at 44khz.. and I'd try to record the reamps at 48khz.. because of I CAN HEAR a HUGE DIFFERENCE between 44 and 48khz trough my apogee rosetta.. i think that distorted guitars and especially the bass guitar MUST be recorded at 48. so I'd try to convert ALL the project (containing ONLY clean tracks to reamp) from 44 to 48khz and then reamping at 48..

what do you think?

No problem.

That won't work - your 44.1 DI tracks won't sound like they've been recorded at 48. If you recorded DIs at 44.1, keep them there and reamp at 44.1. You're using a Rosetta for your AD and DA - they're great converters so I wouldn't worry about it too much.

Do a little bit of reading on sample rates - it might make more sense afterwards. Sure - what you'll be recording from the amp will be at 48, but I'd want the DI to be 48 as well. Might as well stay at 44.1. It's your call though, it will turn out fine in the end.
 
No problem.

That won't work - your 44.1 DI tracks won't sound like they've been recorded at 48. If you recorded DIs at 44.1, keep them there and reamp at 44.1. You're using a Rosetta for your AD and DA - they're great converters so I wouldn't worry about it too much.

Do a little bit of reading on sample rates - it might make more sense afterwards. Sure - what you'll be recording from the amp will be at 48, but I'd want the DI to be 48 as well. Might as well stay at 44.1. It's your call though, it will turn out fine in the end.
so, you're telling me that if i have a DI recorded a 44khz IT'S THE SAME if I make the reamp at 44 or 48khz, right?

mm i don't think so.. even if my DI was recorded at 44khz, now I have THAT signal passing through an amp+cabinet+mics+cables.. I just think about how much harmonics the distortion adds to my original signal.. on this HORDE of TOTALLY NEW information added to my poor 44100hz DI..
I think that recording a 48khz this almost NEW track should be different than operating at 44..

ok, now offend me.:lol:
 
1. If you have DI at 44.1, send the signal to the amp and record whatever sample rate you prefer.
2. Changing sample rate of the DI from 44.1 to 48 hz makes no sense at all.


BTW, as far as I know, most of the AE gurus here record at 44.1Khz. And their albums sounds FUCKING LOUD AND METAL.:headbang:

Even though, 48 Khz is CPU friendly and easier for the converters as Mago stated.

Seriously, the differences between 44.1 and 48 are unnoticeable for most human beings (unless you are batman)... having a cable of 9 metres instead 15 metres makes more difference.
 
so, you're telling me that if i have a DI recorded a 44khz IT'S THE SAME if I make the reamp at 44 or 48khz, right?

mm i don't think so.. even if my DI was recorded at 44khz, now I have THAT signal passing through an amp+cabinet+mics+cables.. I just think about how much harmonics the distortion adds to my original signal.. on this HORDE of TOTALLY NEW information added to my poor 44100hz DI..
I think that recording a 48khz this almost NEW track should be different than operating at 44..

ok, now offend me.:lol:

You can only put a 44.1 signal (ovbiously converted to analog) into your amp since you only recorded at 44.1. What comes out of your amp you can convert to whatever sample rate you want. I stick with 44.1 for everthing.

Does that make sense now? Your DI will always be 44.1, even if you convert upwards, you can't get the detail back.

Don't worry about it too much man, 44.1 is fine.
 
let me start out by saying, unless you have really good reason to, stick with 44.1 (and i'd prefer 88.2 if you *must* operate at a high sampling freq.)

that said, i think it's important to note that, although the DIs are recorded at 44.1, there will still be more information in a 48khz recording of a reamp, as opposed to a 44.1khz reamp. most (if not all) of the 10khz+ signal from an amp is produced by the amp, not the guitar/DIs. basically, what MetalSir said.

seriously. you can record DIs at 22khz and the reamp, recorded at 44.1 will (theoretically) sound the same as 44.1 DIs. guitars have almost no signal content (information) when you get into the khz signal range, so really, why does it matter.

if you actually want a tangible increase in useable information in your DIs and reamps, record at 24/32bit (although low order bits of 32bit are probably gonna be mostly ADC noise, or interference from that cell phone mast 3 miles away..).

thanks,
 
depends what you mean by law of halfs?

conversion from 88 to 44 is mathematically simple, and yea, good algorithms exist for converting 48 to 44 with no added artefacts, but there's no guarantee that's what your daw does, or plugins, or everywhere a sample rate change might occur.

however, my actual point is 48khz is 9% ish of an increase in sample rate (wrt to 44khz) - how can that really have an effect on the audio it samples? 88khz is double the bandwidth, so if you're using higher sampling freq for "audiophile" reasons, at least use one with a bit more science :)

thanks,
 
depends what you mean by law of halfs?

conversion from 88 to 44 is mathematically simple, and yea, good algorithms exist for converting 48 to 44 with no added artefacts, but there's no guarantee that's what your daw does, or plugins, or everywhere a sample rate change might occur.

however, my actual point is 48khz is 9% ish of an increase in sample rate (wrt to 44khz) - how can that really have an effect on the audio it samples? 88khz is double the bandwidth, so if you're using higher sampling freq for "audiophile" reasons, at least use one with a bit more science :)

thanks,

The "mathematically simple" argument for conversion to 44.1 from other sampling rates isn't applicable.

And to reiterate - resampling does not obey any law of halves.

This thread is getting too complicated - read the OP. It's just a common sense thing.