24/48 to 16/44,1 does it make a difference?
Home > Home Recording Forum > Recording Techniques > 24/48 to 16/44,1 does it make a difference?
Posted on Apr 11, 2005 09:54 pm
rkruz
Member Since: Sep 02, 2004
I recorded several songs using 24/48.
When I went to burn the CD the burn app Nero did not recognize the files. I had to convert them to 16/44.1 first. So does it make a difference in the audio of the final cd tracks to record at the higher rates when its then converted back to 16/44.1? Maybe its not worth getting a fancy audio interface after all?
thanks for any tips
[ Back to Top ]
Noize2uCzar of MidiAdministrator
Since: Apr 04, 2002
Apr 11, 2005 11:27 pm The major differance is in the editing. The higher bit rate and sample rate allow for more intricate editing. When doing sound design or sample creation I always use a higher sample rate to allow for much finer editing.
But yes, to burn to CD you will have to dither down to 16/44.1.
In the long run it is worth the money as the day is coming when players will use ghigher sample and bit rates.
Apr 11, 2005 11:28 pm Well, if you ask 10 people you'll likely get 10 different answers.
My thoughts are thusly:
As you process your audio, applying effects, panning, running through busses and all that the computer is constantly performing calculations to make the resulting tone. During all these calculations to figure reverb depth, delay time, ambience, compression, panning and all that, there are always "remainders" to the math, generally these remainders are dropped and filled in with logical sequences between the areas that were not dropped (dithering) and therefore, your file, while still technicall 16 bit/44.1 khz, is a bastardized version of it.
If your original recording is a higher resolutions, it has more data to loose, but still be above a true 16 bit/44.1 khz file...so technically spekaer, higher resolution is better. Also, if it is capable of much higher resolution, and you don't max it out, you are giving your gear that much more breathing room, or headroom as well, which is never bad thing.
Did any of that make sense? I just got back from my 7th night in a row at the gym and spent the last 30 minutes grappling...so I am a little bit groggy...I hope it made sense...
WaltChief Cook and Bottle WasherMember
Since: May 10, 2002
Apr 11, 2005 11:51 pm Personaly, I chould not answer that in any definitive terms.
Record at 44.1/16 vs dither to 44.1/16 at time of burn.
Hmmmm
So I apply Wally logic. That an a buck fifty will buy you a coffee. It goes as follows:
I do hear a difference between 44.1/16 and 48/24 in my raw recordings....so.... If I can hear a difference I can better mix and master at the higher rate. My proficency not necessaraly the computer. I also believe that working closer to the 'native' 32bit resolution of the programs that process the files may be a logical move. So I record 48/24, mix and master 48/32 (allow conversion to 32 by Cubase) and dither down for the final.
Imperical data..... nada ... No way I can do a side by side test with any reliable constants to ever find out for sure.
el musicoFreeleance Producer/Engineer/GtrMember
Since: Aug 11, 2002
Apr 12, 2005 01:31 pm Ok... it is best to record at higher resolution and sample rate and convert later... as long as the sample rate math is easy like 88.2 to 44.1, it isn't worth it to track at 48kHz and convert to 44.1 since some extrapolation of the data is needed, making it less accurate.
So I gues it is best to record for the media format that will be the final product if it's CD or DVD. DVD/DVD-A supports 48kHz, 88.2kHz, 96kHz and 192kHz at 24bit and CD is 16bit 44.1 kHz.
If a client is looking to record an album they want on CD then I record at 24bit 44.1 and dither down to 16bit... if they plan to release the material on DVD later then I would record at 24bit 88.2kHz. Of course I could just record it at 44.1 and convert up to 88.1 and it would sound fine... but it wouldn't sound like 88.1 would normally since the information wouldn't be there. follow?
WaltChief Cook and Bottle WasherMember
Since: May 10, 2002
Apr 13, 2005 01:14 am Yup, I think I have your logic. And you are right, if you record 44.1 and print at 88.1, the sound would be as good as 44.1 can get.
It's the same thing as drawing a curve in trig class. You use 20 sample points to draw a curve and it looks kinda close. You use 40 and it is more accurate. Same as graphics on the computer, at some point as you zoom in it is just dots. The closer together the dots the more accurate the immage especially as you increase scrutiny.
Theoreticaly there may be more loss in printing something 88.1 from 44.1 as there is no dithering up. Every two samples will be the same. When dithering down there is some extrapolation of the waveform considered to 'smooth out' the transition.
