#882998 - 12/20/05 11:06 PM
16 bit or 24 bit
|
mixnut
Member
Registered: 12/20/05
Posts: 2
|
Offline
|
|
I currently receive omf files from avid editors at 48k/16 bit. I mix using protools and bounce to disk my final mixdown at 48k/16 bit for layback to digibeta. Would there be any advantage by going to 48k/24 bit in protools just to have my VO recorded at 48k/24 bit?? I would still have to bounce to disk 48k/16 bit!!! Would I hear a difference?? Would my source sound files from the avid sound better??? Can anyone help???
Thanks
|
|
Top
|
|
|
|
|
#882999 - 01/19/06 07:13 PM
Re: 16 bit or 24 bit
|
AudioMaverick
Platinum Member
Registered: 05/19/01
Posts: 1790
Loc: Outskirts of Big Bear, CA,UNIT...
|
|
Offline
|
|
I can't believe no one addressed this post. I even missed it last week...
Basically, editing a file causes math rounding errors in the lower bits... usually the lower 3 or 4. This will introduce distortion. But, editing at a higher a higher resolution than the final product will truncate off those bits in the final product. It could potentially give a higher resolution or cleaner final product, depending on how the recorded material utilizes the frequency response and dynamic range.
Also, 48kHz/24-bit is minimum DVD quality audio. Having that resolution at the source will make it nicer to transfer over from CD, in the future. You never know what you will be doing with the files in a few years!
I have recorded my earliest files (1998) at 44.1kHz/24.8-bit. If I ever find a reason to go to DVD audio, I'll get that much more satisfaction from the transfer.
_________________________
"It's all about the... um-m-m, uh-h-h..."
|
|
Top
|
|
|
|
|
#1790089 - 08/02/07 12:59 AM
Re: 16 bit or 24 bit
[Re: AudioMaverick]
|
Mezclar_NY
Member
Registered: 07/29/07
Posts: 17
|
|
Offline
|
|
Another way to look at it is this. Say you have an image that has 16dots per square inch from a printer, right? And you have another image that produces 24 dots/square inch, right? Obviously, the 24dpi makes a more accurate image.
Now, you make printed copies of those pages on a high res machine, one for the sheet with the image that produces 16dpi and the other with 24dpi.
When the copies come out of the machine which picture is going to have a better resolution? The one with 16dpi or 24dpi? I guess you could retouch the 16dpi with photoshop or something to make it look like 24dpi but there is a difference compared to the 24dpi image and it won't be as "true" as a 24dpi image.
This is the only way I can describe how audio reproduction can work when you work with 16bit files or 24bit files. Either way when you transfer down to the same pipeline there will be a difference. Would the end result be what you need? Your ears will tell you.
I would go with Audio Mavrick's advice.
_________________________
If she shakes dat azz while you play your bass, you know you got groove: even if it's only a one note samba.
|
|
Top
|
|
|
|
|
#1799463 - 08/23/07 09:09 AM
Re: 16 bit or 24 bit
[Re: AudioMaverick]
|
Neil Wilkes
Gold Member
Registered: 07/06/02
Posts: 506
Loc: London, UK
|
|
Offline
|
|
I can't believe no one addressed this post. I even missed it last week...
Basically, editing a file causes math rounding errors in the lower bits... usually the lower 3 or 4. This will introduce distortion. But, editing at a higher a higher resolution than the final product will truncate off those bits in the final product. It could potentially give a higher resolution or cleaner final product, depending on how the recorded material utilizes the frequency response and dynamic range.
Also, 48kHz/24-bit is minimum DVD quality audio. Having that resolution at the source will make it nicer to transfer over from CD, in the future. You never know what you will be doing with the files in a few years!
I have recorded my earliest files (1998) at 44.1kHz/24.8-bit. If I ever find a reason to go to DVD audio, I'll get that much more satisfaction from the transfer.
I missed it too.
You're off on the "audio for DVD" requirements though.
DVD-Video LPCM support is only actually mandatory at 16/48. 24 bit and 96KHz sample rates are optional only. You get warnings when using 24/48 audio that not all players can handle it, and some will truncate to 16 bits. As the resulting truncation will be in the LSB, it will be at around -95dB (unless the music has been massively overcompressed of course) and close to inaudible.
DVD-Audio LPCM support goes from CD resolution at 16/44.1 and all the way up to 24/192 in stereo, using every combination of bit depth (16/20/24) and both sample rate families (44.1 and 48KHz).
In multichannel it gets more complicated.
CD resolution to 24/48 in 5.1 LPCM.
Up to 24/96 in Quad.
More than 4 channels at 24/96 or 24/88.2, and use of MLP (Meridian Lossless Packing) or PPCM as it sometimes gets called is mandatory.
Personally I prefer the sound of 24 bit audio over 16 bit. It seems more open & yet more detailed at the same time. Probably because whenever you perform any DSP operation, the wordlength expands and the resulting rounding errors all add up, and are noticeable to the ears. The more DSP you do, (Plugins, EQ, whatever) the larger the resulting rounding errors and the more "brittle" the file sounds to me.
LSB at 24 bits is all the way down at -143dB. This is not close to inaudible, it IS inaudible.
|
|
Top
|
|
|
|
|
#1830101 - 10/25/07 10:50 AM
Re: 16 bit or 24 bit
[Re: Neil Wilkes]
|
audiorulez
Senior Member
Registered: 06/20/07
Posts: 493
|
|
Offline
|
|
Audio for video is 16/48, not 24/48.
|
|
Top
|
|
|
|
|
#1833131 - 10/31/07 06:35 AM
Re: 16 bit or 24 bit
[Re: audiorulez]
|
alfonso
Platinum Member
Registered: 11/06/00
Posts: 1032
Loc: Fregene, Italy.
|
|
Offline
|
|
The greatest advantage of working at higher bit depths is the quality of processing. Processing introduces errors, higher the bit depth, less evident the errors. This is the main reason for keep working on higher depths, even if the last product will be dithered down to 16 bits.
The higher samplerates sound much better as far as they are kept high, a not perfect downsampling algorithm will make more damage than keeping the work at the final samplerate. What will benefit very much of an higher samplerate work is all the synthesis math. Oscillators don't alias, filters shine and envelopes are much faster at 96khz. This could be a reason to work this way, but many good synth algos are internally oversampled. That's a particular scenario though...
|
|
Top
|
|
|
|
|
|
|
