lowdown
Max Output Level: -85 dBFS
- Total Posts : 300
- Joined: 2004/04/14 16:07:38
- Location: Essex - UK
- Status: offline
Dithering...?
Hi All.
What does Dithering actually do ?
I noticed several different types in the Export Auido menu..
And do you need to Dither when you bounce / render soft synths down
to Audio [save Cpu].
Just Curious really , [ and proberly stupid ]
Using SP7...
Garry
|
ohhey
Max Output Level: 0 dBFS
- Total Posts : 11676
- Joined: 2003/11/06 16:24:07
- Location: Fort Worth Texas USA
- Status: offline
RE: Dithering...?
2008/03/20 11:48:02
(permalink)
ORIGINAL: lowdown
Hi All.
What does Dithering actually do ?
I noticed several different types in the Export Auido menu..
And do you need to Dither when you bounce / render soft synths down
to Audio [save Cpu].
Just Curious really , [ and proberly stupid ]
Using SP7...
Garry
When going from a higher bit level like 24bit down to 16bit you have to use much fewer loudness values and it gets nasty. So some smart folks came up with a method to smooth things out by adding some low level "noise" to the signal, that's dithering. You would think adding noise would be bad but in this case it's very cleverly constructed noise that really works ! Note: Some methods are more clever then others.. POW-R seems to be the best there is.
post edited by ohhey - 2008/03/20 12:07:22
|
Geokauf
Max Output Level: -72 dBFS
- Total Posts : 912
- Joined: 2003/12/01 20:59:45
- Location: Port Chester, NY, USA
- Status: offline
RE: Dithering...?
2008/03/20 12:12:16
(permalink)
Hello,
Dithering is a mathematic process that is not specific to audio. Dithering can also be performed on digital photos and video. Dithering is a way of resolving a change in bit depth (from higher to lower) while preserving the detail at the lowest volume levels of audio (or the detail in the darkest part of a photo), that is, the very, very tail ends of a reverb decaying into silence (the 1 or 2 bit level). Dithering has no effect on the bright parts of a photo. Likewise, audio dithering has no affect on the loud parts of the audio. So if your tune is a "full-tilt-boogie" then dithering will have no affect on the mix. But if your finished track is a solo guitarist or Yo Yo Ma playing the Bach Unaccompanied Cello Suites, then you would most probably employ dithering. I use PE 5.2 and work in 24 bit within Sonar and mix to 16 bit. I leave the dithering turned on in the options for purely superstitious reasons. If dithering is on when you export from 24 bit to 16 bit Sonar will automatically perform the necessary calculations. Can one hear the effect of dithering? I'm not so sure. I think at best we are talking about an extremely small incremental change.
GK
|
lowdown
Max Output Level: -85 dBFS
- Total Posts : 300
- Joined: 2004/04/14 16:07:38
- Location: Essex - UK
- Status: offline
RE: Dithering...?
2008/03/20 12:27:14
(permalink)
Ok Folks..
Thanks for that...
Garry
|
UnderTow
Max Output Level: -37 dBFS
- Total Posts : 3848
- Joined: 2004/01/06 12:13:49
- Status: offline
RE: Dithering...?
2008/03/20 12:27:41
(permalink)
ORIGINAL: Geokauf
Hello,
Likewise, audio dithering has no affect on the loud parts of the audio.
Quantisation distortion (what you get when you don't dither) has an effect of on the entire signal soft OR loud. On the other hand, the level of the distortion is low so it gets masked by loud signals but it is still there. For any dithering that isn't the last step (when you bring a mastered track or album down to 16 bits) you should use TPDF (Triangular Probability Distribution Function) dither. For the final dither you can use one of the noise shaped dithers like POW-R (included in Sonar). UnderTow
|
robby 
Max Output Level: -17 dBFS
- Total Posts : 5819
- Joined: 2006/05/13 15:53:45
- Status: offline
RE: Dithering...?
2008/03/20 13:00:37
(permalink)
And supposedly different dithers are better for different dynamics.
From WIKI
POW-R algorithms
POW-R users are offered a choice of two or three POW-R algorithms from which to select:
POW-R #1: optimized for simple program of average dynamic range, such as spoken word.
POW-R #2: optimized for low dynamic range program, such as rock music.
POW-R #3: optimized for complex, high dynamic range program, such as orchestral music.
The POW-R Consortium, LLC is an autonomous, non-competitive entity established for the sole purpose of maintaining and supporting the POW-R Algorithms. POW-R is a registered trademark of the POW-R Consortium, LLC (John La Grou, Chairman). POW-R is licensed to Apple Computer, Avid-Digidesign, SADiE, Sonic Studio, BIAS, Cakewalk Sonar, Merging Technologies, Ableton, Magix / Sequoia / Samplitude, and others.
So I use 2 and 3.
|
spindlebox
Max Output Level: -49 dBFS
- Total Posts : 2645
- Joined: 2007/05/30 07:56:11
- Location: Kansas City, MO
- Status: offline
RE: Dithering...?
2008/03/20 13:13:23
(permalink)
+1 on POW-R #2. I pretty much stick with that all the time.
|
Dr. Mac
Max Output Level: -78 dBFS
- Total Posts : 618
- Joined: 2006/07/19 22:50:18
- Location: Upstate New York
- Status: offline
RE: Dithering...?
2008/03/20 14:48:13
(permalink)
I've had great results using POW-R #3 on Rock/Metal music. Haven't tried #2, but now I think I will. Didn't realize the difference between #2 and #3.
RME FireFace 800, 3.4GHz quad-core AMD-64, 8 Gigs RAM Sonar X2a Producer, Fav. Plugs: Ozone 5 Advanced, Waves, Sonnox, Melodyne, Voxengo, SSL Native, Drumagog 5 Platinum
|
lowdown
Max Output Level: -85 dBFS
- Total Posts : 300
- Joined: 2004/04/14 16:07:38
- Location: Essex - UK
- Status: offline
RE: Dithering...?
2008/03/20 16:19:29
(permalink)
Hey Robby..
Good one , and thanks for that.
Garry
|