Playback Processing in the Frequency Domain

3 posts / 0 new
Last post
DSPC_Peaston
Offline
Last seen: 2 weeks 5 days ago
Joined: 10/04/2016 - 4:16pm
Playback Processing in the Frequency Domain

How does one use the frequency domain for playback processing? Are there any advantages/drawbacks to using this instead of time domain processing?

DSPC_Peaston
Offline
Last seen: 2 weeks 5 days ago
Joined: 10/04/2016 - 4:16pm

Disadvantages of doing processing in the frequency domain are:

Increased latency involved in the FFT/Windowing stage (8ms minimum using 256 blocksize, increasing with higher blocksize).

Blocksize=256
 
 
 
 
 
 
Stopband Attenuation(dB)
1_31
32-45
46-59
60-73
74-87
88-102
Latency(samples) added
384
640
896
1152
1408
1664
Latency(ms)
8
13.3
18.7
24
29.3
34.67

Higher memory requirement

Possibly more setup when managing many frequency bins

Advantages of doing processing in the frequency domain are:

Less computations required

Higher resolution filterbanks/crossovers

Linear Phase

Constant, flat group delay

Sharper filter band edges.

I will attach an example system for further help soon.

DSPC_Peaston
Offline
Last seen: 2 weeks 5 days ago
Joined: 10/04/2016 - 4:16pm

Binary Datafilterbankeq.awd
Here is the example file for the FilterbankEQ.