| (4 intermediate revisions by 2 users not shown) | |||
| Line 4: | Line 4: | ||
| − | + | For the 5120 point FFT, can we downsample it by a factor of 5 to 1024 points? and then interpolate the FFT to get the 5120 points? | |
| + | * That's a very creative idea. The question is whether the interpolation gives the same as the original DFT. --[[User:Mboutin|Mboutin]] 15:11, 8 October 2009 (UTC) | ||
| + | * I guess not :( . | ||
| + | * Do you think it might be possible in some circumstances? Can we make assumptions about the signal that would guarantee it is possible? --[[User:Mboutin|Mboutin]] 08:53, 9 October 2009 (UTC) | ||
[[ ECE438 (BoutinFall2009)|Back to ECE438 (BoutinFall2009)]] | [[ ECE438 (BoutinFall2009)|Back to ECE438 (BoutinFall2009)]] | ||
Latest revision as of 03:53, 9 October 2009
hw5 discussion (ECE438BoutinFall09)
For the 5120 point FFT, can we downsample it by a factor of 5 to 1024 points? and then interpolate the FFT to get the 5120 points?
- That's a very creative idea. The question is whether the interpolation gives the same as the original DFT. --Mboutin 15:11, 8 October 2009 (UTC)
- I guess not :( .
- Do you think it might be possible in some circumstances? Can we make assumptions about the signal that would guarantee it is possible? --Mboutin 08:53, 9 October 2009 (UTC)
