I am developing an audio retouch program
-
- KVRer
- 9 posts since 6 Jul, 2026 from Russia
Here is the example: . I ask for the feedback. Thank you.
- KVRAF
- 3615 posts since 28 Jan, 2006 from Phoenix, AZ
-
- KVRer
- Topic Starter
- 9 posts since 6 Jul, 2026 from Russia
Before:
After:
After:
-
- KVRer
- Topic Starter
- 9 posts since 6 Jul, 2026 from Russia
Other variant, I think this one is better:
- KVRian
- 1013 posts since 6 Aug, 2005 from England
To my ears there’s a slight blurring of the percussion sounds, suggesting phase changes?
Dave Hoskins. http://www.quikquak.com
-
- KVRer
- Topic Starter
- 9 posts since 6 Jul, 2026 from Russia
Thank you. No phase changes, just spectral manipulation. Now I am busy with the other method, I'll show it when it is ready.
-
- KVRer
- Topic Starter
- 9 posts since 6 Jul, 2026 from Russia
The results are subtle.
Last edited by alexeykoch on Thu Jul 16, 2026 11:45 pm, edited 1 time in total.
- KVRist
- 497 posts since 23 Sep, 2003 from Hungary
-
- KVRer
- Topic Starter
- 9 posts since 6 Jul, 2026 from Russia
It's hard for me tho describe the first method, but the second one is connected with echo signals.
-
- KVRer
- Topic Starter
- 9 posts since 6 Jul, 2026 from Russia
___Main signal
/ \ _ Echo
| | / \
| |_______| |
1 2 3 4 5 6 7
As you can see on the spectrum an echo signal is the small copy of the main signal. Now I try to filter out the echos, which move closely to the signal. First whe must find the direction of the echo, then we use spectral masking. Here the distance is 6 bins. I made an assumption that short echo signals are connected with ADC, which is not perfect.
/ \ _ Echo
| | / \
| |_______| |
1 2 3 4 5 6 7
As you can see on the spectrum an echo signal is the small copy of the main signal. Now I try to filter out the echos, which move closely to the signal. First whe must find the direction of the echo, then we use spectral masking. Here the distance is 6 bins. I made an assumption that short echo signals are connected with ADC, which is not perfect.
