Insert Synthesis Simple -Sound Forge - PC - MAGIX Community Forums Archive

Subject:Insert Synthesis Simple Posted by: Vic Andersen Date:3/12/2012 7:14:43 PM There seems to be a bug in this feature. When inserting a frequency sweep wave the output file top freq is only half as advertised. Example insert a sweep from 4000 to 8000 the output goes from 4000 to 6000. Exemple insert a sweep from 0 to 8000 the output goes from 0 to 4000. Insert a sweep from A to B the output goes from A to (B-A)/2
Subject:RE: Insert Synthesis Simple Reply by: ChristoC Date:3/12/2012 8:01:12 PM Vic, to which application are you referring? - I just looked at output of SoundForgePro10.0c with spectrum analyser and it's working exactly as advertised.
Subject:RE: Insert Synthesis Simple Reply by: Vic Andersen Date:3/19/2012 11:39:34 AM Sound Forge Audio Studio Version 10.0 (Build 177) not Pro I created a mono test wave combining a ten second 8kHz mixed with a sweep from 0Hx to 8kHz. When displayed on my spectrum analyzer I can see the constant 8kHz signal as well as the sweep. The constant 8KHz is correct but the sweep is a straight line going from 0Hz to 4kHz. Other test wave files show a similar discrepency in the sweep frequency signals generated and no discrepency in the fixed frequency components.
Subject:RE: Insert Synthesis Simple Reply by: rraud Date:3/20/2012 11:12:36 AM Is this before or after saving a file? In what file format. What sampling rate are you using? If it is low, the Nyquist frequency rate would prevent any frequencies above half the sample rate. for instance, a 24kHz sample rate would only reach 12kHz
Subject:RE: Insert Synthesis Simple Reply by: Vic Andersen Date:3/23/2012 11:28:04 PM The 8kHz+(0-8kHz) test wave was created as a mono, 16bit, 44100 sample rate and saved to Test8000Sweep.wav. There is no Nyquest problem and this is proved by the constant 8KHz component being correctly displayed in my spectrograph. It is the sweep from 0 to 8kHz that displays from 0-4kHz. I have run this test using other frequency combinations and always the sweep component falls short. In general a sweep from AHz to BHz ends up going from AHz to (A+B)/2 Hz. I can send you the Test8000Sweep.wav file and a screen of my spectragraph analysizing this file if I had a way to do that. I just checked rraud and found your email address, so I'll take the liberty to send these two files now.

Subject:Insert Synthesis Simple

Posted by: Vic Andersen

Date:3/12/2012 7:14:43 PM

There seems to be a bug in this feature. When inserting a frequency sweep wave the output file top freq is only half as advertised.
Example insert a sweep from 4000 to 8000 the output goes from 4000 to 6000.
Exemple insert a sweep from 0 to 8000 the output goes from 0 to 4000.

Insert a sweep from A to B the output goes from A to (B-A)/2

Subject:RE: Insert Synthesis Simple

Reply by: ChristoC

Date:3/12/2012 8:01:12 PM

Vic, to which application are you referring?
- I just looked at output of SoundForgePro10.0c with spectrum analyser and it's working exactly as advertised.

Subject:RE: Insert Synthesis Simple

Reply by: Vic Andersen

Date:3/19/2012 11:39:34 AM

Sound Forge Audio Studio Version 10.0 (Build 177) not Pro

I created a mono test wave combining a ten second 8kHz mixed with a sweep from 0Hx to 8kHz. When displayed on my spectrum analyzer I can see the constant 8kHz signal as well as the sweep. The constant 8KHz is correct but the sweep is a straight line going from 0Hz to 4kHz.

Other test wave files show a similar discrepency in the sweep frequency signals generated and no discrepency in the fixed frequency components.

Subject:RE: Insert Synthesis Simple

Reply by: rraud

Date:3/20/2012 11:12:36 AM

Is this before or after saving a file? In what file format. What sampling rate are you using? If it is low, the Nyquist frequency rate would prevent any frequencies above half the sample rate. for instance, a 24kHz sample rate would only reach 12kHz

Subject:RE: Insert Synthesis Simple

Reply by: Vic Andersen

Date:3/23/2012 11:28:04 PM

The 8kHz+(0-8kHz) test wave was created as a mono, 16bit, 44100 sample rate and saved to Test8000Sweep.wav. There is no Nyquest problem and this is proved by the constant 8KHz component being correctly displayed in my spectrograph. It is the sweep from 0 to 8kHz that displays from 0-4kHz. I have run this test using other frequency combinations and always the sweep component falls short. In general a sweep from AHz to BHz ends up going from AHz to (A+B)/2 Hz.

I can send you the Test8000Sweep.wav file and a screen of my spectragraph analysizing this file if I had a way to do that. I just checked rraud and found your email address, so I'll take the liberty to send these two files now.