1. George Lungu says:

I need to look deeper into this. It might just be some arbitrary scaling.

2. fatih says:

Hello, George. Why do we divide the “cos” and “sin” totals by 20000?

3. Han says:

Hi, George;
It is very good one, I love it. I am learning how to use it properly. I plug in my own data points and time blocks such that I copy and paste my data points under Custom 4, and time blocks at Column 3, changed stop frequency to 10000, then choose Function #4 and run, I do not get an appropriate fourier transform. My data points is 2000 and time increase is 0.005. what should I do in order to make it work with any my own data points?
Thanks,

4. George says:

I want to build an inverse as well as a DFT both direct and inverse. I’ve been too busy lately with insurance claims and a contract I need to finish.

5. liberty says:

I have since discovered that having a second spread sheet opened slows the process. Once I close the second spreadsheet performance picks up considerably and it becomes very usable.

Its very nice. Ever considered an “inverse transform” option ?

6. Jan says:

Thanks. What was the largest data points you used besides the demo version?

7. George says:

Yes that’s easy to do, except it’s going to be slow. I chose the limits you see there for speed reasons.

8. Jan says:

Hi George, Any thoughts? If considering speed, what would be a realistic data point limit? Thanks.

9. Jan says:

10X of the current limit.

10. George says:

How many points you need?

11. Jan says:

This is a great tool! Thanks!
If I need to include more data points, how would I do it? Thanks.

12. George says:

Thanks Harold.

13. Harold Jang says:

I love this tool. used in work place to demo something.

14. george says:

Ishan, what you have here is not a general solution. The way the algorithm was written, it can only extract a FT from a real function. Right now I don’t have the time to add to it. Maybe later.

15. Ishan says:

I used this workbook and seemed pretty handy. Thanks. But I am having trouble doing an inverse FFT in excel using the values from this workbook(excel is returning complex values instead of real numbers). Can you post a solution to this? If it is possible to an inverse FFT compatible to the algorithm used in this workbook

16. george says:

How long is the data series? I wrote the whole thing for a certain length. You might have to go in the macro and change the number. Let me know the length.

17. J says:

Hi, I am trying to generate fourier transform graph on Unusual excel by using my own data points and time blocks. When I copy and paste my data points under Custom 1, I dont get an appropriate fourier transform. Should I type them in in a form of functional equation? Or can I just plug in numbers ? what do I have to do to generate fourier transform based on my own data points?

18. Geoge says:

Liberty,
That is unfortunate. Try minimizing the chart or the whole Excel window in 2010 while is calculating. It must be 10x faster this way. After it’s finished you could increase the chart or the window. Good luck!

19. liberty says:

Looked like a very useful tool, but is almost unusable. Its so slow in Excel 2010 – even for your examples. I think I will have to resort to FFTW or other package

20. George says:

Good luck!

21. Allan says:

George, Thanks, just as I suspected. I will let you know if I get anywhere, but since I am frustrated, I believe I will write my own C routine. Too bad I am unable to use yours, as it is a really nice tool.

22. George says:

Allan, I tried the old Mac versions of Excel before (2004) and I was never able to run macros. In my case simply it wouldn’t run no matter what and I got so frustrated I eventually gave up. This was not because of the division with zero which it is not uncommon in my worksheets. Windows Excel let you get away with that usually so unless it’s widespread I don’t worry about. Please let me know if you figure it out. Thanks, George

23. Allan says:

George, Thanks again for your prompt answer. I am now suspecting my problem is a Windows-Mac Excel incompatibility, since I am using Mac’s. One has “MS Excel for Mac 2004” (supposedly equivalent to Excel 2003), and another has “MS Excel for Mac 2011” (supposedly equivalent to Excel 2010) (aka Bill Gates strikes again). Neither versions work. I am trying to edit the macros, but so far I have had no luck in getting them run through to the end. I get a divide by zero run time exception with the later version of Excel, and nowhere on the earlier version. BTW, I have never encountered a Windows-Mac incompatibility with Excel previously. Oh well, I guess that I will just have to write my own version of your great tool. Thanks again for your help.

24. George says:

Allan, I am using 2003. In 2003 or earlier go to: Tools-> Macro -> Security -> Low (or Medium). Save then close and then re-open the sheet. George

25. Allan says:

Thanks for your prompt reply, but that is not the problem I am having. It appears that I am using an older version of Excel that doesn’t support the macros you are using. I assume that one has to use MS Office 2010 or later. True?

26. George says:

Allan, the fuctions are defined in column N to S and the index (1 or 2 , etc) just defined the order. You need to go in one of those columns and change the formula yourself to get the right function then change the index so that the model uses your custom function. George

27. Allan says:

Yours looks like a useful tool that I would like to use. However, the version (“Fourier4.xls”) I downloaded from this site doesn’t work. If I change the input “Function #” to anything other than “1”, the output does not change. As downloaded, there are constants in columns H, I, J, K and L. Any suggestions? Is there another site I can download your tool. Any help would be appreciated, as it would save me much time, since I would not have to write my own version. Thanks.

28. George says:

Thanks Paul!

29. Paul says:

Your FFT Excel program is a very nice teaching tool — thanks for making it available. Well done !

30. george says:

Alan, Sure I will help you but I didn’t understand the question. Be more detailed.

31. Alan says:

Hi! I have done some experiments in Neuroscience to see the membrane resonance of neurons. I am applying sine waves {sin(2pi*t/1000)*(t/2000))}. However, in my previous software I just asked it for doing, and now with the new one, I have to find equations… I was wondering if you can help me on it, please!
Thanks a lot!

32. Anonymous says: