[opendtv] Re: Please ignore and delete previous message on "Math of oversampling"

  • From: Jeroen Stessen <jeroen.stessen@xxxxxxxxxxx>
  • To: opendtv@xxxxxxxxxxxxx
  • Date: Mon, 16 Oct 2006 10:14:07 +0200

Dear Mark, 

I can't ignore your message to Tom, because at that time I was also 
involved in generating some "oversampled" images. You have my permission 
to use those too, if you wish. I say this because Tom and I have used 
entirely different methods for bandwidth reduction. Tom used a block-
based (!) DCT and discarded the upper half of the coefficients in the 
frequency domain. (This is asking for blocking artefacts.) I used a 
cascaded downsampling and upsampling (i.e. scalers), which only loses 
the higher frequencies. My method may also introduce a bit of aliasing, 
because during the downsampling the other samples are actually thrown 
away. The perfect experiment would also include an encoding/decoding 
pass on the downsampled signal, which I didn't apply. 

-- Jeroen

| From:     Jeroen H. Stessen   | E-mail:  Jeroen.Stessen@xxxxxxxxxxx |
| Building: SFJ-5.22 Eindhoven  | Deptmt.: Philips Applied Technologies |
| Phone:    ++31.40.2732739     | Visiting & mail address: Glaslaan 2 |
| Mobile:   ++31.6.44680021     | NL 5616 LW Eindhoven, the Netherlands |
| Skype:  callto:jeroen.stessen | Website: http://www.apptech.philips.com/ 

Other related posts: