VST Oversampler 1.4 (with SSE/3DNow! support)
- KVRAF
- Topic Starter
- 2187 posts since 25 Jan, 2007 from the back room, away from his wife's sight (or so he thinks)
thanks 
Cakewalk by Bandlab / FL Studio
Squire Stratocaster / Chapman ML3 Modern V2 / Fender Precision Bass
Formerly known as arke, VladimirDimitrievich, bslf, and ctmg. Yep, those bans were deserved.
Squire Stratocaster / Chapman ML3 Modern V2 / Fender Precision Bass
Formerly known as arke, VladimirDimitrievich, bslf, and ctmg. Yep, those bans were deserved.
- KVRAF
- Topic Starter
- 2187 posts since 25 Jan, 2007 from the back room, away from his wife's sight (or so he thinks)
Bad News...
I just realized that I simply don't have the time, at least in the near future, to do much more with Oversampler. I'm on vacation, I have a project to finish, I have two new projects I need to start, and courses start again on October 15.
I'll release 1.3 with a few bugfixes, and then put it on hold, probably with some source code so that somebody else can continue working on it. I'm thinking of opening up a sourceforge project for the oversampler, and setting up an SVN repository. Known VST devs would be welcome to have repository access, and others that send in a few good patches will get access as well. Thoughts on that?
I just realized that I simply don't have the time, at least in the near future, to do much more with Oversampler. I'm on vacation, I have a project to finish, I have two new projects I need to start, and courses start again on October 15.
I'll release 1.3 with a few bugfixes, and then put it on hold, probably with some source code so that somebody else can continue working on it. I'm thinking of opening up a sourceforge project for the oversampler, and setting up an SVN repository. Known VST devs would be welcome to have repository access, and others that send in a few good patches will get access as well. Thoughts on that?
Cakewalk by Bandlab / FL Studio
Squire Stratocaster / Chapman ML3 Modern V2 / Fender Precision Bass
Formerly known as arke, VladimirDimitrievich, bslf, and ctmg. Yep, those bans were deserved.
Squire Stratocaster / Chapman ML3 Modern V2 / Fender Precision Bass
Formerly known as arke, VladimirDimitrievich, bslf, and ctmg. Yep, those bans were deserved.
-
- KVRAF
- 1627 posts since 3 Oct, 2001 from Thailand
Thank you very much for your works, Chris. 
Just the bugfix that would make it compatible with Tracktion would be more than good enough for me. And I totally support the idea to open source it so that the hard work you have started can live on.
Just the bugfix that would make it compatible with Tracktion would be more than good enough for me. And I totally support the idea to open source it so that the hard work you have started can live on.
Peace, my friends. I'm not seeking arguments here. 
- KVRAF
- Topic Starter
- 2187 posts since 25 Jan, 2007 from the back room, away from his wife's sight (or so he thinks)
Yes, I will make a better filter for the next release.8-bitter wrote:Hi Chris,is there any chance you could fix the high frequency attenuation for the 1.3 release.
Cakewalk by Bandlab / FL Studio
Squire Stratocaster / Chapman ML3 Modern V2 / Fender Precision Bass
Formerly known as arke, VladimirDimitrievich, bslf, and ctmg. Yep, those bans were deserved.
Squire Stratocaster / Chapman ML3 Modern V2 / Fender Precision Bass
Formerly known as arke, VladimirDimitrievich, bslf, and ctmg. Yep, those bans were deserved.
-
- KVRian
- 1028 posts since 20 Jun, 2007
it's a very very good idea. Open source is a way to improve just about anything. And likely...there are many in these parts that will do just that.Chris Walton wrote:Bad News...
I just realized that I simply don't have the time, at least in the near future, to do much more with Oversampler. I'm on vacation, I have a project to finish, I have two new projects I need to start, and courses start again on October 15.
I'll release 1.3 with a few bugfixes, and then put it on hold, probably with some source code so that somebody else can continue working on it. I'm thinking of opening up a sourceforge project for the oversampler, and setting up an SVN repository. Known VST devs would be welcome to have repository access, and others that send in a few good patches will get access as well. Thoughts on that?
- KVRAF
- Topic Starter
- 2187 posts since 25 Jan, 2007 from the back room, away from his wife's sight (or so he thinks)
'k, working on it again, on and off. Most of the code is rewritten because I wrote the code while braindead

The code is back to being 2x oversampling only. I had to reinstall my computer recently so matlab isn't installed, but once it is, I will finally put in some decent up/downsampling filters (right now, I have a simple RC filter for upsampling, and a 9 tap FIR filter I stole from MusicDSP for downsampling). They'll be linear phase FIR filters, should be quite transparent. I will have to see how many taps I need to get the frequency performance I want.
Basically, what all this means is that the high frequency attenuation will be gone, and the aliasing will be a little less than before.
Also, I think I've finally worked out the audioMaster <-> client interaction, which should make things a bit more stable, and I now hook into getParameter and setParameter and give the host and plugin the correct AEffect structures.
What that means is that Tracktion and Orion should work correctly now.
I added a prefix to the name ("[OS] PolyIBlit" instead of "PolyIBlit"), and generally the code is a bit faster at what it does, which means with the better but more expensive filters the added consumption should be minimal.
I'm really sorry for letting this slide for so long but I just haven't had the time...
The code is back to being 2x oversampling only. I had to reinstall my computer recently so matlab isn't installed, but once it is, I will finally put in some decent up/downsampling filters (right now, I have a simple RC filter for upsampling, and a 9 tap FIR filter I stole from MusicDSP for downsampling). They'll be linear phase FIR filters, should be quite transparent. I will have to see how many taps I need to get the frequency performance I want.
Basically, what all this means is that the high frequency attenuation will be gone, and the aliasing will be a little less than before.
Also, I think I've finally worked out the audioMaster <-> client interaction, which should make things a bit more stable, and I now hook into getParameter and setParameter and give the host and plugin the correct AEffect structures.
What that means is that Tracktion and Orion should work correctly now.
I added a prefix to the name ("[OS] PolyIBlit" instead of "PolyIBlit"), and generally the code is a bit faster at what it does, which means with the better but more expensive filters the added consumption should be minimal.
I'm really sorry for letting this slide for so long but I just haven't had the time...
Cakewalk by Bandlab / FL Studio
Squire Stratocaster / Chapman ML3 Modern V2 / Fender Precision Bass
Formerly known as arke, VladimirDimitrievich, bslf, and ctmg. Yep, those bans were deserved.
Squire Stratocaster / Chapman ML3 Modern V2 / Fender Precision Bass
Formerly known as arke, VladimirDimitrievich, bslf, and ctmg. Yep, those bans were deserved.
- KVRAF
- 12615 posts since 7 Dec, 2004
you would probably be better off using windowed sincs for both up sampling and decimation. the process is very uniform and you could offer adjustment of the windowing and other options on the gui. before you spend too much time fiddling you might want to look at my implementation of that stuff (adft.cpp / minblepdata.cpp) and this source here which only does decimation, but should give you an idea of how the process works.
http://xhip.cjb.net/temp/public/renderm.cpp
http://xhip.cjb.net/temp/public/renderm.cpp
- KVRAF
- Topic Starter
- 2187 posts since 25 Jan, 2007 from the back room, away from his wife's sight (or so he thinks)
Yeah, I'm using a windowed sinc for downsampling now, I plan to keep using them too because they're easier to optimize (linear phase for the win). I'll take a look at your source though, because knowing you you have some masterful super optimization trick of doom in thereaciddose wrote:you would probably be better off using windowed sincs for both up sampling and decimation. the process is very uniform and you could offer adjustment of the windowing and other options on the gui. before you spend too much time fiddling you might want to look at my implementation of that stuff (adft.cpp / minblepdata.cpp) and this source here which only does decimation, but should give you an idea of how the process works.
http://xhip.cjb.net/temp/public/renderm.cpp
Cakewalk by Bandlab / FL Studio
Squire Stratocaster / Chapman ML3 Modern V2 / Fender Precision Bass
Formerly known as arke, VladimirDimitrievich, bslf, and ctmg. Yep, those bans were deserved.
Squire Stratocaster / Chapman ML3 Modern V2 / Fender Precision Bass
Formerly known as arke, VladimirDimitrievich, bslf, and ctmg. Yep, those bans were deserved.
-
- KVRAF
- 3150 posts since 22 Dec, 2004
Glad to see you've started up again on this. 
-
- KVRian
- 889 posts since 10 Jul, 2004
people think that it only helps prevent aliasing.....but it also helps improve definition of filters,compressors etc etc(aiding the code i guess?)....although has to be said some of the plugs allready come oversampled nowadays.
i wish i had this years ago (when oversampling the entire sequencer was not feasable in terms of cpu).
well my cpu still don't allow me to run the entire song at double the samplin rate.....so this plug is still very welcome here. thanks....probably plug of the year.
waitin for 1.3
i wish i had this years ago (when oversampling the entire sequencer was not feasable in terms of cpu).
well my cpu still don't allow me to run the entire song at double the samplin rate.....so this plug is still very welcome here. thanks....probably plug of the year.
waitin for 1.3
- KVRAF
- Topic Starter
- 2187 posts since 25 Jan, 2007 from the back room, away from his wife's sight (or so he thinks)
thanks 
I should be working on something else however today I did a bit of work on the oversampler again. Next thing up is improving the filters, and then 1.3 should be just about done.
Any volunteers to do beta testing with a debug version before I give it to the masses? In particular, I'm looking for people to test Orion, Tracktion, and Cubase, as those have given the most problems.
I should be working on something else however today I did a bit of work on the oversampler again. Next thing up is improving the filters, and then 1.3 should be just about done.
Any volunteers to do beta testing with a debug version before I give it to the masses? In particular, I'm looking for people to test Orion, Tracktion, and Cubase, as those have given the most problems.
Cakewalk by Bandlab / FL Studio
Squire Stratocaster / Chapman ML3 Modern V2 / Fender Precision Bass
Formerly known as arke, VladimirDimitrievich, bslf, and ctmg. Yep, those bans were deserved.
Squire Stratocaster / Chapman ML3 Modern V2 / Fender Precision Bass
Formerly known as arke, VladimirDimitrievich, bslf, and ctmg. Yep, those bans were deserved.
