PG8X (inspired by the JX8P): new beta version uploaded
-
- KVRAF
- 2747 posts since 13 Feb, 2012 from Amsterdam
This is interesting, I found that the VST version (the 32 bit one in the 10.6 folder on Dropbox) also works in Live 64 bit. However, this does nog apply to the AU version.
-
- KVRian
- Topic Starter
- 960 posts since 27 Jun, 2009 from Germany
Did you pick up the files before I sent the post about he wrong compilation of the VST's? In this case, you would still have picked up a fat binary (i.e. 32/64 in one).BDeep wrote:This is interesting, I found that the VST version (the 32 bit one in the 10.6 folder on Dropbox) also works in Live 64 bit. However, this does nog apply to the AU version.
Otherwise, it is possible that the host automatically bridges. I don't know about Live, but reaper does that.
Martin
- KVRian
- 727 posts since 30 May, 2007 from Barkhamsted, CT, USA
I believe Live automatically bridges or rather bufers automatically; Running Live 32-bit, hosting 32-bit plugins in a 64-bit OS, so I assume it's not the plguins that are bridged, but the audio output that's buffered. Or am I talking through my butt here?
Martin, unless you really feel the need to cater to the Mac PPC crowd... folks running Macs 2005/6 and older; you can drop the fat binary. Mac has been Intel-only since 2006/7, and the OS has been using Universal Binary (more or less) since then. The Universal Binary portion has been ignored for the last two years, OSX 10.8/10.9 runs only the x86 portion of the code. To best future-proof (in computer terms) the plugins, you can drop the fat binary part. (At least that's how I understand it).
Martin, unless you really feel the need to cater to the Mac PPC crowd... folks running Macs 2005/6 and older; you can drop the fat binary. Mac has been Intel-only since 2006/7, and the OS has been using Universal Binary (more or less) since then. The Universal Binary portion has been ignored for the last two years, OSX 10.8/10.9 runs only the x86 portion of the code. To best future-proof (in computer terms) the plugins, you can drop the fat binary part. (At least that's how I understand it).
-
- KVRian
- Topic Starter
- 960 posts since 27 Jun, 2009 from Germany
Maybe I was using the wrong terminology here. When I was talking about 'fat' binaries, I did not have PPC in mind. Simply the combination of 32bit and 64bit in one binary.lionscub68 wrote:Martin, unless you really feel the need to cater to the Mac PPC crowd... folks running Macs 2005/6 and older; you can drop the fat binary. Mac has been Intel-only since 2006/7, and the OS has been using Universal Binary (more or less) since then. The Universal Binary portion has been ignored for the last two years, OSX 10.8/10.9 runs only the x86 portion of the code. To best future-proof (in computer terms) the plugins, you can drop the fat binary part. (At least that's how I understand it).
To be honest, I am still a bit puzzled about the settings in Xcode. There is the main target, which can be 32bit or 64 bit, and there is the switch in the architecture settings which can be either 32bit, 64bit or 32/64bit (standard). The latter is what I thought to be 'fat' binary.
Please, if anybody could enlighten me more on that...
Cheers,
Martin
- KVRian
- 727 posts since 30 May, 2007 from Barkhamsted, CT, USA
Time to fire off a support email to Apple's bright boys.
-
- KVRian
- Topic Starter
- 960 posts since 27 Jun, 2009 from Germany
OK. New trial... I found some information on problems for the Mac on the wdl-ol forum, and tried to apply the patches to the compiler before building it.
The new builds are again combined 23/64 bit plugins, and are in the Mac-SDK10.6/Mac-32-64 folder.
I also included some smoothing of the parameters. Please, give these ones a try.
Thanks,
Martin
The new builds are again combined 23/64 bit plugins, and are in the Mac-SDK10.6/Mac-32-64 folder.
I also included some smoothing of the parameters. Please, give these ones a try.
Thanks,
Martin
- KVRian
- 727 posts since 30 May, 2007 from Barkhamsted, CT, USA
Checking it out now.
P.S. Martin - need to repost link to dropbox folder per update, otherwise I (we) have to go fishing back through the string to find the link
P.S. Martin - need to repost link to dropbox folder per update, otherwise I (we) have to go fishing back through the string to find the link
- KVRian
- 727 posts since 30 May, 2007 from Barkhamsted, CT, USA
Reporting back:martin_l wrote:OK. New trial... I found some information on problems for the Mac on the wdl-ol forum, and tried to apply the patches to the compiler before building it.
The new builds are again combined 23/64 bit plugins, and are in the Mac-SDK10.6/Mac-32-64 folder.
I also included some smoothing of the parameters. Please, give these ones a try.
Thanks,
Martin
No dice, neither VST nor the AU show in Live 9.1.1 (32-bit) on OSX 10.9.2.
The VSTs show in my audio editor VST list but neither work (same message as I had before, see my screenshot a few pages back), and the Au does not list there either.
Sorry for the bad news. (Now I gotta backtrack and find the ones that worked)
Last edited by lionscub68 on Sat Mar 29, 2014 5:01 pm, edited 1 time in total.
-
- KVRAF
- 3330 posts since 18 May, 2003 from Sweden
Some good news for a change: the 32 bit VST and AU both work in a 32 bit environment (Tracktion 5.2.4 32 bit and Audio Plugin Player) in 10.9.2.
I couldn't get them to work in 10.8.4, but I'll try again later.
The Mac 32/64 bit AU plugs likewise work in 64 bit environments (Garage Band and Tracktion 5.2.4 64 bit) in 10.9.2. Good work!
However, they don't work in 32 bit environments in the same Mavericks System. Again, I'll test later in 10.8.4.
Kind regards,
Joachim
I couldn't get them to work in 10.8.4, but I'll try again later.
The Mac 32/64 bit AU plugs likewise work in 64 bit environments (Garage Band and Tracktion 5.2.4 64 bit) in 10.9.2. Good work!
However, they don't work in 32 bit environments in the same Mavericks System. Again, I'll test later in 10.8.4.
Kind regards,
Joachim
If it were easy, anybody could do it!
- KVRian
- 727 posts since 30 May, 2007 from Barkhamsted, CT, USA
Just a question about the combined 32-64 plugins:
Are they written/compiled like this, intermixed:
If 32 then, if 64 then (process A)
next
If 32 then, if 64 then (process B)
etc
or this (two distinct blocks of 'clean' code):
If 32 then (big block of 32 code only)
if 64 then (big block of 64 code only)
etc
If it's the former then one 'endif' brings the whole house of cards crashing down.
Hope the logic train helps a little here.
Are they written/compiled like this, intermixed:
If 32 then, if 64 then (process A)
next
If 32 then, if 64 then (process B)
etc
or this (two distinct blocks of 'clean' code):
If 32 then (big block of 32 code only)
if 64 then (big block of 64 code only)
etc
If it's the former then one 'endif' brings the whole house of cards crashing down.
Hope the logic train helps a little here.
-
- KVRian
- Topic Starter
- 960 posts since 27 Jun, 2009 from Germany
I am nt sure, to be honest. This fact is happening behind the scenes. In my code, there is nothing which is specific for 32bit or 64bit. As far as I understand, the binaries contain the complete code for both, and when the library is loaded, the correct block is chosen. But I might be wrong here.lionscub68 wrote:Just a question about the combined 32-64 plugins:
Are they written/compiled like this, intermixed:
If 32 then, if 64 then (process A)
next
If 32 then, if 64 then (process B)
etc
or this (two distinct blocks of 'clean' code):
If 32 then (big block of 32 code only)
if 64 then (big block of 64 code only)
etc
If it's the former then one 'endif' brings the whole house of cards crashing down.
Hope the logic train helps a little here.
Martin
- KVRian
- 727 posts since 30 May, 2007 from Barkhamsted, CT, USA
So suffice it to say the Mac translation/compiler is to blame. 
Maybe contact the TAL developer to see what he's using to compile all his OS verisons.
Maybe contact the TAL developer to see what he's using to compile all his OS verisons.
-
AdmiralQuality AdmiralQuality https://www.kvraudio.com/forum/memberlist.php?mode=viewprofile&u=83902
- Banned
- 6657 posts since 10 Oct, 2005 from Toronto, Canada
martin_l wrote:I am nt sure, to be honest. This fact is happening behind the scenes. In my code, there is nothing which is specific for 32bit or 64bit. As far as I understand, the binaries contain the complete code for both, and when the library is loaded, the correct block is chosen. But I might be wrong here.lionscub68 wrote:Just a question about the combined 32-64 plugins:
Are they written/compiled like this, intermixed:
If 32 then, if 64 then (process A)
next
If 32 then, if 64 then (process B)
etc
or this (two distinct blocks of 'clean' code):
If 32 then (big block of 32 code only)
if 64 then (big block of 64 code only)
etc
If it's the former then one 'endif' brings the whole house of cards crashing down.
Hope the logic train helps a little here.
Martin
Ignore this, it's ridiculous. The code is the code, it's the compiler settings that determine whether it becomes 32 or 64 bit. (You might want to post in the Dev board. There's all kinds of wacky misinformation out here.)
- KVRian
- 727 posts since 30 May, 2007 from Barkhamsted, CT, USA
I'm not a programmer beyond simple data processing at work.
Don't think it hurts to help in this community with these simple thoughts, even if my ideas are full of crap.
Just want Martin to clear this frustrating hurdle.
Don't think it hurts to help in this community with these simple thoughts, even if my ideas are full of crap.
Just want Martin to clear this frustrating hurdle.

