Signal Paths - One for the engineers at Sony -Sound Forge - PC

Subject:Signal Paths - One for the engineers at Sony Posted by: Spheris Date:1/12/2004 9:27:02 AM Okay 1. 64 bit res, is it indeed and does it apply only to temp file with the internal fx, not to dx plugins.? 2. side effects of 24 bit plugins as they show in the properties of the plugin display vs on a 64 bit file being edited. any caveats or issues to be known and what is the actual res of the plugin path vs the 64 bit wave file being edited. Oh I know, being a pain in the ass. But I am asking and I do need to know Thanks in advance
Subject:RE: Signal Paths - One for the engineers at Sony Reply by: Sonic Date:1/20/2004 8:06:58 AM Any built-in effect that modifies sample values is done in 64-bit doubles. You can control the processing depth for DX plugs launched via the menus with the little "X-bit" popup on the lower right of the dialog. ("24-bit" and "53-bit" really represent 32-bit float and 64-bit double, respectively, just noting the number of bits in the mantissa to jive with the other PCM depths). The chainer is hard coded to 32-bit float, but will handle conversions internally if a plug-in doesn't support that. We may try to up this to 64-bit in the next major version. Note that if the "Use floating point temporary files" pref is not set, the results of any process will be converted back to the file's current bit-depth without dithering or noise shaping. If the pref is enabled, results for < 32 bit PCM files will be stored as 32-bit float, and otherwise as 64-bit double. In your case, if you are always using 64-bit files, there's no need to set the pref. All relevant processing (with the exception of the chainer), will be performed at that depth, and the results will be stored at that depth until you render the file to something else. Clear? J.
Subject:RE: Signal Paths - One for the engineers at Sony Reply by: Big_Faced_Boy Date:1/22/2004 3:37:56 AM Crikey, it's been a while since I last heard anyone use the word "mantissa"! BFB
Subject:RE: Signal Paths - One for the engineers at Sony Reply by: Spheris Date:1/22/2004 9:48:35 AM All clear, and I THANK YOU, Sir for the in-depth clarity and straightforward answer to the question at hand. You have made my day with this because it was a bit confusing going by the outline of the manual and for some material I'm dealing with, I would like to keep as close to the ceiling process wise as possible. One question remains on the x-bit dialogue, only the 8-16-24 options seem to show, even when using the izotope spectron - a 64 bit plugin( i would assume it would show a 56 bit option or am I mistaken?) This is with the right click on the xbit display, is there something I did miss? or is this correct? Again, my thanks

Subject:Signal Paths - One for the engineers at Sony

Posted by: Spheris

Date:1/12/2004 9:27:02 AM

Okay

1. 64 bit res, is it indeed and does it apply only to temp file with the internal fx, not to dx plugins.?

2. side effects of 24 bit plugins as they show in the properties of the plugin display vs on a 64 bit file being edited. any caveats or issues to be known and what is the actual res of the plugin path vs the 64 bit wave file being edited.

Oh I know, being a pain in the ass. But I am asking and I do need to know

Thanks in advance

Subject:RE: Signal Paths - One for the engineers at Sony

Reply by: Sonic

Date:1/20/2004 8:06:58 AM

Any built-in effect that modifies sample values is done in 64-bit doubles. You can control the processing depth for DX plugs launched via the menus with the little "X-bit" popup on the lower right of the dialog. ("24-bit" and "53-bit" really represent 32-bit float and 64-bit double, respectively, just noting the number of bits in the mantissa to jive with the other PCM depths).

The chainer is hard coded to 32-bit float, but will handle conversions internally if a plug-in doesn't support that. We may try to up this to 64-bit in the next major version.

Note that if the "Use floating point temporary files" pref is not set, the results of any process will be converted back to the file's current bit-depth without dithering or noise shaping. If the pref is enabled, results for < 32 bit PCM files will be stored as 32-bit float, and otherwise as 64-bit double.

In your case, if you are always using 64-bit files, there's no need to set the pref. All relevant processing (with the exception of the chainer), will be performed at that depth, and the results will be stored at that depth until you render the file to something else.

Clear?
J.

Subject:RE: Signal Paths - One for the engineers at Sony

Reply by: Big_Faced_Boy

Date:1/22/2004 3:37:56 AM

Crikey, it's been a while since I last heard anyone use the word "mantissa"!

BFB

Subject:RE: Signal Paths - One for the engineers at Sony

Reply by: Spheris

Date:1/22/2004 9:48:35 AM

All clear, and I THANK YOU, Sir for the in-depth clarity and straightforward answer to the question at hand. You have made my day with this because it was a bit confusing going by the outline of the manual and for some material I'm dealing with, I would like to keep as close to the ceiling process wise as possible.

One question remains on the x-bit dialogue, only the 8-16-24 options seem to show, even when using the izotope spectron - a 64 bit plugin( i would assume it would show a 56 bit option or am I mistaken?) This is with the right click on the xbit display, is there something I did miss? or is this correct?

Again, my thanks