What I posted was pseudocode to keep things brief. My guess is you've forgotten to add what I had omitted and are not actually copying and casting to a different datatype all the actual values by iterating through the arrays but just some pointers, the data behind which then makes no sense to the method expecting another datatype. The parameters with the suffix "F" are temporary versions with float values to and from which you'd have to copy the contents of the input and output arrays
Both versions in this thread break VST plugins for me. I presume you just changed the parameters of "processReplacing" from "**float" to "**double", but the resulting method is not defined in the plugin. There is another function called "processDoubleReplacing" for that in VST 2.4 already. The following pseudocode is what I propose to get plugins working again: if plugin.canDoubleReplacing() plugin.processDoubleReplacing(input, output, samples) else if plugin.canProcessReplacing() inputF = (float)...
@Matthew Faust Power users rejoice! :P EqualizerAPO really strikes me as the best combination of simplicity, flexibility, quality and low latency for every day use so an update to this story is really good news. It seems to work fine so far on my system configuration. Binaries (overwrite EAPO files in Program Files, ideally by shutting down the Windows Audio and Windows Audio Endpoint services then starting them again post overwrite, or just use the Configurator reboot Configurator method)... To...
Power users rejoice! :P EqualizerAPO really strikes me as the best combination of simplicity, flexibility, quality and low latency for every day use so an update to this story is really good news. It seems to work fine so far on my system configuration. Binaries (overwrite EAPO files in Program Files, ideally by shutting down the Windows Audio and Windows Audio Endpoint services then starting them again post overwrite, or just use the Configurator reboot Configurator method)... To my knowledge, killing...
Power users rejoice! :P EqualizerAPO really strikes me as the best combination of simplicity, flexibility, quality and low latency for every day use so an update to this story is really good news. It seems to work fine so far on my system configuration. Binaries (overwrite EAPO files in Program Files, ideally by shutting down the Windows Audio and Windows Audio Endpoint services then starting them again post overwrite, or just use the Configurator reboot Configurator method)... To my knowledge, killing...
Makes sense, your processing 4 64 bit floats instead of 4 32 bit floats, so you need to either half the amount your packing or double the amount you can pack. Yes, that sounds right. Sorry for throwing you for a loop with the incomplete/incorrect instructions. Come to think of it, I remember generalizing that expression into one using 'sizeof(type)' and putting it before the loop. Once this is all set I might upload it somewhere, 1.2.1, optimised 64bit internal processing. Given proper permissions/credits...
Makes sense, your processing 4 64 bit floats instead of 4 32 bit floats, so you need to either half the amount your packing or double the amount you can pack. Yes, that sounds right. Sorry for throwing you for a loop with the incomplete/incorrect instructions. Come to think of it, I remember generalizing that expression into one using 'sizeof(type)' and putting it before the loop. Once this is all set I might upload it somewhere, 1.2.1, optimised 64bit internal processing. Given proper permissions/credits...
Don't remember, didn't end up using it for anything. There might have been some expression containing "/ 4" that needed to be adjusted.