SampleRateConvertion error in plugin

Question

Pablo Panitta 2020-11-23

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/658043-samplerateconvertion-error-in-plugin

评论： Jimmy Lapierre 2021-2-2

Hi

I´m trying to implement a x4 oversampling process but I´m having difficulties on how let the SampleRateConverter sytem object knows about sample rate used.

When compiling the example attached for plugin generation, the following error shows up: “Failed to compute constant value for nontunable property 'SampleRate'. In code generation, nontunable properties can only be assigned constant values.”

I'd really appreciate if someone can give me a hint on how fix/treat this.

Br

Pablo

classdef (StrictDefaults)DistoLab_Test2 < matlab.System & audioPlugin
    
    properties        
        GainDisto=0;
        Input=0;
        Volume=0;    
    end
    
    properties (Constant, Hidden)
        % Define the plugin interface
        PluginInterface = audioPluginInterface( ...
            'InputChannels',2,...
            'OutputChannels',2,...
            'PluginName','DistoLab_Test2',...
             audioPluginParameter('Volume', ...
            'DisplayName',  'Out', ... 
            'DisplayNameLocation','none',...
            'Label','dB', ...
            'Mapping', { 'lin', -15, 15}, ...
            'Style', 'rotaryknob', 'Layout', [4 6]),...
            audioPluginParameter('Input', ...
            'DisplayName',  'In', ... 
            'DisplayNameLocation','none',...
            'Label','dB', ...
            'Mapping', { 'lin', -5, 5}, ...
            'Style', 'rotaryknob', 'Layout', [4 3]),...           
            audioPluginParameter('GainDisto', ...
            'DisplayName',  'Saturation', ... 
            'DisplayNameLocation','none',...
            'Style', 'rotaryknob', 'Layout', [4,5],...
            'Mapping', { 'lin', 0, 10}, ...
            'Filmstrip','knob_67_black.png', ...                                   %<--
            'FilmstripFrameSize',[80,80]), ...
            audioPluginGridLayout('RowHeight', [30 90 10 100 37], ... 
            'ColumnWidth', [30 100 100 20 100 100 20 80 80 80 60], 'Padding', [10 10 10 10]));
   
    end    
    
    properties (Access = private)    
        Up4;
        Down4;
    end
    
    methods
        % Constructor
        function plugin = DistoLab_Test2
  
            plugin.Up4=dsp.SampleRateConverter;
            plugin.Down4=dsp.SampleRateConverter;
            calculateSampleRates(plugin);
            
        end
        
    end
    methods(Access = protected)
        function out = stepImpl(plugin, in)
            inadjusted=in*(10^(plugin.Input/20)); % Incoming signal adjusted by Input Gain
%SampleRateConverter x4 process
    
    v=step(plugin.Up4,inadjusted);
    disto=(v*plugin.GainDisto).^5; % input distorted
        
    outdisto=(disto)*(10^(plugin.Volume/20));
    
    out = step(plugin.Down4,outdisto);  
    
         end
     
        function resetImpl(plugin)
            reset(plugin.Up4);
            reset(plugin.Down4);
        end
        function calculateSampleRates(plugin)
            
           plugin.Up4.InputSampleRate=getSampleRate(plugin);
           plugin.Up4.OutputSampleRate=4*getSamplerate(plugin);
           
           plugin.Down4.InputSampleRate=4*getSampleRate(plugin);
           plugin.Down4.OutputSampleRate=getSampleRate(plugin);
        end
  
    end
    
end

0 个评论
显示 -2更早的评论隐藏 -2更早的评论

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

Jimmy Lapierre 2020-11-23

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/658043-samplerateconvertion-error-in-plugin#answer_553843

在 MATLAB Online 中打开

Hi Pablo,

There are some limitations with codegen and the resampling objects. If you have a hard-coded ratio of 4, the easiest way around it might be to use a nominal sample rate, such as 48kHz. The added benefit is that avoid calls to getSampleRate will be faster.

function calculateSampleRates(plugin)
    sampleRate = 48000; % use a nominal rate
    plugin.Up4.InputSampleRate=sampleRate;
    plugin.Up4.OutputSampleRate=4*sampleRate;
    plugin.Down4.InputSampleRate=4*sampleRate;
    plugin.Down4.OutputSampleRate=sampleRate;
end

The following challenge will be that the input of the sample rate converter will be of unknown size. You can get around that by looping over partitions of a maximum size (like 4096). Something like this might work for you:

% Replaces out = step(plugin.Down4,outdisto);
out = zeros(size(in));
for ii = 1:4096:size(in,1)
    endIdx = min(ii+4095,size(in,1));
    xin = outdisto(4*ii-3:4*endIdx,:);
    assert(size(xin,1)<=4*4096); % Tell codegen xin won't be larger than 4x4096
    out(ii:endIdx,:) = step(plugin.Down4,xin);
end

7 个评论
显示 5更早的评论隐藏 5更早的评论

Pablo Panitta 2020-11-24

编辑：Pablo Panitta 2020-11-24

HI Jimmy

Thanks a lot for your promt answer and detail explanation.

I cannot see how the process porposed is solving the issue. My understanding is the following: when plugin is instantiated in a DAW session, works at the sample rate of the session. As this is not knowing in advance by plugin, there´s no way to predefine the ‘InitialSampleRate’ parameter value (until inserted). So the methodology should be: insert the plugin in a session, get the its sample rate value, and then set up the InitialSamplerate accordingly.

But when I try to implement this concept, in the constructor areas (as always do with SystemObject parameter definition), it is not possible to define anything but fixed values (what is exactly I don´t need!). If I try to make it relative to samplerate of the plugin a “non tunable parameter…” error describe above appears.

E.g if your example is inserted in a 88kHz DAW session, the SampleRateConverter will end up working with initialSampleRate of 48kHz instead 88kHz?

If I'm wrong please let me know, otherwise : is there any way to do it?

Thanks

Pablo

Adrian Alexander 2021-2-2

Hi i realy like the oversampling code.

I want to change the code to some other oversamplingfactor like x8 or x16

what would change in the second part ? i dont realy know how it works.

ut = zeros(size(in));

for ii = 1:4096:size(in,1)

endIdx = min(ii+4095,size(in,1));

xin = outdisto(4*ii-3:4*endIdx,:);

assert(size(xin,1)<=4*4096); % Tell codegen xin won't be larger than 4x4096

out(ii:endIdx,:) = step(plugin.Down4,xin);

end

thanks

Jimmy Lapierre 2021-2-2

Replace all the "4*" by a constant named "OSF" (oversampling factor) and set it to 4, 8 or 16.

请先登录，再进行评论。

SampleRateConvertion error in plugin

0 个评论
显示 -2更早的评论隐藏 -2更早的评论

采纳的回答

7 个评论
显示 5更早的评论隐藏 5更早的评论

更多回答（0 个）

另请参阅

类别

标签

产品

版本

Community Treasure Hunt

SampleRateConvertion error in plugin

0 个评论 显示 -2更早的评论隐藏 -2更早的评论

采纳的回答

7 个评论 显示 5更早的评论隐藏 5更早的评论

更多回答（0 个）

另请参阅

类别

标签

产品

版本

Community Treasure Hunt

0 个评论
显示 -2更早的评论隐藏 -2更早的评论

7 个评论
显示 5更早的评论隐藏 5更早的评论