diff --git a/osl_normalise_sensortypes.m b/osl_normalise_sensortypes.m
deleted file mode 100644
index 66c845fa..00000000
--- a/osl_normalise_sensortypes.m
+++ /dev/null
@@ -1,147 +0,0 @@
-function D = osl_normalise_sensortypes(S)
-% Normalise Elekta sensortypes using minimum eigenvalue normalisation
-%
-% D = osl_normalise_sensortypes(S)
-%
-% REQUIRED INPUTS:
-%
-% S.D        - SPM MEG object filename
-% 
-% OPTIONAL INPUTS:
-%
-% S.method   - normalisation method to use ['min,'mean'], default 'min'
-%
-% S.pcadim   - set the PCA dimensionality to use, default []
-%
-% S.prefix   - prefix to append to file name, default ''
-%
-% S.do_plots - produce diagnostic plots [0/1], default 0
-%
-% Adam Baker 2015
-
-D = spm_eeg_load(S.D);
-
-% Check SPM File Specification:
-try
-    S.D = char(S.D);
-    [pathstr,filestr] = fileparts(S.D);
-    S.D = fullfile(pathstr,[filestr '.mat']); % force .mat suffix
-    D = spm_eeg_load(S.D);
-catch
-    error('SPM file specification not recognised or incorrect');
-end
-
-S.method   = ft_getopt(S,'method','min');
-S.pcadim   = ft_getopt(S,'pcadim',[]);
-S.prefix   = ft_getopt(S,'prefix','');
-S.do_plots = ft_getopt(S,'do_plots',0);
-
-modalities = {'MEGPLANAR','MEGMAG'};
-
-if ~numel(intersect(D.chantype,modalities)) == 2
-    error('SPM object must contain MEGPLANAR and MEGMAG modalities');
-end
-
-normalisation = ones(1,length(indchantype(D,modalities)));
-
-if S.do_plots 
-    figure
-    linecolor = {'r','b'};
-end
-
-% Compute covariance of each sensor type
-for modality = modalities
-    
-    % Good channels:
-    channel_inds = indchantype(D,modality,'GOOD');
-    
-    % Select data:
-    data = reshape(D(channel_inds,:,:),length(channel_inds),[]);
-    
-    % Good samples & trials:
-    goodsamples = ~all(badsamples(D,channel_inds,':',':'));
-    goodsamples = reshape(goodsamples,1,[]);
-    
-    % Remove bad samples:
-    data = data(:,goodsamples);
-    
-    % Estimate normalisation
-    C = osl_cov(data);
-    
-    if isempty(S.pcadim)
-        rankEst = estimate_rank(C);
-    else
-        rankEst = S.pcadim;
-    end
-    
-    eigenvectors = svd(C);
-    
-    switch S.method
-        case 'mean'
-            normalisation(channel_inds) = 1./sqrt(mean(eigenvectors(1:rankEst)));
-        case 'min'
-            normalisation(channel_inds) = 1./sqrt(eigenvectors(rankEst));
-    end
-    
-    
-    if S.do_plots
-        subplot(1,2,1); title('Unnormalised eigenvectors'); hold on;
-        plot(log(eigenvectors),'linewidth',2,'color',linecolor{strcmp(modality,modalities)})
-        stem(rankEst,log(eigenvectors(rankEst)),'--','color',linecolor{strcmp(modality,modalities)},'marker','none')
-        subplot(1,2,2); title('Normalised eigenvectors'); hold on;
-        plot(log(eigenvectors.*normalisation(channel_inds).^2'),'linewidth',2,'color',linecolor{strcmp(modality,modalities)})
-    end
-    
-end
-
-
-% Apply normalisation via SPM montage:
-montage             =  [];
-montage.tra         =  diag(normalisation);
-montage.labelnew    =  D.chanlabels(indchantype(D,modalities));
-montage.labelorg    =  D.chanlabels(indchantype(D,modalities));
-
-[~,indx] = ismember(montage.labelnew,D.sensors('MEG').label);
-
-montage.chanunitnew =  D.sensors('MEG').chanunit(indx);
-montage.chanunitorg =  D.sensors('MEG').chanunit(indx);
-montage.chantypenew =  lower(D.sensors('MEG').chantype(indx));
-montage.chantypeorg =  lower(D.sensors('MEG').chantype(indx));
-
-Sm = [];
-Sm.D              =  fullfile(D.path,D.fname);
-Sm.montage        =  montage;
-Sm.keepothers     =  true;
-Sm.updatehistory  =  1;
-Sm.prefix         = S.prefix;
-
-if ~isempty(S.prefix)
-    D = spm_eeg_montage(Sm);
-else
-    D = osl_spmfun(@spm_eeg_montage,Sm);
-end
-    
-
-
-end
-
-
-function rankEst = estimate_rank(C)
-% Matlab's rank.m and SPM's spm_pca_order.m (Minka's approximation) seem to
-% estimate the rank of Maxfiltered or AFRICAed data poorly... Please feel
-% free to augment this function with a better estimation approach.
-
-minRank = 40;
-
-eigDiff = diff(log(svd(C)));
-
-rankVec = eigDiff<10*median(eigDiff);
-rankVec(1:minRank) = false;
-
-rankEst = find(rankVec,1,'first');
-
-if isempty(rankEst)
-    rankEst = rank(C);
-end
-
-end