diff --git a/src/spikeinterface/sorters/internal/spyking_circus2.py b/src/spikeinterface/sorters/internal/spyking_circus2.py
index a0a4d0823c..db06287f6c 100644
--- a/src/spikeinterface/sorters/internal/spyking_circus2.py
+++ b/src/spikeinterface/sorters/internal/spyking_circus2.py
@@ -6,7 +6,7 @@
 
 from spikeinterface.core import NumpySorting, load_extractor, BaseRecording, get_noise_levels, extract_waveforms
 from spikeinterface.core.job_tools import fix_job_kwargs
-from spikeinterface.preprocessing import bandpass_filter, common_reference, zscore
+from spikeinterface.preprocessing import common_reference, zscore, whiten, highpass_filter
 
 try:
     import hdbscan
@@ -22,7 +22,7 @@ class Spykingcircus2Sorter(ComponentsBasedSorter):
     _default_params = {
         "general": {"ms_before": 2, "ms_after": 2, "radius_um": 100},
         "waveforms": {"max_spikes_per_unit": 200, "overwrite": True, "sparse": True, "method": "ptp", "threshold": 1},
-        "filtering": {"dtype": "float32"},
+        "filtering": {"freq_min": 150, "dtype": "float32"},
         "detection": {"peak_sign": "neg", "detect_threshold": 5},
         "selection": {"n_peaks_per_channel": 5000, "min_n_peaks": 20000},
         "localization": {},
@@ -60,11 +60,12 @@ def _run_from_folder(cls, sorter_output_folder, params, verbose):
         ## First, we are filtering the data
         filtering_params = params["filtering"].copy()
         if params["apply_preprocessing"]:
-            recording_f = bandpass_filter(recording, **filtering_params)
+            recording_f = highpass_filter(recording, **filtering_params)
             recording_f = common_reference(recording_f)
         else:
             recording_f = recording
 
+        #recording_f = whiten(recording_f, dtype="float32")
         recording_f = zscore(recording_f, dtype="float32")
 
         ## Then, we are detecting peaks with a locally_exclusive method
diff --git a/src/spikeinterface/sortingcomponents/clustering/clustering_tools.py b/src/spikeinterface/sortingcomponents/clustering/clustering_tools.py
index 891c355448..6dba4b7f0f 100644
--- a/src/spikeinterface/sortingcomponents/clustering/clustering_tools.py
+++ b/src/spikeinterface/sortingcomponents/clustering/clustering_tools.py
@@ -598,14 +598,17 @@ def remove_duplicates_via_matching(
             "waveform_extractor": waveform_extractor,
             "noise_levels": noise_levels,
             "amplitudes": [0.95, 1.05],
-            "omp_min_sps": 0.1,
+            "omp_min_sps": 0.05,
         }
     )
 
+    spikes_per_units, counts = np.unique(waveform_extractor.sorting.to_spike_vector()['unit_index'], return_counts=True)
+    indices = np.argsort(counts)
+
     ignore_ids = []
     similar_templates = [[], []]
 
-    for i in range(nb_templates):
+    for i in np.arange(nb_templates)[indices]:
         t_start = padding + i * duration
         t_stop = padding + (i + 1) * duration
 
diff --git a/src/spikeinterface/sortingcomponents/clustering/random_projections.py b/src/spikeinterface/sortingcomponents/clustering/random_projections.py
index 1f97bf5201..d7ceef2561 100644
--- a/src/spikeinterface/sortingcomponents/clustering/random_projections.py
+++ b/src/spikeinterface/sortingcomponents/clustering/random_projections.py
@@ -33,7 +33,7 @@ class RandomProjectionClustering:
             "min_cluster_size": 20,
             "allow_single_cluster": True,
             "core_dist_n_jobs": os.cpu_count(),
-            "cluster_selection_method": "leaf",
+            "cluster_selection_method": "leaf"
         },
         "cleaning_kwargs": {},
         "waveforms": {"ms_before": 2, "ms_after": 2, "max_spikes_per_unit": 100},