diff --git a/CHANGES b/CHANGES
--- a/CHANGES
+++ b/CHANGES
@@ -142,3 +142,6 @@
 
 0.2.7.4:
 		Reduce Setup.lhs
+
+0.2.7.5:
+		update for hmatrix 0.18
diff --git a/hsignal.cabal b/hsignal.cabal
--- a/hsignal.cabal
+++ b/hsignal.cabal
@@ -1,8 +1,8 @@
 Name:               hsignal
-Version:            0.2.7.4
+Version:            0.2.7.5
 License:            BSD3
 License-file:       LICENSE
-Copyright:          (c) A.V.H. McPhail 2010, 2011, 2014, 2015
+Copyright:          (c) A.V.H. McPhail 2010, 2011, 2014, 2015, 2016
 Author:             Vivian McPhail
 Maintainer:         haskell.vivian.mcphail <at> gmail <dot> com
 Stability:          provisional
@@ -17,7 +17,7 @@
      .
      Feature requests, suggestions, and bug fixes welcome.
 Category:           Math
-tested-with:        GHC ==7.10.1
+tested-with:        GHC ==8.0.1
 
 cabal-version:      >=1.8
 
@@ -41,10 +41,10 @@
                         array,
                         vector,
                         bytestring, storable-complex, binary,
-                        hmatrix >= 0.17,
-                        hmatrix-gsl >= 0.17,
-                        hmatrix-gsl-stats >= 0.4,
-                        hstatistics >= 0.2.5.3
+                        hmatrix >= 0.18,
+                        hmatrix-gsl >= 0.18,
+                        hmatrix-gsl-stats >= 0.4.1.6,
+                        hstatistics >= 0.2.5.4
 
     Extensions:         ForeignFunctionInterface
 
diff --git a/lib/Numeric/Signal.hs b/lib/Numeric/Signal.hs
--- a/lib/Numeric/Signal.hs
+++ b/lib/Numeric/Signal.hs
@@ -3,7 +3,7 @@
 -----------------------------------------------------------------------------
 -- |
 -- Module      :  Numeric.Signal
--- Copyright   :  (c) Alexander Vivian Hugh McPhail 2010, 2014, 2015
+-- Copyright   :  (c) Alexander Vivian Hugh McPhail 2010, 2014, 2015, 2016
 -- License     :  BSD3
 --
 -- Maintainer  :  haskell.vivian.mcphail <at> gmail <dot> com
@@ -94,7 +94,7 @@
 -----------------------------------------------------------------------------
 
 -- | a broadband FIR
-broadband_fir :: (S.Filterable a, Double ~ DoubleOf a, Container Vector (Complex a), Convert (Complex a)) =>
+broadband_fir :: (S.Filterable a, Double ~ DoubleOf a, Convert (Complex a)) =>
                 Int           -- ^ sampling rate
               -> (Int,Int)     -- ^ (lower,upper) frequency cutoff
               -> Vector a -- ^ filter coefficients   
@@ -108,7 +108,7 @@
                         in standard_fir o be
 
 -- | a broadband filter
-broadband_filter :: (S.Filterable a, Double ~ DoubleOf a, Container Vector (Complex a), Convert (Complex a)) 
+broadband_filter :: (S.Filterable a, Double ~ DoubleOf a) 
                    => Int        -- ^ sampling rate
                  -> (Int,Int)    -- ^ (lower,upper) frequency cutoff
                  -> Vector a            -- ^ input signal
@@ -120,7 +120,7 @@
 
 -- | standard FIR filter
 -- |   FIR filter with grid a power of 2 greater than the order, ramp = grid/16, hamming window
-standard_fir :: (S.Filterable a, Double ~ DoubleOf a, Container Vector (Complex a), Convert (Complex a)) => 
+standard_fir :: (S.Filterable a, Double ~ DoubleOf a, Convert (Complex a)) => 
                Int -> [(a,a)] -> Vector a
 standard_fir o be = let grid  = calc_grid o
                         trans_ = grid `div` 16
@@ -133,7 +133,7 @@
 
 -- | produce an FIR filter
 fir :: (S.Filterable a
-      , Container Vector (Complex a), Convert (Complex a), Double ~ DoubleOf a) =>
+      , Convert (Complex a), Double ~ DoubleOf a) =>
       Int               -- ^ order (one less than the length of the filter)
     -> [(a,a)] -- ^ band edge frequency, nondecreasing, [0, f1, ..., f(n-1), 1]
                         -- ^ band edge magnitude
@@ -195,7 +195,7 @@
 freqzF b a s f = S.freqz b a ((2*pi/(fromIntegral s)) * f)
 
 -- | determine the frequency response of a filter, given a number of points and sampling rate
-freqzN :: (S.Filterable a, Enum a, Double ~ DoubleOf a) =>
+freqzN :: (S.Filterable a, Double ~ DoubleOf a) =>
          Vector a     -- ^ zero coefficients
        -> Vector a       -- ^ pole coefficients
        -> Int     -- ^ sampling rate
@@ -216,8 +216,7 @@
 analytic_power = S.complex_power_
 
 -- | the phase of an analytic signal
-analytic_phase :: (S.Filterable a, Container Vector a
-                 ,Double ~ DoubleOf a) => 
+analytic_phase :: (S.Filterable a) => 
                  Vector (Complex a) -> Vector a
 analytic_phase = (uncurry arctan2) . fromComplex
 
diff --git a/lib/Numeric/Signal/Internal.hs b/lib/Numeric/Signal/Internal.hs
--- a/lib/Numeric/Signal/Internal.hs
+++ b/lib/Numeric/Signal/Internal.hs
@@ -4,7 +4,7 @@
 -----------------------------------------------------------------------------
 -- |
 -- Module      :  Numeric.Signal.Internal
--- Copyright   :  (c) Alexander Vivian Hugh McPhail 2010, 2014, 2015
+-- Copyright   :  (c) Alexander Vivian Hugh McPhail 2010, 2014, 2015, 2016
 -- License     :  BSD3
 --
 -- Maintainer  :  haskell.vivian.mcphail <at> gmail <dot> com
@@ -40,7 +40,7 @@
 
 -----------------------------------------------------------------------------
 
-infixl 1 #
+infixr 1 #
 a # b = applyRaw a b
 {-# INLINE (#) #-}
 
@@ -62,7 +62,7 @@
       , Convert a, Floating (Vector a), RealElement a
       , Num a)
        => Filterable a where
-    -- | conver from Vector Double
+    -- | convert from Vector Double
     fromDouble :: Vector Double -> Vector a
 --       b ~ ComplexOf a, Container Vector b, Convert b) => Filterable a where
     -- | filter a signal
@@ -103,7 +103,7 @@
 
 convolve_vector_double c a = unsafePerformIO $ do
                              r <- createVector (size a)
-                             signal_vector_double_convolve # c # a # r #| "signalDoubleConvolve"
+                             (c # a # r # id) signal_vector_double_convolve #| "signalDoubleConvolve"
                              return r
 
 foreign import ccall "signal-aux.h vector_double_convolve" signal_vector_double_convolve :: CInt -> PD -> CInt -> PD -> CInt -> PD -> IO CInt
@@ -114,7 +114,7 @@
 
 convolve_vector_float c a = unsafePerformIO $ do
                              r <- createVector (size a)
-                             signal_vector_float_convolve # c # a # r #| "signalFloatConvolve"
+                             (c # a # r # id ) signal_vector_float_convolve #| "signalFloatConvolve"
                              return r
 
 foreign import ccall "signal-aux.h vector_float_convolve" signal_vector_float_convolve :: CInt -> PF -> CInt -> PF -> CInt -> PF -> IO CInt
@@ -127,7 +127,7 @@
 
 convolve_vector_complex c a = unsafePerformIO $ do
                               r <- createVector (size a)
-                              signal_vector_complex_convolve # c # a # r #| "signalComplexConvolve"
+                              (c # a # r # id) signal_vector_complex_convolve #| "signalComplexConvolve"
                               return r
 
 foreign import ccall "signal-aux.h vector_complex_convolve" signal_vector_complex_convolve :: CInt -> PC -> CInt -> PC -> CInt -> PC -> IO CInt
@@ -170,7 +170,7 @@
        -> Vector Double -- ^ output signal
 filterD l k v = unsafePerformIO $ do
                r <- createVector (size v)
-               signal_filter_double # l # k # v # r #| "signalFilter"
+               (l # k # v # r # id) signal_filter_double #| "signalFilter"
                return r
 
 foreign import ccall "signal-aux.h filter_double" signal_filter_double :: CInt -> PD -> CInt -> PD -> CInt -> PD -> CInt -> PD -> IO CInt
@@ -182,7 +182,7 @@
        -> Vector Float -- ^ output signal
 filterF l k v = unsafePerformIO $ do
                r <- createVector (size v)
-               signal_filter_float # l # k # v # r #| "signalFilter"
+               (l # k # v # r # id) signal_filter_float #| "signalFilter"
                return r
 
 foreign import ccall "signal-aux.h filter_float" signal_filter_float :: CInt -> PF -> CInt -> PF -> CInt -> PF -> CInt -> PF -> IO CInt
@@ -194,7 +194,7 @@
 hilbert v = unsafePerformIO $ do
             let r = complex v
             -- could use (complex v) to make a complex vector in haskell rather than C
-            signal_hilbert # r #| "hilbert"
+            (r # id) signal_hilbert #| "hilbert"
             return r
 
 foreign import ccall "signal-aux.h hilbert" signal_hilbert :: CInt -> PC -> IO CInt
@@ -207,7 +207,7 @@
        -> Vector Double  -- ^ power density  
 pwelch w v = unsafePerformIO $ do
              let r = konst 0.0 ((w `div` 2) + 1)
-             (signal_pwelch $ fromIntegral w) # (complex v) # r #| "pwelch"
+             (complex v # r # id) (signal_pwelch $ fromIntegral w) #| "pwelch"
              return r
 
 foreign import ccall "signal-aux.h pwelch" signal_pwelch :: CInt -> CInt -> PC -> CInt -> PD -> IO CInt
@@ -221,7 +221,7 @@
     | l == 1          = konst 1.0 1
     | otherwise       = unsafePerformIO $ do
                         r <- createVector l
-                        signal_hamming_double # r #| "Hamming"
+                        (r # id) signal_hamming_double #| "Hamming"
                         return r
 
 foreign import ccall "signal-aux.h hamming_double" signal_hamming_double :: CInt -> PD -> IO CInt
@@ -233,7 +233,7 @@
     | l == 1          = konst 1.0 1
     | otherwise       = unsafePerformIO $ do
                         r <- createVector l
-                        signal_hamming_float # r #| "Hamming"
+                        (r # id) signal_hamming_float #| "Hamming"
                         return r
 
 foreign import ccall "signal-aux.h hamming_float" signal_hamming_float :: CInt -> PF -> IO CInt
@@ -273,7 +273,7 @@
          -> Vector (Complex Double) -- ^ the values
 polyEval c z = unsafePerformIO $ do
                r <- createVector (size z)
-               signal_real_poly_complex_eval # c # z # r #| "polyEval"
+               (c # z # r # id) signal_real_poly_complex_eval #| "polyEval"
                return r
 
 foreign import ccall "signal-aux.h real_poly_complex_eval" signal_real_poly_complex_eval :: CInt -> PD -> CInt -> PC -> CInt -> PC -> IO CInt
@@ -285,7 +285,7 @@
               -> Vector Double           -- ^ output
 complex_powerD v = unsafePerformIO $ do
                   r <- createVector (size v)
-                  signal_complex_power_double # v # r #| "complex_power"
+                  (v # r # id) signal_complex_power_double #| "complex_power"
                   return r
 
 foreign import ccall "signal-aux.h complex_power_double" signal_complex_power_double :: CInt -> PC -> CInt -> PD -> IO CInt
@@ -295,7 +295,7 @@
               -> Vector Float             -- ^ output
 complex_powerF v = unsafePerformIO $ do
                   r <- createVector (size v)
-                  signal_complex_power_float # v # r #| "complex_power"
+                  (v # r # id) signal_complex_power_float #| "complex_power"
                   return r
 
 foreign import ccall "signal-aux.h complex_power_float" signal_complex_power_float :: CInt -> PC -> CInt -> PF -> IO CInt
@@ -306,7 +306,7 @@
 downsampleD :: Int -> Vector Double -> Vector Double
 downsampleD n v = unsafePerformIO $ do
                r <- createVector (size v `div` n)
-               (signal_downsample_double $ fromIntegral n) # v # r #| "downsample"
+               (v # r # id) (signal_downsample_double $ fromIntegral n) #| "downsample"
                return r
 
 foreign import ccall "signal-aux.h downsample_double" signal_downsample_double :: CInt -> CInt -> PD -> CInt -> PD -> IO CInt
@@ -315,7 +315,7 @@
 downsampleF :: Int -> Vector Float -> Vector Float
 downsampleF n v = unsafePerformIO $ do
                r <- createVector (size v `div` n)
-               (signal_downsample_float $ fromIntegral n) # v # r #| "downsample"
+               (v # r # id) (signal_downsample_float $ fromIntegral n) #| "downsample"
                return r
 
 foreign import ccall "signal-aux.h downsample_float" signal_downsample_float :: CInt -> CInt -> PF -> CInt -> PF -> IO CInt
@@ -326,7 +326,7 @@
 derivD :: Vector Double -> Vector Double
 derivD v = unsafePerformIO $ do
           r <- createVector (size v - 1)
-          (signal_diff_double) # v # r #| "diff"
+          (v # r # id) (signal_diff_double) #| "diff"
           return r
 
 foreign import ccall "signal-aux.h vector_diff_double" signal_diff_double :: CInt -> PD -> CInt -> PD -> IO CInt
@@ -335,7 +335,7 @@
 derivF :: Vector Float -> Vector Float
 derivF v = unsafePerformIO $ do
           r <- createVector (size v - 1)
-          (signal_diff_float) # v # r #| "diff"
+          (v # r # id) (signal_diff_float) #| "diff"
           return r
 
 foreign import ccall "signal-aux.h vector_diff_float" signal_diff_float :: CInt -> PF -> CInt -> PF -> IO CInt
@@ -346,7 +346,7 @@
 unwrapD :: Vector Double -> Vector Double
 unwrapD v = unsafePerformIO $ do
            r <- createVector $ size v
-           signal_unwrap_double # v # r #| "unwrap"
+           (v # r # id) signal_unwrap_double #| "unwrap"
            return r
 
 foreign import ccall "signal-aux.h unwrap_double" signal_unwrap_double :: CInt -> PD -> CInt -> PD -> IO CInt
@@ -355,7 +355,7 @@
 unwrapF :: Vector Float -> Vector Float
 unwrapF v = unsafePerformIO $ do
            r <- createVector $ size v
-           signal_unwrap_float # v # r #| "unwrap"
+           (v # r # id) signal_unwrap_float #| "unwrap"
            return r
 
 foreign import ccall "signal-aux.h unwrap_float" signal_unwrap_float :: CInt -> PF -> CInt -> PF -> IO CInt
@@ -368,7 +368,7 @@
                            r <- createVector (2*l)
                            alloca $ \sx -> 
                                alloca $ \sy -> do
-                                 (signal_cross_covariance_double (fromIntegral l) sx sy) # x # y # r #| "cross_covariance"
+                                 (x # y # r # id) (signal_cross_covariance_double (fromIntegral l) sx sy) #| "cross_covariance"
                                  sx' <- peek sx
                                  sy' <- peek sy
                                  return (sx',sy',r)
@@ -383,7 +383,7 @@
                            r <- createVector (2*l)
                            alloca $ \sx -> 
                                alloca $ \sy -> do
-                                 (signal_cross_covariance_float (fromIntegral l) sx sy) # x # y # r #| "cross_covariance"
+                                 (x # y # r # id) (signal_cross_covariance_float (fromIntegral l) sx sy) #| "cross_covariance"
                                  sx' <- peek sx
                                  sy' <- peek sy
                                  return (sx',sy',r)
@@ -397,13 +397,13 @@
 cumSumD :: Vector Double -> Vector Double
 cumSumD v = unsafePerformIO $ do
               r <- createVector (size v)
-              signal_cum_sum_double # v # r #| "cumSumD"
+              (v # r # id) signal_cum_sum_double #| "cumSumD"
               return r
 
 cumSumF :: Vector Float -> Vector Float
 cumSumF v = unsafePerformIO $ do
               r <- createVector (size v)
-              signal_cum_sum_float # v # r #| "cumSumF"
+              (v # r # id) signal_cum_sum_float #| "cumSumF"
               return r
 
 foreign import ccall "signal-aux.h cum_sum_double"
diff --git a/lib/Numeric/Signal/Multichannel.hs b/lib/Numeric/Signal/Multichannel.hs
--- a/lib/Numeric/Signal/Multichannel.hs
+++ b/lib/Numeric/Signal/Multichannel.hs
@@ -6,7 +6,7 @@
 -----------------------------------------------------------------------------
 -- |
 -- Module      :  Numeric.Signal.Multichannel
--- Copyright   :  (c) Alexander Vivian Hugh McPhail 2010, 2014, 2015
+-- Copyright   :  (c) Alexander Vivian Hugh McPhail 2010, 2014, 2015, 2016
 -- License     :  BSD3
 --
 -- Maintainer  :  haskell.vivian.mcphail <at> gmail <dot> com
@@ -312,7 +312,7 @@
 
 
 -- | filter the data with the given passband
-filter :: (S.Filterable a, Double ~ DoubleOf a, Container Vector (Complex a), Convert (Complex a)) => 
+filter :: (S.Filterable a, Double ~ DoubleOf a) => 
          (Int,Int) -> Multichannel a -> Multichannel a
 filter pb m = let m' = mapConcurrently (S.broadband_filter (_sampling_rate m) pb) m
               in m' { _filtered = Just pb }