STI channel-loop based #118

SimonBuechner · 2026-01-16T10:36:09Z

Changes proposed in this pull request:

implements a channel-loop based version of STI calculations and according tests
STI calculation is split into three functions: speech_transmission_index, modulation_transfer_function and a small function handling final STI computations.

f-brinkmann

Generally looks good to me. I made a first round of review without already looking at the tests.

f-brinkmann · 2026-01-16T12:28:38Z

pyrato/parameters.py


-def _energy_ratio(limits, energy_decay_curve1, energy_decay_curve2):
-    r"""
-    Calculate the energy ratio for the time limits from two energy
-    decay curves (EDC).

-    A variety of room-acoustic parameters are defined by energy ratios derived
-    from one or two time-domain Energy Decay Curves (EDCs). These parameters
-    distinguish between time regions using the four provided limits, and some,
-    such as Strength (:math:`G`), Early lateral sound (:math:`J_\mathrm{LF}`),
-    and Late lateral sound (:math:`L_J`), require EDCs obtained from different
-    impulse-response measurements [#iso]_.

-    Energy-Ratio is calculated as:
-    .. math::
-        ER = \frac{
-            \displaystyle \int_{lim3}^{lim4} p_2^2(t) \, dt
-        }{
-            \displaystyle \int_{lim1}^{lim2} p_1^2(t) \, dt
-        }
-    where :math:`[lim1, ..., lim4]` are the time limits and :math:`p(t)` is the
-    pressure of a room impulse response. Here, the energy ratio is
-    efficiently computed from the EDC :math:`e(t)` directly by:
-    .. math::
-        ER = \frac{
-            \displaystyle e_2(lim3) - e_2(lim4)
-        }{
-            \displaystyle e_1(lim1) - e_1(lim2)
-        }.
+


Should be two blank lines before a function only

Please do suggest code changes on unrelated changes in reviews. Unrelated changes should not be included or discussed in PRs

It not unrealted - but the diff above makes it look like. Theres too many blank lines before the start of a newly added function.

Ah, alright. My bad then!

pyrato/parameters.py

f-brinkmann · 2026-01-16T12:39:51Z

pyrato/parameters.py

+        snr = np.asarray(snr).flatten()
+         # Check if SNR has the correct number of components
+        if np.squeeze(snr.flatten().shape)/7 != (np.squeeze(data.cshape)):


You are flattening and reshaping twice. Would this work as well

Suggested change

snr = np.asarray(snr).flatten()

# Check if SNR has the correct number of components

if np.squeeze(snr.flatten().shape)/7 != (np.squeeze(data.cshape)):

snr = np.atleast_2d(snr)

# Check if SNR has the correct number of components

if snr.shape[:-1] != cshape:

# error about wrong number of channesl

elif snr.shape[-1] != 7:

# raise error about wrong number of octave values

# check snr as is already done

# only one reshape

snr = np.reshape(snr, (-1, 7))

f-brinkmann · 2026-01-16T12:41:36Z

pyrato/parameters.py

+        if np.any(snr < 20):
+            warnings.warn(
+                "SNR should be at least 20 dB for every octave band.",
+                stacklevel=2,


should this be stacklevel 1 or 2? If this is the main function, it should probably be 1.

f-brinkmann · 2026-01-16T12:54:04Z

pyrato/parameters.py

+    Parameters
+    ----------
+    data : pyfar.Signal
+        Single-channel room impulse response with shape ``(n_samples,)``.


If single channel is important it should be checked and an error sould be raised if it is violated.

f-brinkmann · 2026-01-16T12:55:19Z

pyrato/parameters.py

+    data : pyfar.Signal
+        Single-channel room impulse response with shape ``(n_samples,)``.
+
+    data_type : {'electrical', 'acoustical'}


Can you please add checks for data_type, level, snr,and amb similar to the way they are checked in the sti function?

f-brinkmann · 2026-01-16T12:57:05Z

pyrato/parameters.py

+        data, num_fractions=1, freq_range=(125, 8000),
+    )
+
+    f_m = np.array(


Can you add a comment that these are the modulation frequencies and ideally link to the equation, section that defines them in the standard?

f-brinkmann · 2026-01-16T12:57:44Z

pyrato/parameters.py

+    f_m = np.tile(f_m, (data_oct.cshape[0], 1))
+    energy = data_oct.time ** 2
+
+    term_exp = np.exp(-2j * np.pi * f_m[:, :, None] * data_oct.times)


The below is hard to understand. Can you please reference equations from the standard that are computed?

f-brinkmann · 2026-01-16T12:59:07Z

pyrato/parameters.py

+    sti : np.ndarray
+        Speech Transmission Index.
+    """
+    with np.errstate(divide="ignore"):


Can you here add references to equations/sections in the standard as well. Makes it way easier to review and maintain.

dcostring spellling Co-authored-by: Fabian Brinkmann <fabian.brinkmann@mailbox.org>

Error message adjusted to include reference to reason Co-authored-by: Fabian Brinkmann <fabian.brinkmann@mailbox.org>

Co-authored-by: Fabian Brinkmann <fabian.brinkmann@mailbox.org>

simon.buechner22 added 8 commits November 21, 2025 16:54

STI Artur functions added to parameters

04ba6b1

STI Artur tests raw added

ceeae1d

ruff linting

4e7ccd3

STI artur version cleaned and added

5f05869

empty line deleted

cda59c7

arthur version cleaned

f5c5f33

basic mtf tests added

8df15d0

mtf WINMF test result added

e77c020

SimonBuechner added this to the v1.0.0 milestone Jan 16, 2026

SimonBuechner requested a review from a team January 16, 2026 10:36

SimonBuechner assigned SimonBuechner and artur-pa Jan 16, 2026

SimonBuechner added the enhancement New feature or request label Jan 16, 2026

SimonBuechner requested review from ahms5, artur-pa, f-brinkmann, hoyer-a and mberz January 16, 2026 10:36

SimonBuechner changed the title ~~Feature/sti channel loop~~ STI channel-loop based Jan 16, 2026

f-brinkmann reviewed Jan 16, 2026

View reviewed changes

SimonBuechner and others added 4 commits January 20, 2026 10:33

Update pyrato/parameters.py

17b051b

dcostring spellling Co-authored-by: Fabian Brinkmann <fabian.brinkmann@mailbox.org>

Update pyrato/parameters.py

800506b

Error message adjusted to include reference to reason Co-authored-by: Fabian Brinkmann <fabian.brinkmann@mailbox.org>

Update pyrato/parameters.py

b529761

Co-authored-by: Fabian Brinkmann <fabian.brinkmann@mailbox.org>

Update pyrato/parameters.py

80c7cca

Co-authored-by: Fabian Brinkmann <fabian.brinkmann@mailbox.org>

-        snr = np.asarray(snr).flatten()
-         # Check if SNR has the correct number of components
-        if np.squeeze(snr.flatten().shape)/7 != (np.squeeze(data.cshape)):
+        snr = np.atleast_2d(snr)
+         # Check if SNR has the correct number of components
+        if snr.shape[:-1] != cshape:
+            # error about wrong number of channesl
+        elif snr.shape[-1] != 7:
+            # raise error about wrong number of octave values
+        # check snr as is already done
+        # only one reshape
+        snr = np.reshape(snr, (-1, 7))

STI channel-loop based #118

Are you sure you want to change the base?

STI channel-loop based #118

Conversation

SimonBuechner commented Jan 16, 2026

Changes proposed in this pull request:

Uh oh!

f-brinkmann left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

5 participants