Software Forum - openfnirs Forum

Splitting a .snirf file in two

Luke Gosse — Thu, 29 Jan 2026 18:33:09 +0000

Hello, I have a .snirf file with two different trigger durations that I would like to split in two. I had a run where I forgot to stop recording after transitioning between short finger tapping blocks and long finger tapping blocks. I was curious if there is a way to split the one .snirf file into two separate .snirf files while removing the noisy middle where I was informing the participant that the duration was going to change. I am using the Homer3 executive version (not with MATLAB). Thank you!

Question about short channel and accelerometer regression in hyperscanning analysis using Brain AnalyzIR (nirs toolbox)

Annabelle Li — Fri, 19 Sep 2025 13:03:15 +0000

I am conducting parent-child hyperscanning analysis using Brain AnalyzIR (nirs toolbox), where I am trying to calculate the coherence (using wavelet transform coherence; WTC) between the neural signals of a parent and a child as they cooperate on a problem-solving task. My question is: if I conduct hyperscanning analysis using the Hyperscanning.m module on the hb data, I cannot include the GLM module in my pipeline, which includes options for short channel regression and accelerometer regression. Is there any way to conduct short channel regression and accelerometer regression in a way that is compatible with hyperscanning analysis in the Brain AnalyzIR/nirs toolbox? Any help would be greatly appreciated. Thank you!

Question about infant altas available in AtlasViewer

will babu — Thu, 28 Aug 2025 07:04:20 +0000

Hello,
I am working with fNIRS on infants aged 6 months to 2 years. I am planning to identify the regions covered by our probe template and am looking for an infant atlas that can be used to identify segmented brain regions in AtlasViewer. Unfortunately, I was unable to find an available infant atlas in AtlasViewer.

Has anyone had experience importing an infant atlas into AtlasViewer and successfully projecting probes onto the atlas? Any advice on how to resolve this issue would be greatly appreciated.

Thank you!

Question about infant altas available in AtlasViewer

Yuta Ariyama — Fri, 01 Aug 2025 04:24:17 +0000

Has anyone had experience importing an infant atlas into AtlasViewer and successfully projecting probes onto the atlas? Any advice on how to resolve this issue would be greatly appreciated.

Thank you!

Any lightweight real-time fNIRS visualization tool for Raspberry Pi?

Jhonn Mick — Thu, 17 Jul 2025 18:26:56 +0000

Hi everyone,

I’m experimenting with a portable fNIRS setup and considering using a Raspberry Pi for low-cost data acquisition and real-time monitoring.

I’m wondering if anyone here has successfully used (or developed) a lightweight real-time visualization tool that runs well on a Pi? I’ve checked Homer3 and AnalyzIR, but they’re a bit heavy or require full MATLAB.

Ideally looking for something open-source, minimal, and easy to customize (Python-based would be perfect).

Would love to hear what others are using for mobile or embedded setups.

Cheers,
— Jhonn Mick

Merge datasets in the NIRSToolbox

Jeremie Ginzburg — Fri, 11 Jul 2025 16:45:41 +0000

Hi everyone, I was wondering if anyone had a solution to merge/concatenate two datasets from two different acquisition datasets. I'm working with NIRx but I guess there could be a solution to do that after importing the datasets into the toolbox.

Thanks to anyone who would have a tip :)

Jeremie

reading in Kernel DevKit data

Zach Rosenthal — Mon, 23 Jun 2025 20:44:25 +0000

Hello,

I'm demoing the Kernel DevKit and am trying to open the snirf outputs in Homer3 on Windows X - I have tried on both Matlab 2017b runtime and also running the homer3 source code in Matlab 2023. In both cases I can open my data files but then immediately get the same error that prevents it from actually plotting the data. My channel geometries also get glitchy and aren't displaying consistently, as shown in the screenshot. Has anyone successfully used Homer3 with the Kernel DevKit? I am not the most computationally literate and can't tell if this is my own ignorance or a glitch about how the kernel snirf files interact with Homer3. Any advice would be appreciated, thanks!

Zach

Here's what I get when I try and run the source code in Matlab 2023 (the text in red is the same error I get from the 2017b runtime execution)

>> Homer3

June 23, 2025, 16:33:06
Loaded config file C:/Users/rosenthalz/Desktop/Homer3-1.87.0/AppSettings.cfg
Loaded config file C:/Users/rosenthalz/Desktop/Homer3-1.87.0/DataTree/AppSettings.cfg

Running Homer3, (v1.87.0), R2023b
Running Utils, (v1.8.3), R2023b
Running DataTree, (v1.22.0), R2023b
Running FuncRegistry, (v1.5.1), R2023b

============
SYSTEM INFO:
============
Platform Arch : PCWIN64
RAM Total : 274.5 GB
RAM Free : 256.5 GB
HD Space Total : 1023.6 GB
HD Space Free : 11.6 GB

MainGUI: June 23, 2025, 16:33:10 - Will load group folder #1 - C:/Users/rosenthalz/Box/Box/Joe Zach/

DataTreeClass: v1.22.0
Loaded saved registry C:/Users/rosenthalz/Desktop/Homer3-1.87.0/FuncRegistry/UserFunctions/Registry.mat

DataTreeClass.FindAndLoadGroups: Searching for data files. Please wait ...

Loaded saved DataFilesClass object: C:/Users/rosenthalz/Box/Box/Joe Zach/derivatives/homer/DatasetFiles_snirf.mat

Saving DataFilesClass object in C:/Users/rosenthalz/Box/Box/Joe Zach/derivatives/homer/DatasetFiles_nirs.mat

DataTreeClass.PrintDatasetFormat: Found non-standard data set (format #2) with 9 data files in 1 folders files.

Added group "Joe Zach" to dataTree.
Added subject "study-sweetcombo250609" to group "Joe Zach" .
Added session "sub-04_study-sweetcombo250609" to subject "study-sweetcombo250609" .
Added run "study-sweetcombo250609_sub-04_desc-da5b6e5_GATES.snirf" to session "sub-04_study-sweetcombo250609" .
Added run "study-sweetcombo250609_sub-04_desc-da5b6e5_HB_MOMENTS.snirf" to session "sub-04_study-sweetcombo250609" .
Added run "study-sweetcombo250609_sub-04_desc-da5b6e5_MOMENTS.snirf" to session "sub-04_study-sweetcombo250609" .
Added session "sub-28_study-sweetcombo250609" to subject "study-sweetcombo250609" .
Added run "study-sweetcombo250609_sub-28_desc-22a6f0a_GATES.snirf" to session "sub-28_study-sweetcombo250609" .
Added run "study-sweetcombo250609_sub-28_desc-22a6f0a_HB_MOMENTS.snirf" to session "sub-28_study-sweetcombo250609" .
Added run "study-sweetcombo250609_sub-28_desc-22a6f0a_MOMENTS.snirf" to session "sub-28_study-sweetcombo250609" .
Added session "sub-30_study-sweetcombo250609" to subject "study-sweetcombo250609" .
Added run "study-sweetcombo250609_sub-30_desc-c97681f_GATES.snirf" to session "sub-30_study-sweetcombo250609" .
Added run "study-sweetcombo250609_sub-30_desc-c97681f_HB_MOMENTS.snirf" to session "sub-30_study-sweetcombo250609" .
Added run "study-sweetcombo250609_sub-30_desc-c97681f_MOMENTS.snirf" to session "sub-30_study-sweetcombo250609" .

Loaded group Joe Zach acquisition data in 5.6 seconds
Derived data output folder : C:/Users/rosenthalz/Box/Box/Joe Zach/derivatives/homer/
Derived data output file : groupResults.mat

DataTreeClass.ErrorCheckLoadedFiles: WARNING - The following files in this data set were NOT loaded:
"random recording from zach" s head/study-test_sub-s001_desc-f0fffbb_HB_MOMENTS.snirf
"random recording from zach" s head/study-test_sub-s001_desc-f0fffbb_MOMENTS.snirf

Group processing stream:
= hmrG_SubjAvg(dcAvgSubjs, nTrialsSubjs);
= hmrG_SubjAvgStd(dcAvgSubjs);

Subject processing stream:
= hmrS_SessAvg(dcAvgSess, mlActSess, nTrialsSess);
= hmrS_SessAvgStd2(dcAvgStdSess, nTrialsSess);

Session processing stream:
= hmrE_RunAvg(dcAvgRuns, mlActRuns, nTrialsRuns);
= hmrE_RunAvgStd2(dcAvgStdRuns, nTrialsRuns);

Run processing stream:
dod = hmrR_Intensity2OD(data);
aux = hmrR_BandpassFilt(aux, 0.000, 0.500);
dod = hmrR_BandpassFilt(dod, 0.000, 0.500);
dc = hmrR_OD2Conc(dod, probe, );
= hmrR_BlockAvg(dc, stim, );

Loaded data set in 13.5 seconds
Loaded 77 colors from SDGcolors.csv.
WARNING: Not all data files loaded successfully. Please see Homer3 GUI for details.
----------------------------------------------
ERROR: Operands to the logical AND (&&) and OR (||) operators must be convertible to logical scalar values. Use the ANY or ALL functions to reduce operands to logical scalar values.
Current Folder : C:/Users/rosenthalz/Box/Box/Joe Zach/
Call stack:
Error in DisplayAxesSDG.m > DisplayAxesSDG (line 141)
Error in MainGUI.m > Display (line 1038)
Error in MainGUI.m > MainGUI_OpeningFcn (line 273)
Error in gui_mainfcn.m > gui_mainfcn (line 220)
Error in MainGUI.m > MainGUI (line 19)
Error in Homer3.m > Homer3 (line 58)
----------------------------------------------

Capture.PNG

Inconsistency with HRF output query

Sally MacGarry — Fri, 13 Jun 2025 15:00:36 +0000

Hello all,

I'm a PhD student who's new to fNIRS and Homer3. In general I've not found it easy to find resources on how to use the latest version of Homer3, so perhaps I've missed something. I've been removing artefacts, filtering (bandpass and PCA), mitigating for SSCs and generating HRF using GLM from my data in Homer3. I'm then exporting to the HRF.txt and doing the next stage of cleanup and statistical analysis in excel and R.

This is a bit clunky, but there's sample randomisation and replication in my data so it was always going to need a lot of manual manipulation at the end. This has been working well and I've processed five out of 16 panellists with this process.

However my issue: The HRT output for one (of four) session for one panellists data is inconsistent to the rest. Most of the .txt files come out in 2 second intervals but the output for this one is n 4 second intervals. I've restarted, reloaded, tweaked the filters, checked all the available settings etc.. and I can't figure out why this would happen or how to change this time output setting for HRF. I've attached some screenshots of the .txt files for reference 1: time column in 2 secs, 2: time column in 4 secs.

If anyone can give me some advice, it would be much appreciated.

When best to run qtNIRS for quality control

Henry Brice — Tue, 06 May 2025 17:34:25 +0000

The wiki for qtnirs (https://github.com/lpollonini/qt-nirs/wiki) is not very clear about which stage is best for the identification of bad channels. The sample code in the wiki actually shows it being run on hb data, but trying to run that, I get en error that it has to be run on raw or OD data.
If I run it on raw, OD, or motion corrected OD data, I get quite different numbers of bad channels. Are there best practices for which stage to run qtnirs?
Also, rather surprisingly to me, I get the highest number of bad channels (by quite some margin...) on the motion corrected data (utilising a PCA correction). Any ideas why this might be?

Tips for data analysis of prolonged data (up to 6 minutes)

Felipe Balistieri Santinelli — Sat, 26 Apr 2025 17:36:37 +0000

Hi all

I have been using fNIRS and Homer functions for 3-4 years, and I have a challenge now in analysing walking data over a 6-minute walking test.

I was wondering if someone has done some kind of similar analysis and has some tips on how I could proceed with it.

Currently, my pipeline involves the functions of Homer3, including:
- hmrR_Intensity2OD

-hmrR_MotionArtifactByChannel

- hmrR_MotionCorrectWavelet

- hmrR_OD2Conc

- Besides these functions, I'm trying some filters approach (either IIR or FIR) with high (0.01 hz), low (0.2 hz) and stop (0.06-0.14 hz), with high orders (5th to IIR- Butterworth and 750 to FIR), based on the recommendations of Pinti et al., 2018.

- Regressing long channels with the nearest short-channel

At this moment, I'm just looking for oxy/deoxyhemoglobin, but I might try some connectivity analysis as well.

I would greatly appreciate it if someone could provide any tips for me to proceed with the analysis.