[FEATURE, BREAKING] Support 1-to-n and n-to-1 channel configurations in Linear models - #335
Merged
Merged
Conversation
dfernandes83
added a commit
to dfernandes83/NeuralAmpModelerCore
that referenced
this pull request
Sep 21, 2026
Brings in NAM Core 0.6.0 from sdatkinson/NeuralAmpModelerCore: - models with variable sample rate (sdatkinson#333): Linear reports SupportsArbitrarySampleRate() and resamples in Reset() - 1-to-n and n-to-1 channel configurations in Linear models (sdatkinson#335, breaking: a Linear model now needs one kernel per output/input path) - Linear models instantiated from .wav files through get_dsp() (sdatkinson#337) - version bump to 0.6.0 (sdatkinson#336) Conflicts: - NAM/get_dsp.cpp: kept all three includes (<algorithm> and <cctype> from upstream, <cmath> from our validation code). - NAM/linear.cpp: took upstream's rewritten direct-convolution loop. Our side only had a static_cast to silence an MSVC conversion warning in the code that upstream replaced. Consumers now have to compile NAM/wav.cpp too: get_dsp.cpp calls nam::detail::load_wav_ir().
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Extend
Linearto support one input feeding multiple outputs and multiple inputs summing into one output. Previously, unequal channel counts silently ignored extra inputs or silenced extra outputs.This is a breaking change to no longer support "N-to-M" with N,M>1.
Both direct and FFT implementations support these mappings, sample-rate conversion, and in-place processing. Document the weight layout and bias semantics.
Validation
Resolves #334.