apart from tomax~ -> listiter -> peak~
You can do it in a non-rt thread. It is covered in the tutorial 11
----------begin_max5_patcher---------- 932.3oc0WsriaaCEcs8WgfV6Nkj5cWkjMYU1jBTfhhh.ZYZGlPQZPQM1oAwe 68xKkjsmwimWJKxrvh5JRdOmCuO3784yhWZ1KZii9in+IZ1ruOe1Lzj2vr92 mE2v2Wq3s3zh0hclkeIdQ3SNwdGZds5FKWuxzbH52UB8F2minDV5v7LcNkv4 91VQvWwwQ+a+m1xc0eVp27Iqn1E9JM6FxBX8gGI3CF6Fx3ZzcMRMrgHjXGMF bCZk1aUtBwGf4eiE6M8i4y8+r30w2kcqWKrGhVZLqiXDB4AY5Zkg6hW.Kgq2 7HzlfDtpm9IWm2zKxa184M8kv6c7aEqM1lCONwtx.krEedj2AgSyaDAcDzu3 GKRfwx8OJBQBz7zGRRxtnjj2acfR0FkwF1exMEY4k4rhzLRUIIoZAXJiVjmV PxJKYIYrKahdBDNUpKmvXLHmpUp+5gn2DDsnSEq6cbzJ2n4piB8TyxGNOsOd kVMQ4oT1zJh6jPgocGdU0hBol4SUoHJ4kPwFQaKei3dbzJZEtEQaslZXFQvw 2UpG8jXK9KiNQrsZZOOkZmvdKWc.awD8FqK5UQ27PY2PQlma81KP2hSoKtWP cP8cayhvwa+bMn0zYqG.eX2fTyQDsRz5jZtSZzmLGVXNWTkep9g8D7CD195c zvlbcOMETp5I3nh64mvIFe61aE119Iit.hO+RnjZ4B7UoN7JFC.Yg2JGleXB bKDv4fnsNana297vMhhaLqflfcRb0y8jCbYq6ap6FkL1n7scqjl+zwccse5C BcWHfGH0ZdmxctPrbyZoRM1.X13Wf7n9Li91CK5MeVyBVUJkV5aHjPSJX43H XPVFjkLHm8qgNrnzrzJByO0TrQCNx2GIwOhbmkwN5KBsJ3AeqozvHvD8rtOf Vp2nPjyJ7oq8VgRdaM1giTXoUiyuyY1X4qjBMpNDzNdDGjaLNxBeEkcTS50z FPPkCq4z3twCiPso2o5DvkqtxAAutFbvYhaB1i0Sy7QEqxaZfvSI.slc5mMB OhqL3LOEuYPxOID9dqP77QHAAWh+O+HJd2loGbeTr5YCshxzDJFESwKagA+m dapIDe+sPoL6tNDWa.GH+u9phTn4VVEgTNVT7BTnjA2CbAxEZIEIPVUlOidB IPWyRg88F0KPg8WQEqXjUwnYHRomVwXR.3VtVn9KoAZy+SrRaxXBFfeRQYnn IDLikOYujZszeoq0F58hWBzeTztkWO7+ItOd9Ol++L+6NOF -----------end_max5_patcher-----------
next time I try it, most probably for the next project in a few months, I’ll add time for the tutorials I did not do yet. More soon-ish, now back to fluidbufcompose options.
There’s no buffer writing object yet - it would be possible but there are some complexities to do with ordering and threadsafety that also might impact on speed for reading (because I’d have to assume that any read object might refer to the same buffer as a write object and the ordering would then need to be correct). You also lose precision when you go to a buffer.
What is the use case? I would be helpful for me to see more examples of how it would be useful to figure out how to support that need best.
The use case would be to not use fluid.bufcompose~ - an ola process in a buffer that is then processed as a new frame itself. This was for sample and hold few 10sec segments on a minute buffer that is then granulated. I did it with the fluid stuff and a ring buffer instead and it worked. but then there was my idea of additive resynth and then I would need to composite my stuff then envelop it… and it became difficult.
I think I’d have to understand the details a bit better to understand this. You can obviously write to a standard buffer with max objects and read in framelib, but the buffer can only be updated once per audio vector.
For some storage needs in framelib you should use fl.store~ and fl.recall~, but that’s not a route to only write part of a larger frame easily (can be done, but could be easier with a different object set).