BufRd with multiple playheads frequently crashes

[jamshark70]

@Thor_Madsen

This all seems quite confusing as (at 48k) b.duration still reports 0.9666, where as if you truncate the the next lower integer the duration is 46396 / 48000 = 0.96658333333333 which I suspect could lead to unpleasant surprises.

The allocation message sent to the server explicitly truncates (see allocMsg: numFrames.asInteger).

The only confusing bit is that Buffer doesn’t change the numFrames value to match the actual server value. Possibly it should? But you can request an update:

s.boot;

s.dumpOSC(1);

// buffer is allocated with an integer numFrames
// but the Buffer object's variable keeps the given, fractional, value (maybe a mistake)
b = Buffer.alloc(s, s.sampleRate * 0.9666, 1);
-> Buffer(0, 46396.8, 1, 48000.0, nil)
[ "/b_alloc", 0, 46396, 1, 0 ]

// if you `query` the buffer then it shows the integer number
b.query;
[ "/b_query", 0 ]
bufnum: 0
numFrames: 46396
numChannels: 1
sampleRate: 48000.0

b
-> Buffer(0, 46396, 1, 48000.0, nil)  // and now the numFrames variable is fixed too

@Thor_Madsen

So I guess the safest thing is to make sure that when you use syntax as eg. Buffer.alloc(t * s.sampleRate), t * s.sampleRate returns an non-fractional number.

Tim Blechmann (supernova author) argued the same thing – supernova throws (threw?) an error if you ask it to allocate a float number of samples in a buffer.

hjh