I am getting different results with identical parameters between command line and Max using bufnoveltyslice~. These are the differences:
This is my Max object:
and these are my CLI calls:
I’ve checked numerous times I am in fact analysing the same file which is the case, but I can’t get the slicing to be the same no matter how many times I try or check.
File in question:
libLLVMAMDGPUDesc.a.wav.zip (229.2 KB)
will investigate. I did not check result integrity in the last one since it was always right in the past… I’ll start with my testfiles
To be certain James can you test again using 2048 -1 -1 for the fftsettings on the CLI (or 2048 1024 2048 in max)? I’m not 100% sure on how those should default, but having exactly the same inputs would be a good starting test.
I get the same discrepancy in across all 4 permutations of those settings.
I can confirm there is a bug, I’ll chase on which version it happened and fill a bug report, nice catch!
Thanks,
I’m getting my head stuck into CLI for batching now - much easier (and perhaps faster?) than Max so if it can be fixed you might find haribo in your office…
For batch stuff I can definitely see something like this being useful.
(his the haribo a reward or a threat?!)
Its a reward for PA but a threat to everyone else around him.
Not to be a nag but is this a simple fix? I only ask because I will start doing my batch processing in Max if this is something that will come out in a newer version of the alpha rather than a casual bug fix. I realise I’m probably the only person using CLI right now and you are on tour so something like this is necessarily down the bottom of all your to-do’s.
This is a priority fix: we’re on a tour to present the tools and your use of the CLI is very valuable. You can code around it now and at one point soon the results will be consistent. You have actually found something that we need to fix, but how simple that is is beyond me. @weefuzzy is on the case. We’ll let you know very soon if we change our mind 
I appreciate the hard work! I will soon have two solid batch processors, so a win in that regard 
