> did you do an upgrade or a reinstall of EigenD and 10.11?
I've done both – an upgrade and a fresh install of El Capitan.
For several months now I've been running my iMac on Snow Leopard (SL) – which I've reinstalled more than once from Time Machine, because each time I lost the use of too many vital apps and peripherals.
But I did a clean install of El Capitan (ElC) onto a Seagate-USB external hard disk, and reinstalled everything EigenD afresh from GitHub. I forget what I did about /usr/local/pi/ – I think I copied it manually under Unix using Apple Terminal from /usr/pi/ on the internal hard drive.
This lash-up worked very well for whatever I tried. Launching ElC from my Seagate-USB didn’t conspicuously poop over the SL disk [apart from occasionally corrupting aliases on my SL Desktop] and performed surprisingly well. I expected ElC to be slow, booted as it was from a USB drive.
Earlier this month I felt that at last I had viable alternatives to all my Snow Leopard (SL) stuff that didn't work on El Capitan (ElC), and (yet again) upgraded my internal hard drive – to ElC this time. Then I bought Logic Pro – which knackered my Camel Audio Alchemy plug-in (for reasons discussed elsewhere, which I was fully expecting).
Throughout all this, the rest of EigenD continued to work under SL as it always did – viz I had drum loops and all instruments including cello [and Alchemy!]. Frequently I kept trying cello: it worked fine under SL but sounded rough under ElC.
[Alchemy is a tale in itself, and I won’t digress into it in this thread.]
So I was led to assume that Cello worked under old EigenD, but didn’t (and never had done) in EigenD 2.1.2-community. That is, until I discovered my fix, of swapping values of wet and dry gain. Unfortunately upgrading my internal hard drive from SL to ElC has wiped a lot of the evidence I’d need to check my assumptions.
Crucial is this question: was I running 2.1.0-experimental or 2.1.2-community under SL? I have to confess I’ve forgotten. I’ve been blithely supposing the former, but the file timestamps and the contents of /2.1.0-experimental/log/ strongly suggest it was the latter.
A lot hangs on whichever it was.
However all is not lost. The old SL installation with its working Cello is still extant on my Time Machine disk, and I propose to reinstall it – but this time on the now-surplus Seagate-USB drive. Thus effectively reversing my two-horse lash-up I was running so successfully a month ago. I have no reason to suppose this won't work – and give me the old installation to compare in detail with the new one. Not side-by-side, but at least in quick succession.
Until I do that, I guess I’m not in a position to deal sensibly with your questions. But for now, here’s the best that I can say…
Yes, I’m satisfied /usr/local/pi is all there (on both my internal and Seagate-USB drives), and is the one EigenD uses.
/usr/pi/ is no longer there – and I suspect my recent ElC upgrade wiped it (I’ve no recollection of manually deleting it). The /usr/ directory now contains only directories: adic/ bin/ lib/ libexec/ local/ sbin/ share/ standalone/ X11/ (plus an alias: X11R6) – and I believe ElC locks /usr/ down for security reasons, and resists you adding things like user-written shell scripts.
I’ve opened Workbench to view the Cello rig, but I confess to being baffled by it. You suggest: "try reselecting the convolution wave and save" – but I can’t identify what entity you mean. When I open "convolver1" – I see (apart from numeric fields) only "impulse" – which has the contents: "None available". This suggests to me that "convolver1" is not what convolutes the pure tone with a cello wave, but only handles an impulse response (potentially, but not in actual use).
Then you go on to address the "buzzing" I mentioned. I might be imagining things, and maybe I’ve introduced a red-herring with this. Maybe I’ve restored the Cello rig to fully working order with my values-swap "fix", and mis-remember the quality of the authentic gritty Cello sound when I hit a key.
Originally when I got Cello back, I thought it might have been a simple typo during manual transcription of code, or manual replugging in Workbench. But from what you’re saying this is impossible. (And from what I’ve said above, too.) So I now feel I’ve stumbled over the tip of an iceberg, which might go some way to explaining other reported problems with EigenD under El Capitan. So I'm keen to get to the bottom of it.
Much will come clear when I bring SL back to life again. Which I need for a number of forensic reasons starting to emerge as I use my upgraded iMac more and more. Then I can compare what I see in Workbench with the Cello rig as it was back in SL days when it worked. I’ll report back when I do.
UK
