Minor UI glitch when setting impulse response?

Hello Martin,

I found, as Don has also reported, that running the development build does not fix the IR UI glitch.

However I did discover that this build seemed to fix another problem, viz., that some of my organs always crash HW5 at the end of loading. All of my organs became loadable with the new build.

Unfortunately, none of my organs has audio with the new build. So for the moment I've returned to HW 5.0.0.333, where some organs won't load but luckily the ones I use most will load and do have audio.

I'm now running the latest MacOS Catalina 10.15.3. If you'd like me test anything else in this OS, I'd be happy to help.

[mdyde]

Thanks, Jim.

I'm sure the build I sent couldn't in itself prevent audio from working -- I think something else must have been to blame for that. Perhaps you were using a different organ, and you're using multi-channel audio, but the organ's ranks weren't routed to the groups you intended?

Using the build I sent, please try loading St. Anne's and see whether you can hear it. If not, do you see audio signal activity on Hauptwerk's 'Audio, MIDI and Performance' large control panel?

[mdyde]

Hello Jim,

P.S. If you definitely hear audio from St. Anne's with v5.0.0, but not with the build I sent you, please use 'Help | Create a diagnostic file' and send it to me as soon as you have a chance. (I'll send you a forum P.M. with a direct email address.) Many thanks.

[mdyde]

Hello Jim,

P.S. I've just remembered: there is one specific case in which the build I sent you intentionally would result in a rank not being heard even though it was (incorrectly) audible in v5.0.0:

"Bug-fix: any pipes' perspectives routed to empty audio mixer bus groups would incorrectly sound through audio mixer primary bus 1."

Does that explain it? (If not, or if in doubt, please do still send me a diagnostic file as soon as convenient, having checked with St. Anne's.) Thanks.

Hello Martin,

St. Anne's running in 5.0.0.333 produces normal audio. In the development build St. Anne's provides no audible output to my headphones nor to the audio meter in the Audio, MIDI and Performance control panel.

No multi-channel routing is configured with either build, just basic stereo.

I'll send a diagnostic file made from running St. Anne's in the development build.

[mdyde]

Thanks very much, Jim.

I can see the reason from the diagnostic file that St. Anne's isn't currently audible:

- On 2019-12-10 you restored your general settings and organ settings from a v4 backup, allowing your v4 audio mixer/group settings to be retained/migrated when asked.

- You then loaded St. Anne's, also allowing its v4 audio routing to be retained/migrated when asked. Hence all ranks were routed to audio group 13 (which was the migrated audio group from v4). At that point in time, audio would then correctly have been audible in St. Anne's in v5.0.0, and/or in the v5.0.1 build I sent you, since the St. Anne's ranks were all routed to non-empty (v4-mirgated) audio groups (i.e. group 13).

- Then on 2019-12-20 you used the General Configuration Wizard to reset your audio mixer/group settings to defaults, (thus discarding the migrated v4 audio group settings, rendering the v4-migrated group 13 now empty).

- However, you didn't use the Organ Configuration Wizard to do likewise for each of your organs (such as St. Anne's). Hence the St. Anne's ranks would still be routed to group 13, which would now be empty, so none of its ranks should have been audible. However, because of the bug I mentioned in v5.0.0 (which v5.0.1 resolves), in v5.0.0 ranks routed to empty groups would actually (incorrectly) have been audible through primary bus 1, whereas in v5.0.1 ranks routed to empty groups correctly produce no sound.

Hence the solution is simple: please just use the Organ Configuration Wizard to reset the rank routing for St. Anne's (and any other organs for which the same applies). They will then routed to the (non-empty) groups that are appropriate for your previously reset general-level audio mixer/group settings.

In summary: given that specific sequence of events (migrated v4 audio mixer/group settings, migrated v4 organ rank routing, then reset general-level audio settings but not organ-level routing) you shouldn't have been able to hear the ranks (in v5 or potentially in v4). However, there was a minor bug in v5.0.0 whereby they would still (incorrectly) have been audible, which v5.0.1 resolves. In any case, the solution is just to use the Organ Configuration Wizard to reset the rank routing.

Martin, I'm happy to report that the development build is now working as expected. All of my organs can now be loaded without crashing; and after using the Organ Configuration Wizard to reset the rank routing for each of my organs, all of them now provide audio signals. Ironically, the minor cosmetic glitch that started this thread remains unresolved, but I suspect that issue is of low priority since the workaround (loading IRs twice) is so easy.

Your explanation of the silent audio problem was as usual both practical and educational. I've been a Hauptwerk user since 2006 and have greatly admired your work ever since. Given your track record, it never occurred to me to delay upgrading to HW5. I really don't understand why there was at first such a flurry of complaints. Being only 79, I still personally enjoy learning about upgrades.

Thanks again for your analysis and resolution,

Jim

[mdyde]

Thanks very much, Jim.

You're very welcome.

Ironically, the minor cosmetic glitch that started this thread remains unresolved, but I suspect that issue is of low priority since the workaround (loading IRs twice) is so easy.

Unfortunately, since that appears to be a compatibility issue between macOS 10.15 and the version of the (third-party) Qt multi-platform library that we use for Hauptwerk, solving it would need a newer version from its makers, and (even if they have since addressed that, which I don't know for certain), moving to a major new version of Qt isn't something that we could risk for a Hauptwerk patch release -- it could have other, much more significant, implications and would be a big project to re-test everything, and need beta testing etc., so it would take too long for v5.0.1 (even if it was technically possible). Hence that one will probably need to wait for v5.1.0 (assuming a fix exists in Qt by then).