Resolved ….HW4 Backup Carryover to HW5

Does note-by-note level voicing carry over from the HW4 backup file, or does HW5 see a new organ number and not associate the old sample-set-specific data to the HW5 version?

One thing I've noticed as I begin the task of reinstalling sample sets in HW5 is that note-by-note level voicing seems to be way off in some cases in quite a few ranks. I have restored a HW4 backup (latest) into HW5. It isn't behaving like one of my output stereo pairs is out of whack, because there's no pattern to the note intervals between too loud and barely hearing the pipe at all.

I probably do need to rebalance my raw audio channel levels since moving to a new machine and installing new drivers anyway. And dig in and learn the new mixer, or course. But something else seems off ...

[mdyde]

Hello Grant,

Yes -- all existing per-pipe voicing adjustments should certainly be migrated to v5. (The exception would be if the sample set producer's new version of the organ definition specifically changed the internal organ/pipe/layer IDs for some other reason, but that wouldn't have been needed or expected purely for v5 compatibility.)

There's no change between v4 and v5 in the way that the pipe level voicing adjustment should behave. (The previous 'mod depth adjustment dB' settings are now specified as percentages [not decibels], but the dB values are converted to percentages during the migration, so there should be no audible change as a result of that.)

Do you see the same issue with St. Anne's?

Also, if you temporarily load a sample set that exhibits the issue in a 'spare' 'alt config N' configuration, and use 'Organ settings | Reset all voicing ...', are the pipe levels of the relevant ranks still unexpectedly uneven?

In case you didn't migrate your existing v4 audio routing settings to v5, could it simply be that the pipes are being heard from different speakers, and thus interacting differently with your listening room?

Thanks, Martin ..

I did import all of the settings, including audio configuration into V5. The configuration is pretty simple — 3 stereo pairs and mixdown to one sub.

I'll load St Anne and check to see if I get similar results. I'm also planning on putting up Trost today also which should be a good test. I was actually testing St Georgenkirche when I first noticed it.

The improved pipe voicing feature that allows adjustments in real time while pipes are playing was a tremendous help in getting the set rebalanced quickly in any case. Thank you for that change!!

[mdyde]

Thanks, Grant.

Martin, et al ….

I'm still trying to figure out what's going on with audio after loading my last HW4 backup into a fresh install of HW5.

I've checked all the cables from my Echo Audiofire 12 to the speakers and headphone preamp. And am using only the headphone mix down to simplify things. I haven't had time yet to dig in to the new mixer and routing (working on it, but Christmas has me buried right now) — but presuming it would have been translated correctly from the HW4 backup, something isn't working right.

I have 3 stereo pairs. In HW4, in addition to being fed to their respective speaker channels, they were sent to 2 different mixdowns — one to the subwoofer channels, one to my headphone channels. Pretty simple.

Using the headphone mix down only in V, what I hear is 2 out of 3 pipes going (essentially) mono to the right channel, and the rest seemingly going to both channels with emphasis on the left. In the first case, it isn't a matter of stereo perspective — the left channel is simply missing. The anomaly is less noticeable on the top end of the keyboard — seemingly — and is most obvious in the bottom 2-3 octaves. This appears to be what was causing my initial concern about voicing of the Georgenkirche. I'm testing now with the Trost and it suffers from the same issues. I did quite a bit of pipe-to-pipe leveling on the Trost in HW4, but HW5, after importing the update from HW4, doesn't reflect any of that — everything is flat, essentially out of the box flat on the HW5 voicing screens. So changes made in HW4 didn't make it across the divide.

The same thing is happening with Steinkirchen — so it isn't sample set specific.

I don't know where to look, frankly. But clearly, something didn't translate correctly from HW4 to HW5. Can you (or anyone) provide some pointers of where to start poking at the beast? For now at least, I'm lost in the weeds …

Thanks ..

[mdyde]

Hello Grant,

Could you send me a diagnostic file? (I'll send you a P.M. with my direct email address.) I can then look at how your v4 audio settings got migrated. Thanks.

[mdyde]

Hello Grant,

Thanks for the diagnostic file and screeshots.

It turned out I got the right and left output channels confused. It’s the right channel that seems to “vanish”, not the left. I had my headphones reversed.

I’ve found a very odd anomaly in the way the HW4 audio definitions were translated when I loaded my HW4 backup. All I’ve done to any of this so far in V is rename the audio channels on the Audiofire.

In any case, the first image below is the Audiofire definition. The three speaker pairs use (L/R) CH 2/1, CH 4/3, Ch 6/5. HW5 appears to have assigned the pairs as (L/R) CH 2/3, CH 4/5, and CH 6/7.

Since port 7 on the Audiofire is not connected to a speaker, that would explain why the Right channel disappears completely on some notes, and why everything else doesn’t sound at all balanced.

Looks a bit odd .. I’ll try to fix it in the morning.

I've had a look at your v4 audio outputs:

- Main1: device channels: 2,1
- Main2: device channels: 4,3
- Main3: device channels: 6,5
- Sub Channels: device channels: 8,7
- Headphone Channels: device channels: 21,11

... and can I can see the problem: for a stereo mixer bus (v4 audio outputs translate to v5 audio mixer buses), Hauptwerk v5 requires that the right channel number be one higher than the left channel. Here's the excerpt from the release notice about that:

Please also note that technically Hauptwerk v5's audio configuration also has the following limitations compared to v4: ... the left and right device channels selected for any given mixer bus (v4 audio output) when stereo should usually be consecutive (although if you really want to use non-consecutive channels then channel orders can be remapped manually via the 'General settings | Audio device and channels' screen).

Hence it wouldn't have been possible to migrate your v4 settings exactly to v5 (given that you had their left and right channels swapped around, thus the right channel number wasn't exactly one higher than the left channel number). The migration did the next best thing it could, and assigned the right channel number to be one higher than your specified left channel number. It would have written some migration warnings to that effect in your Hauptwerk log when the migration was performed, mentioning that you would need to check/correct the relevant settings manually.

My recommendation would be just to swap the left and right cables for each stereo pair that connect from your audio interface, rename the device channels accordingly (e.g. device chan 1=Top L, device chan 2=Top R), and then just adjust the channel numbers on Hauptwerk's mixer accordingly (e.g. select device chans 1/2 for your 'Main 1' mixer bus).

Martin ..

I missed that in the release notes. However I did configure the correct channels this morning and now all is well.

Thanks … Grant

[mdyde]

Thanks, Grant. Excellent. No problem. Sorry that it wasn't possible to handle that particular case in v5.