So I've now completed the "CSV-based RESTORE" of my A16 setup, following the factory reset required with the installation of firmware 2.15. Note that I had previously been somewhat behind, still running 2.11 which did not yet have the original delivered beta CSV processor. So I actually had to go forward with the 2.15 install and factory reset, in order for me to have the first opportunity to try out the CSV method.
And I can confirm that it worked almost perfectly. Yes, it took several days of work to customize the prototype XLSX sheets that were provided to match my configuration, but that really was a great way to review my previous configuration and re-organize all of the BMP screenshots I'd made previously, or take new screenshots for those setup screens I'd overlooked. So now I have a complete graphical collection of BMP images reflecting my setup, as well as the XLSX version of that same information from which the A16 can be rebuilt (mostly) fully automatically (mostly).
As expected there were a few minor clerical typos in my data entry, easily corrected either as the build process proceeded, or once the build process completed and I reviewed everything. Also I did not use the "macro approach" to automatically saving all of the sheets individually as a CSV. I preferred to do each sheet manually, myself, just to be sure I was saving things in the right place and in the right name and NOT re-saving the complete XLSX by accident, etc. A small amount of manual effort, but just a one-time task.
However I also did run across what I would characterize as some problems with the XLSX, as well as the CSV processor inside the A16 which seemed to not build my Auro listening rooms to match the CSV input.
(1) This isn't really critical, but on the Solo/Mute table as delivered in the sample XLSX, the value for the OH key shows as "T", and the keypad group values are shown as follows (which probably reflects a particular Smyth A16 setup):
Actually I have the keypad solo group set to NO because I use those keys for "instant preset selection", so the values differing from true factory default doesn't really bother me. But to be a stickler for accuracy the actual default value for the OH key following a factory reset is "blank". And this is also what the User Manual shows in section 5.3.1, although there is no explicit description of what the OH key means (other than to group-play whichever six overhead speakers are shown in the "OH" boxes) and what its valid values might be if not "blank". I don't know if "T" and "blank" are functionally identical.
Nevertheless I decided to just blank out OH, and specify exactly the values shown following the factory reset, while also setting "Key Pad Group" to my NO preference to activate these keys for instant-preset selection.
(2) There are three different values shown or described in the documentation when making reference to the BBC PRIR for speaker extraction where the explicit user's PRIR does not contain all the specific speakers required to construct the particular listening room:
(a) use factory prir
(b) user factory prir
(c) use_factory_prir
There's no explanation of this in the documentation, so it's not clear if all three are usable valid values or just one or two of them. Anyway, I decided to change ALL of the references wherever they were needed to show the (a) value above, i.e. "use factory prir". That certainly seemed to work just fine.
(3) In defining the presets for each of the 8 users, the documentation says "First, select the username for the preset. This exposes the username list you entered in the Common user tables tab." And then there is an image provided to illustrate what is to be done:
However in reality nothing like this actually occurred for me. I was unable to produce that username dropdown list by clicking on any of the cells in that table, including on the 'Name" title, or any of the 16 cells below it in that column E. I eventually just gave up and manually clicked on the 8th cell in that column below "Name" (i.e. the name value on the same line as the "/1" and then manually typed in the correct username.
(4) Although I'm not doing any measurement so this oversight isn't at all critical, it appears that the "Settings -> Headphones" section (as described in User Manual section 5.2) is missing from the XLSX. So there's no way to automatically install these values. The XLSX and CSV processor needs to be updated to include support for this one additional table.
Not urgent. Just for completion it's easy enough to manually populate the proper values in the A16 once the CSV build process completes.
Note that the XLSX and CSV processor also need to be updated to support the relocation of the Auro settings to be specifically preset within each preset, from where they previously were "globablly" located in "Misc Settings". Support for the Headphones table can be incorporated at the same time.
(5) Now for the one VERY SIGNIFICANT PROBLEM that I did encounter. All of the above items are minor and trivial. But this one did require me to spend a bit of time carefully correcting and rebuilding my Auro listening rooms.
==> Despite my XLSX "Auro listening rooms" sheets (I have two of them, one for Auro rooms 5-8 and a second for Auro rooms 9-12) and corresponding saved two CSV files being correctly coded, the CSV processor's construction of the Auro rooms DID NOT CORRECTLY REFLECT THE AURO ROOM LAYOUT I had coded for my non-skipped actual Auro rooms 5-10 (5-8 in the first CSV, and 9-10 in the second CSV). I believe it was only for my Auro rooms 7-9 that were built wrong, with rooms 5-6 and 10 being built correctly. The problem with rooms 7-9 was that the CSV processor forced "7.1.4h" as the room layout and set up the room speakers that way even though that was not the room layout I'd specified for those three rooms.
Note that the real cause of this problem may be that there is no dropdown list that appears in that room layout cell to specify what is supposed to be the correct value for that cell. Yes, there would be a different set of valid room layout values depending on Atmos, Auro, DTS or PCM rooms. But still, it's not clear exactly what the value in room layout cell should be for Auro in particular when some of the values shown in the documentation (i.e. the 2.15 Install Instructions which provides the Auro room layouts) actually differ from the name of that layout which the A16 offers if you do a manual room construction on the A16 itself.
Although theoretically this same problem could also exist for other listening rooms classes (i.e. Atmos, DTS and PCM) it is only for my Auro rooms 7-9 that I had this problem, most likely because I'd manually specified the incorrect room layout value in the absence of a dropdown list to guarantee I was choosing the correct value.
To be specific, the layouts of my six Auro rooms are: 5.1, 7.1, 9.1, 11.1, 13.1, and 7.1.4h. These correspond to layout numbers: 12, 22, 16, 20, 26 and 24, from the following chart provided in the 2.15 instructions:
However the actual "LAYOUT NAMES" for Auro rooms as offered by the A16 itself (where it is referred to as "Listening Mode") for my particular 6 Auro rooms are: 5.1, 7.1, Auro 9.1, Auro 11.1, Auro 13.1, and 7.1.4h.
==> I should have manually coded: "Auro 9.1" rather than "9.1", and "Auro 11.1" rather than "11.1", and "Auro 13.1" rather than "13.1". That is my hunch.
I'm sure it was no coincidence that the three Auro rooms 7-9 where my manually coded room layout values (as shown in the chart) differed from the values most likely edited for by the A16 CSV processor all got built incorrectly, all three producing 7.1.4h by default error-correcting action. That is my hunch.
I've now changed my three room layout values for Auro rooms 7-9 to be what I think they should be (again, in the absence of a dropdown list of valid values in the XLSX cells to select from). I would like to re-run just the Auro listening room CSVs to see if my hunch is correct, ending up with the properly built Auro listening rooms. I'm hoping that will actually work (even though I've already manually corrected Auro rooms 7-9 using the A16 itself). I'm hoping I can do just this "corrective rerun of only the Auro listening rooms CSVs" and not need another full factory reset and full CSV rebuild to prove my theory.
EDIT: Just updated my XLSX in anticipation of my re-run of the Auro listening room re-build. I WAS able to enter "Auro 9.1" instead of "9.1", and it was accepted and the speakers shown remained just as it was before. Similarly, I was able to change "11.1" to "Auro 11.1" and again the speakers remained unchanged. However when I changed "13.1" to "Auro 13.1" that apparently is invalid (incorrectly so, obviously) because all the speakers got flushed and replaced by "#N/A".
==> It would appear the XLSX must be updated to accept "Auro 13.1".
In the meantime I will manually enter the correct values for all the speakers while also retaining the value oa "Auro 13.1" which I believe is what the CSV processor in the A16 is actually looking for. I should therefore get the correct "Auro 13.1" instead of "7.1.4h" for room 9.
We shall see.
(NOTE: somehow there is a stray "attachment" on this post that I didn't intend to post, of my first screenshot, and I don't know how to delete it. Sorry for the distraction.).
