Bill Roper's Journal

It is sad when the *second* worst thing that happened last night was Ruby getting skunked again.

I had plans to do a lot of recording this weekend. Now, whenever I need to do some recording, I usually wander down to the studio and discover that it is time to install a whole bunch of updates. It was pushing 9 PM, so I figured I could quickly run down to the studio, install the updates, and hit the ground running today.

I already knew that there was a Cubase point release that I should go ahead and install, so I did that. Then I noticed that there were three Intel updates waiting to be installed. Ok, I could take care of that too. No problem. A couple of reboots, but no problem.

The Intel update screen lists (among other things) my motherboard type and the current BIOS version. I looked at that and said to myself, "That BIOS is pretty old. I wonder if there is a newer one that I should install." This was my first mistake.

On the ASUS website, there was a brand-new BIOS for my motherboard, less than a month old. Looked good, so I figured I'd install it. This requires putting it on a memory stick, booting into the BIOS, and then loading it from there. I've done this sort of operation before, so I didn't have too much trouble with it. There was also a newer version of the Intel ME utility, so I installed that too. All good.

Having installed a new version of Cubase, I figured I'd fire it up so that it could inspect all of the plugins, because that sometimes can slow things down on the first restart. Still no problem.

Well, there was no problem until Cubase told me that I needed to pick an ASIO driver. It should just default to the Universal Audio Thunderbolt driver. Except I picked that and Cubase said "What Thunderbolt device?"

Oh, that is bad. Let me start up the UA application and see if it sees the Apollo unit. It does not. And the Thunderbolt cable is plugged in. Ack!

I start searching the Internet. Apparently, this is a problem with older versions of the ASUS Thunderbolt add-in card when the BIOS for this type of motherboard (and its various relations) gets updated. I check the Device Manager and it tells me that there is a problem with the Thunderbolt port. Yes, I had figured that out. Removing the device and putting it back does not help.

Maybe there is a newer driver or firmware for the Thunderbolt card. There's no new driver, but there is new firmware. I set up to flash the card with the new firmware and discover that it won't take it.

Apparently, there are *two* slightly different versions of this card. I have the older one, which will not take this update. There does not appear to be an update for the older card.

Maybe I can get a newer version of the Thunderbolt card. Micro Center does not carry this card. Amazon does. They can get it to me around July 1st, which is not compatible with recording this weekend. Or next weekend.

Ok. How can I get up and running? I *do* have a laptop with Cubase installed *and* a Thunderbolt port, but that is the same port that it uses for charging. About now, I realize that I could probably bodge things together with a Thunderbolt dock, but it was approaching midnight last night and I was running out of brain cells.

The latest generation of PC motherboards has a number of boards that support Thunderbolt directly on the back panel ports. My new office PC is one of those. I am not going to move my freshly-configured office PC to the basement for this. Really not.

I could *buy* a new motherboard. Which will require buying a new processor and new RAM. And a new heatsink. That is going to be annoyingly expensive and a whole lot of work, but is an available backup plan.

Let's try reverting to an older version of the motherboard BIOS. What version had I started with? Eventually, I realized that I still had it on the computer in installable format, so I copied it to the memory stick, rebooted, and installed the older BIOS. So far, so good. Let's boot up the computer.

The computer does not boot up. It beeps eight times. My phone tells me that this is a sign of a problem with the CMOS memory on the computer.

I am old. I know what to do about this. I shut off the power to the computer, pull the plug, and pull the CMOS battery. If I wait until morning, the computer will forget all of the BIOS settings and I should be able to get back into the machine. (Later, I check the manual and find the location of the two pins that I need to short to clear the CMOS. They are inconveniently buried behind the Thunderbolt card. I try fishing at them in the morning with a screwdriver, because why not? I'm not sure if I ever got to them...)

It is now nearing midnight and time to head up to bed.

At some point during this fiasco, Julie comes downstairs to tell me that Ruby has encountered a skunk, so if I smell something when I go upstairs, don't panic. It is apparently less bad than some of the previous skunkings. Gretchen has rubbed the dog down with some odor killer called "Pooph" and the report is that it has improved the situation. Gretchen, meanwhile, has gone off to the bedroom, having had enough of all of this for the night.

When I get upstairs, things are not *too* stinky, so I turn off the kitchen exhaust fan and head upstairs to join Gretchen. It is a *long* time before I can manage to get to sleep (which includes watching another episode of "Leverage" so that we can both wind down).

This morning, I get up, get cleaned up, and head down to the basement. I fish around for the clear CMOS pins, decide that I am not going to remove the Thunderbolt card to try to get at them right now, and put the battery back in. Then I fire up the computer.

Happily, after a mild round of complaints, it boots into the BIOS. I turn the Thunderbolt support back on and reboot. Windows fires up, I start the UA application and it informs me that there is no Apollo unit attached.

Then I plug the Thunderbolt cable back in. And now I have a connection! And there is much rejoicing.

And then I fire up Cubase and it tells me that there is no device connected. So I fire up the UA Console app, see the message "Connecting to Apollo", and now Cubase can see the device and lets me select the Thunderbolt ASIO driver. I open up a song, hit play, and there is sound from the speakers.

My mood is *greatly* improved.

So, kids, this is why you just shouldn't mess with a system that is working. Just ask old Uncle Bill.

In other news, the house still smells mildly of skunk downstairs and we are trying to air it out. Ruby does not seem to be very skunky, for which I am thankful.

At least I didn't have to run out and buy peroxide last night.

I have a file that is loading up one way in one release, but which is loading up differently in a different release. As far as I know, we didn't make any changes to the file loading. However, the good news is that after much effort, I now have debugging versions of both environments to load the file into, which means that with sufficient persistence, I should be able to figure this out.

Assuming that the combination of code and data is deterministic. That's normally a pretty good assumption, so I'm hoping it holds here.

Fingers crossed.

I built myself a new computer on Thanksgiving weekend. Today, I copied all (I think) of my current data files over from the old computer and am switching to the new box as my primary home machine.

We'll see how this goes -- and what I've forgotten!

It's easy to record a track down in the basement studio. Really. Yeah.

Ok, this reminds me of why I decided to record a couple of scratch tracks up here in the office. But it got better.

It didn't start well. I had gone down to the basement earlier in the evening and discovered a fair amount of disarray in the recording booth. I'm not sure why, but I have children, so that's always a leading suspect. I saw that the new mic was mounted on the stand that I wanted it on for vocals, but there wasn't a guitar mic in the room.

Before I started doing final renders for other tracks on the album, I decided to hunt down the guitar mics. I found a bunch of the cheap Marshall mics, but I remembered having bought a better pair of small diaphragm instrument mics some years back. They weren't in the mic locker though. Hmm.

I started poking around on the bookcase and found a box of small reel-to-reel tapes of some origin or another. And under the box was a big wooden box with two still-unused small diaphragm instrument mics in it.

I really need to do more recording in the studio. And I have another album that I should be working on that calls for that. But I digress...

Anyway, I finished rendering all of the other tracks on the album and went upstairs for dinner. After dinner, it was off to the basement with the guitar (in case), the lyric sheet, and my iPad. I repositioned the mics, because I planned to do this standing up, adjusted the music stand, cleared the space for my iPad, and then said "Where is the guitar stand?"

There had *been* a guitar stand in the room the last time I was there. It was not there now and it was something that I really wanted, so I went out and found it disassembled on top of the pool table. Why? I don't know. So I put that back together, put it back where it belonged, and then got the guitar out of the case and took it off to the room.

Happily, the guitar was still nicely in tune for not having been touched since Windycon. Ok, let's go get the iPad remote working.

I had downloaded Avid Control the last time that I had to do recording in the studio, because the Cubase remote app was just broken in so many ways that it was both unusable and unprintable. But I had noticed that a new version of the app had come out, so I figured I'd update the iPad to that version and see what happened.

When I went to the App Store, I was informed that they needed my password, which I provided, which was followed by my being told that I needed to go enter my password in Settings, which was followed by a demand for my phone number, which I entered, which was followed by a text message to my phone with a code that I needed to enter, which I *would* have been able to enter much more easily if the iCloud app hadn't kept coming up on top of that window asking for my password, and then after I entered the code, I was told I would need to make a new password, which I did, and then I could finally go back to the Cubase app page where I discovered that the app had been automatically updated at some point.

But I didn't come here to talk about that. I came here to talk about the draft...

(No, I didn't, but this was starting to feel like "Alice's Restaurant" there.)

Anyway, I downloaded the Steinberg SKI remote software and updated it, went into Cubase, activated the remote, connected the iPad to the computer, and amazingly, everything worked. This was good, because if I had just gone through all of that to see things fail, I would have been very unhappy.

I pulled up the UA Console, powered up the mics, and set some trial levels which at least got signal. Then I had to figure out what was wired up correctly in Cubase, because Cubase was hearing nothing, which turned out to be just a matter of setting up the hardware routing.

Ok, let's go record.

Since the original album had been two tracks direct to tape, no punch-ins, no saving throw, I figured I'd do the bonus track the same way. It's a two minute song. How hard can it be?

It's a two minute finger-picked song. Harder, as it turns out. I had a ludicrous number of false starts, but that's ok, because you just reset and start again. I wasn't trying to synchronize with anything, so I wasn't messing with headphone mixes. I was just playing.

Cubase tells me that I pushed the record button 18 times doing this. I think there were only six or so complete takes and I finally decided that the last one was good. I rendered it, compared levels to the existing tracks around it (all good), and then compared the levels to Clif's bonus track.

Clif's bonus track was substantially less loud than the surrounding tracks. Grump.

I opened Cubase back up, adjusted Clif's levels, compared it to the surrounding tracks, and rendered it again.

At this point, I need to duck over to WaveLab and apply the opening and closing fades. Then I can assemble the album, burn a CD to test with, and make the DDP master.

And if everything is good, I can upload this to the duplicator.

But that will be *tomorrow's* project.

I ran down to Micro Center this morning and picked up the bulk of the parts for the new computer, so I guess I'm committed at this point. I ordered the M2 drives from Amazon this morning, because it saved me a few dollars and let me use a gift card that I've been sitting on since I got my new phone. The case is supposed to arrive from Newegg on Friday, so all of the parts may be in order for a Thanksgiving weekend computer assembly. (Except the OS. Still need to grab an OS.)

There are a silly number of things that we need to do around the house. I think I'll go down and consult with Gretchen and see what we should take a run at. :)

I installed two new drives in my Synology NAS today. They are now getting incorporated into the RAID array and should be available later.

This is good, because I was running out of space. I expect that I will not be running out of space any time soon once this is done. :)

Ozempic is back in stock at our pharmacy, so I went out and picked up the prescriptions for me and Gretchen. This is a good thing.

While driving there, I switched to the Cardinals game on the radio. KMOX has been coming in nicely this summer (other than the usual urban interference), which isn't abnormal for a clear channel 50,000 watt broadcaster, although getting a good signal early in the evening in the summer is not always something you can count on. And then I thought about it for a moment and realized that we are somewhere around the solar maximum, so the ionosphere is nicely recharged and prepared to bounce AM radio signals with high efficiency. If I were more motivated, I'd hook the old shortwave radio up to an antenna and see which of the international broadcasters are still there to pick up.

I am not sufficiently motivated.

The Cardinals are probably motivated, but this didn't stop them from coughing up the lead before I got home.

School resumes tomorrow, so both kids will be out of the house again. This has not been a common phenomenon this summer. K is still planning to go see the new Teenage Mutant Ninja Turtles movie, but it looks like that's going to be on a weekend.

The new monitors in the office are working ok, but are not a cure for the eyestrain problem. Maybe an improvement, though. We'll see.

A shipment arrived from the mint carrying proof specimens of the 2023 Morgan and Peace silver dollars. In case you missed the news, the mint started minting these again for the 2021 anniversary and the issue turned out to be way more popular than they had anticipated, so they are doing this as an ongoing thing. Looking at the proof of the Peace dollar I am struck by just how pretty a coin this is. Maybe not as nice as the way-more-money-than-I-want-to-pay-for-one St. Gaudens 20 dollar gold piece, but a darn sight more affordable.

K has become hooked on White Collar after reading a Batman / White Collar crossover fanfic. She binges faster than we do, but it will take her a while to catch up to us. Maybe not that long, but the start of school will slow her down.

That's probably enough for now.

I've been having some problems with eyestrain with the monitors on my desk lately. Now, I might be able to improve the situation with reading glasses, but I really dislike having to use them. And new contacts might help, but I still have a large supply of the current prescription. But the monitors that I have are both running off a VGA connector at 1920x1200 through a KVM switch and it's just possible that the conversions involved are contributing to the problem.

I have lately been on the email list from a computer reseller that has an enormous number of refurbished products available and they had a pair of Samsung QHD monitors for $170 plus shipping and tax. You know, that seemed like not a huge amount of money to see if they would help out and, if not, I could surely find some other place to use them.

Except that the KVM switch I use to switch between my home computer and the work machine is VGA. The new monitors are display port and HDMI. This set off a lot of research into KVM switches and the discovery that dual monitor KVM switches need to support something called EDID or else Windows will thoughtfully move programs between monitors every time you switch computers. Oh. And getting EDID support with display port turns out to be moderately expensive. I found a KVM with display port for one monitor and HDMI for the other, which is a combination that I can make work with just one adapter for the home computer, which has lots of display port and no HDMI.

And then Amazon put the KVM switch on a Lightning Deal yesterday. So I bought it.

The monitors and KVM are now sitting upstairs in my office, but I am short one display port and one HDMI cable. Another Amazon order will fix this in the morning.

I'll wire it up and we'll see how this looks.

I took a load of laundry to the basement today and figured I would spend a bit of time in the studio today looking at some Cubase features and seeing if I could learn a few things. I learned a few things, but they weren't what I intended to learn.

The video on the studio computer has been a little glitchy in the Chrome browser lately, for reasons that I don't understand, but it's dropping boxes of flashing text onto the wrong spot on the screen. This is not good. After staring at this for a bit, I decided that maybe I should try updating the video drivers. So I went to the Nvidia site and let it download the GeForce Experience app and install it. I also pulled down the latest Windows 10 update in case that was the problem. And then I fired up Cubase.

The screen was now a complete mess (I mean, garbage and bad vector graphics everywhere) and the system crashed, eventually blue screening. This was followed by Windows doing the recovery dance, checking the hard drive, failing a boot, and eventually getting its act back together.

Off to a Google search which suggested that I needed to make sure that I had the latest Cubase version installed (I was one minor -- bug-fix level -- release behind) and that I should uninstall all of the Nvidea software and reinstall *just* the video driver. Ok, I could do that.

Well, no, I couldn't. The Cubase update install crashed the machine again.

So I made sure that the audio interface was turned off, rebooted, and started uninstalling Nvidea software. A couple of reboots later, it was all gone and I was looking at a VGA-sized screen. Then I installed the latest video driver (from 2018). And then I installed the Cubase point release and fired it up, loading the giant-sized demo project.

Which played correctly with no audio or video glitches.

Further examination indicates that the video card in this machine is an Nvidea GTX 560 TI, which originally started shipping in 2011. I guess that explains why driver releases are few and far between now.

I am really starting to believe that the current studio computer is aging out rapidly.

The video card, however, is not a part that I really wanted to replace in a new build.

*sigh*

I found myself with some spare time at the computer today and did some work on a *very* long delayed project.

It's not nearly done, but now it's just a matter of repeating steps until it is.