Audio pops

Hi Les,

A bit off-topic, but I've just read an article about finding problems with RAM.
https://www.theregister.com/2024/01/11/ ... _released/

The tool it refers to (https://memtest.org/) may be able to confirm if there is a problem with the RAM on your computer.

I've never used it, but thought I'd share the info in case it was helpful.

Alex

The pops returned after waiting a day. They are, of course, on different pipes. Clearly one of two things must be causing this:

1) I have a memory chip that has a flat that occurs only every day or so or

2) Something is overwriting a position in RAM - but it only happens approximately daily.

If I could truly convince myself that it is option 1, then I would immediately either buy new RAM or simply buy a new computer. I am looking for tests that could help mje distinguish between the two options. Any ideas?

Since this is a Mac mini, it is non-trivial to remove or switch out RAM - and this is done ion 32 GB chunks. I want to minimize the nudge rod times I try this.

In the meantime, I will force myself to reload organs each time I play them to minimize this effect.

Les

When I went to reload the PCC organ, I discovered two new symptoms. First, there was another Apple process failure message - for "loginwindow". This appears to be a process associated with the Finder - and I had no way to close this window or save the text within it. Also, as a result of this failure, HW was placed into the "Load Organ" dialog.

I restarted the computer and will try loading a reduced size version of my PCC sample set - trying to get it to fit on a single RAM module.

Les

[mdyde]

I don't know of any way to test individual RAM boards, other than by removing/swapping them, unfortunately.

Yes -- it does look rather fiddly for the 2018 Mac Mini:
https://9to5mac.com/2018/11/20/how-to-u ... ini-video/

Apple says its RAM isn't officially user-installable: https://support.apple.com/en-us/HT205041 .

I suppose you could conceivably install a completely clean macOS installation (with new, clean Hauptwerk installation) in a new partition (e.g. on external SSD), as a test just in case the existing macOS installation has become corrupted on the drive. However, it seems fairly unlikely that it would solve it, since the macOS 14 upgrade would probably have replaced most of the macOS executable files anyway.

By loading only four of the seven division, and using 16-bit sampling rather than 24, the custom PCC takes up less than 6 GB! It is not quite as small at St. Anne, but it is still very small.

It sounds fine - though I now have only a three-manual instrument. I'll see if this eliminates the delayed popping tomorrow... If so, this will increase the likelihood that the problem is one of the two RAM modules.

Les

Martin,

I agree that the upgrade to OSC 14 likely did a good job of cleaning any disk problem I might have had. I do not use disk portioning - all one partition. I do not have an external disk drive on this machine. The 1/2 TB size is more than adequate for a HW-only system - and I run all my sample sets in simple stereo with only one perspective.

The RAM on this Mini is indeed not user-installable, according to Apple. However, I purchased these two 32 GB modules and installed them myself anyway. I own the special tools required to do this and I followed a very good video showing how to do this. This was several years ago and my "delayed pop" problem only started last month.

Les

I'm sure you installed them correctly! RAM does go bad though, on occasion. It's one of those computer things.

I found a pop (albeit a much quieter one) in my reduced PCC sample set this morning - on a pipe in the 32' Contra Bombarde. This seems to imply my lower-address 32 G RAM board is faulty - but at an address higher than St. Anne will exercise. I was not expecting this.

I discovered today that Apple no longer sells a Mini that can address 64 G RAM! This makes my replacement option more expensive as I will need to purchase a Mac Studio instead.

There is also a chance that my old HP touchscreen monitor will not be compatible with a newer machine - which would mean purchasing a new monitor and a new driver from TouchBase. This could get expensive, but I may have no other choice. I am unwilling to switch to Windows given my expedience with Mac and Unix. I am glad I have a good retirement package!

Les

Data corruption, whether a hardware or a software error, tends to turn out as a (somewhat) random value at a (somewhat) random place. I don't think a soft pop at another location rules out any cause.

There are more extensive hardware tests than Apple Diagnostics, e.g. MemTest86. It requires you to create a bootable USB stick and then run the mac from there.

BTW, I'm a mac guy, and I bought a PC for Hauptwerk. Still not a fan, but as long as you only use Windows for starting Hauptwerk and such, it's fine. After a few system changes for best performance, a dedicated Hauptwerk PC isn't bad.

Although I have HW installed on a PC, I have also had clicks several times during playback. Going from W7 to W10 was terrible; I had to disable WinSat from W Defender and other services that were clearly starting when the clicks appeared. I had the "Resource Monitor" open next to the HW for a while to see it. Also the standby RAM usage of W10 is almost triple that of W7.
But I have always considered the possibility of reconfiguring the latencies and clock of the RAM modules so as not to make them work at their limits.
Moving the installation folders of those organs that you do not usually use to another disk is a good idea to make room on the hard drive.

Biogons,

Thanks for the new ideas. Since I am Mac-based, some things clearly work differently. I am about to go on a month-long vacation. When I return, I will likely replace my computer on the assumption that this is a faulty RAM problem. Starting with a new installation on a new machine will also clean up other possible problems. I will get a larger disk drive so I stay very far away from disk limits too.

Les