It is currently Sun Oct 25, 2020 6:35 pm


Curious Audio Glitch - SP CAEN

Buying or building computers for Hauptwerk, recommendations, troubleshooting computer hardware issues.
  • Author
  • Message
Offline
User avatar

dw154515

Member

  • Posts: 349
  • Joined: Mon Nov 12, 2012 2:52 pm
  • Location: Indianapolis, IN

Curious Audio Glitch - SP CAEN

PostWed May 13, 2020 7:31 pm

Folks,

I am referencing the organ built for Greenwood UMC, here.

http://forum.hauptwerk.com/viewtopic.php?f=17&t=15009

I have, on multiple occasions now, had a very curious audio glitch that so far (knock on wood) has only ever happened on the Sonus Paradisi Caen set. Now, I do not, in any way, think this issue is due to any fault with the set. But there is something causing this that makes it unique to that set. (Or maybe it's wild coincidence.) However, there has to be some reason that this pesky dropout (or series thereof) hasn't appeared in ANY of my other samples (SP Casavant, Armley Schulze, Silbermann, Paramount 341, SP Reuter, Redlands Casavant....). I am using the wet subset of the Caen, which requires very little RAM (relatively speaking, compared to the others).

Basic computer specs:

Hauptwerk V4
Windows 10
64GB of RAM
MOTU 24ao Interfaces
All SSD

What happens is what I can only describe as the sound of a string of firecrackers going off. It happened once, at full tutti, across all 50ish channels.... and it absolutely scared the heck out of me. It sounded like machine gun fire, or exactly like someone threw one of those strings of firecrackers up into the speaker chambers.

When this happened a while back, I thought I had solved the problem. I went into the BIOS and forced the RAM speed, rather than relying on "Auto". This seemed to have cured the issue..... for a while. Now, it has done it again - albeit, nowhere near to the extreme that had happened in the past. This time around, it is only one or two quick drop outs (which manifest themselves as loud POPS) and then it's stable again.

The audio meter never goes into the red. And it can happen no matter how full or soft the registration.

I also noticed that one of the sample rates for one of the ranks (in the Load Organ Adjusting Rank Audio...., etc.) hadn't been set the same as the others - now they are. I don't have it in front of me, but I do believe they are all set to 20-bit (or 24-bit).

My hunch is that the issues is either RAM or clocking with/between the MOTU interfaces. However, again, I can't help but feel there is some reason why CAEN is behaving differently....

Any thoughts?
Drew A. Worthen
Master of Music in Composition - Butler University
http://www.drewworthen.com
Director of Music & Website Admin - Greenwood UMC
http://www.greenwoodumc.org
Field Engineer - Diversified (Formerly Sensory Technologies)
http://www.sensorytechnologies.com
Offline
User avatar

micdev

Site Admin

  • Posts: 1824
  • Joined: Sat Feb 09, 2008 10:24 am
  • Location: Canada, Quebec

Re: Curious Audio Glitch - SP CAEN

PostWed May 13, 2020 9:03 pm

Good evening Drew,

First thing I would do is to update the Motu drivers; you have a very complex setup (multiple 24ao) and having the latest driver will probably fix some "weird bugs". Also, make sure to update the firmware of your 24ao(s) so they use the same one and the latest one.

Since it is rarely happening, it is hard to diagnose... you're saying that the CPU doesn't peak in the red... what about the "audio meter"? When you had the "bug", were you playing with a large registration? What is your TrimDB setting? On my system (using a 24ao) if I set the trim dB above -13 the audio meter will peak in the red when using Caen with large registration and playing a complex piece (very high polyphony). What is your polyphony setting for Caen. Why Caen and not the others? Maybe because all the other organs have a lot let reverb, requiring less processing power... but you're saying it doesn't peak...

I don't know when your system was installed, but based on the link provided, it was probably in 2016, almost 4 years ago, so quite possible that not only your drivers are old, but your computer BIOS also (and it was probably one of the first BIOS for this motherboard), check and update also.

Why Caen and not the others? Well, all the other organs has a lot let reverb, requiring less processing power. What is your polyphony setting for Caen? When you had the problem,
Best regards
François

Virtually sharing my enthusiasm and experience with you
Worldwide technical assistance, consultation and ready to play system.

http://www.HauptwerkConsultant.com

AND Hauptwerk Support Manager
Offline
User avatar

dw154515

Member

  • Posts: 349
  • Joined: Mon Nov 12, 2012 2:52 pm
  • Location: Indianapolis, IN

Re: Curious Audio Glitch - SP CAEN

PostWed May 13, 2020 11:03 pm

Thank you for the input. That system was completed early last year. The computer never touches the internet, and therefore never updates. (Same for the MOTU interfaces). It is more of an "appliance" at this point. Once I got it stable, I don't want to have to mess with it.

Just to clarify, NONE of the meters go into the red. I'm not sitting at the organ right now, but I do know that the Polyphony is set extremely high and has also never peaked. I had to do a lot of setup and tweaking in the beginning to get this thing firing on all cylinders - followed the "Glitch Free" guide to a tee, and disabled just about everything in Windows that wasn't crucial to Hauptwerk's performance.

This might be helpful to note: It happens "in spells," I've noticed, too. In other words, if I get the dropouts, it's usually on a big registration. And if I keep playing on, it keeps popping, snapping, crackling, etc - and sort of "sputters" out if you let off the keys. The dropouts cary all the way through the reverb tails. If I immediately switch to a soft registration and keep playing, the artifacts are still there - pop, pop, crack, bzzzzzz, bang...... At that point I usual stop, close Hauptwerk, and try again.

I was there a couple of hours ago (after posting the original message) and it played flawlessly for about an hour. No issue whatsoever. It's very unnerving; always fearful of setting off a string of fireworks in a service or performance. :twisted: lol

I will give the firmware update a shot - I suppose that one never hurts.
Drew A. Worthen
Master of Music in Composition - Butler University
http://www.drewworthen.com
Director of Music & Website Admin - Greenwood UMC
http://www.greenwoodumc.org
Field Engineer - Diversified (Formerly Sensory Technologies)
http://www.sensorytechnologies.com
Offline
User avatar

micdev

Site Admin

  • Posts: 1824
  • Joined: Sat Feb 09, 2008 10:24 am
  • Location: Canada, Quebec

Re: Curious Audio Glitch - SP CAEN

PostThu May 14, 2020 7:20 am

Hi,

I understand your "it is more an appliance". This is what I do with all my clients, but when I have a problem, well we need to do something, and this means to upgrade a few thigs (even Windows!)

To protect yourself, since you are tech savvy, you could buy an extra SSD drive (not really expensive) then clone your actually SSD disk. So should the upgrades goes wrong, you will be able to rollback.

Make sure also that your AVB switch is also updated... and wiring well connected, not damaged.

Hope this helps
Best regards
François

Virtually sharing my enthusiasm and experience with you
Worldwide technical assistance, consultation and ready to play system.

http://www.HauptwerkConsultant.com

AND Hauptwerk Support Manager
Offline
User avatar

dw154515

Member

  • Posts: 349
  • Joined: Mon Nov 12, 2012 2:52 pm
  • Location: Indianapolis, IN

Re: Curious Audio Glitch - SP CAEN

PostTue May 19, 2020 8:07 pm

micdev wrote:Hi,

I understand your "it is more an appliance". This is what I do with all my clients, but when I have a problem, well we need to do something, and this means to upgrade a few thigs (even Windows!)

To protect yourself, since you are tech savvy, you could buy an extra SSD drive (not really expensive) then clone your actually SSD disk. So should the upgrades goes wrong, you will be able to rollback.

Make sure also that your AVB switch is also updated... and wiring well connected, not damaged.

Hope this helps


I did the MOTU Firmware updates, just for good measure, but as it turns out, I had a bad stick of RAM. Seems to be batting 100 again.
Drew A. Worthen
Master of Music in Composition - Butler University
http://www.drewworthen.com
Director of Music & Website Admin - Greenwood UMC
http://www.greenwoodumc.org
Field Engineer - Diversified (Formerly Sensory Technologies)
http://www.sensorytechnologies.com
Offline
User avatar

nrorganist

Member

  • Posts: 44
  • Joined: Mon Sep 26, 2016 7:22 pm
  • Location: Northern Colorado, USA

Re: Curious Audio Glitch - SP CAEN

PostTue May 19, 2020 11:39 pm

Drew,

Given the bad stick of RAM - this made me wonder whether Hauptwerk systems in public performance spaces would benefit enough from error detection/correction of a computer with ECC (Error Correcting Code) memory enough to justify some extra cost?

Since some of the best sample sets are only getting larger, a few years back, I searched for very large memory PC-based systems. I found refurbished HP Z-series workstations with single or dual Xeon processors and ECC memory were minimally more expensive than Core i7 or i9 alternatives without ECC. Z-series runs Hauptwerk tested and supported Windows Professional (doesn't need Windows Server), and are quiet (29 dB) due to a multi-fan / multi-airflow-path design. On the downside, they are full-sized towers which start weighing in at 50 lbs. But I haven't regretted ending up with a Z840 and populating a few of the 16 slots with 32GB modules.

Just a thought.

Mark
Offline
User avatar

dw154515

Member

  • Posts: 349
  • Joined: Mon Nov 12, 2012 2:52 pm
  • Location: Indianapolis, IN

Re: Curious Audio Glitch - SP CAEN

PostFri May 22, 2020 9:01 pm

nrorganist wrote:Drew,

Given the bad stick of RAM - this made me wonder whether Hauptwerk systems in public performance spaces would benefit enough from error detection/correction of a computer with ECC (Error Correcting Code) memory enough to justify some extra cost?

Since some of the best sample sets are only getting larger, a few years back, I searched for very large memory PC-based systems. I found refurbished HP Z-series workstations with single or dual Xeon processors and ECC memory were minimally more expensive than Core i7 or i9 alternatives without ECC. Z-series runs Hauptwerk tested and supported Windows Professional (doesn't need Windows Server), and are quiet (29 dB) due to a multi-fan / multi-airflow-path design. On the downside, they are full-sized towers which start weighing in at 50 lbs. But I haven't regretted ending up with a Z840 and populating a few of the 16 slots with 32GB modules.

Just a thought.

Mark


No doubt, what you just said makes a heck of a lot of sense, but it is over my head, if I'm honest. My knowledge of PC components is marginal - I know just enough to get by. Your suggestion sounds great in theory but I'm afraid I'm not smart enough to know any pros and cons in using ECC.
Drew A. Worthen
Master of Music in Composition - Butler University
http://www.drewworthen.com
Director of Music & Website Admin - Greenwood UMC
http://www.greenwoodumc.org
Field Engineer - Diversified (Formerly Sensory Technologies)
http://www.sensorytechnologies.com
Offline
User avatar

engrssc

Member

  • Posts: 6735
  • Joined: Mon Aug 22, 2005 11:12 pm
  • Location: Roscoe, IL, USA

Re: Curious Audio Glitch - SP CAEN

Offline
User avatar

nrorganist

Member

  • Posts: 44
  • Joined: Mon Sep 26, 2016 7:22 pm
  • Location: Northern Colorado, USA

Re: Curious Audio Glitch - SP CAEN

PostSat May 23, 2020 6:11 pm

Drew,

No worries - I found an article on the Puget Systems (a successful high performance custom computer build company for nearly 20 years) website from a few years back which appears to summarize ECC memory clearly, concisely and fairly. Alternatively, to gain higher confidence in non-ECC memory, one could extensively test memory cards before use.

I see Ed started a new thread (viewtopic.php?f=16&t=18696) on it, since my question was admittedly a tangent. I will post details on both in Ed's thread.

Mark

Return to Computer hardware / specs

Who is online

Users browsing this forum: No registered users and 5 guests