Report on sequencer problems due to RAM

elvisjohndowson · Post by **elvisjohndowson** » Fri Oct 19, 2007 5:37 am

Hi Dan,
This bug used to be there early on, don't remember which O/S version exactly, the one that was current in Sept 2007. At that time, the LAC-1 upgrade wasn't there and I think a similar type of crash occurred in the sampling mode or disk mode. I'm not sure which.

The only reason I stumbled onto this one was because I was trying to see which buttons were mapped to the Polysix EXi, and it was reminiscent of a similar type of lock up that I had in the past.

Elvis Dowson

danatkorg · Post by **danatkorg** » Fri Oct 19, 2007 5:42 am

elvisjohndowson wrote:Hi Dan,
This bug used to be there early on, don't remember which O/S version exactly, the one that was current in Sept 2007. At that time, the LAC-1 upgrade wasn't there and I think a similar type of crash occurred in the sampling mode or disk mode. I'm not sure which.

The only reason I stumbled onto this one was because I was trying to see which buttons were mapped to the Polysix EXi, and it was reminiscent of a similar type of lock up that I had in the past.

Elvis Dowson

I believe that you saw something before, but this particular freeze appears to be very specific. If you find another reproducable example, please let us know!

Best regards,

Dan

elvisjohndowson · Post by **elvisjohndowson** » Fri Oct 19, 2007 6:31 am

Will do Dan!

Mag66 · Post by **Mag66** » Sat Oct 20, 2007 6:25 am

Anashwaran wrote:
Mag66 wrote: Yeah.. the errors I got were quite intermittant.. I think I got a couple of errors every 25-30 passes.. (Corrected number of passes)
Do you mean you had your test running for 15-20 hours?
Because it took almost 2 hours for 2 passes.

IIRC.. I ran the tests over a weekend.. so yes, it took about 36 hours in total...

I should also add that I don't think that I had a bad stick of Korg installed memory. I think my issue was fixed by using matched pairs of memory sticks. Maybe the stick I added originally, had some form of timing issue when used with the Korg stick so using two sticks that are exactly the same, having replaced the Korg stick.. fixed the issue for me

EJ2 · Post by **EJ2** » Sat Oct 20, 2007 2:18 pm

It would be interesting to know if indeed "matched" pairs of ram are "part" of the solution to some of the issues being reported. Simply switching the position of my sticks "seems" to have corrected my sequencer problem.

I still have to re-open my OASYS to copy down the particulars written on my sticks and report back to Dan and Brandon. Sorry, guys, I've been rehearsing for my upcoming mp3 demos, so I'm a bit reluctant to upset things for the moment. I will get to it though.

It seems strange, however, that a good many of the issues being reported, including mine, have not been reproduceable at Korg R&D. I guess the etherial world of computers might be preventing many of us from tracking things down. Certainly, the fix (switching ram stick locations) suggested by Andy, Daz, and you has worked - I think. But as Daz has pointed out, I may have just shifted the digital pothole elsewhere. I'm suspicious that he may be right. Last night while pounding out some improvs, I had a stuck note in one of my lead sounds. That's never happened to my O before. So, I'm tempted to follow up with your suggestion to get a match for my Kingston. But then again, will it be a 100% identical match?

Cheers,
Eric

Lou · Post by **Lou** » Sat Oct 20, 2007 4:35 pm

For those of you that had some luck with shifting ram around, you're fortunate, but I'm skeptical of this. I've done the switch and notice absolutely no change. Identical failures.

While Korg R&D have their hands full right now, trying to locate these issues, some repeatable and some not, I wouldn't want to be in their shoes. Thank God their here for us. Yet, I can't help but feel that I don't want to be in my shoes either.

It's understandable that OS systems will have problems and that a great deal of rigor must have gone into the beta testing. I just don't understand how so many issues that have been reported could have gone unseen. Let's face it, there are many. I'm not knocking anyone here, I'm just wondering how many O's were in the beta line? I'm becoming so disgruntled with my O, and loosing faith in her performance. (Love/Hate) Praying for a fix... and standing by

danatkorg · Post by **danatkorg** » Sat Oct 20, 2007 5:20 pm

Lou wrote:For those of you that had some luck with shifting ram around, you're fortunate, but I'm skeptical of this. I've done the switch and notice absolutely no change. Identical failures.

Have you tried removing the additional RAM module, returning to the stock configuration?

- Dan

Lou · Post by **Lou** » Sat Oct 20, 2007 7:03 pm

Hi Dan

No, I haven't done this. I figured I was running 123 without problems. Are you suggesting to remove the additional ram and place the factory back to the front slot, leaving just the 1g? I will try this if you think there is some relevancy. I think I'm going to leave out the 2 screws for a while, it's becoming a daily thing, opening up the hood.....

Why would the additional ram be working fine in 123 and then cause issues with 1.3? This is not my territory and would like to understand this a little.

Thanks

danatkorg · Post by **danatkorg** » Sat Oct 20, 2007 11:14 pm

Hi Lou,

Lou wrote:No, I haven't done this. I figured I was running 123 without problems. Are you suggesting to remove the additional ram and place the factory back to the front slot, leaving just the 1g? I will try this if you think there is some relevancy.

It's a long shot, but it's a good troubleshooting step.

- Dan

Anashwaran · Post by **Anashwaran** » Sat Oct 20, 2007 11:17 pm

Lou wrote:Hi Dan
Why would the additional ram be working fine in 123 and then cause issues with 1.3? This is not my territory and would like to understand this a little.

Thanks

I want to ask the same question.
Why not admit there is some bug in this update?
And check on that side? There is no error in my Mem-testing and there was no problem before 1.3! (I have posted an answer to this which has disappeared from the Forum. Strange

)
I reported my trying to go back to 1.2.3 and what happened to me in the Karma-Lab Forum. I will not repeat it here. But as I have not yet bought the Update with MOD-7 there would not really be some interest to keep 1.3 except for Karma2. But I cannot work anymore at the moment, and am very frustrated by that.
SO I think after my return in November from Spain these problems will be solved and I can again use the sequencer as before.
One really gets afraid by these updates. So much time and frustrations
Anashwaran

elvisjohndowson · Post by **elvisjohndowson** » Sat Oct 20, 2007 11:56 pm

Just a thought, I have not yet authorized MOD-7. Could it be that the new 1.3 update is causing a freeze for those who have not authorized MOD-7? In my case, I have authorized LAC-1 only.

danatkorg · Post by **danatkorg** » Sun Oct 21, 2007 12:38 am

Anashwaran wrote: Why not admit there is some bug in this update?

I have already acknowledged some bugs, and noted that we are working on them. I will also continue to work to help people to solve problems which are *not* related to 1.3, including situations which turn out to not actually be bugs.

Since we know that one person has reported that their sequencer-related problems were solved by removing third-party RAM, and since we also know that bad or mis-matched RAM can be a source of various woes in any computer system, it makes sense to try that as a troubleshooting step.

- Dan

EJ2 · Post by **EJ2** » Sun Oct 21, 2007 1:49 am

Hi Anahswaran, Dan, and others following this thread,
Here's what I wrote over at KLF

Originally posted by Eric J Sawyer Hi Anashwaran,
As a workaround till you open up your OASYS (It's easy enough. Check near the back of the PG: page 1036 of the PDF version), try hitting the disk mode button, then go back to combi or sequencer mode and see if it works better. That's what I had to do, till I changed position of my ram sticks.

Now, Nigel and I are of the opinion that having identical ram sticks may be one of the keys to some of the strange things happening with our OASYS. I have a Kingston stick and the factory stick in reverse position to how my OASYS came. However, I'm going to purchase an identical Kingston this Monday and swap out the factory ram.

While shopping for a stick, one of the computer geeks in town here mentioned that it is not unusual for systems to prefer identical ram because, quote, "Ram can sometimes be finnicky and kick up a fuss if its partner is not from the same clan." Anyway, I'm going to do this as a test and report to Dan and Brandon from R&D.

Dan, Brandon, I'll be able to check the factory stick, take a pic and email you the details. I'll do the same with the Kingston ram as well. Following all this, I'll monitor my system closely to see if the same quirks persist.

Cheers,
Eric

PS: Anashwaran and others, when I removed one of my ram sticks and rebooted, my sequencer problems disappeared. So, when Dan asks you to try doing that, I would follow up and report the results before reinstalling or swapping positions for with your other stick.

Lou · Post by **Lou** » Sun Oct 21, 2007 2:36 am

EJ2 wrote: PS: Anashwaran and others, when I removed one of my ram sticks and rebooted, my sequencer problems disappeared. So, when Dan asks you to try doing that, I would follow up and report the results before reinstalling or swapping positions for with your other stick.

Eric,
Maybe, just maybe the removal of the ram stick isn't all that is happening here.
Quite posibly selecting to load just one of the EXs and not both is the key? After all,
removing the additional stick only allows the loading of one EXs.
So whether there is one or two sticks may not be the issue,
rather one or both EXs may be? I think the puzzle lies somewhere here.
Memory or the way both EXs are loaded. The loading process is is different now in 1.3.

I've sent you an email, please check...

EJ2 · Post by **EJ2** » Sun Oct 21, 2007 3:07 pm

Hi Lou,
Got your message. And replied. I'm not having any problems with both EXs check marked in my Global settings to load on boot up. There may be some other issues to be rooted out, but for now, I'm almost certain my problems were due to a mismatch of my two ram sticks. Either my Kingston doesn't like living beside my Factory ram or vice versa. After pulling the factory ram, leaving the Kingston in, my sequencer problems disappeared. When I pulled the Kingston and returned the factory ram, the sequencer problems were still absent. Everything worked properly.

But, as I wrote, I'm going to be installing 2 Kingston ram only under the assumption that matched pairs of ram sticks needs to be looked at as a possible recommendation. I will monitor everything very closely during my combi devlopment, audio demo work, and just plain jamming with my combis and I'll be reporting back.

Cheers,
Eric