Terrible nightmare session from hell

PipelineAudio wrote on 6/11/2003, 12:57 PM
Ok last nite I had a guy bring in his keyboards, same as always with this guy, then after it would be time to throw " vocals" on it

Project is at 48k 24 bit, about 3 minutes long

Audio card is RME HDSP 9652
Converters are RME ADI DS
Word Clock is Lucid Genex6

I needed to have vegas generate midi clock so that I could get all the tracks from the keys in sync. First 5 tracks, no problem, keys chase just fine. Track 6 it was ahead or behind, soi hit stop to try it again, then the loop bug would happen as I tried to play. If I hit stop ANY time that audio was playing it would just make this horrendous looping noise until I played from somewhere else.

Gave up and recorded track 6 then slid it in by hand. Interesting to note that while it started at the wrong time, all the timing was correct, so sliding it over worked

Track 7 worked fine, synced right up, so did track 8

Track 9 same as track 6

and on and on it went thru the night, some tracks fine, some tracks not, loop bug left and right

Finally we got all the keys in and started doing vocals. Everytime I would hit stop BUZZZZZZZZ theres the loop noise in the singer's headphone, and shes getting all pissed off. Card kept dropping audio clock ( tried using word clock from a lucid genex6, tried card as master, tried converters as master all same thing)

Finally had to call the session till I get to the bottom of this. The loop thing has gotten worse and worse since vegas 4.0c

Comments

JoeD wrote on 6/11/2003, 4:36 PM
Pipe, will you get your ass over to ProRec and talk to the MANY people using their RME cards? First make sure you're setup\using your shit properly. Or dump it and get a delta 1010 or Layla 24 or something. I'm starting to wonder if you're using it properly.
I'm starting to wonder if you even know what you're doing.

Session time is NOT the time to be trying out, but rather getting to work with WHAT WORKS. You're supposed to be capturing some art here - remember? That's your role here. Are they paying you for this session?

Well, You're obviously not getting any help here, I HAVE to assume you're not getting any help from RME (which is VERY strange because I hear from everybody their support is beyond the call of duty). You did speak with RME, correct?

THEN come back and talk\rant about A BUGGY V4 (smarter: switch to V3 and get work done, did everything worky with V3?).

Sorry, but all I see from you are problems left and right. I'm begginning to wonder. Do you seriously run a studio now?

and...why 24\48?
why not 24\44.1? Look into it.

JoeD
PipelineAudio wrote on 6/11/2003, 5:44 PM
If this always happened of course I wouldnt be running it this way, thats the wierd part about it. It is very random

Ill take your advice and head over to prorec with this tho

yeah Ive talked to RME, but their tech support is some of the worst in the business. If its an answer they know or have a fix for they are great, if not you are screwed

I cant imagine any way of running this stuff that would be significantly different than I do. Which areas do you think I could possibly have wrong?

One setup which works fine 80% of the time is Lucid Genex 6 as master word clock, then all devies in external ( three RME ADI-DS's and a RME HDSP 9652 ), Vegas in ASIO or MME set to the projects correct sample and bit rate. Last night this setup decided not to work. Period

where am I wrong in this ?
kilroy wrote on 6/11/2003, 8:08 PM

I can think of a few things you might look into based on an experience we had.

I am assuming that your word clock generator has distribution network capabilities, i.e. it has multiple outputs, one for every slave in your system. If so that is good...but imagine what would happen if one of those outputs was periodically flakey. Or something else in that clock network decides to breakdown. Really screws with audio sync big time, and you can get some absolutely ghastly noises along with it. We certainly did, and it was waaaay more spectacular than those annoying little clicks you get from poorly clocked systems.

We *finally* traced our troubles to a baulky power supply in the clock generator, (a very, very good one incidently) but boy, was it frustrating trying to figure it out cause it didn't always manifest itself...well...except in the company of high profile clients, naturally.

Sync cabling all good and...um...expensive?

You might also check to see if something is abnormally hot. Some DACs get so hot you can cooks eggs on them. Do your problems start after the system has been up for a while or does happen at any old time? Does it ever happen when there is nothing clocking anything from anywhere?

I don't really get this loop bug thing you describe but the fact that your losing word sync makes me inclined to lay the blame on some component in that system. If that were the case then removing/swapping/replacing each component one at a time might be another route to take.

The thing with any audio sync network (especially really complex ones) is that the slightest borderline behavior from any little component can bring your whole studio grovelling to its knees. It's what I hate about digital. It can be absolutely torturous to bug shoot.

PipelineAudio wrote on 6/11/2003, 9:23 PM
cool Ill look into the stuff you wrote and think about it..I can say this though, it also happens when the card is master and nothing but the card's spdif is being used

same exact deal
PipelineAudio wrote on 6/11/2003, 10:14 PM
"You might also check to see if something is abnormally hot. "

I will leave the stuff on tonite and be looking for this

" Do your problems start after the system has been up for a while or does happen at any old time? "

It can be right from the start or after a while or not at all

"Does it ever happen when there is nothing clocking anything from anywhere?"

not sure, but I know that the clock source doesnt matter, it can happen even internal

" don't really get this loop bug thing you describe but the fact that your losing word sync makes me inclined to lay the blame on some component in that system."

The loop bug happens to enough people to know its not just my system

" If that were the case then removing/swapping/replacing each component one at a time might be another route to take.
"

every component has been changed at least three times. One thing tho, I pretty much build the RME Reference PC each time, and RME themselves will NOT shed any info on how it is setup, even though they tend to steer you towards that setup... SO it very well could be the specific type of components

JoeD wrote on 6/11/2003, 11:53 PM
I don't know...but try some others at ProRec as I've mentioned. I'm tired of seeing you in pain with this. It's got to be TOTALLY pissing you off. Time to solve it, or move to another interface.

Ya know...
I like Kilroys post (pwr supply in clock gen) and you really should look into his past findings. Why exactly do you need the Genex (my suspect in your chain I would guess btw) though? Seems like there might be a better route to accomplish all this.

but...I think I saw you state you've exchanged all parts already (correct?)

You mentioned your system as a possible culprit, please elaborate (describe the system - everything. If you posted this before I missed the system info).

<<<Finally had to call the session till I get to the bottom of this. The loop thing has gotten worse and worse since vegas 4.0c >>>

Well, eliminate all the hardware as the culprit, then it may be time to try vegas V3 (not v4).


JoeD

PipelineAudio wrote on 6/12/2003, 5:45 AM
" I like Kilroys post (pwr supply in clock gen) and you really should look into his past findings. Why exactly do you need the Genex (my suspect in your chain I would guess btw) though? Seems like there might be a better route to accomplish all this."

Right, I am interested in that as well, however the same thing happens even when it is clocked internal

"but...I think I saw you state you've exchanged all parts already (correct?)"

I have, and built the exact " RME reference PC" three times. I also tried the Intel board and chip that one of the SF guys recomended. THAT was awesome, same problems but everything was faster. I switched back to the RME reference PC as RME implied that would be the only way they could help

"You mentioned your system as a possible culprit, please elaborate (describe the system - everything. If you posted this before I missed the system info)."

Ok right now I stripped a few things down to eliminate suspects:
ASUS P4 T-E
Intel P4 2.4ghz ( 400mhz, not the hyperthreading model )
512 Megs Samsung RAM ( also tried Corsair)
Matrox G550 Dual Head
RME HDSP 9652
3XRME ADI DS ( though most of these problems will happen with these hooked up as well
Seagate Barracuda System drive
Western Digital WDJB 160 audio drive
TEAC CD Burner
WinXP Pro
UA UAD-1 card ( though taking it out doesnt help )

Now I guess its time to remind everyone of some history here. time and time again, I have had one problem or another, first SF blames me for it, then an audio card manufacturer blames SF, and the buck ges passing all over and later it turns out it WAS the fault of the hardware or software.

This carried into the Vegas 4 beta. After the internal beta testers deemed it time to turn it to the public beta, within HOURS I posted quite a few BIG stop shipping type problems. Right here in this forum for a few weeks we watched SF blame me, and then later it turns out it was a fault in vegas itself. More than once, more than twice, quite a few times. Some of the problems we are still dealing with now, I mentioned first, and mentioned them in the public beta days. Im just an unlucky person that happens to find bugs FAST. I dont know why but thats the truth

Everyone can hemm and haw all they want but I feel with this track record, I should be given the benefit of the doubt this time that MAYBE its not my fault? The problems I had this session were very similar to the other problems we all know about right now. Im not the only one who reported the loop bug ( though most liely I was the first), and the edit glitches are also well known. It was the INTENSITY of the known glitches that made this session go crazy.

About the MIDI bugs, I had tried and tested the day before with a Yamaha Motif with no problems for a few hours. It wasnt until the customer brought in his Triton and after about 5 tracks where the problem really showed up

Mathius from RME is trying to help out right now as well. He is aware from a week or so ago, when I asked him to check if the " flush ASIO buffers " bug was back in the newer drivers and he confirmed they were. This COULD be part of the problem, so I am going to try one driver back and see what happens.

The problem is these problems are totally random. I can be fine for days on end
kilroy wrote on 6/12/2003, 10:08 AM

Ok, this whole deal has me intrigued now. And I certainly do not rule out the whole application bug thing, Pipeline. Honestly, I am the first person in here to blame the app when something messes. For the life of me, it is difficult to slam a piece of pro hardware that you have just dropped really serious money on when the margin for screwups seems so much more in favour of the app itself.

Anyway...

Got another sound card kicking about that you can try in place of the RME?

Does that motherboard have on board sound?

Maybe try removing one stick of RAM and/or try turning ECC off in your bios if the RAM is error checking stuff. You tried different RAM so this may be a bit of a flyer. We once put an end to some pretty spastic behavior on one system by pulling one stick as opposed to just swapping the whole wad.

Modem in this computer? Not hardly likely but had to ask.

One thing I thought of. I notice you have a Teac drive. Is it a CD-512EB per chance? These are very good drives by the way, fine burn performance, although we don't like internal burners or IDE as a rule. Anyway, we had an issue where the firmware that came with the drive insisted on having auto insert notification turned on. A completely silly notion to say the least, but in one system it caused all kinds of realtime issues. The solution was the latest firmware update *and* taking the Adaptec software that came with the drive off the system (??!!?!?!) All well after that. Go figger. If it makes you feel any better the Teac guys thought we were on ludes or something, totally loopy.

I will check our "screw up logs" to see if there is anything else that someone has encountered. We have this logging system (appropriately named "SNAFU") where if something fouls up you record what the symptoms are and if you finally nail down the solution you log that under the problem for (hopefully not) future reference. That way you can compare notes, as it were, if/when one or another system boffs for some reason. It's a pretty scary read on your time off but it can be a brilliant time saver if things are really going down the pan in a hurry.

I will try to repo your problem as well, now that my curiosity is perked. We don't use Vegas in a real time recording environment though, mostly assembly editing and occasionally mixing, no internal summing, very little realtime processing. I will have to change a few things to approximate your setup, obviously won't be exact but we know its outboard DAC and EIAJ fiber-optic I/O right so I will see what I can throw together.

Refresh my memory here...this problem with all Vegas versions? Just the 4 version?
PipelineAudio wrote on 6/12/2003, 1:27 PM
Ill look into the burner. The problem is with vegas 4 all versions. The loop bug happens only in playback, NEVER in recording....I guess I should have told SF that, as its probably a good clue