• Welcome to the General Discussion forum for UAD users!

    Please note that this forum is user-run, although we're thrilled to have so much contribution from Drew, Will, and other UA folks!

    Feel free to discuss both UAD and non-UAD related subjects!

    1) Please do not post technical issues here. Please use our UAD Support Forums instead.

    2) Please do not post complaints here. Use the Unrest Forum instead. They have no place in the the General Discussion forum.

    Threads posted in the wrong forum will be moved, so if you don't see your thread here anymore, please look in the correct forum.

    Lastly, please be respectful.

Crackling....yeah, I know, it's been discussed....but ......

Animix

Member
Well, I'm thinking this is likely a PCI latency problem. My system is as
follows:

ASUS A8V-Deluxe
AMD 64 4200 X 2
4G Corsair CAS2.
Matrox G450
Magma
2 x RME HDSP 9652 PCI
1 x RME Multiface PCI
version 2.94(3), 12/15/2005
4 x UAD-1's (v4.1)

All PCI cards are in the Magma with the Magma host card using PCI slot #4 of the mobo which doesn't share IRQ with anything else. All UAD cards are on IRQ 18 & 19, all RME cards are on IRQ 17. These IRQ's aren't shared with any other devices.

I'm getting crackling in the audio when a use over 40% of the resources of
my UAD-1 cards (13 -20 plugins M/L depending on which are used). Increasing buffers doesn't really do much to solve the problem. Lowering the number of UAD-1 plugins eventually cures the problem. I have also read the article by Tom Freeman and Joe Bryan on the UA website regarding PCI latency so I know the UAD-1 card (and probably the Magma) is involved in this for sure.

I have loaded the DoubleDawg PCI latency utility and I have tweaked the
various latencies. Maybe it did a little good, but it certainly wasn't the
*for sure cure*,,,,,,,,but........while I'm trying to sort out the possible
issues and variables......is there any possiblilty there could also be a
contributing HDSP driver issue as well? Anyone here using RME drivers who has any insight on this before I start breaking things down to rearrange what is otherwise a rock solid system (if plugin count is carefully managed)

Thanks,
 

Animix

Member
MASSIVE Mastering said:
What program(s)?
Sorry.....guess that would help, huh?

Cubase SX v3.1.1.944.
Wavelab v5
Ivory
GPO
Halion Player
NI B4 II
BFD
Drumagog 4.03
Jamstix
Groove Agent II
 

rydan

Active Member
Not sure it helps, but try to disable all your input channels in cubase and then enble them again. This solved the exact same problem on my setup. (A8V dlx, A64 3700+, RME 9652, 2xUAD). Had to run 12ms or more, but after disable, enable, I can run 1.5ms, 20-something stereo tracks with my two cards running at 90% without a glitch...

You can also try to delete (or rather, move) everything in the cubase settings forlder, to have cubase create new, clean versions. They sometimes get corrupt, with strange fenomenon as result...
 

MASSIVE Mastering

Active Member
One common CubEndo fix is changing your driver to \"Generic Multimedia ASIO Driver\" unless you're tracking.

Works almost every time if everything else seems proper.
 

Animix

Member
MASSIVE Mastering said:
One common CubEndo fix is changing your driver to "Generic Multimedia ASIO Driver" unless you're tracking.

Works almost every time if everything else seems proper.
Hmmmm....thanks. I'll have to check this out but my mix methodology may limit this since I lightpipe 48+ tracks 1:1 out of Cubase SX over a digital matrix for panning, aux FX and summing in Paris. I'm not in the studio right now, I don't know if the generic ASIO multimedia driver which is available to the RME HDSP hardware allows access to each individual ADAT output for routing elsewhere. I will check this out later today. If it's possible, I can just set this up as a mix template and my problem could be solved.

Certainly worth a try.

Appreciated,
 

Felix

Member
mh dunno but for me its sounds like some kind of

X-Treme OverKiLL!!!
 

Tungsten

Member
Animix,

I'm having almost the exact same problem. I have a Magma w/ 4 UAD-1 cards & 2 RME cards. I'm running Nuendo 3.2.

Some of the finer points:

- All cards are in the Magma using PCI slot 4 from the computer
- 2 RME cards are on IRQ 19 (not shared)
- 2 UADs are on IRQ 16 (shared w/ video card)
- 2 UADs are on IRQ 17 (not shared)
- Running latest drivers for everything
- I've also tried setting the PCI latency for all devices, it seems to make no difference
- If I'm running @ Buffer 256 (6 ms) most of the crackles go away, all other buffer settings are unusable
- the song I'm testing shows a usage of about 21% on the UAD meter
- If I disable the UAD plugins (using the UAD meter) the crackles seem to disappear at all buffer settings (except 3ms and below)
- Every once in a while I get an aweful sound that only goes away by resetting the RME cards
- I was also getting lock-ups (& horrible squelching sounds) when I had the RMEs inside the computer instead of in the MAGMA

* one other problem I'm having (possibly completely unrelated but) is this particular project that I'm using for testing has a Roland CE-1 plugin on a mono channel. When I open the project I get an aweful squelching sound. If I move the plugin to a stereo channel, save, and re-open I don't get the sound.

Sorry to hi-jack your thread I just thought the more info we get on this the easier it might be to solve the problem...I'm almost at wits end with this, I've spent well over a month trying to get to the bottom of it with no luck. I've even gone as far as ordering another Asus A8V Deluxe to try (even though I'm almost positive this isn't the problem)


Thanks,
Tungsten


PC Specs:
=======
Asus A8V-Deluxe (rev.2)
AMD X2 4400+
2GB Corsair XMS Ram
(4) UAD-1 Cards [in MAGMA Chassis] (using 4.1 drivers)
(2) RME Digiface Cards
Nuendo 3.2
 

Tungsten

Member
Animix,

A couple other things I noticed last night:

- Disabling the option \"Multi Processing\" in Nuendo seems to make the problem go away
- Moving the Magma Host card to other slots (I tried all them) does not fix the problem (it doesn't even seem to make it worse)

Also, I was wondering if you are running SATA or PATA drives and which controller you're using if you are using SATA drives? I am using the onboard VIA SATA controller with 2 SATA drives. I've read that the SATA drives can be problematic so I'm thinking about trying a PATA drive to see if it fixes the problem.


Thanks for listening,
Tungsten
 

jcat

Active Member
Re: Crackling....yeah, I know, it's been discussed....but ..

Animixproductions said:
Well, I'm thinking this is likely a PCI latency problem. My system is as
follows:

ASUS A8V-Deluxe
AMD 64 4200 X 2
4G Corsair CAS2.
Matrox G450
Magma
2 x RME HDSP 9652 PCI
1 x RME Multiface PCI
version 2.94(3), 12/15/2005
4 x UAD-1's (v4.1)

All PCI cards are in the Magma with the Magma host card using PCI slot #4 of the mobo which doesn't share IRQ with anything else. All UAD cards are on IRQ 18 & 19, all RME cards are on IRQ 17. These IRQ's aren't shared with any other devices.

I'm getting crackling in the audio when a use over 40% of the resources of
my UAD-1 cards (13 -20 plugins M/L depending on which are used). Increasing buffers doesn't really do much to solve the problem. Lowering the number of UAD-1 plugins eventually cures the problem. I have also read the article by Tom Freeman and Joe Bryan on the UA website regarding PCI latency so I know the UAD-1 card (and probably the Magma) is involved in this for sure.

I have loaded the DoubleDawg PCI latency utility and I have tweaked the
various latencies. Maybe it did a little good, but it certainly wasn't the
*for sure cure*,,,,,,,,but........while I'm trying to sort out the possible
issues and variables......is there any possiblilty there could also be a
contributing HDSP driver issue as well? Anyone here using RME drivers who has any insight on this before I start breaking things down to rearrange what is otherwise a rock solid system (if plugin count is carefully managed)

Thanks,

The Matrox G450 is not properly AGP 2.0 complient, and has been known to cause lots of problems. I had a crackling system which was all caused by my G450. But I only discovered it after many days of trial and error :cry:

Try swapping iot for another one (if you have one) and see what happens.

:)


Cheers,

jcat
 

Jake68

Member
Sincerely,

You must ditch that generation of Matrox card.

But also reading about your light pipe shenanigans, is it possible there is a clock issue somewhere?
 

Jeraz

Active Member
If all else fails, you might try COMPLETELY UNINSTALLING 4.1 and rolling back to 3.9. It worked for me. :roll:
 

rydan

Active Member
After getting a bit desperate, I've tried most of everything. On a whim, i enabled the \"AMD-8131\" mode (I have a A8V deluxe, so it shouldn't really be enabled...), and my troubles seem to be gone. Haven't tested it for a long time yet, so it might be a bit early to promise anything, but seems promising...
 

rydan

Active Member
And of course, that didn't work either. Now, everything runs without crackles and pops, but I get a complete system freeze every once in a while. Not usable. This is actually getting to the point where I consider either throwing the computer out and buy another one (even though i don't know what to buy, I upgraded to get the recommended A8V to get rid of this exact kind of troubles) or throwing the UAD:s out and trying to get native plugs with similar performance for similar price (hard...)

This really sucks!!!!
 

KAyo

Member
I had problems with crackles and pops etc.. I did everthing asked, nothing cahnged.

Then, Brett from UAD helped me with some info:

Switch off one of your cards via the UAD meter, and boom!!! All crackle and pops etc... gone.

But, that left me with only one UAD card being used, the other just sits there, doing nothing. What a shame, especially when you paid for two.

Anyway, I've left it like that for now. Until I decide to dive back in, hopefully with a bulging wallet. That always seems to help, unfortunately.

Regards,
KAyo
 

rydan

Active Member
I actually think this might be due to high traffic on the PCI bus. Running 26 input channels from my RME, together with plugins to two UAD cards, there is some stress on the PCI bus. I have noticed that if i run like 22 or more plugins with my input channels to cubase active, i get crackles at 6ms latency, but fewer plugs than that, I'm usually fine. Also, if i run like 22 or more \"low dsp\" plugs, i still get crackles even though the dsp load is low, but if i run the card hard with fewer hi dsp plugs, like say 9 pultecs + 9 1176, no crackles, so it actualy seems to be channel number dependent, at least in my case.
 

rydan

Active Member
Actually, AMD-8131mode helps for me! (the freezes turned out to be due to a bad IDE cable. Took me a while to figure that one out... =/)

I have a A8V, so no AMD-8131 in sight really, but as I understand it, that mode moves control af the data flow to the UAD:s from the bussmaster controller to the processor. In my case that means a verry small increase in CPU load, and actually a decrease in UAD DSP load.

An interresting experiment. If you have two cards, try creating a project with say 50 or so EX-1. Run it with one card disabled, you probably get something like 80% UAD load. Now turn on both cards, do the DSP load drop to 40%? Mine doesn't, it ends up more like 60%, unless I run in AMD-mode... Strange...
 

BENBLEASE

New Member
Hi Guys.

Think I've spotted a pattern:

Most of us seem to be running UAD + nuendo/cubase with an RME (mine's a fireface)

I got it to stop doing it when i tried using my pod xt as a usb soundcard.

Could it be the RME?

Whats going on?!!!
 

Animix

Member
Thanks for all of the replies and suggestions.


I've been in the middle of a big mix and haven't revisited this for a while.

The thought about it being the Matrox graphics card is interesting. I was actually using two of these cards (PCI and AGP) in my older Cubase SX host on an ASUS A7V8x mobo with no crackling whatsoever. I don't know if the A7V8x is more compatible with the older AGP cards but it's certainly worth looking into.

The fact that I'm using 3 x RME PCI cards on the PCI bus along with all of the other stuff rather than a single MADI could also be an issue I guess. Lots of devices on the PCI bus, though the RME cards are all sharing one interrupt. If I was streaming 64 tracks from a MADI, would this be less PCI load?

When I'm mixing, I'm really not using RME inputs, except for external processors so I'm going to just ditch these from the mix template and see if it helps.

I just ressurected my old ASUS A7V8x rig and bought two Intel 10/100/1000 NIC cards and a Lynksys Gigabyt switch. I'm going to make it into an FXT server. Since I have no pops and crackles whatsoever when running 2 x RME cards, even at 99% DSP load, I'm going to offload two of the cards to the old A7V rig and try using FX Teleport. From what I've read here and on the FXT forum, the PDC in SX should work OK except perhaps on the busses, but I will still have 2 x UAD-1 cards in the SX host machine to cover the busses and another benefit to this might be that I can add two more UAD-1 cards to the FXT server and therefore have 6 x UAD-1 cards running during a mix.

One can never have too many Pultecs :lol:
 

rydan

Active Member
Hi again!

After trying to analyze the problem a bit further, I have found my system to behave a bit strange when using more than one card (regardless of wether I run my RME or the onboard audiocard on the mobo (haven't had time to test with another \"real\" audio card yet)). Apart from crackles and pops, which doesn't appear when running one card only, i get strange UAD dsp loads when using two cards.

A little benchmark test:

http://www.cutpaste.org/temp/LoadTest.zip

Cubase projcet using 20 instances of EX-1 (stereo, 48kHz, 16bit). When I run that project on one card, I get 98% dsp load. When I run it on two cards, I don't get the expected 49%, but 63% instead, which I find odd. A frind of mine, on a P4 computer, gets 50% when using two cards...

What do you guys get?
 
UAD Bundle Month
Top