Digital Pops & Clicks problem.... possible solutions?

[Mongoo]

I'm using a computer like an external effects box via ADAT Optical cables between my Hardware recorder and an RME 5496 PCI card. The computer is also setup to send audio of VSTi's over the adat for recording. My comp is an AMD 64 2.2 or 2.3 Ghz single core, a few years old with 1Gb of ram.

Anyways, I'm having a problem with digital pops and clicks every now and then. Usually from making sudden changes to the audio processing, most notably, A/B'ing effects or bypassing them and turning them back on to compare what the track sounds like with and without. I've tried syncing the two pieces via Adat (each as master or slave), spidf, and now word clock. I will try Adat Sync soon too, but that hasn't resolved the issue yet. Also I've tried setting different buffer sizes and that only helps so much. I can get rid of some clicks that way, but most all the time I still get clicks with A/B'ing & bypassing and also when sweeping eq with a medium gain on it, though this is not as much an issue. Some more intensive Vsti's create pops/crackling too, but I'm thinking that could be a ram issue. The noise is not a continuous noise, it's more every now and then without reason but reoccurring enough to make it on the recording within a takes length, or by doing something specific that triggers it.

The end goal out of this all is to be able to route 24 tracks to and back from the computer in real time via Adat for signal processing, plus the vsti's. More less the audio seems fine when I'm not adjusting things. Like faders aren't a problem and most eq adjustments aren't either. The quicker I change something the more likely I am to get a click or series of light digital noise. So would this mainly be a CPU power issue? The trouble with that is, I could be running one track just the same as several tracks at once, and more likely than not still get clicks/pops doing the same types of things that trigger them. I also don't notice spikes in windows XP task manager when A/B'ing. I do notice sharp spikes when adjusting faders, and hear no audible reflection to that.

I'd like to upgrade my computer to a quad core system, it's just not in the cards yet. What can I do to get rid of the distortions or minimize them as much as possible for now? Do you think this is about CPU power or Sync, or Buffering, Ram etc or some other type of issue? Thank you much.

... Another question too about sync. What is the strongest most reliable sync method out of Adat, Spdif, Wordclock, & Adat Sync for a signal to go from a hardware recorder to a computer and back to the recorder? Which should be Master/Slave?

 

[dgatwood]

You're pretty likely to get a pop turning effects on and off unless your app does near perfect latency compensation simply because the signal will shift in its timing between the processed signal....

If you're getting pops in other situations, though, perhaps some other driver is spending too much time in an interrupt handler or deferred procedure call. Try disabling other drivers one at a time and see if the problem goes away.

The most reliable clock is word clock. Sync over ADAT or S/PDIF is going to introduce more jitter due to the clock recovery process even in an ideal world.

 

[Mongoo]

Thank you Dgatwood,

So getting pops & clicks is common place and accepted then when turning on/off effects, A/Bing or turning on the "in"'s of eq's?

I just thought this would be bad for the monitor speakers, and am surprised most software hasn't done something to compensate for this yet. Even if the sequencer software didn't have perfect latency switching, I'd think they could have any switch that would cause such a change in the signal, routed to go through a process of quick but gradual volume change (a dip) to mute or lessen the audible effect, that would be potentially dangerous to monitoring equipment, not to mention extremely annoying to have to hear. As soon as the signal would resolve itself the volume could be brought back up, again so, gradually. The whole thing wouldn't take any longer than the actual Pop or distortion itself, or it shouldn't I should say.

What applications have near perfect latency compensation then? I'm using Sonar primarily on the computer, but have also tried reaper. Both show the same issues with the digital distortion pop/clicks on the On/Off/In/Out type switches for plugins. There must be some way around these noises. My internal effects on the hardware recorder for example are very smooth with out "clicks" when I turn FX on or off.

How can I check if a driver is spending too much time in an interrupt handler or deferred procedure call? Where would I turn off drivers? or are you talking about uninstalling things in Windows XP control panel's add/remove software or Device manager, or something thats actually in the sequencer software itself? These are real foreign terms to me right now, so I'm not sure quite what you mean or how to trouble shoot for that if it is the problem. Is there any testing software that can give me spec's to look for possible problem spots in my computers processing of these drivers?

Thats what I suspected with word clock. What about Adat Sync, not through the Adat fiber optic cable, but the 9 pin D-sub cable which carries no audio (that I know of)? How does word clock compare to that?

Also, in the RME DPS settings tab, it will show the status as either "No Lock", "Sync" or "Lock" for each connection cable, but the only time I've seen it say Lock is when using the RME card as Master via Wordclock, and even then it would switch back and forth from Sync to Lock to Sync to Lock again in the status. It would do this for both the word clock and Adat 1 status. If I use the Hardware recorder as Master via wordclock both the Adat 1 & Wordclcok status's in the RME DSP tab are showed to be at "Sync" and remain so entirely.

I guess my question is, what is the difference between when it says "Sync" or "Lock". In my mind I think they're probably the same thing but then again, should I be shooting for one over the other? I've never seen the RME card show a status as a solid "Lock". If ever I've seen lock it would be switching periodically between lock and sync and showing this in two of the status displays for different cables.

Thank you again for your help

 

[BushmasterM4]

duplicate post

 

[BushmasterM4]

Might be as simple as a Network card causing the problem. If this pc has an internet connection, disable the nic card, firewall and virus protection. Give it a try. When programs access a nic card the interference can be quite bad.

 

[dgatwood]

What applications have near perfect latency compensation then?

Don't know. I'm a Mac person.

Thats what I suspected with word clock. What about Adat Sync, not through the Adat fiber optic cable, but the 9 pin D-sub cable which carries no audio (that I know of)? How does word clock compare to that?

That just contains a time code stream, AFAIK. It's suitable for keeping two tape decks close to each other, but I would think it would be even harder to do precise clock recovery from that than from ADAT. I could be wrong, though.

Also, in the RME DPS settings tab, it will show the status as either "No Lock", "Sync" or "Lock" for each connection cable, but the only time I've seen it say Lock is when using the RME card as Master via Wordclock, and even then it would switch back and forth from Sync to Lock to Sync to Lock again in the status. It would do this for both the word clock and Adat 1 status. If I use the Hardware recorder as Master via wordclock both the Adat 1 & Wordclcok status's in the RME DSP tab are showed to be at "Sync" and remain so entirely.

I read one comment somewhere that suggests RME uses the terms Sync and Lock backwards, i.e. that Sync doesn't mean Sync detected and Lock doesn't mean that the hardware has successfully locked to the sync as one would expect, but that the reverse is true and it should stay steadily in the Sync state. *scratches head* Dunno. I don't use RME gear.