Michael McBroom

When I bring up the "Sound, video, and game controllers" tab in Device Manager, it shows 4 instances of a NVidia driver (0 through 3 Internal High Definition Audio Bus), plus one Virtual Audio Device (Wave Extensible) (WDM). So, which do I disable? One of the above -- and if so, which one? -- or all five?

With this same file open, the performance bars are bouncing around more than I typically see, but now they're not showing me anything. When I click on the Play transport button, the distortion has gotten so bad now that it won't play for more than about a second before and Audio Engine Dropout occurs. But before it does, I'm seeing the bars bouncing up to about mid-height, which is a lot higher than I see when CW is running right. With the Windows Resource Monitor Performance Tab, if I continue to click on Play to ignore the dropouts, what I'm seeing is the resource graphs are indicating about 50% CPU tasking. When I'm not doing this the usage is close to zero. No Cubase. Never had it, never will.

Yep, set to high performance, 100%. CPU speed is 3.4 GHz. It's a quad-core AMD processor, getting rather long in the tooth now. 16GB of RAM. Here's the thing, though -- I'm also running CWbBL on an old 2.8GHz laptop with 8GB of RAM and it runs it flawlessly. No hiccups of any sort. I just tried it again with buffers set to 256. Didn't make any difference.

Thanks for your response, David. In one example file with MIDI content that I'll use for this note, I'm using TTS-1 as the only instrument set. I've found that TTS-1 has a very low level of CPU demand on my system. The Cakewalk performance module rarely shows spikes above 50%. Watching it just now, I'm getting an occasional spike or two up to about 50%, but with plenty of audio dropouts. ThreadSchedulingModel in my config file is set to 1. I'm assuming that if I bump it up to 2, I'll have even more problems? I've used only ASIO with the UMC. Until yesterday, Behringer's ASIO drivers have worked very well. My Playback and Record I/O buffer settings have always been at 128. I'll give 256 a try though. I just d/l'd and ran LatencyMon. It reports: "One or more DPC routines that belong to a driver running in your system appear to be executing for too long. One problem may be related to power management, disable CPU throttling settings in Control Panel and BIOS setup." I don't recall seeing a "throttling" setting either in CP or BIOS, but I'll look for it. In the bar graph below this message pertaining to DPC, it shows my NVIDIA driver as being the culprit, with a run time of 2786 microseconds. This driver is for a graphics card I had to add into my system because I bought a monitor that had HDMI ports only and my machine didn't have any, so I bought a card with HDMI ports. And it has an NVIDIA chip. Since this chip also handles audio for the HDMI video I've wondered if a conflict might occur here, but I don't know what I can do about it other than get another card. Which, incidentally, I probably will be doing in the very near future because I think that this relatively new video card is starting to crap out on me. Another bar graph displayed that I'm somewhat concerned over is the "Highest measured interrupt to process latency" one. It's pegged all the way to the far right, deep red, reading 31657 microseconds. I take it that isn't good either. On the other hand, the "Highest reported ISR routine execution time" is "only" 238 microseconds (DirectX Graphics Kernel) The top graph, "Current measured interrupt to process latency," is bouncing around between a low of about 26 microseconds to as high as 1200 microseconds, averaging, I'd say, about 200 microseconds.

This problem I'm having appears to be the way my machine is choosing to handle MIDI files in Cakewalk. I need to give you a bit of background so you can have a better understanding of what's going on. Periodically, my machine has a problem handling large Cakewalk files that have lots of MIDI tracks. I get distortion and audio dropouts, for example. But there are other occasions where this doesn't occur, even when working on the same exact file. Now, audio files in CW are a different matter -- it handles them without issues. It's always been a mystery to me why my machine behaves this way, but it does. Incidentally, I also use Band in a Box with my compositions, and BiaB handles MIDI files without problems. In fact, BiaB is just the opposite. It tends to choke up if I load files that have a lot of audio tracks. I suppose I should point out that these problems are restricted to audio playback. When I record any of these tunes, they always record without any hiccups. Recently, my machine has been very well behaved. It has handled large files with lots of MIDI tracks without any problems at all. I don't know why sometimes it's like this and other times, there is much distortion, pops, clicks, and what have you, when nothing has changed. But anyway, I was enjoying a clean sounding machine until yesterday morning when we had a power outage that knocked my machine out. Power outages around here are not all that uncommon, so I'm used to this happening, and they usually don't affect my machine's performance. But yesterday morning, something changed. When I booted CW and tried to play a file with 12 MIDI tracks, the distortion was the worst I've ever heard, and it couldn't play for more that a couple of seconds before audio dropouts occurred. I tried restarting the machine, it didn't make a difference. I even shut down the machine and then booted it back up, still didn't make a difference. I haven't yet tried shutting it down and switching off the power supply. I'll do that next, but I won't be surprised if it doesn't make a difference. So I'm wondering now if somehow something got corrupted within CW? It was not resident when the power outage occurred. In fact, no programs were up -- just the OS, which is (still) Win7. I've run the Windows Task Manager to see if the CPU loading is uncharacteristically high -- it isn't, really. It's about the same amount of loading I see when I'm streaming a Netflix video, for example, which my machine handles easily. Also, if I load a file into CW that is audio only, it plays through it without any issues at all. So for whatever reason, MIDI has become a big problem right now, and I sure would like to understand why and figure out what to do about it. Oh, I should mention my hardware. I'm running a Behringer UMC404-HD audio interface, which has worked flawlessly with my system, so I don't suspect it. Before I had the Behringer, I was using a M-Audio Delta 66 PCI card, and I was having similar problems with it, so I really don't think it was this hardware.

After putting that one tune to bed -- finally -- and reflecting on it for a while, I thought I should add one more comment to this thread. First of all, let me explain that I'm just a Cakewalk user, I don't claim to be an expert. Heck, I don't even claim to be very proficient at it. I just know enough about this DAW to get done what I want to get done and I continue to find it to be superior to any other I've tried. Okay enough of that. The problem I'm still seeing with this one file in particular is it seems that the MIDI system -- if you will -- is somehow broken. I've noticed that, if I try to change patches on any other track, it does the same thing -- reverts that track to Piano 1 and then I have to manually reassign it using the TTS-1 UI. So it isn't just that one track. It appears to be all of them. Or I should probably say, all MIDI channels. It's worth mentioning that this problem is affecting only that one file right now.

No, I'm attempting to change instruments from the Tracks window in Cakewalk, same as I always do. This tune has six MIDI tracks, one instance of TTS-1 and one instance of another soft synth. Obviously I took care to make sure each track was assigned a unique channel in TTS-1. TTS-1 is handling five of the instrument assignments, while the other synth is handling the sixth. I went ahead and booted up the TTS-1 UI (normally I keep it minimized). Sure enough, even though I was indicating the correct patch in CW, TTS-1 was showing Piano 1. So I changed it in the UI and now things are back where I want them. So thanks for the tip. Glad I could put this behind me.

This problem crops up occasionally and usually resolves itself -- for some unknown reason -- but this time it isn't. In fact, the problem has begun to infect adjacent tracks. So I'm turning to the forumind for hopeful answers. Here's the current situation. I have a MIDI track configured with a TTS-1 patch and I decided to try another. But when I did, it loaded Piano 1 and won't unstick from Piano 1. Now, this is the sound only. The patch I selected appears in its little window; it's just I'm not getting its sound, I'm getting Piano 1's sound. I checked to make sure I hadn't changed the MIDI channel or the bank (I hadn't) and even called up the Event List to see if something had mysteriously appeared there. It hadn't. In fact, no patch notices were loaded at all in the Event List for that MIDI channel. Then I noticed that in adjacent track (with a different channel number of course) I had selected Piano 1, so I tried to see if I could change it off Piano 1. Initially it took the change, then it reverted back and now it is also stuck on Piano 1. This is less of a problem for me because Piano 1 is the patch I want for that track, but it bothers me that it's stuck. It also bothers me that this appears to be some sort of creeping condition. I don't know what else to try at this point. Maybe remove and re-install TTS-1? I have a lot of tweaks set in that synth so far, so I'd really like to avoid removing it, if at all possible. What about installing a second copy of TTS-1 and transferring all the tracks over? Hrm, that should work, but would it do any good?

Poor choice of words on my part. I didn't mean "daisychain," in that comment. I just meant basically to port it into my system, which would require the use of a MIDI to USB adapter, I reckon. I don't think I've ever used MIDI "thru" before, so I had to check my devices. I have one -- an old Roland JV-1010 synth module. It has a thru port. Currently I don't have it hooked up to my system, but I guess it's worth a shot now, especially since I can daisychain other synths from it.

Actually, no. I was talking about legacy MIDI synths that have the old 5-pin DIN ports only. I own a few -- four, I think. Not including the Soundblaster Live! cards I own that have MIDI/game ports. These old legacy synths I daisychained to my DAW because it had only one MIDI port -- the game/MIDI port of the Soundblaster card.

Nope, it's toggled on.

I dunno, maybe. Shoot, I don't even know what most of that stuff is. I have a couple of MidiSport 2x2s and a Roland UM-One, plus my audio interface has MIDI ports, so that gives me six MIDI to USB ports. More than enough for my needs.

Way back when, I was a Pro Audio user, and my DAW had two sound cards, one of which was audio only, but the other had MIDI capabilities via its Game port. Back then I had daisychained at least three different MIDI devices together off that game port, and things worked well. These days, it's all USB. And when you plug the USB port into your system, it responds with the name of the USB host or adapter, not the MIDI instrument that's plugged into the USB adapter (or host). So in Cakewalk, that's where the Instrument Definition files become useful. But I've wondered if I could still daisychain instruments -- although I don't see how I can, since I have to assign a single instrument to that adapter, or at least that's the way it seems to me. So, since I have no MIDI ports in my DAW, I guess that means the days of daisychaining are over, yes? My DAW has PCI slots and I still have a couple of PCI sound cards that have the MIDI game ports. Only problem is nobody's written drivers for them for Win7 or later. So they're of no use. And I've looked for drivers, with no luck. Recently, I've been thinking about building an obsolete system from leftover motherboards and RAM and hard drives I have laying around -- one that'll run Win98 or Win XP -- one with an OS for which these cards still have drivers. And then daisychain this system into a MIDI/USB adapter or port on my current DAW. I wonder how that would work out. Probably wouldn't, I guess. It would still show up as a single USB entry, wouldn't it. I have plenty of USB ports in my current DAW to handle the few legacy pieces of gear I have, so daisychaining isn't a real issue. The question is mostly from idle curiosity. However, I still might build an obsolete system just because of one of my old sound cards -- a Soundblaster Live! card that has a synth chip on board. Some of those instruments on that synth chip really had a cool sound and I wouldn't mind being able to access them again.

Got another troubleshooting question for you guys. After a few days absence, I decide to fire up the GR-33 again and try to lay down a track or two. Nothing's changed on the GR-33 and nothing's changed with the piece of music in which I've already set up a track for the GR-33. I switch it on and I'm getting GR-33 audio through my speakers and headphones -- this because I'm running the GR-33's audio outs into the audio interface. Then I set it to record and hit the big record button in the transport area --- and nothin. I hear the audio fine, but it isn't laying down any MIDI. I've got the track record button set and I'm clicking on the main transport's record button. Last time I did this a few days ago, it was recording MIDI just fine. The audio interface's channel volumes are set to 12 o'clock and Line is selected (Inst distorts way too much). Mains out are also set to 12 o'clock. Something I just noticed. My UMC404HD audio interface has two LEDs associated with MIDI -- one sez "IN" the other sez "OUT." The "IN" LED blinks just about constantly, even when I'm not playing. The "OUT" LED isn't blinking at all, even when I am playing and trying to record. So, now I'm wondering if this might be a problem internal to the GR-33. What do you think?

Thanks, guys. I'm on it.

Several days ago, I ran across a YouTube video where the presenter was discussing the merits of applying a gain control prior to the controls in the track -- I think I'm recalling this correctly. Much of what he said only partially sank in, and he was discussing a control in Pro Tools, so I had to hunt around a bit to discover the same control -- I think at least -- in Cakewalk. You know, when you're in the Console view, the Gain control at the top of the track? Cakewalk calls this Input Gain, and sez it's a pre-fader input gain for fine-tuning a track's volume. MIDI channel strips have a similar control, which is called a Velocity Offset, but I won't be discussing it here. Anyway, this Input Gain amounts to this pre-gain this guy is discussing, right? Sounds like it to me, at any rate. He was suggesting that a much better way to control the volume of a track was to use this "pre-gain" control than using the track's gain. I don't recall what he called this control, so "pre-gain" is my term, not his. "Pre-fader" is Cakewalk's term, so whatever. So anyway, I'm pretty hazy on this topic, and I didn't subscribe to this guy's channel, so I'd probably have a tough time finding this video again. So instead, I'm turning to the Forumind here, hoping that one or more of you can clearly explain the reasoning behind using a pre-fader input gain, when to use it, its benefits, etc. I'm always looking for ways to improve my mix, and if a better understanding of how this control works will help in that regard, then I will be most appreciative of your response. I should add that I have selected one tune of mine with several audio tracks, and have attempted to balance them using this pre-fader instead of the tracks' main volume faders. After playing around with this tune a bit more, I guess I'd have to say at this point that it is simply just what Cakewalk sez it is -- a way to fine tune a track's volume.

Yeah, the UMC404HD has a choice of Line or Inst. I selected Line. Now that I'm thinking about it, though . . . Just looking at it, I see the GR-33 labels its outs as Mix Out. I've always assumed these outs were Line, but who knows, maybe Instruments? I guess I'll check the next time I fire up my rig. There is also a (mono) Guitar Out, but I'm pretty sure that's just a clean guitar signal, without the synth. As I dimly recall, I ran the Guitar Out to my guitar amp.

Okay, well I don't know why that Roland interface wouldn't work, but plugging the GR-33 into my audio interface DID work. Yay! I'm running a Line Out cable from the GR-33 to my audio interface so that the audio can be recorded. The only problem I'm running into now is my audio interface is showing the GR-33 clipping even at low to moderate volumes, plus its audio is distorting in the playback. This might be because I have both the GR-33 and the pickup's volume turned up to maximum. But if I turn these levels down, it becomes very quiet. Okay, so anyway, that's an audio problem, not a MIDI one. I'll get it figured out -- soon I hope.

Thanks for the responses, guys. You know, that's one thing I didn't think of trying -- the MIDI ports on my audio interface. I might as well give them a try and see if I might have any better luck. I'll try the rest of your advice as well, see if it makes a difference. And I'll report back.

Back when I used Pro Audio, I was able to get my GR-33 to work fine. Yes, it's been just a few years, and I've forgotten just about everything I had going on back then, but I have an occasion where I'd really like to use it again. So, time to dust it off and see if I can get it going. Back then I was able to interface my MIDI gear to my DAW directly with the round MIDI plugs. Now I have to use an adapter that converts the MIDI plugs to USB. At the advice of another GR-33 user, I bought a Roland UM-ONE MIDI to USB interface. Prior, I'd been trying to get it to work using a Midiman MidiSport 2x2 USB interface. Didn't have any luck with it. Too bad. I have two of them, and it would be nice if I could use them for some of my other old MIDI gear. Anyway, the problem I'm having is I'm not getting any MIDI data coming through. I installed the UM-ONE driver, which happened when I plugged it into my computer. No issues. I already have an instrument definition file for the GR-33, so I assigned it to the UM-ONE in Cakewalk's "Assign Instrument Definitions" screen. In the MIDI channel I set up for the GR-33, I've indicated the UM-ONE in both the In and Out sections at the bottom of the track, I selected a channel, and so with the instruments, I'm able to select whatever I want from the available GR-33 instruments. The pickup mounted to my guitar has a volume control and a selector switch. I have the switch selected to Synth and I turned the volume all the way up. But when I click on the track's record button, what I expect to find is some activity with the track's VU lights, but there's nothing. If I click on Record to get Cakewalk rolling, there are still no VU lights and the track is not laying down anything. The GR-33 has a headphones jack. So I plugged my 'phones into the jack and, yep, I'm getting synth sounds sounding just the way I expect them to sound. So there's something screwy with the output of the Synth or the input of Cakewalk. One thing I'm wondering about is -- do I need to have the GR-33 showing up in the Synth Rack? Because I don't have anything there and I don't even recall any sort of dll for it, although I can look, I suppose. Or even look at Roland.com. I've only used the Synth Rack for plug-ins. So any GR-33 users out there? What have I missed?

Another UMC404HD user here. I've owned it for quite a while but I've just started using it on a regular basis after my great old M-Audio Delta 66 finally tanked. It will be missed. I haven't had any issues such as you describe, but I have it set for 44.1k mostly because I use Band in a Box a lot and it will only do 44.1k. I agree, it's a great interface for the money. I researched audio interfaces extensively before I bought it and I felt quite confident -- still do -- that it's the best audio interface out there for the money. Only things I've noticed about it so far are it doesn't seem to have the same amount of gain capacity as my old Delta before noticeable distortion sets in.

Heard that! That's usually been my approach.

Nope, I'm using CW's TTS-1 for GM, and just its default setup protocols. The audio trace is there before I do any bounce to combine tracks. And yes, since it's actually a chord that's bleeding through, I suspected it might be a MIDI track, so I checked the most likely culprits -- MIDI guitar and keyboard tracks. Nope, nothing there. It was definitely in the audio. I ran into this again with the same piece of music, but in a different spot with a different guitar. This guitar is a Gibson with P90 single coil pickups that can be rather noisy. There's a four bar intro to the piece, before the guitar part starts and, during playback, I couldn't help but hear a rather pronounced bit of line noise in the guitar's audio track while I waited for the Intro to complete. So I highlighted the noisy "dead" area before the guitar started, then I used the Edit>Cut command to remove the snipit from existence -- or so I thought. Now, the way it sits, the audio channel has no signal from the beginning of the tune to the end of the fourth measure, but if I solo that track, I still hear those four measures of line noise. Strange, eh? I got to thinking that it may actually be the Auto Save function causing this, since a save was done right before I deleted that section, and I was wondering if that, somehow, was messing with things. So I deleted the Auto Save file and tried again. Nope, line noise is still there. Not a huge issue, but an annoying one. I used the "Write" function to reduce the audio track's volume to "0" prior to the start of the 5th measure.

That's just it, though. I deleted the audio track -- or at least I thought I did. It doesn't exist anymore, far as I know. It certainly doesn't appear anywhere. I've never hidden a track before -- I'd have to consult the CW online Help just to find out how to do that. Frustrated over this whole business, I learned a lesson. I should have saved the file before I recorded the audio. But I didn't. And in fact, I'd done a lot of work on the project before recording the guitar part, to make things worse. I ended up having to rebuild the entire piece. Fortunately I had clean components of it that I was able to reassemble, but it was still a time consuming process. Took me all afternoon yesterday getting the project back to where it was before I made the recording. Lesson learned. I'm still baffled and really curious as to why this happened, though. Hopefully it doesn't repeat itself.

Thanks for the additional info, guys. Well, for starters, I did try WASAPI but it was giving me problems. The piece of music I'm working on right now is fairly large and, even though all tracks are MIDI, there are 22 of them, although only 10 and 12 instruments are playing at any one time, respectively. Nonetheless, at least about half the time, roughly halfway through the piece, it would fall into a state of total distortion. If I paused and restarted it, the distortion would clear up and usually go away for the remainder of the piece, but this was unacceptable to me, so I went back with ASIO. And yes, my interface has ASIO-specific drivers that have always worked very well. I've been running the amp's audio out into the interface. I prefer this method over a microphone because of the ambient noise level I have to deal with here. Some CPU fan noise and also a cockateel in the background who likes to chirp at unpredictable times. And occasionally the dog decides to bark. The downside to running straight from the amp to the interface is I'm picking up some line hum now. Nothing very serious, but it's there, raising the audio floor. I can probably nail the hum with a parametric EQ -- haven't tried yet. Next time I'm working on a smaller piece, I'm gonna give WASAPI a try and see if CW will recognize the amp as audio. I never did try and see if that would work, since I was running into problems with that driver mode.