Jim Roseberry

+1 on the suggestion of using gear with (real) analog transformers while recording. This is part of why I like Neve preamps. The transformer gives the sound some "girth"... but in a different way from a tube (not as "soft/squishy"). You can over-do harmonic distortion from either tubes or transformers... making the mix muddy. Newer Neve models (Portico-II or Shelford Channel) have a knob that lets you dial in the desired level of transformer distortion (they call it "Silk"). The "Red" Silk setting adds harmonic distortion that enhances higher frequencies. The "Blue" Silk setting adds harmonic distortion that enhances lower mid frequencies. As with a sonic "Enhancer" or "Exciter", it's all too easy to over-do the effect. I make use of the Silk function if the track needs some extra "sparkle" on the high end... or needs filled out in the lower mids. For many tracks, I leave the Silk setting off.

That's the ProTools equivalent of your "VST Plugins" folder. You'll be fine without it... as long as you're not wanting to run ProTools.

I'd start at the source: Guitar Preamp AmpSim If you get the sound pretty close "up front", it's a whole lot easier to mix. A little high-pass filter on the bottom end (to keep from competing with the kick/bass)... and maybe a minor EQ boost/cut. If you're EQ'ing the guitar heavily at mix, the signal (up front) isn't right. A quality DI can make a huge difference in the final result using software based AmpSims. Something like the Neve Shelford channel (albeit expensive) can make a massive difference in guitar/bass recordings. I was just talking about this on The Gear Page. The Shelford channel combines a world-class Neve DI/preamp, EQ, and versatile compressor. If you've ever tracked a passive Fender bass thru a cheap DI, it sounds weak/anemic. That same passive Fender P or J bass sounds great straight off the Shelford channel's preamp. If you've ever compared playing a real amp vs. playing thru software AmpSims, the real amp responds much more smoothly to transients. The Shelford channel's compressor can be used to smooth out the transients... making the AmpSim respond/sound more like a real amp. I know Craig Anderton has written about this subject (using a compressor before AmpSim) in numerous articles. He's also the reason why Gibson Les Paul Standard HP models have the dip switch position to reduce these transients. Finally, there's the EQ section... which is perfect for tone-shaping on the way into the AmpSim. Dial up your favorite Marshall tone. Now, engage the Shelford's Mid band and give a slight boost at the 1.8k setting. Perfect for that "pushed Mid" Marshall tone. Running out and getting a world-class channel-strip isn't practical for every situation... but it's one of the few things that can make a very significant difference. As with all things recording, get the sound as close to "right" as possible... up-front (at the source).

Hi Neil, I've been using TRacks 5 plugins for a good while (in particular the 1176, LA2A, Pultec, Tape Delay, and Stealth Limiter). FWIW, I haven't encountered any CPU load issues. When choosing a CPU for DAW purposes, CPU clock-speed is the single most important factor. Having more cores is beneficial... but not at the expense of significant clock-speed. This is why Xeon CPUs (although expensive) aren't a great choice for a DAW.

That's a MIDI performance that you dragged into Cakewalk's timeline. I'd open Addictive Drums and click on the individual "kit" elements (drums/cymbals)... to make sure you're hearing them. If you can't hear them, it's almost surely a signal routing issue. ie: lf you load AD on an audio track (instead of an Instrument track), you have to put the MIDI performance on a separate MIDI track... and route that MIDI track to that particular AD instance. More complicated... but more flexible. An "Instrument track" (specifically for using virtual-instruments) combines an Audio and MIDI track. You assign AD... and drop the MIDI performance on that track.

When using plugins like Falcon (that don't use multiple cores), you may achieve better performance using multiple instances... rather than a single instance playing all parts. This would help spread the total load across cores.

There are numerous potential reasons for a single core being under heavier load than the others. Not all processes in a DAW can be multi-threaded. ie: Playing/monitoring in realtime using a 32-sample ASIO buffer size is not something that lends itself to being heavily multi-threaded. Some plugins don't use multiple cores (ie: UVI Falcon). If you've got a heavy load running in Falcon it's going to result in one core being heavily loaded. The lower the clock-speed on your CPU, the more single-core spikes will be evident.

The issue isn't Cakewalk. Sounds like you're using the onboard sound on the motherboard for your audio interface. You can lower buffering (in Cakewalk) to minimize the latency, but it's not going to be ideal. For responsive playback/recording, you want to use a dedicated audio interface that has a proper ASIO driver. A dedicated audio interface will also have significantly lower noise-floor, better A/D D/A, etc. Also, though the RTX-2070 is a great video card, it has been causing high DPC Latency. High DPC Latency can cause glitches when working at low (audio) latency settings.

Windows 10 is a fine mature DAW platform. We've got many DAW using clients running it (including professional composers/engineers/musicians). You need to be diligent with backups (should be doing that anyway)... and it certainly helps to have the Pro version (Group Policy Editor and Registry tweaks to stop all Automatic-Updates. Once reined-in, there's no issue at all with Win10 for DAW (or video) purposes. If you want to run Thunderbolt (using "PCIe via Thunderbolt" for PCIe level performance), you have to be running Win10. Microsoft doesn't support "PCIe via Thunderbolt" under Win7.

Hi Bill, Was tied up yesterday. Give me a call today... and I'll remote-in and get the situation resolved. 😉

Win10 is an excellent DAW platform (once reined-in). The Pro version is helpful in this regard... as the Group Policy Editor makes it quick/easy to disable Cortana, OneDrive, etc. With the Pro version, you can also add two Registry entries that'll stop all Automatic Updates (including notification). Win10 boots extremely fast here (not slower than Win7). Modern Z390 and X299 motherboards boot fast. Use an SSD as boot drive... and the machine boots extremely fast. When applying major updates, check to be sure DAW tweaks are maintained. Settings like Fast Startup... and power-management features (that were disabled) can be re-enabled by some updates.

We were going to stay most of Saturday... but the wife and I were exhausted after being there for 12 hours Friday. Stayed until about Noon on Saturday... and headed home. This year was more crowded than the past several. I'm guessing 15,000 people. Paul Alfery (one of my clients) was doing demos for Music Group. Was great to meet him and talk shop. Wanted to get to both of Craig Anderton's seminars... but unfortunately missed them both. Got to speak with Craig for a few minutes while we were outside perusing the Presonus tent. Even tech-savvy folks learn from Craig; always something to add to your work/methods. Sweetwater puts on a great event. This year was no exception.

I wasn't referring specifically to the 3950x... (just meant in general that's what I'd like to see from either company) 😉 Unless there's been a radical change, there's no way the 3950x will do 5GHz across all 16 cores. Thus far, Ryzen has had limited over-clocking ability. I doubt the 3950x will do full boost clock speed (4.7GHz) across all 16 cores. If it can... then we've got some serious competition. If the 3950x averages ~4GHz for each core, (to me) that's not overly exciting. I want to see the performance envelope pushed. 🤪

The opportunity to meet and talk with Craig Anderton, Roger Linn, Paul Reed Smith, etc.... is something I look forward to each year. If you decide to make the trip, hit me up Saturday and we'll meet up for a moment.

Competition between AMD/Intel is good for all involved. It'll drive performance up... and costs lower. I want to see 5GHz clock-speed across all 16+ cores... and (with proper cooling) I'd like it to run near dead-silent. 😁

Of course it depends on the software (and scenario)... but we've got numerous professional composer clients running the i9-9980xe... and those 18 cores are getting used. 😉 These users are running massive scoring templates... which can be significantly more demanding than more traditional recording/mixing scenarios. The perfect scenario is to have highest clock-speed , maximum number of cores available, and Hyper-Threading (Simultaneous Multi-Threading in AMD speak). BTW, Nice guitar in the avatar!

To elaborate a bit more... CPU core performance gain doesn't scale 1:1. IOW, Doubling the number of CPU cores doesn't double performance. Having more cores is certainly beneficial... but not at the expense of significant clock-speed. This is why Xeon CPUs are typically a significant performance hit (due to their significantly slower clock-speed). If the 3950x can run all 16 cores at max boost speed, that's when it gets exciting. At $500, Intel has a great performer in the i9-9900k. Since its release, I don't recommend socket 2066 i9... unless you're going for the 9940x or better. The exception would be hard-core composers who specifically need 128GB RAM.

1. Clock-speed 2. Cores Your disk speed will determine the number of simultaneous tracks you can run. If you're working at 44.1k or 48k, a conventional HD can sustain over 100 solid/contiguous tracks. IOW, It's probably not much of a limitation. 😉 If you're working at 192k, you'll want the extra speed of SSD. As a point of reference: Conventional HD sustains ~200MB/Sec SATA SSD sustains ~540MB/Sec M.2 Ultra SSD sustains ~3500MB/Sec With SATA SSD and M.2 Ultra SSD, we don't configure many RAID setups these days. Last RAID setup we did was for a client who was using EWSO (allowed only a single drive location for the entire library). This client needed heavy disk-streaming polyphony... so we put a pair of SATA SSDs in RAID-0. Net result was a single drive location... that sustained ~1000MB/Sec. Keep in mind that M.2 Ultra SSDs (using four PCIe lanes) weren't available at that time.

Hi Todd, I believe most of the presentations are ~1 hour long. There are so many great seminars (many simultaneous), you won't be able to catch them all. Craig Anderton's seminars are always great. You'll definitely walk out having gained some new insights. The all-star concerts are a great time.

😂 I can relate. Last year, I splurged on a Private Stock PRS. The single best guitar I've ever touched. It was quite an experience. Had them pull three of the best... and put me in a private room where I could actually hear them (impossible on the floor with all the kids banging on guitars/amps). Don Carr walked in while I was playing. I should have stopped and had him play (exceptional guitar player). 😉 Checkout on Saturday was actually quite stressful (even with heavy discount from GearFest, it was a substantial chunk). Told the wife I wouldn't be repeating that scenario this year.

Keep in mind that the $500 i9-9900k (socket 1151) also beats the i9-9980xe (socket 2066) in single-core performance. The i9-9900k can be run with all 8-cores locked at 5GHz (highest boost frequency). At that point, the 9900k is an extremely formidable DAW CPU. CPUs with higher numbers of cores (16+) typically can't run all cores locked at the highest boost frequency (AMD and Intel). The base clock-speed of the 3950x is 3.5GHz. the base clock-speed of the 9900k is 3.6GHz. The base clock-speed of the 9980xe is 3GHz. If the 3950x can't run more than one core at 4.7GHz, we'll pretty much see similar DAW performance (scaled up a bit) to that of the original Ryzen. IOW, It'll be particularly good for heavily multi-threaded applications (video rendering)... but won't best the 9900k for most DAW applications. Not all processes in a DAW can be multi-threaded. This is why highest clock-speed is still the single most important factor. Things like playing/monitoring in realtime thru an AmpSim plugin using a 32-sample ASIO buffer size (extremely low round-trip latency) don't lend themselves to being heavily multi-threaded. It's great to see some real competition from AMD. We'll certainly test the 3950x. If all 16 cores can run rock-solid stable at 4.7GHz (and do so quietly), AMD has a contender. If it's essentially one or two cores at 4.7GHz... then it's much less exciting (for DAW purposes). If you're running a Thunderbolt audio interface, Thunderbolt-3 is new for X590 motherboards.

Do you have everything in the studio powered from a single outlet? If not, I'd start there... to try and avoid ground-loops.

August Zadra is a really talented singer and guitar player. Sounds great as Tommy!

The wife and I are going. It's a great time (especially the all-star concerts)... and it's pretty cool to talk to folks like Roger Linn, Craig Anderton, Paul Reed Smith, etc. Last year, Paul Reed Smith thought I was Tommy Shaw (that was awkward for a moment). 😉 My wife told Neal Schon that story... and he laughed and told her he's been telling Paul he needs new glasses.