Jim Roseberry

FWIW, I've never encountered a scenario where I couldn't get a Thunderbolt-3 audio interface functioning with a Thunderbolt-4 controller. The Apollo series requires certain BIOS tweaks. If these aren't in-place... the Apollo won't work and/or you'll hear glitches in audio. There's not much difference between Thunderbolt-3 and Thunderbolt-4. All current Thunderbolt audio interfaces are Thunderbolt-3. There's little difference between Win10 and Win11. Myself and many clients use Thunderbolt-3 audio interfaces under Win11 (and Win10). I used an Apollo 8 and Satellite for a while (both connected via Thunderbolt). I moved on for a couple of reasons: The Apollo's DI's imparted a "tubby" quality to the lower mids with DI guitars (using amp-sim plugins) The Apollo's round-trip latency (even though it's Thunderbolt) is about the same as a good USB-2 audio interface Basic function of the Apollo/Satellite was always solid. I wanted to push the low RTL envelope... and that's not Apollo's forte'. After the Apollo/Satellite, I migrated to using a combination of RME Fireface UFX+ and Presonus Quantum. UFX+ was used as the primary audio interface Quantum was used when I wanted super low round-trip latency I currently use an Antelope Orion Studio Synergy Core as my main DAW's audio interface (connected via Thunderbolt-4). Absolutely love it. To me, it's the best of all facets. Can achieve sub 1ms round-trip latency (with a fast/well-configured machine) Converters, DI's, Preamps sound better (to my ears) than the Apollo and UFX+ Rock-solid performance For secondary machines, I use a Presonus Quantum (connected via Thunderbolt-4) Can achieve 1ms round-trip latency (96k using a 32-sample ASIO buffer size) Converters, DI's, Preamps sound good for the cost Rock-solid performance

For ultra low latency performance, the Threadripper 7xxx series is about the same as an Intel i9. With the Ryzen 5xxx series, AMD *finally* got their ultra low latency performance together. A well spec'd AMD "Storm Peak" DAW is going to be ~$10k+ (parts alone). 7960x = $1500 7970x = $2500 7980x = $5000 7995WX = $10000 The target market is pretty small (the most extreme of professionals and enthusiasts).

I'm heading out of town until the 21st. If you need assistance, I can certainly help. One thing about Reaper, it's using a different application to bridge. Bit-Bridge is used for Cakewalk I believe JBridge is used for Reaper Some 32Bit plugins cope well with one... but not the other. In short, to make your life less complicated... avoid 32Bit plugins. 🤪

Shure recently released the SM7dB... which has an onboard preamp (like a Cloud-Lifter). This allows much healthier levels... without cranking your mic preamp (which may get noisy at higher gain settings). If you have a higher-end outboard preamp, you likely don't need a Cloud Lifter or the SM7dB. These preamps are usually much cleaner at higher gain settings. The wife is on the local Classic-Rock morning show. Sometimes the station needs a VO recorded... after she's left work. Instead of her having to drive back across town, I grabbed SM7b, RE20, RE320 mics (already had Neve preamps). For station VOs, I like to use the RE20. The tone of her voice sounds better than using the SM7b. The RE320 sounds similar to the RE20... but with more of a presence boost. The RE20 is more flat... giving the production engineer more room to EQ.

"In this case, I'm getting pops on some (but not all) of my VSTs when playing back but if I take the vst track and bounce to track, the recorded audio is fine. It's only when live playing the midi, and again only on some vsts (32 bit ones I believe)." This line (blue) from you original post... is why I mentioned "bounced-down" audio. Since the bounce (whether a single track or an entire mix) is done offline, it's not subject to realtime processing demands. Thus, high DPC Latency wouldn't cause glitches in the bounced audio... but would cause glitches when playing back the project in realtime. If you're using 32Bit plugins, those are being "bridged" to function. Some plugins cope just fine being bridged... others can be completely unstable. If this only happens with 32Bit plugins, it's most likely the culprit. When doing basic trouble-shooting, you want to eliminate variables. Latency Mon will either rule-in... or rule-out DPC Latency as a potential culprit. Definitely worth the couple minutes of time/effort.

The driver would have no idea... But here's the thing: High DPC Latency would affect realtime playback (glitches). Audio that's bounced-down (not in realtime) would thus not be affected by high DPC Latency.

I ran into that issue yesterday... testing Windows 11 23H2. Don't have the issue with Windows 11 22H2.

Sounds like a problem with DPC Latency (not to be confused with audio latency). Download and run Latency Mon https://www.resplendence.com/download/LatencyMon.exe It'll tell you the highest DPC Latency... and list the culprit. Many times, high DPC Latency is caused by a poorly written/behaved driver (monopolizes the CPU).

Glad the machine is working well for you, Max!

Another potential issue with RAM and PCIe cards: The connection area (points of contact) are quite small. If the case is moved or "flexes", it can cause those contact points to be "misaligned" just enough to cause issue. On some builds (especially trying to get a 420mm radiator in a mid-tower case), the tolerance can be *super* tight. A single millimeter can mean the difference between something working... or not. ie: RAM with LED lighting can sometimes be too large to install after radiator fans have been mounted. If the RAM is installed first, there's just enough clearance.

Given the lead-up with the bearing-edges not being flat... I thought they were going to sound terrible. The final result (recording) sounds pretty good.

If you want to get excited about an AMD CPU, checkout the Storm Peak 7960x. Cinebench R23 scores are above 50k. Two downsides: Cost: CPU is $1500, motherboard is ~$600, 128GB DDR5 Registered is ~$1200 Heat/Noise: TDP=350w

FWIW, If you benchmark the 14900k, it comes out exactly the same as the 13900ks. Intel shouldn't have called it 14th Gen... but rather "13th Gen refresh" or "13th Gen mk2". The 13900ks in most benchmarks (and running audio), handily bests the 5800X3D and 7950X3D. ie: If you run Cinebench R23 (Multi-Core)... the 13900ks scores ~40k The 5800X3D... is nowhere remotely close (less than half at ~14.5k).

Polybrute is also on sale: https://www.proaudiostar.com/arturia-polybrute-analog-polyphonic-synthesizer.html

I'm assuming it'll have the new triple-sensor key-bed. It's a nice improvement over the dual-sensor. The 88-weighted key action is one of my favorites. One of the few weighted keyboards that doesn't get my tendonitis flared up.

It would be nice to have more "traditional" presets. Many instruments (hardware and software) are heavily populated with presets for EDM, House, Dub-Step, Trance, etc. Nothing against those sounds... they're just not what I gravitate toward. I just assume that (most likely) I'll have to program what I need. Even with a more traditional hardware instrument like the Nord Stage 4, I have to program/tweak nearly everything (for live use with the band).

I'm a Rock guy... Picked up Avenger 2 earlier today. Just spent a few minutes tinkering with a stacked 4-voice SAW waveform (single oscillator)... and it was quick/easy to mimic the thick/bright stereo analog sound I use live for Turn Up The Radio, Don't Tell Me You Love Me, Let's Go Crazy, etc. Though it's nothing like a Nord Stage 4, the final result sounded remarkably similar. Edit: Though this sound isn't particularly complex, it's a whole lot easier to achieve in Avenger vs Falcon. I have Falcon 3... but rarely use it.

If you're looking to capture nuance/detail, then there is no debate. TLM is a much better choice. If you're working in a room with less than ideal acoustics... the SM7B will capture far less of the room. This is why so many YouTube videos have folks using SM7B mics. If the singer really belts... the SM7B can be the better choice.

Can you make a single M.2 SSD work? Yes. Is it ideal? Absolutely not. For logistical reasons, you don't want large sample-libraries, audio, or video projects on the boot drive. ie: When you go to back-up the C drive, you don't want HUGE redundant files being included in each backup image file. That makes backup files incredibly large... and takes much longer to create/restore. As Tom pointed out above, it's best to have a drive dedicated to each specific task (OS/Apps, Audio, Samples, etc). If you're wanting maximum disk-streaming polyphony from many large sample-libraries, it's common to use multiple "Samples" drives. We have some TV/Film composer clients who are running 6+ SSDs for disk-streaming sample-libraries. This would be over-kill for more typical singer/songwriter scenarios.

Don't assume that all is well because the drive is new. Some new parts are defective... Benchmark the drive... and you'll have a better grasp on what's happening.

If the D drive is properly functional, it's not a problem to record to it. I'd benchmark the drive... to make sure it's working properly.

I'd definitely go with the Pro version. Not for performance reasons... but because it gives more control. Completely shut down Automatic Updates Group Policy Editor (Shut down Cortana, One Drive, etc)

I run Native Access... and the update doesn't show. 🤷‍♂️ EDIT: Had to update Native Access. Now the updates are available.

The Friedman IR-X is also excellent. True high-voltage tube preamp (Plexi and BE channels) Simulated power section Cab IR Separate boost for each channel (adjustable gain/level for each) EFX Loop

It's almost exactly the same CPU as the 13900ks. Nearly identical performance/features. Specs on the 14900k: https://www.intel.com/content/www/us/en/products/sku/236773/intel-core-i9-processor-14900k-36m-cache-up-to-6-00-ghz/specifications.html Specs on the 13900ks: https://www.intel.com/content/www/us/en/products/sku/232167/intel-core-i913900ks-processor-36m-cache-up-to-6-00-ghz/specifications.html Don't get me wrong, it's an amazing performer... But it's really a slightly refreshed 13900ks. No major architecture change. No major change in TDP. If you're not running a 13th Gen CPU, it makes sense. If you have a 13900k, it's not worth the cost/effort to upgrade (200MHz difference).