Jump to content

Jim Roseberry

Members
  • Posts

    1,097
  • Joined

  • Last visited

  • Days Won

    9

Everything posted by Jim Roseberry

  1. It's actually the Output that you're most concerned with when playing a virtual-instrument (3.7ms in your case). You can actually combine monitoring via hardware (dry vocal) and monitoring via software (just for the reverb). Setup a track with your preferred vocal reverb... and set the plugin to 100% processed/wet signal Enable input monitoring on this "Vocal Reverb" track. You can adjust the level of the reverb by adjusting this track's level. Monitor the dry vocal directly off the audio interface (or hardware mixer) This way, the dry signal is near zero latency... and you get to hear your choice of reverb. The 6.7ms round-trip latency won't be distracting on the 100% wet reverb signal. It'll sound like 6.7ms of Pre-Delay. You can use this some process if you have vintage keyboards (that don't have onboard effects). It's nice to be able to monitor with chorus, delay, reverb... while still being able to record the dry signal.
  2. When using the word "overclocking", I want to clarify: I'd recommend running RAM at (or near) it's maximum listed clock-speed. This assumes the motherboard supports the higher RAM clock-speed. I would absolutely not recommend trying to achieve RAM clock-speed above what the RAM is rated. That's almost guaranteed to cause instability. If you're in Cakewalk's preference settings: "Effective Latency" shows that it's listing the audio interface driver's Input latency. This can get confusing... I'll try to explain: When working with a DAW, you're dealing with two types of latency. Playback (or one-way) - when playing virtual instruments Round-trip (or two-way) - when playing/monitoring in realtime thru software based effects/processing Round-trip latency is the sum of the following: ASIO input buffer ASIO output buffer The driver's (often hidden) "safety" or "streaming" buffer Latency of the A/D and D/A converters Let's say you have your audio interfaces set to 44.1k with a 64-sample ASIO buffer size. ASIO input buffer = 1.5ms ASIO output buffer = 1.5ms So you're already at 3ms... without counting the latency of the driver's safety/streaming buffer and latency of the A/D D/A converters. Total round-trip latency is going to be 4-5ms (or much higher if the interface uses a large safety/streaming buffer size). To complicated matters, some interfaces don't report their latency accurately. Ideally, you want to use a loop-back utility like RTL Utility... to measure the actual round-trip latency. If an audio interface reports its ASIO input/output latency accurately, Cakewalk will list it accurately under ASIO Reported Latency. If an audio interface driver doesn't report its ASIO input/output latency accurately, recorded audio maybe slightly offset (out of time). In this case, you can take a short transient and loop-back record it to a second track... and zoom-in and see if there's any difference in time (between them). If so, you can take the time difference between them (in samples), and enter it in Manual Offset - which is under Record Latency Adjustment (samples). Hope that helps clarify!
  3. I'm definitely running RAM above 3600MHz (all four slots populated). It's more complex than just setting XMP. Initial support for DDR5 (both X670e and Z790 motherboards) has been less than ideal. Initial Z690 support for 13th Gen CPUs has also been less than ideal (~25% lower performance).
  4. Hi Jeffrey, IMO, The reason to have a CPU like the 7950x or 13900k is ultimate performance. While undervolting is fine (just need to test for absolute stability), I wouldn't go in expecting an undervolt to be a substitute for proper/robust cooling. You're still going to need a large AIO and a large case to accommodate. A successful undervolt may buy you a few degrees. Don't expect it to be a night vs day difference. If you're talking a significant undervolt, you're probably going to need to (significantly) reduce clock-speed. That's not what you want in a DAW. The 7950x doesn't need "fixed". It simply needs proper cooling. The 7950x sitting next to me is blazing fast... and very quiet.
  5. IMO, The Air plugins (for MPC Key-61) are pretty decent. Some of the instruments hold their own vs. a traditional keyboard workstation (Montage, Fantom, etc). They're not going to best the likes of Pianoteq, Keyscape, Omnisphere... and other top-tier VIs. I bought the Mini-D (for MPC Key-61). It's OK. Nothing more... nothing less If you compare it to something like a Moog Matriarch, it comes up very short. 😉 If you're composing with the MPC Key-61, it's useful in that environment.
  6. Hi Grem, No, you won't get the same performance (13900k) using a Z690 motherboard. I tested that with high-end Z690 motherboards (dual Thunderbolt-4 ports). Performance was significantly slower (to where it was more in-line with the 12900k). Thunderbolt-3 devices work just fine with Thunderbolt-4 (Apollo, Antelope, UFX+, and Quantum all work well).
  7. Huge update for Helix. Cabs sound significantly better. ie: Take the MESA 4x12 SM57 - edge-of-cap with distance of 2" M160 - center with distance of 12" Sounds exactly as it should. No high/low cuts necessary Maybe it's just me... but I think the Placater (BE-100) model now feels more responsive/lively.
  8. Your audio interface will determine the lowest latency you can achieve. The machine obviously has to be able to sustain the load (glitch-free). The X-factor with audio interfaces is the (often hidden) "safety" or sometimes called "streaming" buffer. A small number of audio interfaces allow you to change this buffer size. The best audio interfaces (best drivers) use a small safety/streaming buffer. Mediocre units use a larger buffer... and have significantly higher round-trip latency. If your audio interface uses a large safety/streaming buffer, it'll never yield low round-trip latency. This is separate from the ASIO buffer size.
  9. FWIW, I run my own benchmarks... in addition to Cinebench R23. I'm particularly interested in ultra low latency audio performance... so I push the machine to find absolute limits. Based on previous CPU performance, Intel/AMD reported performance gain, etc... you can pretty well guess where a new CPU will fall. That said, I always like to see/experience it first-hand. ie: I had heard the 13900k was breaking 40,000 in Cinebench R23 multi-core. It's nice to actually have the machine (in front of you) breaking 40k... without thermal-throttling... and without CPU temps getting in the 90s C.
  10. I've actually got both 7950x and 13900k machines in the studio. I've tested both extensively. The 7950x is less prone to thermal-throttling than the 13900k. The 7950x can run (without damage) at 95 degrees C for extended periods. Just because it can run without damage doesn't mean it's a good overall scenario. There is a *massive* difference (under heavy loads) when using a NH-D15 (140mm air-cooler with dual fans) and a top-tier 420mm AIO. Running Cinebench R23, I've watched the 13900k thermal-throttle with a quality 360mm AIO. To talk about using something like a Hyper 212 (120mm air-cooler) is pure comedy. Yes, some YouTuber can piece it together... and it runs "works" (under light load). Is it going to perform the same as robust cooling? Absolutely... no... way Render some video with said $30 cooler... and watch the wheels fall off (thermal-throttle). As I mentioned before, you don't want the CPU hitting 95 degrees C running DAW apps... as fan RPMs (noise) will be extremely loud. I've seen YouTube videos where someone assembled a Threadripper based machine using a NH-D14S. Threadripper has 280w TDP. "Runs nice and quiet!" Yeah, now render some video. I've tested the 10980xe (165w TDP) with a D15 (dual fans). Fine if you're just tinkering (light loads). Render some video... and watch clock-speed plummet. If you're wanting to save money: Instead of buying a $700 CPU (and pairing it with an inadequate $30 cooler), buy something like a 12700k and pair it with a quality air-cooler. You'll save ~$250... and the machine will run cool/quiet (no thermal-throttling under load). As someone who's built DAWs professionally for nearly 30 years: If Noctua air-coolers were adequate for the 7950x and 13900k, I'd be using them.
  11. Yes, I can verify that CbB does indeed use the Efficiency cores (in addition to the performance cores). I have the Efficiency cores running at 4400MHz. While not 5.8GHz (like the Performance cores), they're still significantly useful.
  12. Waste of time... If you buy a 7950x... and cool it with a $35 air-cooler, you're going to get absolute crap performance. Under substantial load, there's no way even a D15 (dual fans) is going to prevent thermal-throttling. If you're going to let it thermal-throttle, there's no point in having a $700 "workstation" CPU. Not only do you need to prevent thermal-throttling, but you also need to keep the machine quiet. If the 7950x runs at 95 degrees C, guess what happens to your fan RPMs?!
  13. Absolutely. It's currently a pain to have to use a totally separate instance of ToneX (for a single drive pedal). By comparison, Quad Cortex can run up to 9 simultaneous Captures... making it extremely easy to setup multiple boost/drive options (or multiple amps with multiple boost/drive options).
  14. I liquidated a couple of my nice tube-amps... but I still have a 20th Bogner Ecstasy and a H&K Triamp mk3. Also have a nice collection of boost/drive pedals. I'll get around to Capturing those as time permits. For testing the waters (easy/silent), I've been Capturing some of my favorite Helix amp tones.
  15. The 13900k can run on an existing Z690 motherboard (instead of the newer Z790). However, I wouldn't recommend this... as the 13900k runs significantly slower ie: Using an Asus ROG Maximus Z690 Formula (~$800) motherboard, the P cores are locked at 5500Mhz and E cores are locked at 4300MHz. Doesn't matter what settings you actually choose.
  16. Just demo'd the Big Sky plugin. It may be a bug... but at anything less than a 64-sample ASIO buffer size, it **hammers** the CPU. This is running a blazing fast i9-13900k. By comparison, this same machine can run ToneX and Sonsig Reverb-A at 96k using a 16-sample ASIO buffer size (glitch-free). Aside from the poor performance, Reverb-A sounds just as good.
  17. Back when digital consoles were the "new thing"... Many digital consoles (if they lost digital clock)... would immediately output full scale white noise. Aka: The tweeter killer!
  18. I'll 2nd the Squier by Fender J or P bass suggestion. Sterling by Musicman is also a good lower-cost bass. You might also want to check out Sire basses (Marcus Miller is associated). Ibanez also makes decent lower-cost basses.
  19. Got Helix Native to load on the 13900k. Oddly, I had to Deauthorize the plugin... then reauthorize it. 13900k can play a substantial Helix Native patch at a 24-sample ASIO buffer size (96k). The 7950x and 13900k perform nearly identical. Both are somewhat of a, "go big or go home" scenario.
  20. Both the 7950x and 13900k will allow you to run IK's new ToneX plugin at 96k using a 16-sample ASIO buffer size. That's ~0.5ms total round-trip latency (providing your audio interface allows setting buffers that low). Amplitube 5 is a much heavier load. Both 7950x and 13900k have to be set to a 32-sample ASIO buffer size (96k) to playback glitch-free. Helix Native doesn't currently load with the 13900k... but it does load with the 7950x. The 7950x can run a substantial Helix Native patch at a 24-sample ASIO buffer size (96k). That's still sub 1ms total round-trip latency... which is better than any current hardware guitar processor.
  21. Rebuilt the 13900k with a larger case and 420mm water-cooler. Cinebench R23 Multi-core score = 40,658 (no thermal-throttling) Single-core score = 2284 To run the 13900k at full potential, 420mm water-cooler is absolutely necessary. Same is true with the 7950x.
  22. It's absolutely analogous to high-performance auto racing Under super heavy loads, you'll see clock-speeds drop (from max Turbo). Cinebench is an easy way to watch/experience. As you'd expect, the Performance cores do the heavy lifting... and Efficient cores handle lesser tasks. What's somewhat amusing is that (though the architecture isn't the same), the experience of working with both the 7950x and 13900k is similar. I'll have more to say about the 13900k's ultra low latency audio performance (hopefully later this afternoon).
  23. Not if you're chasing high-performance 😁
  24. As expected, the 13900k does slightly outperform the 7950x in Cinebench R23 (both single-core and multi-core). SC=2210 MC=38701 Testing with a 360mm AIO. Picking up a 420mm AIO today. If you're thinking air cooling (a la NH-D15), I'd quickly forget that notion. If you're wanting to hit a score of 40k with Cinebench R23, you're not going to do that with a 360mm AIO water-cooler. I'd consider a top-tier 360mm water-cooler as bare minimum. Ideally, you want a top-tier 420mm water-cooler (or better). That means a larger case. Haven't run low latency audio tests (will do that today). Like the 7950x (due to heat), configuring the 13900k is a lot more complex than a 12700k. It's a delicate balance finding the sweetest spot for performance... while keeping noise to a minimum.
  25. 8 Performance Cores 16 Efficient Cores 32 Processing Threads 5.8GHz Max Turbo This is going to be interesting (to me) vs. the 7950x. The 7950x is the first CPU to be able to run things like IK's ToneX at 96k using a 16-sample ASIO buffer size. That's ~0.5ms total round-trip latency. 12900k/s and 5950x can get down to a 32-sample ASIO buffer size. I expect the 13900k to (slightly) best the 7950x in Cinebench R23 (both single-core and multi-core). Most curious to see if the 13900k can run ToneX (glitch-free) at 96k using a 16-sample ASIO buffer size. Have to say that once completely dialed-in, the 7950x has been rock-solid and super quiet. Thunderbolt performance has been exceptionally robust. Can't wait to compare ultra low latency audio performance. AMD and Intel are slugging it out... and we benefit.
×
×
  • Create New...