mettelus

There are more and more "AI pose generators" coming online that are descent and will manipulate/merge subjects/settings from input image(s). It seems that is part of your request in that you want to re-use a precise subject elsewhere. With 2D this can be more limiting/challenging (massive pose changes require 3D detail of the subject), so I suspect this will spill over into 3D AI tools going forward, but the processing power for such can be pretty hefty.

First, iZotope has a lot of great material explaining audio processing that are universal (you don't need their plugins specifically as long as the plugin you are using has the same features). This link has a pretty thorough walk through of compressors, limiters, expanders, and gates for your reference. That being said, what you are looking for is in an "expander." An expander is like a compressor rotated 180 degrees, and often you will see compressors/expanders in combination (everything above the knee is a standard compressor, everything below the knee is the expander). Although "expander" implies increasing dynamic range that is not always the case. You can also use the expander section to raise the levels of the quieter parts (called "upward expansion") just as easily as lower the levels to increase dynamic range. This is the video from that link above, but what you are seeking is an upward expander (also called an "upward compressor" which is really more accurate since you are still decreasing the dynamic range). Traditionally the "expander" area is what is going on below the knee (regardless of it being up or down).

Just be aware that for hardware drivers (doesn't apply to ASIO4ALL), Windows updates will very likely re-install and re-activate them. It is better to disable them in the Sound Control Panel (leaving them in the registry is fine), since disabling there tells Windows, "Even if you see it, don't use it."

Yeah, the banner is gone now. Unfortunately a lot of deals never specify the end date, but I suspect it will come back in a couple of weeks.

Boy, its been a while since I used it in SONAR, but selecting "Split Clips" is supposed to delete the silent sections. I thought that deletion also butted them together, but my memory might be flaky. If that actually doesn't work, a dedicated wav editor may be a better choice (just not sure of a good free one with this functionality). This is also default edit behavior in video editors (basically ripple edit based on removing silence from the audio track), which may be another option.

Try the Remove Silence tool. I took all of the samples from an electronic kit by recording them, but used Audition, which allowed them to also be split as clips and saved as discrete files (macros rule). Sonar can remove the silence at least with their embedded tool and will butt them all together for you.

There are a handful of plugins that just had a VST3 wrapper applied to the VST2 version that caused issues when used as VST3 (a lot of posts on the old forum regarding this). GPO 5 is old enough that it may still fall into that category. When the VST3 fails, checking the VST2 version (if available) is always something to check.

**** Edit: The deal finally did expire, but I suspect it will come back in a couple weeks. **** I was going to post this late Monday, but then did a web search on the end date and it said Monday also, but it is still active. It may go through Christmas, so not sure. Click on the "Black Friday" banner on their home page while logged in to see your available pricing. Big picture items: The Pro upgrade is cheaper than the early release was ($60 vs $75) but is still no comparison to the steal that Rebelle 7 Pro was to new users (at $30!!). The upgrade path is now Identical to what SONAR Platinum was (free updates for 1 year, with initial buy in getting you everything you missed), so this also gives you the 3 months of updates since Rebelle 8 was released. Some of the features are not as earth-shattering as Rebelle 7 was (i.e., metallic paint), but the engine, brushes, and detail (now Nanopixel 2, with shading/shadows from light source, and the ability to upscale to a whopping 20x increase in size, although only tiff files are "unlimited"), and the ability to control FX by layer are in the Pro version. While Rebelle 8 does have a "basic" version, the Pro version includes all of the features that set it apart from other painting programs. As with most painting applications, a stylus is highly recommended to get the most from the application. Below is the initial New Features video from early release back in August, with the Pro feature set starting around 14:30 if interested. Side note: Rebelle is the only app being actively updated. Flame Painter 4 and Amberlight 2 have been the same for like 5+ years now, so keep this in mind with any "bundles" as you may already own them. A lot of new papers, BUT... you can now also preview new papers from within the app prior to purchasing (another new feature).

Also try using the other Shift key on the keyboard (or both in a text app) as it may also be a keyboard issue. I tend to be abusive to keyboards and the left Shift and left Ctrl keys tend to go first because they are wider and I hit them on the extreme left end and torque them to death.

That video rocks... the quote at 0:45 sorta sums up the sentiment of this thread (and others too)... "You don't need expensive tools to do good work, you need skill!" That comment alone is the nemesis of a large portion of the marketing industry these days (especially now that "AI" is in the picture). People never gain skill being lazy

Another thing is setup and usage should be the same as with any other mic (arm with input echo on, set gain levels, etc., before recording)... did you do that? Another thing you may find in that exercise (I noticed this with the Revelators) is that the gain knob on the mic and the gain knob in the mixer app are completely independent of each other. You will catch this when setting things up properly, but the physical gain knob was zeroed out on the dynamic when I got it. The Blue Yeti has a physical gain knob on the mic as well.

Yeah, the OP is from July 2022 (v6 wasn't even released then). Take lanes were fully incorporated in July 2023 (v6.2), so #4 is N/A now. Melodyne and SO are pretty much joined at the hip, so the tempo issue in his recent posts has simpler handling in SO as well. While there are a few who can speak to the current version of both apps, they are not big in number anymore (and I cannot speak to the current version of either one).

Confused me on this... I thought if you canceled an account or let it expire that you could renew to take advantage of a sale? That is not the case?

Quick clarification for me here. Are there two versions of that interface? When I was looking for specs I was only getting returns on the "hybrid" model, so I assumed there is only one. I guess another easy check is to look at the device and see if it has a FireWire A port on it (one end is squared off and the other end has beveled corners). Comments about drivers above are definitely worth noting (i.e., an older version may be better if you have them or can still download them). The very last Saffire driver for my Focusrite was just a "Mac-compliant update" but broke the Windows drivers, so I downloaded extra copies of the earlier version just so I have them when the driver page disappears. Also, have you reached out to MOTU directly? If it is truly a driver issue, they would be the best resource since they should already know issues and workarounds for their hardware.

That is the case here... the MOTU Ultralite Mk 3 touts itself as the "the very first "hybrid" audio interface to offer both FireWire and hi-speed USB2 connectivity" from its spec page. Another comment of note from the bottom of that page is "If you use USB2, you you'll need to use the DC power supply included with the UltraLite-mk3 Hybrid, as USB2 does not supply enough power by itself."

Out of curiosity, have you checked system files on that machine? SFC and DISM are two things to check when a system gets flaky. Sometimes a system issue only manifests itself for certain tasks.

There are also icons embedded into log entries if they are warnings or errors, which helps sift through things. The ones that are "informational" are not really of interest in your situation, and the vast majority of them should have the informational icon.

Windows has an "Event Viewer" (can type that in the search bar to find/open). If it is being caught at a system level it should be in the "Applications and Services Logs." You can click on that specifically first and it will give a summary of how many events are in each bin. I am not sure if this is strictly a "Hardware Event" though, so may be in the "Windows PowerShell" folder. It may also be in the "Windows Logs" tree, so having a date/time for an event you are seeking helps if there is a lot going on in the machine. Checking the Event Viewer right after something occurs is often easiest because of this.

Quick side comment, this issue was also on the list of bug fixes in the 2025.11 release post. "ASIO Hang Fix: Resolved an issue where Sonar could hang indefinitely if the ASIO driver was in use by another application at a different sample rate."

Also check Windows Sound properties for the device and make sure both exclusive mode options are unchecked. Windows itself can deny access to apps if another is already using it when those exclusive mode options are checked.

Take a quick spectrogram comparison between the live mic setups and the TH3 bounces, it sounds like there is significantly more high end from TH3 (artificial content) than the live mic captures. An EQ matching algorithm can sometimes assist with a post-FX-chain EQ, but they are definitely useful to visualize the frequency differences (the curve required to get the "match"). I suspect that what you are seeing may be more than just the mic shift, but the fact that TH3 is exposing and processing more harmonic content than the live setup does. I have not used TH3 in a while but thought that had more control over mic placement and orientation, but I may be confusing that with another sim.

Mike Greene had sent out a couple of emails regarding this library recently, but I haven't seen it posted here that I could find. There are two variants to this, the Ultimate (113GB) and the Lite (18GB), with the comparison differences near the bottom of either page. Currently on sale for $399 and $199 respectively (versus $599 and $299). His emails detailed a lot of the setup and production, but the synopsis on the pages is simpler to read! He did a lot of work on this to get as many vocal articulations (300) from the Ensemble as he could (very bottom of either page, with 4 tabs to gander through), and took 3 years to finish (there are top/bottom layers possible for each patch). Pretty thorough review below (and a much better walk through of what I said above in the first few minutes).

So this was the exact same setup with the only difference being the amp versus the modeler? That reminded me of another consideration with modelers (though doesn't seem to be a factor here), which is the signal in. It is far easier (and more successful) to remove content than to add non-existing content in, so pickups that output the cleanest set of harmonic content (the more, the better) will have the best chance to mimic any tone. However, having said that, DSP can enhance (in a bad way) those harmonics in a way that analog circuits cannot, which is why using a LPF/EQ to tame that input signal prior to modelers/amp sims is common practice (or put those first in the modeler/amp sim FX chain). As the entire frequency spectrum is being worked with, many analog components do not have a high response rate to exacerbate harmonics in the way that DSP can.

I upgraded to Win11 last week and in dealing with all of the nuances of the update, I noticed today that the system had essentially purged all sound devices not actively connected to the machine during that update. At first I thought it wasn't displaying disabled/disconnected hardware, but they were being shown, just not there. It also enabled everything I had disabled on Win10, so if you have done an O/S change that is something to check, maybe even go as far as uninstalling/re-installing drivers to be sure. Upgrading to Win10 didn't do this (just to clarify), so it may be a TPM artifact of Win11, but is worth noting in case you have shifted O/S to do another run-through of drivers/settings to be sure things are still what they were prior.

+1 to this entire post, in the analog realm there is no way to "prompt jump" (step function) from point A to point B without some transition/smoothing effect, but in the digital realm you very much can. While individual components are often well modeled, their interaction may be lacking. This also leads into the traditional analog medium (magnetic tape), which adds more variables, with hysteresis being one of them. It is the interaction of all those variables that yields the analog result, so the more "piece parts" there are, the harder (or more complex) the modeling will be. Adding more variables onto the pile would be an amp sim accurately modeling a tube amp, especially with complex circuits... my Carvin XV 212 is one of the more complex ones, and LINE6 put a lot of effort into modeling it in their "Legendary Drive." Even though that has the highest CPU usage in HELIX/Native (not sure if that is still the case), it is not fully modeled nor 100% accurate, but as mentioned above, to the average user that will not be an issue/concern. If they never used the hardware, they will be none the wiser. This did give me a good chuckle though, thanks for that! When I slip up and let my sarcasm loose, I catch myself saying, "Congratulations, you Googled something... now you're an expert." from time to time.