Is it a game changer? Does the "cross-instance talking" really help separating instruments, and help have spacious mix without mudding it?

It's a CPU-heavy reverb? as CPU-demanding as Waves H-Reverb? or more demanding?
(when you use the "cross-instance talking" between a few instances of smart:Reverb2)

Thanks!

 

Spoiler: Source: Grok KI

I haven’t personally tested Sonible smart:reverb 2, as my capabilities are based on analyzing information rather than running real-world plugin tests. However, I can provide insights based on available data and user feedback to address your questions about its features, the "cross-instance talking" functionality, and its CPU usage compared to Waves H-Reverb.

Is smart:reverb 2 a Game Changer?
Sonible smart:reverb 2 is designed to stand out from traditional reverbs by offering AI-powered, content-aware processing and intuitive controls. Its key features include:

• Content-Aware Reverb Design: The plugin analyzes the input signal’s spectral and temporal characteristics to create a tailored reverb profile, avoiding issues like masking or clutter. This adaptability is praised for producing natural, source-specific reverbs without relying on presets or impulse responses.
• Reverb Matrix: A 2D interface lets users blend between Room, Hall, Plate, and Spring reverb types, offering flexibility to craft custom spatial textures.
• Perception-Based Parameters: Controls like Distance, Size, Width, Color, and Clarity allow for intuitive adjustments, focusing on musical context rather than technical settings. For example, the Distance parameter mimics natural depth by adjusting diffusion and high-frequency damping, acting like an advanced dry/wet control.
• Group Mode (Cross-Instance Talking): This feature allows multiple instances of smart:reverb 2 to communicate across tracks, coordinating reverb effects to maintain clarity and spatial coherence in dense mixes.

User feedback suggests it’s a significant step forward for specific use cases, particularly for orchestral composers and those working on complex, multi-track projects. For example, a user on VI-CONTROL described it as “fascinating” for its ability to position instrument groups spatially by adjusting the Distance parameter in Group Mode, streamlining depth staging. However, opinions vary—some users find it innovative but not necessarily revolutionary, especially if they already own high-quality reverbs, with one noting it’s “nice” but not compelling enough at full price given existing alternatives.

Whether it’s a “game changer” depends on your workflow. If you value AI-driven adaptability, intuitive controls, and multi-track coordination, it could significantly enhance your mixing process, especially for genres requiring precise spatial placement like orchestral or cinematic music. For simpler applications, its advanced features might feel less essential.

Does "Cross-Instance Talking" Help with Instrument Separation and Spacious Mixes?
The “cross-instance talking” feature, part of Group Mode, is a standout aspect of smart:reverb 2. Here’s how it works and its impact:

• Functionality: In Group Mode, multiple instances of smart:reverb 2 across different tracks communicate to share data, managing overlapping frequencies to prevent masking and maintain spatial clarity. This allows for cohesive depth staging, where you can control the spatial positioning of multiple sources (e.g., placing strings further back than brass) from a single plugin interface.
• Benefits for Separation and Spaciousness: Users report that this feature effectively reduces muddiness in dense arrangements. For instance, a VI-CONTROL user noted that the adaptive nature of Group Mode, where each instance is “aware” of others, helps clean up frequency clashes, resulting in a clearer, more defined mix. Another user on Gearspace highlighted its usefulness for creating density without crowding, suggesting it excels in maintaining a spacious yet uncluttered sound.
• Practical Application: The ability to drag tracks to designated spatial positions within a single instance simplifies depth management, making it particularly useful for orchestral or multi-layered mixes. This perceptual workflow is described as intuitive, allowing users to focus on the mix’s spatial image rather than tweaking individual parameters.

However, there are caveats. One user reported that Group Mode made a track sound “flat,” though this could be due to improper setup or subjective taste. Additionally, issues like inconsistent naming across instances in Group Mode have been mentioned, which could disrupt workflow if not addressed. Overall, the consensus leans toward Group Mode being a powerful tool for achieving a spacious, clear mix, particularly when managing multiple tracks, but it requires proper configuration to avoid potential pitfalls.

Is smart:reverb 2 CPU-Heavy Compared to Waves H-Reverb?
CPU usage is a critical factor, especially when running multiple instances with cross-instance communication. Here’s a breakdown based on available information:

smart:reverb 2 CPU Usage

• User Reports: Feedback on smart:reverb 2’s CPU performance is mixed. Some users experience significant CPU demands, particularly when using multiple instances in Group Mode. For example:
  • A VI-CONTROL user reported pops and clicks with just one instance on a five-year-old processor, with Cubase spiking during adjustments. Four instances caused playback issues, suggesting high CPU demand.
  • Another user on Gearspace noted that smart:reverb 2 was “pretty heavy” on CPU in Studio One and FL Studio, though specifics weren’t provided.
  • Conversely, a user with a decent rig reported no issues running 10 instances without noticeable CPU overhead, pops, or clicks, indicating that performance may depend on system specs.
• Context: The original smart:reverb (v1) was reported to use 0.8% to 1.2% CPU per instance on a 4-core Intel i7-4770K (3.5 GHz, 16 GB RAM) with zero latency, which is relatively light for a single instance. However, smart:reverb 2’s enhanced AI processing and Group Mode likely increase CPU demands, especially when multiple instances are communicating.
• Group Mode Impact: The cross-instance communication in Group Mode involves real-time data sharing and frequency management, which logically increases CPU usage compared to standalone instances. Users with older systems (e.g., five years old) reported significant issues, while those with modern, powerful rigs had fewer problems.

Waves H-Reverb CPU Usage

• General Performance: Waves H-Reverb is a hybrid reverb plugin known for its FIR (Finite Impulse Response) technology, offering detailed control over reverb tails and early reflections. It’s considered moderately CPU-intensive due to its complex processing, but specific CPU benchmarks are less documented in recent sources.
• User Feedback: In discussions on forums like Gearspace, H-Reverb is often described as manageable on modern systems but can become demanding when running multiple instances, especially in dense sessions with other plugins. No direct comparisons to smart:reverb 2 were found, but H-Reverb’s CPU load is generally seen as moderate compared to convolution-based reverbs like Altiverb or high-end algorithmic reverbs like FabFilter Pro-R.
• Comparison Context: H-Reverb doesn’t feature cross-instance communication, so its CPU usage is tied to individual instances rather than inter-plugin coordination. This makes it less likely to spike as dramatically as smart:reverb 2 in Group Mode, but its hybrid processing (including time-domain shaping and EQ) still requires significant resources.

Comparison

• Single Instance: For a single instance, smart:reverb 2’s CPU usage (based on v1’s 0.8–1.2%) is likely comparable to or slightly lighter than H-Reverb, assuming similar system specs. However, smart:reverb 2’s AI analysis may add overhead, especially during initial signal processing.
• Multiple Instances with Cross-Instance Talking: When using Group Mode with multiple instances, smart:reverb 2 appears to be more CPU-demanding than H-Reverb, as the cross-channel communication and real-time frequency management add computational complexity. Users with older systems reported issues with as few as four instances, whereas H-Reverb is less likely to cause such problems unless heavily stacked.
• System Dependency: Both plugins perform better on modern, high-spec systems (e.g., recent multi-core CPUs with ample RAM). For smart:reverb 2, users with older systems (5+ years) or less powerful rigs may encounter issues like pops, clicks, or DAW crashes, particularly in Group Mode. H-Reverb, while demanding, seems to have a more consistent performance profile across systems.
• Mitigation: For smart:reverb 2, users suggested workarounds like disabling CPU core parking to reduce pops and clicks. Increasing buffer size (as noted with smart:reverb v1 in FL Studio) may also help, but this applies to both plugins.

Conclusion on CPU Usage
smart:reverb 2 is likely more CPU-intensive than Waves H-Reverb when using multiple instances in Group Mode due to its cross-instance communication and AI processing. On a single instance, they may be comparable, with smart:reverb 2 potentially slightly lighter if Group Mode is disabled. However, performance varies significantly by system—modern, powerful rigs handle smart:reverb 2 better, while older systems struggle, especially with Group Mode. If CPU efficiency is a priority and you’re using an older system, H-Reverb may be safer, but for cutting-edge spatial control, smart:reverb 2’s demands may be justified.

Recommendations

• Try the Demo: Sonible offers a 30-day free trial for smart:reverb 2. Test it in your DAW with your typical project setup to assess CPU performance and the effectiveness of Group Mode for your mixes.
• System Check: If your system is older (e.g., 5+ years), expect potential CPU challenges with smart:reverb 2 in Group Mode. Consider upgrading hardware or limiting instances if issues arise.
• Workflow Fit: If you work on complex, multi-track projects (e.g., orchestral or cinematic music), the Group Mode’s ability to manage spatial clarity could be a significant advantage. For simpler mixes, H-Reverb’s detailed control and potentially lower CPU load might suffice.

If you have specific system specs (e.g., CPU model, RAM, DAW), I can tailor this analysis further. Let me know if you’d like more details or assistance with testing strategies!

 

Iv purchased it and found it too demanding on my current setup. Mac Mini M2 Pro.
Mainly tested in Ableton Live 12. The concept seems ok though.
Currently I like ADPTR Utopia Spectral Reverb more for my current work.
But give the 30 days trial ago yourself and see if it fits your needs.

 

"cross-instance talking" doesn't compensate what for me is simply not a successful reverb. Just doesn't sound right to my ears, in fact I would vastly prefer doing a little extra work on placement and use such folks as Liquidsonics, iZotope (Exponential Audio), Valhalla et. al.

 

If you don't mind using sends and spending 30 seconds on dialling in settings, then a spectral compressor with sidechain input will give you a similar result, but with freedom of choice for reverb.

 

I tried it, but I have two issues with how it works:

1- uses way too much CPU for a reverb that is supposed to be an insert. UVI Bloom uses 0 to 1% CPU, so that technically could be used as an insert as much as you want, but smart reverb 2 takes 6 Blooms for a single instance, which is too much supposing that is how you are supposed to use it.

2- I don't like how it sounds. It might achieve what it is supposed to do, but I don't personally think the reverbs sound good. I can't really put my finger on it, but I'm sure that almost every other reverb I have sounds better. Even the UJAM Reverb, that most people would not really consider a serious mixing company, sounds so much better than this. So much better that I paid for both versions.

I love Sonible but that's two misses in a row. They should focus instead on smarteq and smartlimit, those two are the greatest they have. Hell, even smartcomp 3 would be a better use of their time.

 

Tweaking reverb is a religious experience for me. Smart reverb (both versions) or Neoverb were total misses for me.

Like @xorome says, it's really easy to put a spectral compressor after your reverb, and you have so much more creative control.

What was the previous miss?

 

Prime vocal. It's an interesting idea that is very poorly executed mainly for the fact that you need to process the entire audio to start working (which can take minutes), meanwhile their own plugins + something like dxRevive/Clear can do the same in just a few seconds, with more control.

 

Personally what a reverb does to a sound or more importantly doesn't do in some cases, is more important - For me.
Seventh Heaven sounds lovely for me as my go to i.e. and this is missing something. The Lexicon LXP sounds nicer too. being more clever doesn't always mean better.