The No.1 Website for Pro Audio
 All  This Thread  Reviews  Gear Database  Gear for sale     Latest  Trending
sonible releases smart:EQ 2
Old 3 weeks ago
  #331
Gear Head
 

Quote:
Originally Posted by Eric Stravinsky View Post
Still waiting for the Vocals preset @ sonio
Yes - I know. The vocal profile is actually ready, but we have to pack it into a new release. Due to the success of our newest plug-in smart:comp, the plug-in needed quite a lot of attention over the last couple of weeks (initial release, bug fixes, etc.) - but we didn't forget about smart:EQ 2!

We're also planning to implement some minor bug fixes for smart:EQ 2 and we want to integrate some of the optimized processing modules of smart:comp into the plug-in. So - again - stay tuned! We'll try to catch up on this soon.
Old 1 week ago
  #332
Gear Addict
 

Quote:
Originally Posted by sonio View Post
Yes - I know. The vocal profile is actually ready, but we have to pack it into a new release. Due to the success of our newest plug-in smart:comp, the plug-in needed quite a lot of attention over the last couple of weeks (initial release, bug fixes, etc.) - but we didn't forget about smart:EQ 2!

We're also planning to implement some minor bug fixes for smart:EQ 2 and we want to integrate some of the optimized processing modules of smart:comp into the plug-in. So - again - stay tuned! We'll try to catch up on this soon.
Oh yeah smart:comp sounds amazing btw.
Old 11 hours ago
  #333
My biggest gripes with the current version are:

1) It only offers linear phase. Linear introduces pre-ringing. Sonible hasn't mentions ways that they've fixed this, that I know of.
2) The states switch instantaneously (jumping from one, to the other). Jumping from state to state, can cause a full mix to not sound right (the bass kick will become noticeably different, and not in a good way... all though this can sometimes be good, when done in extremely small amounts).
3) The solo band button should solo just the sides signal, if a band is set to side mode (or solo the mids, in mid mode).
4) CTRL + click & drag (or shift) should allow finer tuning of the gain knobs for each band. Same with que. And, both gain and que should be accurate to the nearest hundredth or thousandth of a decimal point.
5) We should be able to press record in the DAW, play around with parameters like the HP/LP Crossovers, bypass on/off for each filter point, and the Hz/strength/width knobs.

Skip my work arounds / methods (next), and read my 10 suggestions for smart eq 3 (after the next bold lettering.

My work around for it being linear phase only is to lower the amounts of changes, and to lower the amount of low frequency attenuation, via the low point crosover slider.

For the sudden jumping, from state to state (for mastering), is to learn as many profiles as is needed (I find 10 to be good, even having 2 or more states saved from one particular section of a song)...
then open a second instance of Smart EQ 2 (instance B) . This one will stay in the same state, during the entire song. For this curve, use the one that the AI created (and you adjusted by ear), for the main drop or chorus (the most loud part of the song, usually).
Instance 1 will transition from each state, through the song. However, it will be turned WAY down, so that there are changes happening, but the scene point endings/beginnings are smoothly transitioning (so that its not something you can easily notice).
Look at each of the other states, on instance 1 and drag this main one (instance B) around until it compensates a TINY bit, for differences between its over-all curve, and the others.
On the first instance of smart eq 2, go and adjust each of the individual states, so that their cuves are shaped slightly more like the main curve that you're using through out the whole song.
Then, just make sure that you lower the amounts of strength on both instances a lot... I would suggest dialing them in until you can hear a clear difference with just instance 2 & bypass signal, and that the difference is an improvement... then A/B until you can barely hear that difference... and then cut that strength in half... A/B it and confirm that you can't hear a difference... and then decide on whether the halved value or full value (no percievable difference vs barely perceivable difference) is best... maybe meet some where in the middle.
Do the same thing with instance 1, but make sure its doing a lot less (at least, half the amount of change... probably closer to 1000% less).
Can be done, at this stage... but I like to follow up with Gullfoss, DeEdger, DeFilter, or Soothe.
Check suggestions of dynamic EQ, with TDR Nova (match pink and deresonate), Ozone 9, and/or others... and then use those suggestions possibly to do last bits of small adjustments to the curves on smart eq 2 & other tools that are before/after smart eq 2, like the pre static/dynamic EQ, coloring EQ, post static/dyanmic EQ, and finalizing EQ/harmonics generating tools.
With TDR Nova, look at suggestions on the sum (mid) and difference (side) modes, and adjust your other tools if necessary. Optionally, check left and right, as well. Verify that the stereo image is good, with tools like Insight, Master The Mix Levels, etc.
With Ozone, try the MA on both vintage and modern modes, and try making various adjustments.
With final adjustments, always decide what to keep, by your ears, but take suggestions from various tools, such as AI suggesting ones, or balancing / level suggesting tools (like Ozone Tonal Balance, etc.).
After the whole thing seems done, add another smart eq 2, learn the same sections of the song; and look at the suggestions... again tweaking instance 1 and 2, and deciding by ear.
Also, if the artist wants a mix to be similar to other artists' material, try EQ matching techniques, to find other suggestions.
Each tool that you use, to find suggestions, is another perspective... which might cause you to make different final decisions... but ultimately, taste & ears win the battle.
Always listen on different speakers, and try to listen to club mixes in various clubs... before going back and doing any re-mastering adjustments.
Also, I suggest trying out other combinations of EQ tools. One that I highly suggest is EQuivocate. Try taking a look at what Smart EQ 2 does, when you've adjusted it until it sounds best to you... and then mimic the curve, with EQuivocate... dialing down the changes in both smart eq 2 and equivocate.




I have ten strong suggestions for Smart EQ 3!!!

+1+ Add various additional individual bands to the over-all filter curve.
**** Keep the original bell-ish type of curve (shaped as it currently is), but add the ability to double click to add individual bands, of various shapes.
**** Bell, tilt, HP, LP, high shelf, low shelf, tilt, "flat top bell", and "air band", should be included. See air band, listed below.

+2+ Variable phase filter types, as requested by others.
**** This should include zero-latency, and "natural" (or "mixed", meaning linear on the highs and closer to zero-latency on the lows), and 100% fully linear (meaning linear filter on the pre and post stages).
**** Add the ability to adjust the phase type, seamlessly from linear to zero-latency, for each filter point.
**** If each filter band had its own strength, width, and smoothness options, for further fine tuning... that would be fantastic.
**** We should have 3 buttons on the top of the interface, with 3 main options for both strength, width, and smoothness: universal, mixed, and individual band. Universal would let only the main strength, width, and smoothness parameters to control it. Mixed would let the universal parameters control it, with each band having the ability to go from 0% to 100% (100% being the value set by the universal parameter). Individual would bypass the universal parameters, letting only the individual band parameters control it. The individual band controls should pop up on the screen, when the mouse is hovered over one filter point.
**** Bell shapes. with flat tops, would also be super great, for dialing in the strength of specific sections of frequencies. These "flat top" curves have been said to be great, for various issues involving pre-ringing (and possibly for latency). The flat top curve types could be improved by adding the ability to adjust the corners' sharpness levels (where it goes from bell curve shape type to suddenly flat)... So have a sharpness knob (or corner "knee" value) make them go from totally sharp corners, to slightly rounded & smoothed out with a small curve.. Speaking of knees... wouldn't it be nice to be able to adjust knee values, in the same manner that SmartComp2 is able to adjust the attack/release curve shape types?
**** I suggest adding a very soft and transparent air band, at 34 kHz... allowing the AI to draw in the shape of the air band boost (or, in rare cases, cut). Allowing the user to flatten it out, to the shape of a standard air band curve, would be great. Side note on air bands: I'm of the opinion that an air band boost of +1.0 dB should require some sort of a low subs shelving boost (perhaps +0.1 or maybe +0.02 dB, for an air band boost of +1.0 dB... at 0 or 10 Hz).

+3+ Add the ability to turn on or off, both the positive gain changes and the negative gain changes (independantly).
**** For example, turning off the cuts would result in a flat curve, where ever there would be cuts.
**** This way, users could dial in either only cuts or only boosts, and specific stages of their mastering (or effects) chain.

+4+ Add a global blend knob for user/AI drawn curve types, going from -100% (user programmed curves) to 100% (AI drawn curve), and a blend knob for user/AI, for each additional individual band.
**** This would let users drawn in actual curves that they would normally draw, and then blend in whatever ratio of what they're drawn curve is vs. what the AI decides to make.
**** The curve drawn by the user would be the ACTUAL shape of the filters that they've drawn in.
**** This would allow for some very advanced types of shaping methods.

+5+ Adding some sort of dynamic fluctuations of various sorts would be pretty epic.
**** Both the main bell-ish curve, and each additional individual band, should have a button to turn on mode for dynamic mode to states of on, off, or mixed.
**** Dynamic mode turned to "on" would mean that the drawn in curve would be triggered/attenuated by the threshold(s)/attack/release/hold, set by the user. Expand would push towards the curve and compress would push away from it.
**** Dynamic mode set to "off" would allow the plugin to function as it currently does, with a static curve.
**** Dynamic mode mixed would allow the static curve set to remain applied, with additional and adjustable dynamic behavior types to be added by the user, if desired. Compress would push the curve away from the drawn curve, towards the flat line; and cause the effect to lessen, whenever selected thresholds are crossed. Expand would push the curve outward, making the effect stronger whenever thresholds are met.
**** Users should be able to select both global and individual band settings for threshold, attack/release/hold (to nearest +/- 0.001 ms), expand/compress mode, and off set +/- 0.001 dB.
**** The static +/- offset would have two modes... to either apply an additional offset curve, of the filter shape that the AI has drawn (for each individual filter) or apply the user-drawn curve shapes. This would allow users more in depth control, which combines the AI curve with curves that they want to manually create. It would be able to offset the functioning of the global/individual user/AI blend knobs (mentioned in +4+), in interesting ways.
**** The attack and release shapes should be able to be adjusted, in an expanded advanced section, which would give the smartcomp2 feature for adjustable knee shapes. This would be the first dynamic EQ with the ability to draw knee shapes, and it would give us new ways of creating dynamically flowing mixtures of sounds.
**** Each band should also have a sidechain input option, allowing any given band to be attenuated by sidechained audio sources. The sidechain sources should be able to be routed to the static behavior (AI learning function), dynamic behaviors, or both.
**** This would allow users to (for example), allow the plugin draw in its normal static curve(s), and turn on dynamic mixed mode, and then allow the plugin to compress the boosted/cut signal amounts (bringing their values closer to the flat 0 line), at certain thresholds which are set by the user. This way, at loudness targets above the threshold, the EQ curve will flatten out a bit closer to the original bypassed signal. So, louder stuff would sound more like the original mix, and quieter stuff would sound more like the EQ'd signal.
**** Alternatively, users could use mixed mode, with dynamic mode set to expand, to let the drawn filter shape to be increased, whenever thresholds are met.
**** In dynamic only mode, the entire curve would only be triggered, when thresholds are passed. Expand would go towards the curve, and compress away from it. The offset +/- dB function would allow dynamic only mode to function similarly to shared mode (when its set a certain way). But, shared mode would allow the offset to bounce back against the globally drawn curves, in a different manner, as well.
**** With the advanced attack/release shapes, this could produce various styles of womping/ducking types of effects, with filter shapes for determining where things are wobbled, ducked, lessened or strengthened.

+6+ I suggest widening the frequency scale, to allow viewers to see down to 0 Hz (or 10, at the most), and up to above 40 kHz.
**** This gives users access to a wider amount of information, and even though frequencies are perceived as being inaudible, they're still important because they are both felt by our bodies (in various ways), and they have a direct relationship with the audible frequencies. In other words, if you took a signal that contains information across the full spectrum (from 0 to whatever Nyquist is); and then completely remove frequencies from 0 to 20 Hz, and from 20 to Nyquist (say, with a brick wall HP and LP, for example), and playback the result. It'll be drastically different in various aspects of pitch, frequency, and time! Not only will things sound slightly different, in pitch and frequency aspects, but also the TIMING and placement of sounds across the timeline will sound audibly different (effecting all sorts of differences in wave form textures/characteristics, such as timbre, transient energy, thump, thwack, muddiness, boxiness, click, variability of BPM [or swing], darkness/brightness, etc).
**** Whether our DAW and plugins allow those inaudible subs and highs to play through or not, is a big deal. Sometimes lobbing that information is more helpful than it is harmful (as it allows us to then gen more perceived loudness from using only the audible frequencies, while slightly damaging the integrity of the original mix). And, being able to have visual representations of them, is also huge (for me, at least). Plugins that just lob off that information and cause it to be mute signal (all bits turned to zero) are a BIG nono, for me (so I assume that the current Smart EQ 2 does pass that information through, untouched, since the selected band cutoffs are at 20Hz and 20kHz.

+7+ I'd like the plugin to have the ability to morph from one state to another, smoothly (or seamlessly). In other words, don't just force it to change directly from one curve to another, instantly. This is a huge one for mastering, for me (personally).
**** Either allow the automation values (I believe they're from 1 to 100) to select values like 2-10, which would select steps of curves that are gradually morphed from one state, to the next...
**** OR, allow users to set a time period, in which the plugin morphs smoothly from one state, to the next. In the latter option, users would have to make their automation value go from 1 to 11, a bit before the transition of one section of a song, to another (either set to be the same as the time period that is set in the plugin, or to half of that value). This way, if your kick drum is the same, from one section to another... but mostly everything else is drastically different... then it won't be so obviously clear, when the plugin jumps from one state, to the next. As the plugin is, now... I'll either dial the strength down to below 10%, so that the jumping of states (from verse to bridge or chorus, for example) is not so noticeable... or, I will let the plugin learn each section of the song... and either manually dial in a curve that is an average of all states (with an emphasis on which ever curves are drawn from the most loud sections of the song), or load up one sonible smart eq 2 for each section, leave them all on, and bring the strength down to 1-5%, for all of them.
**** Even better, let the user draw in an attack/release curve (like in smartcomp2), for how the morphing the behavior of transitioning from one state to the next. This would suggest that the "time period" suggested above be called hold. Attack / Release would be adjustable.
**** See current version gripes and methods, at the top.

+8+ I think that adding something along the lines of manual shaping of the curve, based on the Mel Spectrum (featured in EQuivocate, amongst possibly others), would be a good idea.

+9+ I'd like and option to add even only, odd only, and mixtures of even and odd harmonics, to only the difference of the bypassed signal and the EQ'd signal. Different forms of saturation/distortion/overdrive types of things, or harmonic excitement/enhancement... which would be applied to ONLY the added/subtracted audio (post effect).
**** To be clear, what I'm suggesting can be achieved, if one were to duplicate track A (call the duplicate track B), add smartEQ2 to track B (post EQ effect), flip the polarity of track B, sum them together on a new track (track C), in order to cancel out frequencies that are shared (obtaining only the difference), duplicating track C (to make track D), applying excitement/enhancement of various types to track D (perhaps, with the ability to apply the various types of excitement/enhancement to multiple user selected bands), flipping the polarity of track D (post effects), summing C & D onto a new track (E), to obtain the difference... and then summing the resulting audio from track E, with track A.
**** The result would be generated harmonics, created from only the difference of the bypassed signal and the EQ'd signal. To take it a step further, one could add the ability to flip the polarity of track D, before the excitement is added (this would add harmonics that are in opposite polarity of the original difference... which I've heard can yields some very nice results in creating beautiful stereo images, and help with methods of separating things so that everything sounds more crispy, isolated, and clear).
**** Here, I'd also like the ability to add the harmonics to either only the boosted signal difference, or only the cut signal difference. And, I'd want a global knob to control the percentage of effect, from 0 to 100% (or a wet/dry harmonics knob)... and I'd want each filter point to have a knob for dialing in the harmonics wet/dry (percentage), and a button to select various types of even/odd/mixed harmonics generator types.
**** This would be great... because adding harmonics to removed signal would cause the illusion of those frequencies still being present... making it possibly be perceived to be sounding more similar to the original mix.
**** Also, it could help to make the over-all EQ spectrum to be more balanced, which would allow the EQ filters less necessary, at higher frequencies. With that being said, the dev's at Sonible could take it to a whole new level... if they told the AI filter learning to learn variable curves that are based on the strength and amount of the added harmonics (from harmonics added to both the globally set band curve, and the individual bands of additional bell/shelf/pass filters), band selection (or filter curves selected), and types of harmonics. In other words, if a user selected a section of the curve, which had muddiness (or boxiness) on the snare or kick, then one could generate harmonics from those areas, before doing variable cuts, and dial in amounts of harmonics from those very frequencies that they're cutting. Furthermore, they could flip the polarity of those added frequencies.

+10+ I am also unsure if this has been added or not, but the ability to change the strength for each state, independently, is a must.

+11+ -- (Edit, added) -- If this plugin was a VSTi capabilities, it could allow various knobs & parameters to be driven by LFO's, Envelope followers, sidechain audio inputs, and envelope generators. A good example of this functionality is within some of the Sugar Bytes plugins... as well as fabfilter Volcano. This would be great, to allow the Hz/Strength/Width knobs, crossover LP and HP points, and even the individual filter points' parameters of que, amount, and there could even be an option to let the LFO or w/e to drive the selection of Mid, Side, or Stereo... like a signal over X amount causes a given point to switch from stereo to Mids; and then, after an adjustable hold time goes by, it switches from mid to side... and stays on side until the remaining adjustable release time goes by... and then back to stereo. You could then use sources to cause boosting textures of mids & sides to flip around each other, in weird ways.



I may have over-complicated some things, or you may be able to come up with more simple ways of getting the above things done. Look forward to any possible replies!!!
Topic:
Post Reply

Welcome to the Gearslutz Pro Audio Community!

Registration benefits include:
  • The ability to reply to and create new discussions
  • Access to members-only giveaways & competitions
  • Interact with VIP industry experts in our guest Q&As
  • Access to members-only sub forum discussions
  • Access to members-only Chat Room
  • Get INSTANT ACCESS to the world's best private pro audio Classifieds for only USD $20/year
  • Promote your eBay auctions and Reverb.com listings for free
  • Remove this message!
You need an account to post a reply. Create a username and password below and an account will be created and your post entered.


 
 
Slide to join now Processing…
Thread Tools
Search this Thread
Search this Thread:

Advanced Search
Forum Jump
Forum Jump