Jump to content


Advanced User
  • Posts

  • Joined

Everything posted by Sascha

  1. Here's something fresh & new from my band. Not the world's most happiest song, though, but it's also not the time for party right now. We've recorded most parts just before the coronavirus lockdown but had to finish up on it in isolation. The outcome is not perfect, but nothing is these days. https://www.youtube.com/watch?v=XK_cLbi5amA Equipment: - drums recorded with just 4 mics (was just a test take, but erm...) - guitar amps the usual sm57 way - vocals with Beta57a - Samplitude ProX3 - VSTis: 3x u-he Hive 2, 1x Repro-1 - processing: many instances of u-he Presswerk, Uhbik-A, Samp's eFX VocalStrip, eFX Gate, eFX De-Esser, eFX Limiter, Samp's EQ, Multiband Stereo etc. Enjoy, like & share - and whoever uses Spotify, help us by putting us into your (public) playlist. https://open.spotify.com/album/2SfG6boTmSgyhSd5tnx4UG
  2. Hm, I doubt so. Not sure who at samdev has the will to wade through that code & do additions there. To be honest, developing the speaker sim in Vandal had been one the most time-consuming and tedious tasks in its development, with many incarnations and back&forth. It's basically a (nonlinear) filter bank, paired with two reverb models (one for the acoustic space, one for the enclosure), and it feeds back on itself. AFAIR it even feeds back into the amp, as to simulate damping factor, but not sure if I disabled that later on... the whole speaker sim is a beast, and the amount of grey hair I'm developing must have been because of that I guess - in hindsight - a more flexible choice would be to throw IRs at it, although I'd do multiples and sort of do it dynamically (although not sure if that would be subject to patent infringements), or combine it with the filter-bank approach somewhat. But, you see, it's still quite challenging to do when you're basically developing a DAW as the main project, and I don't see it justifying a man-year of a Dresden DSP guy on-the-side while potentially being needed for the core app.
  3. While you're at it, you may want to test drive it against the eFX limiter, which follows a similar concept. It's limiting and saturation in conjuntion. And it's got the same hard-vs-soft clip thing. Might be that the eFX limiter gives better results on difficult material, don't know. It's a newer design, at least I think I had some parts of its algorithm layed out very differently, like the adaptive release for instance. I do mostly indie rock, and it's still my go-to limiter, and - since I do listen to a lot of metal - I'm pretty sure it'll work great for harder stuff, too.
  4. Thing is, am-munition's output stage can push levels really high. I mean it. You can easily create loudness levels that surpass most other brickwall limiters by 3 to 4dB. But it comes at a price, and needs a cautious and wide-awake approach. Too soon and you've overcooked and burned a track. Added treble beyond a pleasent level can easily be the result of too much distortion/saturation induced. Although, when using the dual-band saturation and output-clip stage, you can adjust it a bit. If the saturation sliders are more towards 'soft' rather than 'clip', less upper harmonics will be produced, e.g. the harmonic series takes on a smoother decay towards the top end. This is easily visible in Samp's output meter, especially when using its 'true peak' (intersample peak) detection mode: the softer the saturation curve, the smoother the approach towards the ceiling will be, and less intersample overshoot. IMO hard clipping should only be tolerated if you aim for a general transparency up to a well-defined threshold. This often works with electronic/sequenced music as levels stay well-behaved and predictable. With acoustic genres it's often unnatural. Comparable to a compressor with soft vs. hard knee. One could also try balance the spectral behaviour with the dual-clip stage. To retain clarity and impact on the bass, allow for more hard clip in the low band (or like 50:50 or so), and set the upper band to all the way 'soft'. This will lead to a more 'wooly' character in the highs, gently smearing the top end, but retain bass punch. Similar to what tape does. You might have to play a bit with the crossover frequency, as when it's too high the hard clip from the lower band might create too much harmonics that bleed into the treble band. Of course, it all depends a bit on the incoming signal. The 'purer' the sources (especially the less odd-numbered harmonics), the more added saturation/distortion will change their character.
  5. This is weird, and, frankly, a bit absurd. Let me explain. 1) The OP, Johnny, utters he is faced with a 'cold' sound. That, IMO, needs explanation. As I'm more on the science side of things, let me ask: what is 'cold' exactly in here? Cold & warm are often attributed for brittle/bright vs. mellow/lowpassed, but everytime nonlinear effects come into the equation (thereby messing with the spectral balance), things become complicated very quickly. Nonlinearity greatly affects our perception, hence people are tricked by 'aural exciters' sounding 'warmer' although they're just distortion boxes. Phase is also an importand matter, yet more difficult to grasp. And, of course, signal envelope. The actual progression of how the amplitude is modulated over time by the process can be very different, yet impart a 'signature' sound for a specific device. Some designs have more 'attack' than others, depending on the shape of the envelope. For instance, feed-back designs are more 'intrusive' towards transients since the envelope is treated more exponentially because of signal roundtrip. Dynamics processing on complex material is more than signal-envelope manipulation. It needs skill, training, and quite an objective view the more a mix approaches a 'final' stage. Our hearing sense is so easy to deceive, and it easily adapts to our own manners, listening experience, preference, and preconceptions. Every cold-vs.-warm debate should take these aspects into account, in a search for the objective. Otherwise we'd be talking esotericism. So, audio snippets, frequency plots or anything else helping to clear up the 'issue' would help. 2) By the time I designed am-munition and layed out its basic concept, I started with a simple idea: create the typcial behaviour of opto-electric compressors. I once build my own HW one based on a simple LED & LDR setup based around a dual OP-amp setup. Was my go-to comp to track vocals and bass for the band I had at that time. I then took its behaviour directly to design ammo's opto section. The 2nd step was to create a brickwall limiter to take care of the transients that an opto design usually lets slip through. But: I'm a strong defender of transient information. I can't stand brickwall limiters that suck out transient information just for the sake of theoretical signal integrity. It's unmusical, in my view. Transients are good, and important for our perception. So the idea in ammo's limiter stage is to convert 'transient energy' into 'spectral energy'. What sounds esoteric here is just what our hearing sense does upon detecting and processing a loud, plosive signal. 'Linearity' is a man-made, artificial concept. Nonlinearity is what happens in nature, thus in our nerval systems as well. Therefore, the ingredient's found in the plugin should - by design - already help in achieving a 'natural' sound. Although, it's a beast, not easy to tame with its numerous parameters and capabilities. But again, give an example of what sounds 'cold' to you, and one can try counteract it. 2) A couple of seemingly alternative suggestions have been given here. Provided the original 'issue' is unknown, this is just pointless and only leads to endless circular debates of personal preference. I'd suggest to first sort things out, then discuss if and what alternatives to go for.
  6. I don't like editing much if I can do a better take. Generally only edit technical faults like double triggers or velocity issues. A lot on my kit is highly modded DIY (normal shells with mesh heads, real (perforated low-volume) cymbals with my own sensoring underneath etc., so it plays mostly like an acoustic) but sometimes there's some garbage in the data stream that is definitely not from my limbs. But the e-kit is only a temporary solution (or for bar gigs sometime), we'd go for an a-kit as soon as we find a right (permanent) space.
  7. Just wanted to share this, as we've just finished up our debut album and put it online. My band 'The Board' is into alternative rock & indie, we're Berlin-based (although I'm the only German guy, the others are from Cape Town & Wales). Just formed the band in 2017, wanted to write 3 or 4 songs, put them online to get gigs. That was the plan. In the end we had 10 tracks, so we went for an album. Done entirely in ProX 3, I played all drums through my E-kit into Addictive Drums, bass is processed through Vandal, guitars are just miked up as-is. Only used internal plugins, along with u-he Presswerk & Satin, and Voxengo Span. If you like it, let us know on FB, subscribe & spread the word (we could need it...) https://www.facebook.com/theboardmusic/ https://www.youtube.com/channel/UCGfmXGuJbW2KyK6JaQJnA8A
  8. Ob Tellerrandmusik das ursprüngliche Problem gelöst hat, weiß ich nicht (es gab ja keine Rückmeldung), aber vielleicht kann ich als der Ex-Entwickler etwas zur Namensverwirrung beitragen...: Die allererste VariVerb-Version hatten wir damals für den MusicMaker entwickelt (ohne 'Pro'-Zusatz, weil es halt nur 1 Version zu der Zeit gab). Wenn ich mich recht entsinne, müsste das um 2004 gewesen sein. Das war intern auch kein VST-Effekt, sondern für die eigene Schnittstelle im MuMa. Dazu zählt auch eine eher rudimentäre Oberfläche mit 5 oder 6 Reglern, die einzelnen Algorithmen wurden ausschließlich per Preset gewählt, und Effekte hatten immer dieses XY-Macro-Control. An Algorithmen gab es im VV1 nur Room/Hall/Plate & Spring, jedenfalls nicht die HQ-Algos. Die Room&Hall-Modelle hießen später im 'Pro'/VV2 'Retro'. Nicht wg. Version 1, sondern weil sie auf ein Design von John Dattoro zurückgehen, das auch in frühen Lexicon- & Ensoniq-Geräten verwendet wurde, allerdings beim VV etwas weiter geht und (wg. der potenziell verfügbaren DSP-Leistung) auch dichter und reichhaltiger klang. Für Samplitude wollten wir das Ding ein bißchen weiter aufbohren und entsprechend der Zielgruppe feiner parametrisiert anbieten. Die dortigen Room&Hall-Modelle waren dann komplett neu und verwendeten eine Delaymatrix, die keine Entsprechung mit anderen Designs hatte. Dann gesellten sich auch noch die HQ-Algorithmen dazu, die das ganze nochmal auf die Spitze trieben (64 Delaylines in getrennten Matrizen für links & rechts und mit freier Lokalisation). Das ganze war dann 'VariVerb Pro', und es wurde erst in Samplitude & Sequoia gebundelt (v9, glaube ich, 2005 oder 2006). Allerdings waren es technisch VST-Effekte. Samplitude verwendet (wie viele andere Windows-DAWs) den dll-Namen zur Identifizierung. Theretisch kann man auch die 4-stellige Plugin-ID nehmen, mit der sich ein VST am Host anmelden soll, aber leider hat es Steinberg nie wirklich hinbekommen, konsequent - wie in der Dokumentation zur API versprochen - eine herstellerübergreifende Datenbank zu verwalten (und auch konstant zugänglich zu halten...). In der freien Wildbahn gibt es eine Reihe von Produkten mit identischen IDs, weil jeder dann seine persönliche Glücks-Kombi wählte... kurzum: ein VST-Host tut gut daran, den Namen der Binärdatei (des Plugins) zu nehmen. Dies wird dann auch in der Projektstruktur (-> VIP) durchgereicht. Natürlich ist es dann nicht trivial, mal eben ein Produkt später umzubenennen. Etwas später (vielleicht 1 Jahr, weiß nicht mehr) kam dann auch die seperate VST-Version. 'VariVerb Pro' machte dann später auch keinen wirklichen Sinn, weil 'VariVerb' in der Kommunikation im MusicMaker nicht mehr groß auftauchte (eher nur 'Hall' oder 'Reverb'), und Pro-Kunden stutzten ob des 'Pro'-Zusatzes im Namen, wahrscheinlich auf der Suche nach dem 'Nicht-Pro', vielleicht um auf Nummer sicher zu gehen. Mit ProX hatte dann 'VariVerb II' seinen Einstand. Die Algorithmen waren grundsätzlich identisch, allerdings bekamen alle Räume den Modulationsparameter, und das GUI wurde komplett überholt (neuer Grafik-Unterbau, Grafiken mit Alphakanal etc.). Die DSP-Algorithmen waren abwärtskompatibel. Nur, umbenennen konnten wir das Ding nicht so einfach, denn das hätte bestehende VIPs zerschossen, und was draußen bei anderen Hosts in der seperaten VST-Variante passiert wäre, ließ sich schwer vorhersagen. Also, alles halb so wild mit den Namen, zumindest außerhalb des MusikMaker/MusicStudio ist immer das gleiche VariVerb zu finden, entweder als V1 (alte GUI mit den kleinen schwarzen Knöpfen und Room/Hall ohne Modulation) oder V-II (neue GUI, weiße Regler m. LED-Kranz u. Modulation bei Room/Hall). Das mit den Parameternamen auf dem GUI und den englischen Tooltips liegt übrigens daran, dass man in einem VST2-Effekt/-Synth nicht immer zielsicher bestimmen kann, unter welcher Sprache der Hosts aktuell gerade läuft, demzufolge sind die Ressourcen nicht lokalisiert. Es ist auch nicht 'Standard' in der VST-Welt, so dass wir von einem aufwändigen Lokalisations-Unterbau unter Zuhilfenahme von Betriebssystemfunktionen (an der DAW vorbei) abgesehen hatten.
  9. Ich hab ja gesagt, dass es Absicht war, zumindest experimentell. Wir hatten schlicht kein UI Control dafür vorgesehen, und der Default-Wert ist ziemlich ungünstig. Zumindest der sollte im Code auf 0 gesetzt werden. Dazu braucht es kein VIP mit Audiobeispiel, es sollte nur beizeiten jemand bei samdev diesen beknackten Default-Wert ändern. Da ich damals alleine für VV verantwortlich war, hat die Möglichkeit der Bitreduktion in den Retro-Algorithmen gar nicht groß die Runde gemacht. Bei bestehenden Presets kann man das auch selber in einem Texteditor erledigen, es sind xml-Dateien: (Preset-Ordner, z.B. C:\ProgramData\MAGIX\Samplitude Pro X3 Suite\MAGIX Plugins\VariVerb\Presets) <param id="num_bits" value="0" /> (Dieser Wert steht normalerweise auf 0.5) Vielleicht kennt ja jemand ein Tool, das sowas per Suchen+Ersetzen im Batch-Job auf allen Presets erledigt?
  10. Dithering hab ich da nicht in Betracht gezogen, es war auch eher experimentell und vor allem im Feedbackpfad des Hallmodells. Jedes zusätzliche Signal erhöht die Gesamtverstärkung und das System wäre nicht mehr stabil, das ist ach schon bei Interpolationsrauschen von modulierenden delay lines sehr tricky.
  11. Mir ist das kürzlich bei einer Albumproduktion selber passiert, ich ich hab versucht mich zu erinnern, was da damals getrieben hab... erst hatte ich die Allpass-Interpolation im Verdacht, aber dann fiel mir ein, dass ich irgendwann mal auf lineare umgestellt hatte (Delay-Modulation via Allpassinterpolation ist bei Chorüssen ok, aber bei Hall wg. der langen Delaylines großer Mist, weil die Koeffizienten zu weit auseinander liegen; irgendwelche anderen Effekte waren damals davon betroffen und ich dachte ich hab's hier verpennt...) Aber dann fiel's mir ein: ich hab da mal - eher testweise - eine Bitreduktion in den Feedbackpfad gebaut, um das Verhalten dem von alten Wandlern anzunähern. Kann sein, dass das standardmäßig auf 12 oder 14 bit steht. Im Plugin gibt es dazu einen versteckten Parameter. Hier ist der Workaround: - Im Plugin-Fenster den GUI-Modus umschalten: Plug-in->Parameterdialog - Parameter 33 auswählen ('Bits'), Wert auf 0% ziehen - Wieder auf Plugin-Dialog wechseln - Ggf. Preset neu abspeichern - Weitersagen! Mea culpa.
  12. That small red dot always belongs to the 'other' channel. You exchange knob & dot by selecting the channel to edit. This is not so intuitive with remote cc, but there was no way around it.
  13. Thanks, @Kraznet. For the UI, credits go to Basti (https://u-he.com/about/team/sebastian.html). A 3ds Max & V-Ray superhuman. He was the one who designed the majority of Samp plugins (AM/VE/Revolta/BB2/Vandal) back in his days at Magix.
  14. Which is still correct when you read it like a Roman: VIVII = 11
  15. Not new, we already released it in Dec. 2014. https://u-he.com/products/presswerk/releasenotes.html
  16. The 'Int:Ext' control balances the incoming source that the detector circuit 'sees'. Turned fully to Int only takes the plugin's own signal into account, while Ext will make it only listen to the sidechain input you select in the plugin dialog's menu of possible external inputs. 'Level' is indeed that simple, it controls the external sidechain volume.
  17. To make that clear, that was neither a port nor a succession of ideas. I hate repeating myself through projects, and it wouldn't do past work and people involved any favour. While I did both plugins' DSP code (and - hehe - Sebastian made both UIs...), the two only share some features like M/S and external SC input, along with a central metering section, where I considered M/S important, and the centrical layout was quite obvious for Basti when we layed out the concept (he even did a 3D model of the virtual 'inside', as can be seen here). But looking more closely probably reveals important differences. Ammo is basically an opto comp in series with a limiter & dual-band saturator, which already is quite a strange combination, while Presswerk is more a compressor construction kit designed to mimic pretty much everything known from the hardware world. PW is more versatile, and its only 'signature sound' comes from the saturation section and optional phase rotator, while Ammo has more flavour right away, but is also more aggressive & up-front. You just can't substitute one for the other.
  18. Just checked, but not handy at all. Sigh, it's overly crippled, only 8 tracks and doesn't allow 3rd-party VST plugins (not even VST2, unless I'm too dumb). I personally don't have a use for it. But maybe something for my daughter (who regularly uses the Android app).
  19. Oh, there we have something in common You've been an invaluable outpost for the company, IMO. Sometimes balls & upright attitude might clash in job life, I hope that wasn't the cause of decisions. Anyhow, please keep up contributing and spreading the word.
  20. Die SE-Version kannst Du nicht umwandeln, dem Binary fehlen entsprechende Teile der 'großen'. D.h. die Version einfach weghauen (evtl. eigene Presets vorher sichern) und die VST-Variante installieren. Natürlich können beide koexistieren, aber Sinn macht das nicht wirklich.
  21. Who's being naive? And about what? Your post makes no sense. Sascha in this case. Oh, I'm perfectly fine with that. Already 43, but still to young to play the old wise man...
  22. It might be an all-too-simple pop song, even mediocre in terms of writing and production. But it managed to really touch a lot of people all around the globe, made them dance and - in the purest sense of the song - feel happy. Making people feel great through one's music is - in my view - the biggest reward as a musician, and outshines everything else. Therefore, in this case, mission accomplished.
  23. Sorry to say, Seb, but to me that's bold and arrogant. Artistic quality is nothing to argue about in an objective way. It's in the eye of the beholder. You might seem bored or offended from a musically-educated point of view (I'm a rocker, therefore uneducated at best, but hey, so what...), but it's always been like this: music makes people happy, or sad, or whatever... but it does something with people. We choose music we like and what clicks with us. My 9-year old daughter likes bubblegum girl pop, which sounds awful to my ears, but it makes her happy. So what. (Well, she plays piano for 4 years now, has a Russian teacher and knows 'real' music, but apparently can't dance to it with her friends...) As a musician, I'm more than happy that equipment is so cheap to obtain. I grew up with 4-track cassette recorders as the only thing the band could afford to record on. We even didn't have enough mics, so I dissected old radio recorders for their electret mics and taped these to the drum kit and the guitar cabs. When we booked a studio (guessed it: to make 'demo tapes' for clubs to book us...) , I had to gather my last pocket money (40 Deutschmarks or so) to overdub my bass tracks for one (!) hour there. And finally we had to collect again for the guy to do a mixdown, on Tascam's MSR16 and a M-3500 desk, which was nothing to argue about at that time. Results were embarassing, of course, given today's standards. No, I definitely don't want these days back. Were we more talented then, having those technological barriers in front of us? Certainly not. Did those barriers stop the lamers (or the people we hated) from trying to make music? Not a bit. I love recording at home, I'm perfectly fine with that. Of course I'm not that fine with the fact that the guys and me have less opportunities to play live, as clubs vanish or show little interest in getting unknown bands up-stage. But that's a complete different issue, connected to other cultural and social aspects, let alone public or local finances.
  24. Well then, Sebastian, what is your solution?
  • Create New...