This article could have been an esoteric one but then it would probably be titled as “the magic is where the change happens” or something like that. Don’t worry, this is just about some findings and myth on audio transient processing and it’s, errm, reincarnation in the upcoming TesslaPRO VST audio plug-in.
When once finished the very first fully functional prototype of TesslaPRO and tested it on a whole mix I had that instant big smile on my face one rather see on some really rare occasions. And this is why.
Beside some other neat stuff TP features a new developed circuit which lets you control the amount of the applied effect related to the signals transient flow. There is a dedicated so called TRANSIENTS dial which determines how long the effect is somehow “delayed” when an audio transient occurs. Turning the knob straight to the left means all transient information gets processed. Turning the knob to 12 a. m. position means 7 miliseconds of the transient remains unprocessed and straight to the right even 30 ms is passing through untouched.
“Save our transients” – Bob Humid
Applying this to a whole mix or stereo submix or so can lead to some unexpected “side effects” beside that obvious effect that some amount of the transient information survives the saturation process:
- The audio signal appears to be more focused but natural and relaxed at the same time
- The audio signal appears to have some more stereo width content
- The audio signal appears to be more “3-D”
- You will have some big smile on your face
I was really amazed that there is this subtle but nice widening effect when relaxing the very first transient information from the saturation process (there is no M/S processing performed here!) but on the other side this observation is correlating to some physics and perceived effects in analog devices. The simplistic and basic waveshaping theory in the digital domain handles saturation as a static and instantaneous sort of effect while in some analog components such as transformers and tape devices this ain’t true at all.
Those analog devices are truly capable of capture and reproduce significant amounts of audio transients but just are limited energy wise (which briefly means the sum of audio level over some time). This way very short attacks are reproduced faithfully and just slightly polished while when constantly driven hard the well known distortion occurs. And that is a big part of all the magic which is linked that much to highend analog audio gear. Forget all the rumble about noise, crosstalk and all that stuff – the transient is the place where the magic is going to happen (or not).
So, after all those theory back to practise and implementation. TesslaPRO will feature some new and exciting approach on signal saturation modeling beside some other way cool stuff. One really important note on the implementation is that even though there is some dynamics processing introduced there will be no audible artifacts implied by intermodulation distortion which is a common problem in e.g. compressors. Using TesslaPRO one just will hear the saturation processor if it’s applied over the audio flow but no modulation artifacts by the dynamics!
The saturation which is actually applied is output level compensated and there is a much more accurate and faithfull transformer style bass distortion simulation which was already introduced in BootEQ mkII. The audio input can be boosted +12dB and the core algorithm is four times oversampled to minimze further aliasing artifacts. Apart of the mandatory VU display TesslaPRO renders a second (red) needle into it which gives some visual feedback on how much transient information is actually passed or not.
TesslaPRO is going to be released in May 2009 and will be available as freeware for PC and VST capable hosts.