Gearslutz.com - View Single Post - Is there a sound difference in various DAW's ?
View Single Post
Old 11th September 2008   #58
Timur Born
Lives for gear
 
Timur Born's Avatar
 
Joined: Mar 2008
Posts: 1,429

Quote:
Originally Posted by Dean Roddey View Post
One issue in these discussions is what 'the DAW' is. 'The DAW' is really just the summing engine. It has nothing to do with plugins, which are completely separate things. If you limit yourself to just the DAW and its summing engine, then the differences are going to be very small.
I'd like to ignore the summing engine (except for one question later in my post) discussion and concentrate on the playback and plugin part of DAWs.

I am very confident that the quality of the playback and plugin-engine can make a very audible difference when comparing DAWs. And I do not mean different plugins/effects which are bundled with different DAWs, but using the same external plugins with different DAWs.

My main DAW is Ableton Live at the moment and I know quite alot about its many quirks. One of the most problematic issues that I am trying to have reproduced and acknowledged by Ableton for more than 8 months is about timing/latency-shift problems on both VST + AU and both PC and Mac. Depending on various factors such as song-tempo, looping-methods, number of times and method of starting song-parts and size of Plugin-Buffers (not Audio-Buffers) Ableton Live can shift the delay/latency of both audio-clips and external plugins' audio-return (or maybe their midi-input, the result is the same). I can make that audible both with and without using cancellation tests!

The resulting effects are varrying from having whole audio-clips shifted by a set amount of latency or having only the Attack phase affected while Hold, Decay and Sustain are not. Even worse depending on the size of the Plugin-Buffers external plugins will suffer from varrying timing offsets instead of fixed ones. As you can imagine this all leads to audible sound-modulation, in some cases grossly audible modulation (like quick succession of 1/16 notes sounding as if some "swing" beat has been applied to them). On top of that this does not only affect online-playback but also offline-rendering to disc.

These are weaknesses of the Ableton Live engine I know about, but I can well imagine that other DAWs have other similiar weaknesses within their engine that affects their sounds when compared to the competition. And yes, in principle these weaknesses and differences should not exist at all and must be considered "broken" parts. The difficult part is to identify those mostly very subtle flaws and afterwards getting the manufacturer to understand the issue and fix the engine.

Quote:
Different floating point engines will be basically the same for the most part, despite claims above.
When thinking about floating-point engines I wonder about one thing: What are they internally calibrated to and does it make a difference? I know that with Ableton Live 0 dB is internally represented by 1.0! So any track's gain/volume going higher than 0 dB is represented by floating-point numbers above 1.0 and anything below 0 dB is represented by floating-point numbers of 0.xyz quality.

As far as I know only a limited number of integer values can be represented exactly by floating-point. So does it matter what 0 dB is calibrated to when talking about rounding-errors and exact representation? I don't know, but I wonder wether this makes any difference?
Timur Born is offline   Reply With Quote