Robert Henke wrote:Well, Timur,
we had long long internal discussions about this, since indeed a lot of energy went into the improvement of the audio engine. However, there was a huge pressure to improve things, even if I still believe the reason why people assumed Live sounded bad had nothing at all to do with the summing or the sample rate conversion. It was routed in the notion of Live as a "loop tool", it was routed in the artefacts created by timestrech and those factors. But, for the psychology of the users it is very important to know that the application we deliver is state of the art. It does not matter if this has any effect even on a single bit in a renderd file. If the changes in the audio engine give our users a better feeling and helps ending some of those previously endless discussions which mainly create *fear* amongst users, it will pay off.
I wasn't around when these discussions where happening, but I think I have read enough on this forum to get the point. I don't think many professional producer/musician shares this fear though, it's mostly the masses (of customers?) of amateurs. A friend of mine finished his last record with Live 6 alone (plus Mastering by another gifted hand with other tools), he finished the record before with Fruity Loops. The Live made record was reviewed by german magazine Keys (see below, sorry for the advertisement, but it's worth listening to his tracks anyway

) and you wont find any comment on substandard sound-quality. Actually my collaboration with him is the reason I have started to use Live at all. And the reason why I am so active on this forum atm is because I spend alot of time on learning my new digital tools including Live and try to learn as much from the experienced users on the forum as possible. Taking part in the 64 vs 32-bit discussion is partly out of curiosity and partly out of the desire to learn the facts (aka learn the tools).
Autoaggression pleases electronica lovers with a new release. "Artefacts" contains what the title promises: digital soundfragments, sample-grains and sharpedged rhythmnloops, which are put together with the precision of a clockmaker by chief-aggressor Lukas Schneider, until they come to life as very unique electronic compositions. The finely graven soundtextures call for concentrated listening - but only untill the third title pulls the surprised listener to the dancefloor. "Artefacts" refuses to be just a minimalistic Ambient album, but offers industrial fans, friends of glitch and even techno disciples music to enjoy. The connenction between those variing eleven tracks is the precise detail work by Lukas Schneider, who puts with "Artefacts" an exclamation mark behind the word "electro". (wus)
I am not so sure about samplerate-conversion, that can be a delicate matter and proper SRC eats lots of CPU (more than half of Creative's million-transistor X-Fi chip is only busy doing SRC). Concerning 64-bit vs. 32-bit it should be clear that the main practical benefit can only be seen when summing lots of tracks/effects. I do wonder however if the higher resolution aka more in-between steps shouldn't lead to less aliasing/quantisation noise in the audible range even with only a few signals when complex effects are used. Digital steppings can be tricky for something that is so much bound to "feeling right" as Audio/Music is. After resampling to 16-bit without dithering or even MP3 it wont matter anymore anyway I guess.
It is not that i cannot understand the motivation to aim for best possible sound quality, it is just that in my experience good sound has nothing to do at all with 32bit versus 64 bit. Good sound is possible with 12 bit and analog equipment from the 1950s.
I agree on the analog equipment and also agree on the 12 bit part as far as electronic music is concerned. It does become lots more difficult with acoustic instruments and vocals though. Not so much because of missing headroom, but because of really ugly artefacts (or nice artefacts depending on what sound you seek

).
The reason, why it became so difficult to achive good sound is the fact that we have so many tools. Instead of spending half a life practicing how to place a microphone best in front of an instrument we assume that some tool will do the work for us and that we need at least five EQs and three compressors in series to achive a good sound. The risk that we do something wrong here is incredibly high. If we had no compressors or EQs all we could do is listen, move the microphone and listening again. This is what's missing. So, after we all know now that Live 7 sounds okay, we could try to explore what we can do with it.
Well, most of your customers expect the tools (Live, Virtual Instruments, Effects, the digital realm in general) to be easy to handle, like owning a car that can drive mostly on its own as long as you don't try to win races with it. They don't realize that there is a reason to call audio-craftsmen "engineers", even in the digital age.
On the other hand that was my point with the last post, as you have noticed yourself alot of people on this forum seem to be quite happy with the overal quality that Live had delivered before 7, but are now very happy that they don't have to buy third-party Compressors and some other stuff. But especially those people who already knew of Live's quality and also knew "what we can do with it" would have prefered to see the energy spend on 64-bit summing to be put in workflow improvements.
Personally - I still own no licence - I think that I have come just at the right time to start working with Live. The change to 64-bit is right and fit, Live is upto-date as far as the numbers are concerned and since 64-bit floating-point computing does not come with any considerable drawbacks anymore we should use the best precision we can get (96 kHz sampling still poses a problem though, so its a welcome addition to have oversampling on plugin/effect basis instead of the whole signal path). The Drum Racks and External Instruments, good Compression and good Oversaturation are the real deal for me (most likely I will buy a VST Compressor or keep using an analog compressor for recording anyway). These add possibilities for adding external Gear (physical and virtual) and affecting external sound like no other Live version before. And that is
very nice for all of us who do more than just 4-to-the-floor electronic music.
Thanks for your elaborate answer!
