ilia wrote:to 3phase:
ok, I'm gonna try this.
I think most sane people will agree that as long we are talking about digital audio, same sequence of numbers sent to the same converters at the same sampling rate will produce the same sound (ok, during listening tests there will be changes in temperature, humidity, head position, and your hormonal balance, and you'll be a bit older during the second test, but it'd be pretty tough to hold Ableton accountable for that). So from here we can continue as follows:
1. you disagree with the above statement --> end of discussion
2. you agree --> then when you are hearing a difference in the way Live and another software are playing a particular file there must be some difference in what is being sent to your card's DA converters. You can then route the sound to your card's digital outs and record them digitally to another computer or DAT. This will bypass issues like rendering and platform/driver interaction and should give you the digital footprint of the output before it hits the converters. Compare these recordings (yes, phase cancellation).
a. If they are identical, then what you are hearing is purely psychological.
b. If they are different, then measure the relative volume of the residue. if it's below -130dB (compared to the original signal), than this might be the round-off error that Robert was talking about (although it shouldn't manifest itself when playing just one file). If you can hear that, you have very unusually sentisitive ears (or again, it's something psychological); this may or may not get fixed in the updates.
c. If the residue is significantly louder than that, then perhaps you stumbled upon something that Ableton need to take care of.
Doing these simple tests will likely convince most people on the board and push Ableton to do something about this if they show the difference. Otherwise it's mostly blowing hot air.
A/B listening tests are prone for psychological preconditioning...
I cant tell such listening tests a proove either and i wouldnt absolutely claim that a singel track test will show the problem i am talking about..
FAct is that i learned to dont mix in live...and even when V6 seems to be better..cant tell for sure because i stoped using live since the crapy version 4 and only switch it on ocasionally..
I imideatly get the audio issue in my ears when i import my audio files from logic. The same track gets this dullness... avoiding abletons mix bus and miss on master plugins by using the hammerfalls processor to mixdown the tracks helps a bit..but as you know.. plug ins in the master can do quite something for a track.
What is going on with my files pushes me back to logic...and i hate logic..and logic is far from sounding good in the mix.. but at least its better than the dullness i experiance in live...
Isnt it funny that so many people complain about a certain dullness and this over years ..and all you can answer is that theese people are trolls?
If there are 2 realitys here we might look for a reason and the only one i can imagine that it might be computer operating system related or missalignement in the drivers of very popular audio interfaces..
i dont know..but there are the only areas where such phenomens could be caused if we dont want to see Live itself as the cause.
Thats the job from ableton to look for that..
Regarding mr henkes post where he admits the rounding errors thru gainstructure as i expected..
I informed avleton 1 or 2 years ago about the intersting publikations of the company rane regardig such issues and asked them to look into that as a possible cause for the audio issue... They offered me a free update when i can proove my case and i send the link to the rane publikation instead.
I never got an answer on that mail but as it looks now it had some results..
mr henke claims that the 120 db range is out of interest because we cant hear that...true..but we well can feel it..
we talk about errors..and errors sum up in each mathematical operation..
You see this with the noise floor in mixing desks and related gain structure..
its easy to turn a ssl desk into a noise generator by having all channels up and some heavy gain in the preamps...
same happens to rounding errors of daw with not optimized gainstructures... bit depth structures.. internal jittering might be an issue.aso..
The rounding issue well might be the reason for the phenomen and wehn developments in this area are not backed up by listening tests it well might be from little fx... but as i said..when you can measuer it it will show up in the mix...there ayre things showing up in a mix that arent measurable allready... an ssl has better technical data than a neve desk...but nobody would claim the ssl sounding better..maybe cleaner... neutral...but the coloration of a neve are hard to measure... standart audio analyzers will show that the specs are very similar in booth brands..but the sound is totaly different... I am no expert in the digital domain but because we deal with audio we experiance similar results in little changes of the code as we experiance in the analoge domain by using another brand of capaciators...
little changes in a channel that sum up to a total different sound in the mix..
I dont see why coding a daw´s audio engine shouldnd be easier to optimize than a mixing desk... That all daw´s in fact sound different prooves that as a fact allready..its not so easy!
its complex and still very much in development for any digital audio company..
the arrogance ableton shows is just a marketing issue...
reading between the lines tells it all...there is something to improove..wether mr henke belives it or dont, that this will be hearable...
some people at ableton seam to belive that this changes will have an effect otherwise they wouldnt spend expensiv programmers time to implement the changes... As i said..that allready prooves the case well enough
Lives audioquality can be improoved..thats a matter of fact
And its good to know that they plan to do so even when they keep their politics of denying the problem...
Additional to that ableton should invest in possible performance differences in certain setups..