Live 9 NEEDs to fix SRC algorithms

Please join the new beta forum at https://ableton.centercode.com/login.html
Locked
djcl.ear
Posts: 23
Joined: Fri Jul 02, 2010 5:12 pm

Live 9 NEEDs to fix SRC algorithms

Post by djcl.ear » Sat Sep 07, 2013 11:21 am

Live 9 does deviate the sound output of SRCs (Sample Rate Conversions) in unacceptable levels as it is.
Check http://src.infinitewave.ca/ precise measures comparing Live 9 side by side with most other DAWs and most importantly, comparing it with the expected behavior of an ideal filter, at least in terms of phase.
Live 9 measured error is too gross as it stands now.

violetpow1
Posts: 107
Joined: Sat Jan 19, 2008 10:35 pm
Location: Germany

Re: Live 9 NEEDs to fix SRC algorithms

Post by violetpow1 » Sat Sep 07, 2013 11:47 am

Live 9 is quite good in comparsion with other big DAWs like Logic 9, Cubase 6, Wavelab 6, Sadie 6, Samplitute 11 if you think about that Live 9 is made for fast realtime work.
When you compare it with Audition or Protools, i guess you have never worked with it. The realtime functionality are pain in the as.

But in the linear phase filter category is Live 9 very distressing.

djcl.ear
Posts: 23
Joined: Fri Jul 02, 2010 5:12 pm

Re: Live 9 NEEDs to fix SRC algorithms

Post by djcl.ear » Sat Sep 07, 2013 6:31 pm

violetpow1 wrote:Live 9 is quite good in comparsion with other big DAWs like Logic 9, Cubase 6, Wavelab 6, Sadie 6, Samplitute 11 .
Live does compare well with some DAWS, but not so favorable if we go to minute levels (over -110 dB), and compare it, for instance with Izotope in some other (minute again) areas.

violetpow1 wrote:if you think about that Live 9 is made for fast realtime work.
When you compare it with Audition or Protools, i guess you have never worked with it. The realtime functionality are pain in the as.
But in the linear phase filter category is Live 9 very distressing.
I am far from questioning Live functionality or ease of use. You may go up and read the title and content of the post that originates this thread, and the one you are replying to: you will see that there is no comment on that, nor a critique of the software itself except for its SRC implementation with the aim to bring this issue to the users and developers attention, in hope of a rapid solution.
Peace.

djcl.ear
Posts: 23
Joined: Fri Jul 02, 2010 5:12 pm

Re: Live 9 NEEDs to fix SRC algorithms

Post by djcl.ear » Mon Sep 09, 2013 4:26 am

Well, Angstrom at the Feature Wish forum has shared that the incoming new Live version 9.1 does in fact replace the SRC algorithms for brand new ones, and he cites this sources:

http://sox.sourceforge.net/SoX/Resampling

http://createdigitalmusic.com/2013/09/l ... more-30457

Which mostly for sure not only indicates that the SRC phase issue was permanently fixed, but some other improvements could be already included in 9.1, which -in my view- is simply great.

//Any improvement in the quality of Live basic algorithms will have and additive effect in whatever other element we utilize while creating, recording, processing music when we utilize Live as the basis of our DAW.
Thank you all involved.

locojohn
Posts: 487
Joined: Mon Feb 19, 2007 8:42 am
Location: riga, latvia

Re: Live 9 NEEDs to fix SRC algorithms

Post by locojohn » Sat Sep 14, 2013 8:32 am

Hi,

As far as I understood, the mentioned SoX algorithms will only be employed for disk rendering, but not realtime rendering. It's could be valid assumption considering that for better quality resampling you would need to spend more CPU cycles resulting in higher latency, and for realtime audio processing increase in latency isn't a welcome addition.

Andrejs
/*
  • the basic tone of life remains the same,
    and in it there are some happy melodies
    and some sad melodies
    - sekito kisen
*/

djcl.ear
Posts: 23
Joined: Fri Jul 02, 2010 5:12 pm

Re: Live 9 NEEDs to fix SRC algorithms

Post by djcl.ear » Sat Sep 14, 2013 1:59 pm

Many different mixing, editing and processing software provide several algorithm options within the same GUI menu. Live could offer the same versatility, so we -as users- could decide whether low latency or better sound quality is preferred.

Live is used both as a traditional DAW and as the center tool for live performances, clearly having to excel at both. However this is facilitated by the fact that both usages are, almost always, needed at different instances.

djcl.ear
Posts: 23
Joined: Fri Jul 02, 2010 5:12 pm

Re: Live 9 NEEDs to fix SRC algorithms

Post by djcl.ear » Sat Sep 14, 2013 2:48 pm

Actually, now that I think about it this way....
Live could come out as the first to actually make explicit (and useful) the design differences developers have been implementing in the different DAWs.algorithms. Bear in mind that the SRC measures cited above not only show differences in the amount of accompanying distortion or spectral re-images of the filters, but differences in the milliseconds the start of the slopes are applied or the length of decays, etc. which might very well be the result of design rather than lack of.

If you compare the several DAW´s algorithm's response in terms of how a pass-band filter behaves, you will notice that several produce slopes with differing exponential tails. A few are very sharp and squared angled (notably Izotope's ones) and others show a slower tail decay.
And these algorithm differences could be taken as an additional feature. Let me exemplify; If I am producing a track layer and I want to divide it -via a pass-band filter- in order to separately process the lower freqs and latter on reunite both high and low freq content. In this case, I need the summing point to be as precise as possible, hence the square angled linear cut filter is best.
If, on the other hand, I am applying the pass-band filter to eliminate certain freq band from the track, the resulting track will latter be better mixed if if has a slope (even if minute), particularly in live situations..

Therefore one could organize choices by grouping a set of algorithms by several categories, by example: surgical, for live mix, fast, forgiving, and the like.

//On third thought, this all could be over the board and unnecessary. Those Algorithm differences have kept me coming back to them lately... This is not the 100% clear cut (selfpun intended) world I imagined digital DAWs would be...

Locked