LIVE 9 : PDC IMPROVED OR NOT ?

[siliconarc]

here you go.... not an exotic problem at all, you can replicate it with a single plugin: https://www.youtube.com/watch?v=Tstw68U-24w

No one is arguing that automation isn't compensated

[theophilus]

Also all time based fx in generel, 3rd party or native are not very well compensated if I remember corectly.

this is the other issue... time based fx that have reference to the bar will be out of sync... i.e. if it's just using the BPM to set the delay time - fine. if you're trying to do trancegate effects, though, e.g. where the 1st beat is suppressed, that will be out of sync. this is arguably worse than the automation issue, because automation you can move manually; there really isn't any workaround for this one, though a lot of third party tools like LFOTool and VolumeShaper/FilterShaper will let you offset it there.

but all audio should be compensated... think someone found a bug with some of the Live devices not being compensated properly if they had delays (i.e. lookup in the compressor) but pretty much with only a couple exceptions, all audio anywhere should be compensated.

[eyeknow]

here you go.... not an exotic problem at all, you can replicate it with a single plugin: https://www.youtube.com/watch?v=Tstw68U-24w

No one is arguing that automation isn't compensated

Exactly. It's all cool and all that people are dedicated to live, but eventually you just have to face unmistakable issues.

[leisuremuffin]

I just think it's important for people to really understand what's happening.
I too hope that ableton adds compensation for automation and beat position effects at some point. (although understanding the problem allows me to work around it with only mild inconvenience) But i also want to make sure people understand that their audio is not being compromised by some PDC boogie man and the issue isn't going to create "muddy phasey mixes".

[pencilrocket]

Live is more likely to get muddy phasey mixes. Why? Because there are two obvious problems. 1. Automation isn't compensated. 2. Live has odd behavior that adds up buffer delay every time you insert vst. No other host does this. 2nd issue aggravates 1st issue more than the other hosts.

That's what people are complaining.

[theophilus]

2. Live has odd behavior that adds up buffer delay every time you insert vst. No other host does this.

umm... it *better* add up buffer delay every time you insert a VST... otherwise, the *audio* will be out of sync, not just the automation. however, from what i have heard, some other hosts don't adjust while playing - they don't recalculate PDC until you stop. live may be unique in that it attempts to keep things sample-synced while it's playing... either way makes sense, there's no way to insert a latent effect and not get at least a hiccup somewhere, so stopping is not a bad requirement if you're doing studio work.

i think what you are really referring to is that the automation delay is dependent on the overall latency, so as you add and remove effects, the effect of the automation latency can change. that is true, and like you said, makes the pain much worse if you're in that situation...

[leisuremuffin]

I don't see how automation being delayed would create phase problems...

[pencilrocket]

umm... it *better* add up buffer delay every time you insert a VST... otherwise, the *audio* will be out of sync,

All hosts add plugin latency to compensate.

[Jakob / Cableguys]

Also all time based fx in generel, 3rd party or native are not very well compensated if I remember corectly.

this is the other issue... time based fx that have reference to the bar will be out of sync... i.e. if it's just using the BPM to set the delay time - fine. if you're trying to do trancegate effects, though, e.g. where the 1st beat is suppressed, that will be out of sync. this is arguably worse than the automation issue, because automation you can move manually; there really isn't any workaround for this one, though a lot of third party tools like LFOTool and VolumeShaper/FilterShaper will let you offset it there.

But it's even worse..: If you manually offset the waveforms, it can get out of time as soon as you add other plugins or channels, our change routings.

But: Interestingly enough, if you trigger our plugins (VolumeShaper/FilterShaper/PanCake) via MIDI in Ableton Live, it's perfectly sample-accurate in sync and will remain in sync in all cases. It's a bit of extra work to set up, but you're then save in all cases. For more info, see our FAQ page and search for "PDC".

[ze2be]

Thanks Jakob!

From Cableguys FAQ:

Q: FilterShaper or VolumeShaper work fine in Ableton Live when being the only plugin on a track. But they can show noticeable latency in other cases — this can (but does not have to) happen when there are several plugins on the track, or when used on a group or on an audio track with other tracks routed to it.

A: This is a PDC (Plugin Delay Compensation) issue with Ableton Live. FilterShaper and VolumeShaper are synced to the timing information that is provided by the host. Sadly, Live does not take PDC into account here and thus the timing information provided by Live can be wrong. We've already talked about this with Ableton in 2011 and are glad about anyone who bugs them about this issue.

[kayhel]

I don't know about that but I think working with consistent settings start to finish is the way to go. And I also think it's best to use the highest sample rate that you can handle. Not for fidelity but for lower latency. 512 samples is obviously faster at a higher rate.

Hmm, these may be good practice in live for you. But its not good practice for Music Production generally. I don't want to open another this DAW that DAW thing here, but in S1 changing buffer has no side-effect to the audio being produced. Also -for me- setting higher samplerates in live requires much higher buffers to avoid artifacts, i guess because it consumes more CPU - ending up in more delay over all. Because of that it also doesn't scale easily to bigger projects.

A word to bigwig: I know it is not perfect, there are still many features i miss - i.e. rewire to be able to integrate it into S1 for easy submix (and ARA usage...). But the PDC discussion between users and devs in KVR-forum was exemplary and ended up in a very fast solution after days, not years. Still, PDC is not fully done, but if it would be implemented in live as it actually is in Bitwig, i would be the happiest live user. What i miss ist that live is simply ignoring this - there's not even a discussion possible with the live guys regarding PDC.

And come-on... Timing is _the_ crucial thing in music production these days. If we are talking about best practices, my suggestion is to talk about improving the root cause of these never ending threads instead of talking about workarounds. Because: none of these so called yet-another-best-practice-workarounds did work out for me, they all ended up in having other side-effects. And it looks that I am not alone here...

[kayhel]

Live is more likely to get muddy phasey mixes. Why? Because there are two obvious problems. 1. Automation isn't compensated. 2. Live has odd behavior that adds up buffer delay every time you insert vst. No other host does this. 2nd issue aggravates 1st issue more than the other hosts.

That's what people are complaining.

1. All PDC Algorithms need to add delay on other tracks to compensate for a new plugin that has a plugin delay.
Nothing wrong with that.

2. Phasing problems related to PDC in live are not causing typical phaser sound artifacts but have the LFO FXs out of sync - which is the same like an LFO phase adjustment (i.e. filter sweeps starting wrong, but this sounds different from phasing)

[leisuremuffin]

I don't know about that but I think working with consistent settings start to finish is the way to go. And I also think it's best to use the highest sample rate that you can handle. Not for fidelity but for lower latency. 512 samples is obviously faster at a higher rate.

Hmm, these may be good practice in live for you. But its not good practice for Music Production generally. I don't want to open another this DAW that DAW thing here, but in S1 changing buffer has no side-effect to the audio being produced. Also -for me- setting higher samplerates in live requires much higher buffers to avoid artifacts, i guess because it consumes more CPU - ending up in more delay over all. Because of that it also doesn't scale easily to bigger projects.

A word to bigwig: I know it is not perfect, there are still many features i miss - i.e. rewire to be able to integrate it into S1 for easy submix (and ARA usage...). But the PDC discussion between users and devs in KVR-forum was exemplary and ended up in a very fast solution after days, not years. Still, PDC is not fully done, but if it would be implemented in live as it actually is in Bitwig, i would be the happiest live user. What i miss ist that live is simply ignoring this - there's not even a discussion possible with the live guys regarding PDC.

And come-on... Timing is _the_ crucial thing in music production these days. If we are talking about best practices, my suggestion is to talk about improving the root cause of these never ending threads instead of talking about workarounds. Because: none of these so called yet-another-best-practice-workarounds did work out for me, they all ended up in having other side-effects. And it looks that I am not alone here...

changing buffer settings in live *doesn't* change the audio in live, but it will change the latency and therefore where the automation and beat position effects are. If you are having issues with that stuff, changing the buffer size mid project is the last thing you want to do. I work in live a lot. pretty much every day. I've found that for me, 88.2 24bit 1024 size buffer works on my largest projects. So those are the settings i use start to finish. it will be different for different users with different computers, audio interfaces, etc...

That's nice that bitwig was able to get this working for their users, but i had a look at bitwig, and in it's current state, it's not really a replacement for live for me. But if it had been better for me in any way, i would have cheerfully switched.

Timing is crucial, yes. And since i know how live works, i have literally ZERO issues with this. I'm not interested in waiting for the next feature to be implemented (or having the next piece of gear) i'm interested in writing music right now with what i have. And to anyone else out there that wants any level of success or just plain satisfaction I suggest the same mindset.

[kayhel]

I've found that for me, 88.2 24bit 1024 size buffer works on my largest projects

Thanks leisuremuffin, I'll give it a try and report back soon.
tbh I have to say that I didn't use live often after discovering this issue. I do not perform live at all. It's really been frustrating for me to do songwriting in live without timing accurate automation. Sometimes I am playing around in live and then I record it immediately into S1 via rewire. I'm producing in S1 now which has perfect timing, but i really miss the session view workflow of live, especially when using different loop lengths This is a no go for all these old-fashioned linear DAWs.
However when recording vocals, ARA integration with Melodyne Editur is a perfect thing in S1.

Lets see what future brings. And a bit of competition may have positive influence on the "big ones", so I bought Bitwig as well and while it's not ready for production now, I have to say that there's a lot of potential and the workflow is really perfect.

But I will not invest into a product that does not respect timing in a way that I do personally expect.
So, I always buy S1 updates
I'll buy Bigwig updates
But no money for Live updates til PDC is working fine - still stay on live 8, which is by far enough for just getting song ideas.

cheers
Kay

[leisuremuffin]

just to be clear, i'm not advocating using the same settings as me, i'm just saying pick settings that work and don't change them.