DP9 set to get an update, claiming 4X CPU performance...

[Dewdman42]

You've been getting 1.4ms round trip latency for quite some time? Lucky you. Not me. And on my system logic has worse latency then DP for whatever it's worth. Not by much. Logic also routinely drops audio and pops up an annoying dialog box, even with larger buffers. DP has been much smoother operation that way. I like Logic for other reasons.

I don't think it's accurate to say that DP had double the latency of everyone else and now they are finally catching up by cutting theirs in half. Cutting latency in half is a big deal. It's a big bold statement. I personally don't think that 1.4ms round trip latency at 32 samples is common place as you are making it sound. We shall see when it's released i guess.

[ObsoleteAcc99]

[DELETED]

[machinesworking]

Regarding the latency, I have to correct you again. Lowering the cpu use doesn't automatically lower latency. It just frees up some cpu time and the latency remains fixed to whatever buffer size is currently configured.

You're not really correcting me, you're just being a pedant about it, that's OK, but I've always said faster CPUs can handle lower latencies, not that a faster CPU changes the buffer setting in your DAW.

Ok great but what motu is saying as the other new feature is that the latency of 128 samples is now the same as was previously the latency for 64 samples, which is about 3.5ms one way. They did a test at 32 samples with 1.6ms latency! If you were previously using 256 with around 16ms one way latency and it drops to 128 samples with the new buffer performance at 4ms one way latency, that is 4x reduction and I absolutely think you will notice!

OK so I get what you're saying, they didn't word it well, but you being there while they presented it clears it up. So they're not claiming it works as in you can now lower your buffer setting, they're saying the same buffer setting will be cut in half.

That tells me that DP was doing something other daw's didn't previously do which made the realtime latency higher than normal. They are talking about equalling other daw's here. That sort of RTL has been available in Logic and PT for many years with 96khz at 32 samples.

Are you certain you're looking at round trip latency? I don't recall Logic stating round trip, just the input latency.

For example Live at 96khz 32 sample buffer reports input latency at 1.47ms, output at 2.33 with a round trip of 3.8ms.

Remember dewdman42 mentioned a 256 buffer getting 4MS, at 256buffer and 44.1khz Live reports input of 7.56ms and round trip of 15.5ms!

[machinesworking]

Motu's drivers don't go that low in RTL. So either they mean they are doing something so DP Can interface with motu hardware at half the latency or something else.

They state it's not a matter of their hardware, any hardware will see the improved latency in DP.

[Dewdman42]

Time will tell when it's released. I will test the DP9 demo when the update is out, and if my apollo through DP (bypassing apollo console in other words) has half the RTL of live/logic/PT/Cubase, then that's something to talk about.

To be clear, their announcement is that the upcoming update will have half the latency of the current DP9 product. They did not make any claims one way or the other against other DAW's only to say that they believe DP will be as good or better then other DAW's. They did not make the claim that the next update will have half the latency of all other DAW's. So its quite possible DP has had inferior latency performance up until now, I doubt double everyone else though.

Regarding my system, its less then 5 years old, plenty of juice to the machine, it routinely throws up that stupid Logic dialog box even with the CPU only at 50%, with certain plugins. I never have that problem with DP, ever. Just sayin'.

But then again, maybe now that MOTU is cutting the latency in half some how, I may start to get some of that Logic-itis in DP also...we shall see...

[Dewdman42]

Motu's drivers don't go that low in RTL. So either they mean they are doing something so DP Can interface with motu hardware at half the latency or something else.

They state it's not a matter of their hardware, any hardware will see the improved latency in DP.

Correct, MOTU's update is not in the driver, its DP. Whatever improvement they are making should probably effect all core audio device performance.

Yes latency is a direct result of the buffer setting. More efficient CPU will not lower the latency. The feature update, which is separate from PreGen, is that the latency for a given buffer size has been cut in half. Take it as you may...

Any improvements to CPU performance will simply give some more room to lower the buffer even more if you so desire, but that is not what MOTU announced, they announced cutting the latency in half as an actual enhancement to the software.

[machinesworking]

Time will tell when it's released. I will test the DP9 demo when the update is out, and if my apollo through DP (bypassing apollo console in other words) has half the RTL of live/logic/PT/Cubase, then that's something to talk about.

To be clear, their announcement is that the upcoming update will have half the latency of the current DP9 product. They did not make any claims one way or the other against other DAW's only to say that they believe DP will be as good or better then other DAW's. They did not make the claim that the next update will have half the latency of all other DAW's. So its quite possible DP has had inferior latency performance up until now, I doubt double everyone else though.

Regarding my system, its less then 5 years old, plenty of juice to the machine, it routinely throws up that stupid Logic dialog box even with the CPU only at 50%, with certain plugins. I never have that problem with DP, ever. Just sayin'.

But then again, maybe now that MOTU is cutting the latency in half some how, I may start to get some of that Logic-itis in DP also...we shall see...

One thing that's going to be a PITA about this is DP doesn't state it's round trip latency like Ableton Live does, it just states buffer settings and "work priority". So to test it you would have to run a hardware sequencer into DP. Luckily the Beatstep Pro is sitting to my left.
From my own experience with Live I haven't found that DP has more latency at high buffer setting than Live. DP can also as it stands go lower than Live without issues. I'm running the current project at 64 in DP, and I pretty much can't go bellow 128 in Live without it being a problem while composing or mix down.

[machinesworking]

The first update is out now, the smaller one with bugfixes and track selection features.

[Dewdman42]

Frankly I don't know why everyone hasn't been doing it this way all along.

Wellllllll, did some more researching. Turns out my original hunch about why haven't they always been doing this, was closer to the mark then I realized.

They have been!

Since 2008 in fact, DP6 until now has had pre-gen technology already, and by default turned on. whenever a plugin window is closed and the track is not record enabled, its supposed to be using pre-rendered version of it...since DP6 (2008).

So when MOTU calls this NextGen PreGen...they mean this is the next generation of their already existing technology.

That being said, I did a quick test with DP9.02 (FirstGenPreGen) to see if the FirstGenPreGen was any good. I instantiated 10 tracks of Omnisphere2. I played back the ten tracks with PreGen enabled, and without it enabled. No plugin windows open, no record enabled tracks. Zero difference in CPU use that I could detect.

This brings us back to their product announcement. They say This is the NextGen PreGen. Ok.. hopefully that means better in some way. They don't really say how its better because what they describe that its supposed to do sounds a lot like what the FirstGen was supposed to be able to do. They do say that it enables 10x as many instruments tracks compared to real time. I don't know what the FirstGen was supposed to be able to do..maybe it never worked that well and now they are finally making it work better?

Let's hope it works better then the FirstGen. I can't really tell if the FirstGen is really doing anything... it certainly did not reduce my CPU use. Its possible that if I tried to max out the machine to see how many instrument tracks I can get going I might find one mode can do more tracks then the other...but I don't have time to do that test. In simple case with a few instrument tracks, I didn't notice any less CPU use either way. Its also possible that Omnisphere2 is somehow not enabling pre-rendering to function...don't know if that will be any better with NextGen, etc.. lots of variables there.

In short, what I can say is that this new feature is not as new as originally thought...DP has had it already since 2008. As I said earlier, I honestly would have thought that everyone would be trying to do something like this a long time ago.

Hopefully they are announcing it as "NextGen" because its substantially superior in some way that we shall find out in a few months. In light of this information, the formal announcement from MOTU doesn't really tell us what they've done, sounds like it will be functionally the same in terms of usability... Hopefully more efficient and more seamless in some ways. We'll see....

[Dewdman42]

see this article from 2008: https://www.soundonsound.com/sos/dec08/ ... rkshop.htm and search for the word "rendering" to read all about it.

[machinesworking]

Are you new to DP? Just kinda shocked, I've mentioned this in replies to you at least twice, that DP already had pre-rendering. We all skip by some things in a big DAW like DP for sure.

It does do some good. From what I could tell doing 'to failure' tests, DP would initially get to about 60% CPU quick and hang there for nearly three times more plug ins; so initially it doesn't look good, but after you start really loading it on it seems to prevent major issues from happening. Nothing like what they're promising with the NextGen PreGen thing though, mostly it meant at the time of testing, that DP could keep up with Logic, although Logic looked a lot better (CPU reading wise) under a lower load.

[Dewdman42]

I did a test just now between DP9.02 and Logic 10.2.1.. The test used 7 instances of Omnisphere2 using the same midi file, same buffer settings, etc.. everything as much the same as possible.

DP9.02 played back those 7 instances in real time mode using 27% cpu. In pre gen mode, it also used 27%. Exactly the same. I froze the tracks and it used 8% to play it.

Logic 10.2.1 used 22% to play real time, 5% less then DP. Frozen it also used 8%.

So what we can generally say is that Logic 10.2.1 is a bit more efficient at playing back unfrozen instrument tracks then DP9.02.. Whether they are doing some of their own pre-rendering or not is unknown. But I can say that DP9.02 showed no difference whatsoever between playing them back pre-gen'd or not. And I can say that DP9.02's so called pre-rendering doesn't seem to be rendering, at least in this test with Omnisphere, and certainly its not even close to frozen track performance.

MOTU Feels their NextGen version is better and made some big claims about 10x compared to real time...which would certainly be better then what I just experienced. We shall see. It will be interesting to see how it compares to Logic after the update as well. We'll see in a few months.

[Dewdman42]

and as you say, perhaps I need to setup a bigger load test to really find out whether DP's existing pre-gen works. Here I saw absolutely zero difference with 7 instrument tracks of Omnisphere2 playing from a midi track

And no I'm not new to DP at all. I just never paid attention to this issue. My presumption would have been previously that of course they are doing something like that. I would have assumed everyone was actually, that is the whole point I made in my first post on this thread. When they announced this NextGenPreGen, my first reaction was along the lines of "you mean it didn't already do that?" Sounds like they had tried to do it, but near as I can tell it didn't work very well before, and now maybe they are making it work better.

Its hard to say what Logic is doing during playback to get 5% better performance then DP. I kind of doubt they are fully pre-rendering, or else performance would be a lot more significant then just 5% better then the real time performance I got with DP just now. But something they are doing is 5% more efficient, with 7 tracks of Omni anyway.

[machinesworking]

and as you say, perhaps I need to setup a bigger load test to really find out whether DP's existing pre-gen works. Here I saw absolutely zero difference with 7 instrument tracks of Omnisphere2 playing from a midi track

Yeah that's why I never do tests that aren't to failure. Every DAW seems to have different curve in their CPU meter. Live craps out at 80% anything above that and crackles. Older versions of DP post pre rendering anyway would load to 80% then take in almost as many plug ins to get to 95%, i.e. you load 10 instances it gets to 80, then 10 more it's at 95.
I'm assuming you mean I know that a 'to failure test' is where you load multiple versions of a heavy plug in with the same MIDI file in the timeline until it crakles then back off until the audio sounds good.

Its hard to say what Logic is doing during playback to get 5% better performance then DP. I kind of doubt they are fully pre-rendering, or else performance would be a lot more significant then just 5% better then the real time performance I got with DP just now. But something they are doing is 5% more efficient, with 7 tracks of Omni anyway.

The consensus was that Logic can dynamically set buffers so that only record enabled tracks are running at 128 say, every non record enabled track is running at twice or higher buffer settings. This was mentioned by Ableton developers and plug in developers during a conversation about why Logic completely skunks a real time performance DAW like Live. (it's almost twice as efficient)

[machinesworking]

and as you say, perhaps I need to setup a bigger load test to really find out whether DP's existing pre-gen works. Here I saw absolutely zero difference with 7 instrument tracks of Omnisphere2 playing from a midi track

Yeah that's why I never do tests that aren't to failure. Every DAW seems to have different curve in their CPU meter. Live craps out at 80% anything above that and crackles. Older versions of DP post pre rendering anyway would load to 80% then take in almost as many plug ins to get to 95%, i.e. you load 10 instances it gets to 80, then 10 more it's at 95.
I'm assuming you mean that a 'to failure test' is where you load multiple versions of a heavy plug in with the same MIDI file in the timeline until it crakles then back off until the audio sounds good.

Its hard to say what Logic is doing during playback to get 5% better performance then DP. I kind of doubt they are fully pre-rendering, or else performance would be a lot more significant then just 5% better then the real time performance I got with DP just now. But something they are doing is 5% more efficient, with 7 tracks of Omni anyway.

The consensus was that Logic can dynamically set buffers so that only record enabled tracks are running at 128 say, every non record enabled track is running at twice or higher buffer settings. This was mentioned by Ableton developers and plug in developers during a conversation about why Logic completely skunks a real time performance DAW like Live. (it's almost twice as efficient)