How is it?

Great, and it has full PDC for my PoCo and UAD-1

Here's a recent thread on the same subject:
OSXAudio

i thought the delay compensation was fine for POCO in SX2? mine seems to be? well, used as inserts, certainly

I've got SX3 now, and I'm having serious problems getting external plug-ins to work properly with it on my dual 2.5 g5 and MOTU 828mk2.

Using ADAT connections to EMU is fine, it works perfectly.
Using an analogue send-return on the MOTU both to a Compounder and an RNC I get unusable delay resulting in serious flamming, even with the delay set at various settings up to 100ms and with sample buffer sizes up to 1024 - the latency becomes terrible but the flamming stays put. I also get some nasty humming when I don't send any signal for a while.

I went back to SX2 and set up an "insert" by routing an output track directly to the Compounder as described in MW's article about integrating hardware FX in March SOS, and this worked perfectly well - when I set the sample buffer size to 512, the delay was not noticeable (by me anyway), and there was no hum when I turned the gain up a bit and left it there. Does that sound like a reasonable test to ensure that it's not my setup?

Anyone else tried this? I've not been able to find anything similar on the cubase forums, except a big list of bugs they want us to test the fixes for when the "maintenance" release comes out in November ...

hmmm..thats scared me off.... my SX3 is sitting here untouched because i am TOO afraid...keep me posted , mr igor

OK I put my hands up: I haven't got a clue what's going on with that, maybe I'm getting direct monitoring when it seems OK. I'm not sure what's happening because the "Direct Monitoring" button in the VST Multitrack settings is greyed out and appears switched off, but I think that it must be on 'cos I tried the external plugin thing again after a couple of reboots, switching the MOTU on and off as well, with 64 sample buffer size and it was smoother, less flam, but still not great. I think I've not understood the way that SX interacts with the direct monitoring on the MOTU.

This is all too boring and complicated, I just want to get on with the tunes! Ho hum.

What I am fairly sure of is that SX3 has got lots of bugs in it, including all the ones on the bug report threads on the cubase.net Cubase Info Pages forum, and it won't work with the MOTU properly to sync to the EMU's ADAT clock, which SX2 does quite happily. But I'm not a happy ugrader right now - hopefully they'll fix this all up in the November release and then I will be happy.

I'll shut up now.

Hey nimhbus, crossed in the post.

Well, I'm not super-impressed at the moment, as you can tell. Maybe I'm being a div about some of this, it wouldn't be the first time, but there's definitely other stuff not right in this. I'd hold back if I were you. Maybe I'll learn (yeah, right).

What I/O do you use with your Mac? Is the "Direct monitoring" checkbox in Device Setup -> VST Multitrack enabled? Have you tried switching multi processing off? The CPU load seems to drop considerably when I do that in SX3 which seems like a bug ...

Igor

> Using an analogue send-return on the MOTU both to a Compounder and an RNC I get unusable delay resulting in serious flamming, even with the delay set at various settings up to 100ms and with sample buffer sizes up to 1024

> I went back to SX2 and set up an "insert" by routing an output track directly to the Compounder as described in MW's article about integrating hardware FX in March SOS, and this worked perfectly well - when I set the sample buffer size to 512, the delay was not noticeable (by me anyway), and there was no hum when I turned the gain up a bit and left it there.

Igor your description is a little confusing. Do you mean under SX3 you patched your compounder on an aux send, while under SX2 you used it as an audio insert?

If you're using it as an aux send, you will get flamming/phase shifts unless you get the latency setting (for the insert) exactly right, as you're hearing back both the direct signal from the track, and the signal from the compressor. You wouldn't normally use dynamics based effects on an aux send, it's (usually) best to patch them in (via an insert) on the channel that needs them. If you do want to work this way, there are ways to calculate the exact latency of your convertors/outboard.

Or do you mean you tried them as inserts on both SX2 AND SX3 and only got problems on 3 with the new hardware insert feature?

Also, if you're using live inputs, remember SX can't compensate for latency in this situation. It needs the audio recorded before PDC becomes active, and if you're using direct monitoring and somehow also hearing your signal back from SX, you'll be seriously confused!

Edited by kris (15/10/04 11:05 PM)

Hi Kris,

Quote:

---

your description is a little confusing

---

I can imagine - I think it's confusing 'cos I'm confused!

Also I didn't explain properly - on SX3 I tried to set the Compounder up as an "external pIugin" which I used as an insert on a drum track. I wasn't using it as a send.

Quote:

---

Or do you mean you tried them as inserts on both SX2 AND SX3 and only got problems on 3 with the new hardware insert feature?

---

Exactly. But I'm not 100% sure how to tell whether I'm just getting the Motu's hardware monitoring, see below.

Quote:

---

Also, if you're using live inputs, remember SX can't compensate for latency in this situation. It needs the audio recorded before PDC becomes active,

---

Well, what I'd really like to do is to use live inputs and monitor them through VST FX, but I (think I) understand that I can't do that. What I'm trying to do here is to use VST instruments or audio tracks and monitor them through hardware inserts, which is what I thought the compensation would allow. Either it doesn't, or I've set it up wrong. I'm quite willing to accept that it's the latter!

Quote:

---

if you're using direct monitoring and somehow also hearing your signal back from SX, you'll be seriously confused!

---

I think that's what's happening, and I'm definitely getting confused. I don't know why the "direct monitoring" checkbox is greyed out, and I'm not sure whether it's on or not because the checkbox is definitely not ticked. I don't want to have to record everything to audio, it completely destroys the immediacy.

i thought you were supposed to be clever n stuff

Nah, thick as pigshit, me, all my mates'll vouch for that

I think I have a bad case of understanding the computer representation of the problem domain but not understanding the problem domain itself. I think that's the thing with all these johnny-come-lately wannabe-'studio' types like me; I can work software no problem and I've got a reasonable musical understanding with plenty of musical ideas, but I've got no real-world production or engineering experience. Thing is I just want to make the music and not have to fuck about with the fiddly bits, I want a producer and an engineer to do that for me, but I don't have a record company throwing them at me, so I just bash my head at the tech in my spare time, learn what I can and hope it'll gradually come together

"I just want to make the music and not have to [ ****** ] about with the fiddly bits, I want a producer and an engineer to do that for me, but I don't have a record company throwing them at me, so I just bash my head at the tech in my spare time, learn what I can and hope it'll gradually come together"

Same here. I end up getting sidetracked with the process of making music, instead of the music itself. I think we all do.

You still used the words "problem domain" though

-2 gangster points for that man.

Good luck mate.

Igor

> Well, what I'd really like to use live inputs and monitor them through VST FX but I understand I can't do that - what I'm trying to do here is to use VST instruments and monitor them through hardware inserts, which is what I thought the compensation would allow.

Unfortunately it can't m8

Let's imagine you set your buffers/latency to one second, without a hardware insert:

You hit a key on your board, it's recognised by your computer which starts to fill up it's buffer with the output of your VSTi. One second later you hear the result though your audio interface.

With a hardware insert:
You hit a key on your board, and the computer again fills up it's buffer. One second later the output of the VSTi goes to your outboard (via your audio interface), then the output of the outboard goes back in SX, where it gets additional latency . It then has to go somewhere (to another output), so it gets more latency on output (via buffering again).

Just by adding the insert you've greatly increased latency and there's no way round this, other than by setting your latency extremely low (if we assumed 10ms is playable, then you'd need a latency of around 3ms in this example [minus whatever OsX safety limit your audio drivers use if SX doesn't take this into account]) or by using your MOTUs direct monitoring to hear back the output of your outboard. I'm not sure of the specs on your particular MOTU unit but I know some of the later ones allowed for individual direct monitoring mixes.

Delay comp can only works on already recorded tracks (whether they be audio or MIDI tracks routed to VSTi). Here, the machine can playback these tracks early to compensate for this latency (if a track had a 1 second round trip, its playback would start 1 second earlier).

One of the advantages of DSP based cards (like the Soundscape Mixtremes I use), is that you can bypass this delay (I could route the output of my outboard directly to the output to my monitors when playing VSTi live). The cuemix on other cards isn't so sophisticated, but you should still be able to achieve what you want.

I'm a little confused as to what you are trying to do here myself, and not being familiar with what v3 has to offer I'm not sure my comments will be useful, but:

If theres something you can do with SX2, surely SX3 can do the same thing too?

And PDC is indeed something that can cause problems when being used live - it will add a lot of latency to everything. However, you can switch it off (there should be a 'Restrict PDC' button somewhere) when playing your VSTi - but its bloody essential when mixing down as added latency isn't so much a problem there.

Maybe I'm missing the point here, but then I'll be waiting for Nuendo 3 before I get to try these new features out.

--------------------
http://www.jeremycorbett.co.uk

bloody steinberg and their 'maintenance releases'

i think users who upgrade straight away, to V.1.00 bug central shoudl be given a beta testers discount -cos this is what steinberg are using them for.

you couldn't make it up. etc.

i gor, i use an 828 mk2 like you -i can't really follow your problem, as i haven't tried to hard..looks complicated and i don't use much outboard - however i do occasionally route stuff out thru my focusrite, and i do that via a new mono output bus and monitor it coming back in on a record channel. this may not be the best way to do it, but when i record the result, the latencey seems to have been compensated for, and the audio does not require lining up. ( in SX2 )

I've been on it for a week and had nothing but trouble. I'm finding it hard to pin down the problems and all I'm getting from tech support is, "It's very buggy at the moment". I'm running it on a Dual 1.8 G5 2.25Gig Ram with an RME card.

I have to say that my machine is generally less stable since I put SX3 on there. It never crashed before and now freezes completely 2 or 3 times a day. It's possible that the drivers for my RME hammerfall 9636 are not ready for 10.3.5 but I can no longer run 10.3.4 because - get this - Apple have written a patch into 10.3.4 that makes running a Formac gallery 19" impossible on a G5. The monitor just doesn't work with the ATI card - so I'm in a corner really.

I moved up because they were bragging that the freeze function is faster and it is, but that's useless if the whole thing is unusable in a pro' setting.

I'm fed up of being Steinberg's bug tester and being charged for the privilege.

lets get together a 'mass Invoice' for beta testing and bug fixing, and march to Hamburg carrying it on our backs...or something....

c'mon chaps! over the top! looks like the hunn need another hiding!

did our grandfathers really fight and die so the germans could flood their geeky grandchildren with slightly flaky music software?
NO! i say...NO!

Hi folks

Quote Mahoobley:

---

If theres something you can do with SX2, surely SX3 can do the same thing too?

---

Sure, but I thought the whole point of the external plugins in SX3 is that it makes the same thing a whole lot easier with a lot less faffing about.

Kris, all that makes sense, thanks. I'll just have to get used to recording external stuff as audio - or just give up using external stuff and get a Powercore or UAD.

nimhbus, that's effectively what I do, I think I was just hoping that the external plugin business would use the direct monitoring but I guess that doesn't make any sense as the direct monitoring all happens in the MOTU.

Increasingly inclined to say "stuff the hardware".

As for the generaly stability / bug-free-ness of the software, I've been away for the weekend and haven't used it since my last couple of posts and hence don't have any more to report, other than that the other day I decided definitely not to bother with it until the maintenance release is out and addresses loads of that sort of stuff. Certainly doesn't feel like I've spent my 100 quid on anything worthwhile at the moment.

If my software forced me to sell my hardware, I'd sell the software instead...

-Paws

--------------------
Signature (up to 200 characters). You may use UBBCode in your signature

> Kris, all that makes sense, thanks. I'll just have to get used to recording external stuff as audio - or just give up using external stuff and get a Powercore or UAD.

Igor, a Powercore or UAD would work in exactly the same way. If you record a guitar into SX3, and monitor with a UAD/PC plugin, you'll still get this added latency.

I understand why you'd want to use your hardware while tracking (I don't personally work that way, but know lots who do), but why can't you just connect it to your output source (guitar/mic/keyboard/whatever) while recording and use the MOTU cuemix. That will completely solve your problem while tracking. Once you've done the tracking, just connect the outboard to whatever you've assigned as external plugin inputs/outputs and use them again.

Out of interest, what bug/stability problems are you having, it's been rock-solid here.


> If my software forced me to sell my hardware, I'd sell the software instead...

Paws, the problem with that argument is that the alternative software is far WORSE than SX in this respect. Using UAD/Poco/External plugins on busses/groups/aux sends under Logic/DP/PT LE is horrible (although certainly possible), and one reason I wouldn't want to go there.

The SX delay compensation is fantastic, but it can't do the impossible!

Fair enough Paws, but not me. I want automation and recall and ease of integration above all else because I know damn well that I'm only making hobby or demo-quality music that will demonstrate my musical abilities rather than my sound engineering prowess, and I'm sure that the sounds etc I can get out of software instruments are more than good enough for my purposes.

Hi Kris,

It might help explain what I want if I explain what I'm trying to do. I want initially to knock tracks together for my own pleasure and that of friends, and eventually, when I've had enough practice doing that, to make demo recordings of my music that are of a good enough quality to demonstrate my musical ideas and my skill (such as it is) in writing, arranging and constructing music. I don't kid myself that I'll make good enough quality recordings for release; I'm not going to try and master stuff because I don't have a clue about it and am not really interested; I just want in the end to make demos which I could use to try to persuade A.N. Other record company to let me take my ideas into a proper studio with a proper producer and a proper engineer, maybe using better singers or guitarists than I am and certainly with better equipment than I have, and do it properly with their help and guidance. Until then I want to make the most straightforward "composing" setup that I can.

Given that, what I want to do is to have my tracks consist (apart obviously from things that you just have to record like voice and guitar) of MIDI and VST automation information controlling synthesizers and samplers, whether they be hardware or software, and I want to be able to route the sound that comes out of the synths and samplers through a load of FX, similarly either hardware or software. I absolutely want to avoid having to record bits and bobs of tracks or even whole tracks down to audio wherever possible, because that completely and utterly destroys spontaneity and control in my mind, and because I regularly and often change stuff and having to go back and re-record is just too boring and computer-y and the more of that sort of stuff I have to do the more I get distracted from making the music, which is what I'm interested in. So I'm trying to get to a situation whereby whether with software or hardware (I really don't care which) I can just lay down MIDI information and occasionally record a voice or guitar line, hear the output of the samplers and the synths the MIDI information is controlling, as effected by the FX I'm routing that output through, and not have to mess about with bouncing down to this and that and all the rest of it. I'll get the track finished as a whole, mixed in Cubase and playing live until it's ready for a mixdown, and then I'll mix it down and take the resulting audio file.

You may ask why I don't just carry on doing what I've done previously, which was to use exclusively hardware instruments and an analogue mixer and do it track by track as this worked fine for me and I was knocking stuff out with it. The reasons are automation and recall. Last year I had time, several months, out of work and so I could just there working on a track until it was done, but work and life mean that that just isn't possible now. Things are really hectic, I'm often shattered after work, and that convenience is realistically going to make the difference between getting on with the music and not. I have a few hours at a time working on something, and if I have to go through the rigmarole of recording all those setting manually I find it demotivating and annoying. I want to have all the settings saved in a Cubase project file, and be able to make the most of the few hours at a time by just loading up whichever track I'm working on or have just had a new idea for, and getting on with it.

The reason I'm talking about Powercore or whatever is that if as you say the inherent latency in computer audio systems in which AD/DA stages take place means that to do all that live (as in monitoring the audio from hardware instruments as routed through software FX, and vice versa) is simply not possible, then I'm thinking that if I forget entirely about hardware instruments and FX, then I should be able to do what I want to do, which is just to make the music by controlling software isntruments with MIDI tracks, use FX sends and inserts when I want to, and hear what I'm doing without having to freeze or bounce all the tracks down to audio, record stuff dry and then have to test it out with different FX, etc, etc. Is that right? Could I do that, or am I asking for the moon on a stick?

I don't particularly care about the hardware or software debate; I'm trying to do it using my existing hardware because I already have it and know how to use it more than I do Reaktor etc, and because I'd like to avoid spending more money if possible, but if it means flogging all the hardware kit and buying Powercore etc to get to that situation then fine, I'll do it. I'm much more interested in making the environment in which I'm working immediate and suited to my way of working if I can.

Thanks for reading (if you didn't give up halfway through this rant!) and thanks for taking the interest.

Quote:

---

which is just to make the music by controlling software isntruments with MIDI tracks, use FX sends and inserts when I want to, and hear what I'm doing without having to freeze or bounce all the tracks down to audio, record stuff dry and then have to test it out with different FX, etc, etc. Is that right? Could I do that, or am I asking for the moon on a stick?

---

you certainly can do that, and i do -i have a powercore in my g5. and i sold my hardware for the same reasons -i want it all there, integrated, saved, sample accurate etc etc.

i find the combo of POCO in a dual G5 really is like having power to burn -i can get the access virus running, and a shed-load of great reverbs on the send busses BEFORE i even touch the huge power of the G5 cpu.

its a great way to work -i recommend it.