@DG said:

Jules, thanks for the reply. it does help to clarify things for me, so it's all good. I have no intention of trying to work at a higher buffer than 256, as even this makes things very uncomfortable. I don't mind muting parts from time to time, but I think that I'm going to need to come up with a different workflow, as there is no way that any of the players I use for live stuff would like to work at a buffer of 512, even when using Direct Monitoring.

DG

Hi DG

Well, it sounds like there is a bypass button, and Christian should certainly know.  I don't recall seeing it, although there are independent dry/wet controls for each instrument, so perhaps that enables the system to bypass MIR.  I'd have to run some tests to establish how much overhead this really saves though.

If you're monitoring directly, I'm not sure I understand why latency should be an issue for live players.  I guess I'm missing part of your workflow, but if all you'd be doing is playing your arrangement back from MIR and recording live audio into your host, with the players monitoring themselves directly from your host or mixer, then latency should not be a factor at all.   Ahhhh (penny drops!) you're probably imagining a system where your host is on the same rig as Mir, in which case if you have to monitor via your host then you have the same master latency, set by your audio hardware.  Personally I never monitor via host software, I use ProTools HD (which has zero latency) on a separate machine for mixing, and Logic Pro on another separate machine for my arrangements, running at low latency, and can record on either system without any significant latency.  I think running big arrangements in Mir, and a host with a load of plugins, and recording live, all on the same system would be pushing things a little.

I will work at 256ms on the Mir machine and I doubt I will have many problems, but then my arrangements are rather less complex than the kind of template you've described.

Jules

Hi Jules, you have described my situation correctly. I have no interest in using more than one machine any more; I've done enough of that.....! As MIR is only outputting one stereo channel (unless one is doing surround), using a separate machine just for recording and mixing would be a waste of effort and hinder my workflow from what I would like it to be.

My pleasure DG.  

I'm afraid I've given up on the one-machine solution, although it's a totally worthy aim.  I just always find that one system to be too heavily taxed to be totally reliable.

Good luck and all the best

Jules

@Trailerman said:

My pleasure DG.  

I'm afraid I've given up on the one-machine solution, although it's a totally worthy aim.  I just always find that one system to be too heavily taxed to be totally reliable.

Hi Jules

I'm not sure why you mean "If you're monitoring directly, I'm not sure I understand why latency should be an issue for live players." ? 

I plan to use a separate machine for MIR, but the latency issue is still in play here too as I see it - when I play on my main keyboard the midi goes though the main DAW then via MOL or similar to the MIR host's instrument, which then introduces an extra 1024 samples delay before sending the audiosignal back via eg. ADAT to the main DAW and finally to the speakers for me to hear. So there is at least 25 ms delay (more likely 35 ms) btw I press a key and I hear the sound playing.

But offcause having a bypass function will help a lot here - running 128 samples buffer instead while recording midi, and then raise the buffer when switching on the MIR functions.

Hi Bjarne

My statement related specifically to DG's comment about latency when working with live players, not when recording MIDI.   If you're working with live musicians and have your host running on a separate system, then as long as the musicians are monitoring themselves 'direct' or via a low latency host buffer setting, MIR latency will not be an issue.  In fact DG intends to run everything on one system, in which case it  clearly IS an issue, because MIR and the host would share the same buffer settings.

I'll have to check into the bypass function, I wasn't aware of it, but have not been using MIR for the last few days.   When I ran some stress tests I found 512 samples (sorry, I keep quoting buffers in ms, brain not working) not  to be a problem, and that involved MOL, and it also invovled audio being routed back to Logic on a separate machine and then from Logic onto another separate machine running ProTools HD via lightpipe.  1024 could become probelmatic, no doubt.

I would have thought that on a well specified i7 machine you would get very good performance at 256 samples, indeed it may deliver close to the Xeon spec I use.  The only reason I ended up with a Xeon machine is because of the price of 4GB Ram modules for the i7 config, because I wanted the system to have 24GB.  When I compared specs, I could get the dual Xeon system for the same price as a 24GB i7 rig because of the savings on Ram cost.

Cheers

Jules

Hi Jules - get it now [:)]

 - thx for your input, much appreciated.

By the way I do think that your dual XEON W5580 setup is significantly more powerfull than the i7 platform. At least when it comes to raw computing power in the form of integer and floting point operations. 

@Hamish said:

the guy recommending that you buy a good graphics card

harmish, it appears you _are_ rude ... though i don't know who *the guy* is i can assure you directX10, openGL 2.1 _is_ a system requirement - we had a hard time to get a slightly older directX9 card to work properly ... read through the posts (especially mine) and you will understand ... would you really like to throw your performance resources on the grafic?

christian

@cm said:

the 5520: we currently have one for testing, 24 GB RAM, overclocked to 2.9 GHz, not yet finetuned ... very impressive ... seems to be a tick less performant than the 5580 (3.2 GHz)

 personally i'm still not sure what the most important part is - CPU speed, memory speed, number of calculation units, level2/level3 cache ... the amount of possible testing scenarios is pretty huge ... we will have to find out.

christian

Hi Christian

I suppose the memory bandwidth is increased also when overclocking the 5520 ?

I've noticed that the X and W types have 6.4Gb memory bus, while the E types "only" have 5.8Gb

Hamish - the easiest way to get some first-hand experience would be to download the demo version of Vienna MIR and try. Christian (cm) has already pointed out the needs of its GUI.

It seems that you don't own any VIs by now, but you could use its graphical features anyway.

@Another User said:

We have made bad experiences when the graphic card doesn't adhere to some minimal standards.

... but of course the requirements are no the same as for full-fledged PC-games! 8-)