MachFive 3.0.3 update available

[b-righteous]

Thanks for the fast reply. The duplicate is what I am looking for. Hard without seeing a manual or having the software but how do you move to another key? Is this simply the root note value? Can the key be set via MIDI input?

[hugy]

Thanks for the fast reply. The duplicate is what I am looking for. Hard without seeing a manual or having the software but how do you move to another key? Is this simply the root note value? Can the key be set via MIDI input?

well, there is a mapping view so you just dnd the keygroup to another place, the root key by default follow the position changes.
You can learn values using MIDI, but I think it's quicker using the mouse and moving the keygroup around.

[rhythm_kitchen]

Check the bonuses thread and call your rep about the competitive upgrade. About the same as upping from K-7 to K-8. I need the IRCAM algorithms for timestretching. You can (I've been told here) drag interleaved or dual mono mixes, AIF, WAV, SD2 even broadcast wave and timestretch or shorten... plan to use it on a Quadcore W-7 Pro X64 workstation. At first in Pro Tools 9. Not upping to 10 but waiting on DP-8, which has more features than PTs 10, including sample accurate MIDI, ADC, 16,24,32 bit files mix and match as long as sample rate is constant.

I used Serato P n T back when it was $1,800 and to me all those mixes (about a 1-2 per increase) tempo wise lost their groove to my ears (composer).

Many years ago, I noticed my Vision MIDI sequence speeding up and slowing down on a MAC SE. It was driving me crazy. Yet no other musicians I' point this out to could hear it. I mean world class musicians.

So I called D'Bennedictus (a composer at Opcode) said he noticed it also and put Dave O on the phone. Dave told me matter of fact that the Mac SE does do this with large seqmented MIDI files so try to consolidate regions. I am a cautious MIDI composer always editing sustain events 'thinning' in list event window. Never used pitch bend and no volume events except ramping. But even with consolidated MIDI regions the SE chugged away like the proverbial Chu Chu Charlie... I can do it. Yet few could hear it.

The IRCAM timestretch algorithm is probably worth more much more just by itself and comes with the Mach V 3.

Also you can set it to stream off 15K raptors and not eat RAM. Which is cool then K-7 is using additional RAM (DP can use up to 2 gigs) and Mach 5 samples are then streaming off a Raptor or a pair of 10k raid striped Cheetahs. The new 2.5 inch WD Raptor with a 3.5 inch cooling footprint is very quiet. If anything I would go with those not for power but for lower noise. Yet they require 33% less current.

I initially chose K-7 because of tuning limitations with the older MOTU VIs not wanted to use Sys Ex.

16 bit & 24bit in a 32 bit mixer a +/ DP-8... running WAVES as VST another + besides freezing / unfreezing tracks. Maybe the new Springamabob plugin.
Sorry for the long winded rant here.

[b-righteous]

Long winded yes but I agree with your points. The ircam stretch is certainly on of the reasons for interest. I also love the easy and functional GUI along with it being a central place to load all my UVI/MOTU sounds.

I was hoping it would also suit my needs in chopping and stretching audio phrases in a fast way for production. I did break out my old M52 and I can see the parallel better now. It still seems to fall short in a few specific places but I might pick it up anyway for all it offers.

[MIDI Life Crisis]

I was hoping it would also suit my needs in chopping and stretching audio phrases in a fast way for production.

That's what DP does - well, not so much the stretching, but hopefully in DP 8 that will change.

It still seems to fall short in a few specific places but I might pick it up anyway for all it offers.

And what specific places does it fall short? Not to be confrontational, but to seriously try to understand how it can be improved.

[b-righteous]

well, there is a mapping view so you just dnd the keygroup to another place, the root key by default follow the position changes.
You can learn values using MIDI, but I think it's quicker using the mouse and moving the keygroup around.

I think a better approach would be to use the slice oscillator first then assign to a new part. Then I can treat each keygroup separate.

I have a specific question about the slicer. I hear it now allows you to audition and select slices from MIDI. Does it also allow you to overlap slices like BPM's clip window? What I mean is if you select a slice via MIDI and move the right slice marker for that chop past the next marker will it allow slices to have overlapping regions?

Also, does it expose the macros to the host as automation?

[b-righteous]

And what specific places does it fall short? Not to be confrontational, but to seriously try to understand how it can be improved.

Posted at the same time. Really prefer if things can integrate with Maschine or automap where parameters show on the controller for instant tweaking. MOTO products don't expose parameters to the host so it is mouse city. Don't want to hassle with making MIDI learn assignments every time either. My Maschine still can't handle MIDI cc data anyway. You could also duplicate M5 in Maschine to different pads and adjust start end regions for fast chopping but each instance would need a very small idle CPU load for it to be workable. Just some specific things for my workflow.

[hugy]

In your case, your best bet is to use the slice oscillator.
Adjust the the sensitivity slider to match the correct slices

Hit Map button (the little keyboard icon of the right of the slice toolbar next to convert), which transform each slices to a dedicated keygroup with the same sample but with the right sample start/stop.
Then you can adjust to your taste those start/stop in each keygroup, so it means that indeed you can overlapping zone in each keygroup.
You can remove the slice layer if you don't need it afterward.

HTH

[b-righteous]

Sounds good. So after you use the map from the slice oscillator, can the resulting zone star/end regions be expanded the full range of the original sample?

[hugy]

Sounds good. So after you use the map from the slice oscillator, can the resulting zone star/end regions be expanded the full range of the original sample?

Yes.

It just create regular keygroup with the standard oscillator only with the right settings for sample start/stop markers, but you can move those to your taste like other regular keygroups with the standard oscillator

[b-righteous]

Thanks hugy. Probably going to check it out. I know a windows user with a solid setup so I will confirm if it is working with him.

Would be nice if MOTU added some host automation parameter support for some of the static parameters. They could even do like Geist or Battery that lets you assign parameters global to 32-64 parameters ID's in a preference setup. That or just do it for macros. Sample start and end should be in there to modulate too. Past that I like what M53 is about.

[b-righteous]

So I finally grabbed the update. Lots of good stuff. Overall it is stable and has a low idle load. Some great instruments included. The Telematic and upright are sick. I did not have the F-grand and did not know it was so good. One of my favorite pianos. Just had to tweak the vel sensitivity to be more responsive. You can dig in and swap out some of the lovely IR verbs for more variation. Did not like the J-base and the Biosphere was a mixed bag (as expected)

I do love the GUI layout. It begs you to do sound design and it is always simple and fun. Love the those new Xpander filters!

The slicing is great and improved from v2. The granular is fun and inspiring.

One huge let down is the IRCAM stretch. Heard lots of praise but the CPU spiking and distortion at low latency is way over the top. Never had a VST cause dropouts so easy on my 3.6 i7 DAW. Even the most CPU hungry synth to date (DIVA) does not cause these periodic CPU spikes and drop outs on most patches. It's unusable imo and while it sounds great (when it works) the new pro stretch in Kontakt 5.02 is just as good without all the artifacts and spiking. edit: At least 4x more efficient. This was a big reason I was wanting the upgrade but I still don't regret purchasing for all the other great things it has.

[b-righteous]

After spending more time and testing in different host, the IRCAM algo seems to be inefficient. The way the CPU spikes jump at random intervals tells me it may be something wrong that can maybe optimized. I also noticed that making adjustments can cause spikes until you toggle another parameter on/off then it stabilizes so I think there is room for improvement.

I noticed that it is much better if you turn the window parameter all the way down to 40. Anything over that creates spikes. The padding at 2x makes it sound great but it is oversampling and naturally causes even more cpu load so it should be left off. The transients options cause slightly more load but is usually needed on most material. It still seems to give some distortion at the beginning of transients when doing fast re-triggering at different pitches though. Another thing that seems can be optimized to react faster without distortion.

This and the lack of host automation parameters are my only real gripes. Still like the program overall. Got my eye on that new Wurly from Accousticsamples.