Yamaha E.S.P - as it pertains to Montage M

[ElmerJFudd]

24 minutes ago, AnotherScott said:

 

$995 might be high, but not necessarily unmarketable. Though I'd be interested in paying more like half that. 😉

 

But as for the comparison with Komplete... I don't think Komplete is a plug-and-play environment is it? I mean, you can't just plug in a controller, and immediately have all the sounds in a coherent performance-oriented interface. Maybe if you buy one of NI's Kontrol-S controllers? But even then, I"m unconvinced that those are well-equipped for live performance, i.e. for dealing with complex splits/layers, mid-song patch changes, etc. Yamaha could be offering something very different here.

 

 

I've been tempted to pick one of those rack units when they have popped up now and then. But realistically, I don't think I have a justifiable use for it.

I’ve always wanted one as well.  But not for live, just to experiment with.  I have the PLG150 AN and DX cards - but it is not an easy task to edit them.  Not to mention how similar other software sounds.  
 

https://www.plogue.com/products/chipsynth-ops7.html

 

For shame, Yamaha,  for waiting so long to get into software.

[AnotherScott]

57 minutes ago, ElmerJFudd said:

For shame, Yamaha,  for waiting so long to get into software.

 

They may have figured they had that covered with their ownership of Steinberg/Halion. And in a way, they're still not any more in software, unless they choose to actually make this product available apart from the hardware.

 

[ElmerJFudd]

1 hour ago, AnotherScott said:

 

They may have figured they had that covered with their ownership of Steinberg/Halion. And in a way, they're still not any more in software, unless they choose to actually make this product available apart from the hardware.

 

Quite true, quite true.  

 

Yamaha seems to have been dabbling in music production for better than the last decade.  Strategically the plan seems to have been, purchase Steinberg and develop inegration with their hardware instruments to give Yamaha customers a reason to choose Cubase/Nuendo over PT, Logic, DP, etc.  Bundling a starter version of Cubase with their keyboards.  Conceptually providing a good reason for integration, making it easy to use and developing a rock solid connection has been challenging.  But USB and Thunderbolt have improved their speed and bandwidth a lot since 2010.  CPU performance and RAM has also improved.  So, here we are with a Montage M in software.   I certainly hope their vision is bigger than just adding value to the hardware Montage M.   

[Theo Verelst]

For me, equivalence of the sound is paramount, like if I'd subtract the SW version "live" from the digital Usb connection from the M, do I get zero? Also, if, like with CP4,I can "improve" the sound on the basis of certain processing, can I do that in SW as well?

 

T

[DeltaJockey]

5 minutes ago, Theo Verelst said:

For me, equivalence of the sound is paramount, like if I'd subtract the SW version "live" from the digital Usb connection from the M, do I get zero? Also, if, like with CP4,I can "improve" the sound on the basis of certain processing, can I do that in SW as well?

 

T

So if I'm reading you correctly, are you, like many people, still under under the impression the ESP will only be equivalent as long as the hardware is connected?

 

Your paramount equivalence is already a reality. The plugin IS a software Montage M, (allowing for them to "finish" it's editing functionality this year), you will be able to do anything with the plugin that you can do with the hardware, (sound wise). I already load my old MODX, and Montage 1 user backups, samples and all, into the ESP. When playing it on my laptop with any old hardware controller, I forget sometimes I'm not using my real Montage.

 

I'm not sure how much more equivalent you can get👍

[Analogaddict]

10 hours ago, ElmerJFudd said:

Quite true, quite true.  

 

Yamaha seems to have been dabbling in music production for better than the last decade.  Strategically the plan seems to have been, purchase Steinberg and develop inegration with their hardware instruments to give Yamaha customers a reason to choose Cubase/Nuendo over PT, Logic, DP, etc.  Bundling a starter version of Cubase with their keyboards.  Conceptually providing a good reason for integration, making it easy to use and developing a rock solid connection has been challenging.  But USB and Thunderbolt have improved their speed and bandwidth a lot since 2010.  CPU performance and RAM has also improved.  So, here we are with a Montage M in software.   I certainly hope their vision is bigger than just adding value to the hardware Montage M.   

 

What are the specs on Your computer? I just stumbled over a Mac Mini M2 with 8GB Ram, wondering if it's a good start for a racked Montage...

[ElmerJFudd]

4 hours ago, Analogaddict said:

 

What are the specs on Your computer? I just stumbled over a Mac Mini M2 with 8GB Ram, wondering if it's a good start for a racked Montage...

2013 MacBook Pro Retina  - it has an i5 and 8gb of RAM.  I'm thankful it runs Big Sur decently.  

It's not an Omnisphere power house.  At some point I'll get over to Apple silicon. 

[Mighty Motif Max]

 

15 hours ago, DeltaJockey said:

15 hours ago, Theo Verelst said:

For me, equivalence of the sound is paramount, like if I'd subtract the SW version "live" from the digital Usb connection from the M, do I get zero? Also, if, like with CP4,I can "improve" the sound on the basis of certain processing, can I do that in SW as well?

 

T

So if I'm reading you correctly, are you, like many people, still under under the impression the ESP will only be equivalent as long as the hardware is connected?

 

Your paramount equivalence is already a reality. The plugin IS a software Montage M, (allowing for them to "finish" it's editing functionality this year), you will be able to do anything with the plugin that you can do with the hardware, (sound wise). I already load my old MODX, and Montage 1 user backups, samples and all, into the ESP. When playing it on my laptop with any old hardware controller, I forget sometimes I'm not using my real Montage.

 

I'm not sure how much more equivalent you can get👍

 

I think what Theo meant is how identical they actually sound. As in, if you reversed the phase of the E.S.P. software and routed the audio through the USB input (to use the same output converters in the Montage M), would the two cancel each other out perfectly when set to the same patch?

[DeltaJockey]

1 hour ago, Mighty Motif Max said:

 

 

I think what Theo meant is how identical they actually sound. As in, if you reversed the phase of the E.S.P. software and routed the audio through the USB input (to use the same output converters in the Montage M), would the two cancel each other out perfectly when set to the same patch?

In a roundabout way, I did think he might have meant that. I find it hard to imagine it is sounding different. I already play the ESP plugin via the hardware Montage USB audio ,and out through the Montages DAC's. That's how I actually have my gear hooked up, because I play the Montage onboard performances, as well as DAW plugins through surround impulse reverbs and out to 4 monitors via the 4 Montage audio outs.

 

I can't hear any difference onboard or external. I think I'm effectively doing what is described.

 

 

[Mighty Motif Max]

38 minutes ago, DeltaJockey said:

In a roundabout way, I did think he might have meant that. I find it hard to imagine it is sounding different. I already play the ESP plugin via the hardware Montage USB audio ,and out through the Montages DAC's. That's how I actually have my gear hooked up, because I play the Montage onboard performances, as well as DAW plugins through surround impulse reverbs and out to 4 monitors via the 4 Montage audio outs.

 

I can't hear any difference onboard or external. I think I'm effectively doing what is described.

 

 

 

I think the true test would be to reverse the phase of one; in theory there should be no audible sound then. But I've only done that in DAW settings, so I'm not sure how that plays out when using audio routing with a hardware source as well.

[jazzpiano88]

22 minutes ago, Mighty Motif Max said:

 

I think the true test would be to reverse the phase of one; in theory there should be no audible sound then. But I've only done that in DAW settings, so I'm not sure how that plays out when using audio routing with a hardware source as well.

 

Also, compare the Montage M Analog Circuit. (new analog low distortion outputs) to the VST.   I'm not paying for a Sampled Montage M (USB Digital Out).   Montage M USB Digital Out is not the Analog Circuit. 

[AnotherScott]

23 minutes ago, jazzpiano88 said:

Also, compare the Montage M Analog Circuit. (new analog low distortion outputs) to the VST.   I'm not paying for a Sampled Montage M (USB Digital Out).   Montage M USB Digital Out is not the Analog Circuit. 

It's not a sampled Montage M. The sound generating engines in both the hardware and software versions are digital (whether AWM sample engine, or the VA engine, or the FM engine... none of it is analog). However it is possible that the hardware could sound better because of its DAC and the post-DAC analog componentry. Conversely, if you have a high-end audio interface on your computer with particularly good DAC and analog components, I suppose that could sound as good or better.

 

Some info about PAC at https://sandsoftwaresound.net/yamaha-montage-quick-thoughts/

 

[jazzpiano88]

That allays my concerns because I had not seen a block diagram from Yamaha that shows  their USB digital output to be identical to the input to their high fidelity analog circuit.   Thanks for clarifying that it is. 
 

Now that means that the standalone Vst is bit for bit identical to the Montage HW internally and the vst consumer can produce an even better analog circuit if so desired, as you mention.   Makes you wonder why the Montage isn’t just a keybed and a PC chipset with an audio interface.   I guess it is?

[DeltaJockey]

8 hours ago, Mighty Motif Max said:

 

 

I think what Theo meant is how identical they actually sound. As in, if you reversed the phase of the E.S.P. software and routed the audio through the USB input (to use the same output converters in the Montage M), would the two cancel each other out perfectly when set to the same patch?

Oh.....I just realized Theo was being literal! (As I smack myself on the forehead Sorry to sound like a broken record, so we're talking about actual waveform cancellation! Yes, or course that in theory would confirm. To actually do it in practice in the digital domain may be a challenge.  Knowing nothing about the Yamaha digital format, and my digitaI processing knowledge is getting a bit long in the tooth but I would assume there's pcm interleaving, parity and all sorts of error checking, along with jitter management etc. You'd have to make sure you had rock solid clock syncing, so you'd have to line all that up if it's possible.  But, I suppose the most practical way would be to capture it straight out of the DAC.

 

The way I'm comparing them now, they are both being generated digitally, and both ending up converted to analog through the same onboard Montage DAC to the speakers.

Comparing them with my mortal ears back to back, they certainly sound identical to me.

 

Perhaps a quick and dirty would be to take a few cycles of one of the raw ANx sinewave  waveforms, with no effects processing from hardware and software and then import into a sample editor, invert one and see what you come up with? Not sure whether you could pin down any other system differences that could affect the result.

 

I guess, with a bit of thought, some creative thinking could devise a smart way of comparing them successfully.

 

The only other thing I might look into, is to mix the hardware audio in on the DAW, and the ESP audio, with a waveform inverter plugin, then mix them out to audio, and see if I get anything but a distorted mess👍

 

I'm sure someone else may have an even better idea. (Ideas and suggestions are always the easy part) 🥴

 

 

Also as I have commented in the thread way back, and as Scott has also said above, using ESP on other USB audio interfaces will render the output only as good as the hardware you are using. I have listened to ESP from my laptop through my RD88 and Numa X GT audio interfaces, along with a couple of standalone DAC's. While it sounds very good, and at least as good as any of my other vst's,  the Montage M DAC has a noticeable improvement of whatever sound comes through it. In particular, sound field imaging and phase distortion. It all just sounds smoother, and with better imaging. The Montage M DAC is the best audio interface I now have.

[Analogaddict]

Ok, my 2015 MBP with 8GB RAM is not enough to run this… at a buffer of 256 I’m hitting the ceiling as soon as I start layering things - oddly enough, the AWM parts are the problem. Layering multiple instances of FM X and AN X is ok though. 8 parts of FM and AN gives me approximately 70% while two layered pianos + a string pad sends me well over 100%. I need a more modern computer for this. I want to 100% clone my Montage M for flying and such, so I only need one instance running. 

[DeltaJockey]

4 minutes ago, Analogaddict said:

Ok, my 2015 MBP with 8GB RAM is not enough to run this… at a buffer of 256 I’m hitting the ceiling as soon as I start layering things - oddly enough, the AWM parts are the problem. Layering multiple instances of FM X and AN X is ok though. 8 parts of FM and AN gives me approximately 70% while two layered pianos + a string pad sends me well over 100%. I need a more modern computer for this. I want to 100% clone my Montage M for flying and such, so I only need one instance running. 

Have you set the Lite mode in the utility? Even my 2021 M1 Pro Macbook stutters in full mode with some multi layers and lots of notes. But a more powerful computer on the whole will allow you to do what you want to do. I would hope that with further updates they may improve the resource efficiency.

[Analogaddict]

On 2/10/2024 at 12:04 PM, DeltaJockey said:

Have you set the Lite mode in the utility? 

Not yet, I actually don't trust computers on stage at all so I'm looking for headroom even when it's using many parts in 'full' mode.

The way things look, I'll be doing some traveling with a computer based rig (and a clonewheel in case things go down) from now on so I figure I'm going to switch right away to be on top of things. This week I'll try ESP on a Mac Mini M2, hopefully that will be enough. I'd prefer to not use a laptop since a rack mount computer is more protected. Last summer I had everything ranging from hot sun to rain, thunder and lightning during gigs. 😅

[DeltaJockey]

An m2 Mac Mini should be fine overall. Interestingly, it's not cpu maxing out that I've found to be the issue, but rather the plugin being limited to one thread in Logic/Mainstage when playing it in realtime. I wish there was a way to tell it to share across multiple threads when doing this. Not sure if other DAW's/vst hosts do this on the Mac, though I tried playing it with one other host and it still did it.

 

As I said, maybe future updates of ESP will be more efficient,

 

Good luck!

 

 

[ElmerJFudd]

3 hours ago, DeltaJockey said:

An m2 Mac Mini should be fine overall. Interestingly, it's not cpu maxing out that I've found to be the issue, but rather the plugin being limited to one thread in Logic/Mainstage when playing it in realtime. I wish there was a way to tell it to share across multiple threads when doing this. Not sure if other DAW's/vst hosts do this on the Mac, though I tried playing it with one other host and it still did it.

 

As I said, maybe future updates of ESP will be more efficient,

 

Good luck!

 

 

Most third party plugins are not multicore enabled - only a problem when the plug-in is a resource hog, ie. Montage M ESP, Omnisphere, etc.  

 

 You can tell Logic how to allocate threads - “Choose Logic Pro > Settings (or Preferences) > Audio. Click the Processing Threads pop-up window, then choose an option. The options available depend on the number of CPU cores in your Mac and if your Mac supports Hyper-Threading.” - but the plug-in itself has to support it. Otherwise, assigned to one core. 

[ElmerJFudd]

20 minutes ago, ElmerJFudd said:

Most third party plugins are not multicore enabled - only a problem when the plug-in is a resource hog, ie. Montage M ESP, Omnisphere, etc.  

 

 You can tell Logic how to allocate threads - “Choose Logic Pro > Settings (or Preferences) > Audio. Click the Processing Threads pop-up window, then choose an option. The options available depend on the number of CPU cores in your Mac and if your Mac supports Hyper-Threading.” - but the plug-in itself has to support it. Otherwise, assigned to one core. 

In a Logic Project what I would probably do is turn off effects in the ESP itself and send the audio out to a bus.   Insert other fx on the bus.  
 

Little by little ESP will get some optimization and CPUs will continue to improve.  

[Analogaddict]

Tested in "lite" mode yesterday on my 2015 MBP. If I layer three AWM pianos and a string patch, I'm well over 80% as soon as I start playing. It works significantly better though, and I believe I could make it work for one of my bands with a little careful programming. Still, that's not what I want, I basically want my whole Montage in a portable format. I will try with a M2 Mac mini later today if there's time. Very exciting stuff!

 

BTW, we are basically reinventing the rack module here. 🤣 Come on manufacturers, just release your stuff in rack format already!

[ElmerJFudd]

8 hours ago, Analogaddict said:

Tested in "lite" mode yesterday on my 2015 MBP. If I layer three AWM pianos and a string patch, I'm well over 80% as soon as I start playing. It works significantly better though, and I believe I could make it work for one of my bands with a little careful programming. Still, that's not what I want, I basically want my whole Montage in a portable format. I will try with a M2 Mac mini later today if there's time. Very exciting stuff!

 

BTW, we are basically reinventing the rack module here. 🤣 Come on manufacturers, just release your stuff in rack format already!

Like when Persing released Keyscape. Montage M ESP is 2024 software written to take advantage of 2024 hardware.  Lite mode likely starts with sample thinning (stretching) and poly reduction. They say some of these Roland Cloud synths are also hogs. 

[Theo Verelst]

Ome can ask if the "Lite" mode makes compromises or magically makes the same quality sound whilst using less resources.

 

T

[ElmerJFudd]

2 hours ago, Theo Verelst said:

Ome can ask if the "Lite" mode makes compromises or magically makes the same quality sound whilst using less resources.

 

T

I love magic! But it doesn’t exist.   

[DeltaJockey]

There's no mention of any changes to the samples, only polyphony limits.

Certainly, I can't hear any difference. Sometimes I have to double check I have Lite mode on, as there's no way of telling by listening to the quality of the sound.

 

Info from the user manual.....

[Analogaddict]

It runs like a champion in light mode on the M2. Eight piano and pad parts and not over 80%. That's good enough for me. Next step will be loading a backup file to see how it handles my gig sounds. 

[DeltaJockey]

Just for those of us who own the hardware, there's another firmware update. Not related to ESP this time, but some minor bug fixes all the same.....

 

V1.21 to V1.22

Fixed problems;

Fixed a problem in which some Performance information is not loaded properly when loading the following model files.
* MOTIF XS, MOTIF XF, MOXF, MONTAGE, MODX, MODX+

[Theo Verelst]

Limiting the polyphony does not in every case limit the workload. Often, it will only through the note reassignment logic, or soft release note reuse, be that actual workload is reduced, what would be possible reasons for 20 notes being played with a 64 max polyphony be any different workload than the same 20 notes with a 128 polyphony?

 

Sure, there can be organ algorithms that continuously simulate the whole set of primary harmonics, and there is memory management which can be sensitive to claiming twice as much fast memory for immediate use, so there could be that.

 

If the indicated polyphony load is as some have reported a fast notebook should be able to run the whole machine, and when getting nice and hot (...) maybe a some more, under the condition that that remains properly responsive and sound and sample accuracy is maintained, that is an interesting proposition.

 

T

[ElmerJFudd]

8 hours ago, Theo Verelst said:

Limiting the polyphony does not in every case limit the workload. Often, it will only through the note reassignment logic, or soft release note reuse, be that actual workload is reduced, what would be possible reasons for 20 notes being played with a 64 max polyphony be any different workload than the same 20 notes with a 128 polyphony?

 

Sure, there can be organ algorithms that continuously simulate the whole set of primary harmonics, and there is memory management which can be sensitive to claiming twice as much fast memory for immediate use, so there could be that.

 

If the indicated polyphony load is as some have reported a fast notebook should be able to run the whole machine, and when getting nice and hot (...) maybe a some more, under the condition that that remains properly responsive and sound and sample accuracy is maintained, that is an interesting proposition.

 

T

Poly is a factor in sustain pedal use and sound layering magnifies the issue when using a low audio buffer setting to keep latency down.  Newer CPUs are of course much better at this, but as you suggest, there may be other efficiency factors in play.  One of which, brought up earlier in the thread, is having the ESP running on a single core.  

[DeltaJockey]

5 hours ago, Theo Verelst said:

Limiting the polyphony does not in every case limit the workload. Often, it will only through the note reassignment logic, or soft release note reuse, be that actual workload is reduced, what would be possible reasons for 20 notes being played with a 64 max polyphony be any different workload than the same 20 notes with a 128 polyphony?

 

Sure, there can be organ algorithms that continuously simulate the whole set of primary harmonics, and there is memory management which can be sensitive to claiming twice as much fast memory for immediate use, so there could be that.

 

If the indicated polyphony load is as some have reported a fast notebook should be able to run the whole machine, and when getting nice and hot (...) maybe a some more, under the condition that that remains properly responsive and sound and sample accuracy is maintained, that is an interesting proposition.

 

T

Well, for whatever reason you say it shouldn't limit the workload, it does seem to work. Have you found Lite mode is not effective at reducing load yourself?

 

Perhaps there's more voice management going on also. I had my most problem on performances with lots of long sustain voices. I could watch the load go up as I pressed more keys, until the sound would break up. So in Lite mode I'm guessing there's a modified algorithm managing voice stealing a little better. Also, on the Apple Silicon Mac at least, the limitation was because the playing plugin was limited to one cpu core. I didn't even get any difference tweaking buffers. The system overall was sitting at around 20 percent, so plenty of cpu left over! That's why I'm hoping there will be some sort of software tweaking to overcome this unnecessary limitation. What it does seem to indicate is that the problem is not worse with multiple instances. So using the plugin to create multiple tracks and playback a song, would not be any worse.

 

As I mentioned before, there's no audible change in the playback or sound in Lite mode. I can't even detect any change in the voice management with my dynamic playing.