What's new

Help me to get to the right decision about Audio Modeling and Sample Modeling

hag01

Active Member
First of all, as far as I concerned there is no serious demo by Audio Modeling for any of their instruments while Sample Modeling has great demos.
Any demo where a keyboard player play an instrument live is not serious.
I want to hear pre programed recordings and not only of one solo instrument but full ensembles and orchestra.

I want to get a modeled instruments package mainly for jazz mock ups.
I also would be glad to layer them with EastWest Hollywood Orchestra for orchestral music, if I’ll have them.

The dilemmas about getting those virtual instruments or not are:

1. I’m afraid those instruments need insane amount of programming. The number of MIDI CCs that have to be programmed on both companies products, and Audio Modeling not having sections patches so you have to program separately every instrument in the sections.

2. I don’t have experience on any wind instruments, so I don’t think breath controller is for me.

3. I can’t understand why you have to play those instruments live and can’t just program them and automate MIDI parameters on a daw, but if the case is that you must play them, I guess those instruments aren’t for me.

If I decide to get those instruments, except from audio modeling woodwinds which are the only modeled woodwinds package from both companies, what strings and brass package is better and the recommended from the two?
 
If you have to get modeled strings, Sample Modeling Ensembles/Solo is the better one and the less difficult to work with in terms of programming, CCs, etc. However neither company has convincing strings. For brass, I haven't used AM but I have no doubt SM is the superior one. It doesn't take a lot of programming if you set the instrument up right (be aware of the release setting, get the sound where you like it). It's pretty natural once you get used to using CC 1 and 11 next to each other.
 
3. I can’t understand why you have to play those instruments live and can’t just program them and automate MIDI parameters on a daw, but if the case is that you must play them, I guess those instruments aren’t for me.

Don't let the live performance videos give you the wrong impression. I think those videos are just to show what kind of sound the library is capable of achieving right out of the box. But you're more than welcome to play or program the notes first, and then add controller data later in your DAW.


I’m afraid those instruments need insane amount of programming.

That's not necessarily true. You can get a really nice sounding performance with just note velocity, expression, and vibrato. But the ability to edit much more than that is where modeled instruments shine.
 
If you have to get modeled strings, Sample Modeling Ensembles/Solo is the better one and the less difficult to work with in terms of programming, CCs, etc. However neither company has convincing strings. For brass, I haven't used AM but I have no doubt SM is the superior one. It doesn't take a lot of programming if you set the instrument up right (be aware of the release setting, get the sound where you like it). It's pretty natural once you get used to using CC 1 and 11 next to each other.
Why CC1 and CC11? One is for expression and the other is for?

Other votes for Sample Modeling Strings vs Audio Modeling Strings?


That's not necessarily true. You can get a really nice sounding performance with just note velocity, expression, and vibrato. But the ability to edit much more than that is where modeled instruments shine.

Yeah, that’s what I wanted to approve before buying them.
If it’s just velocity, expression, and vibrato for the basic shaping of the sound, then that’s OK.
I’m not expecting to get all those wild articulations from a modeled instrument without additional programming☺

What is the general opinion about Audio Modeling woodwinds vs the old Sample Modeling woodwinds?
I remember back then before the split that everyone praised Sample Modeling woodwinds, especially the saxes.

How difficult is it to create and program full sections with Audio Modeling instruments?
There are no section patches...

And what about the necessity of breath controller?
I already asked this question in an older thread, and the answer was that there are many other controllers which are sufficient, but Audio Modeling highly recommend in their manuals not to use other controllers than breath controller.

By the way, I saw demos of live players control their strings instruments via breath controller and I have to say it looks kinda weird.
 
if you want to see the absolute best the audio modeling instruments can do, check out this video and channel,
I swear this person is a maestro at these instruments. They a leapmotion that uses the signal from them waving their hands around to input midi. From other videos you can see that they use a ton of different midi CCs to shape the sound. Because of this, I think this person creates some of the best non-orchestral mockups I have ever heard. Despite my efforts I'm not good enough at using SWAM instruments to create a sound as good as this person.

Personally, I think that if you want to do jazz I would recommend using Aaron Venture's infinite series (which sadly doesn't have a strings package yet) and using the studio reverb. These are the most flexible instruments that can sound good with only a small bit of midi programming (only using the dynamics and vibrato controls can get you most of the way there). The infinite woodwinds have saxophones that I think are really good for both jazz and orchestral contexts. The best example that I know of for showing off the infinite series in jazz is this: But I think these aren't the best at showing the capabilities of the instruments.

For modeled strings, even though I quite like the Sample modeling strings, I still think that the swam strings are a little bit better, but good luck trying to make them an ensemble. There is an example of swam strings in an ensemble here:
There is also a video on the audio modeling channel which shows how the dude does his process here:

I think getting the best results out of swam require some type of real time input as the little bit of human irregularities in the midi signal is a lot of what makes the SWAM instruments realistic.
 
CC1 is for vibrato and CC11 controls dynamics.
The big advantage of SM strings vs AM, is that with SM strings you can control the size of your ensemble with a CC value. This goes from solo- chamber- full string section, whereas AM only has solo players.
 
Yeah, that’s what I wanted to approve before buying them.
If it’s just velocity, expression, and vibrato for the basic shaping of the sound, then that’s OK.
I’m not expecting to get all those wild articulations from a modeled instrument without additional programming

When I create tracks with modeled instruments, I focus mainly on playing the part. Sometimes I simply keep my expression pedal about 75% cocked and only ride the mod wheel for vibrato. After I record the part, I go back and start sculpting out the performance using my mouse to adjust note velocity, timing, and length, and to draw in expression and other CC data. I can spend hours doing this on a solo instrument track, but the results are worth it.

In case you haven't seen this ad, since you're looking to do jazz mockups, keep your eye on the new trumpet library coming from Straight Ahead Samples:

 
As the owner of SM S&E Strings, I would tell you to go to Audiomodeling Strings.
SM is good, but for my part I didn't manage to make it sound on my songs, and I think that AM strings is much more malleable.

AM's fault is that it does not offer string ensemble, it must be tedious to create them instrument by instrument.

Now I have an outsider to offer you, you should also look at Reason Friktion.
It's a bit more complicated to tame than the other two, but it's "easier" to make your own ensemble as you wish.
Here is for example a sound I am working on, creating with a pack of strings that I have built:




If you're interested, check out my thread on Friktion.
 
Last edited:
Personally I don't understand hearing SM as fake but finding AM acceptable. It's definitely worse. Malleability after that point is cool but... can't get past that sound. But yes my god that one guy on Youtube has nailed SWAM strings! I replicated everything in his setup and it's honestly difficult like trying to play violin for the first time. Maybe I need more practice.
 
Why CC1 and CC11? One is for expression and the other is for?

Other votes for Sample Modeling Strings vs Audio Modeling Strings?

Sorry I missed this. On all SM instruments, CC 1 is vibrato and CC 11 is expression, basically dynamics like most sample libraries assign to CC 1. For the brass, high CC 1 results in wild jazz/mariachi vibrato, but with strings you can ride all the way up to 127 and it is still pretty moderate. Then there's CC 99 for 'synchronous vibrato' which is just giving it a little extra romantic-style push.

In generall, riding 1/11 is pretty important to most libraries.
 
Personally I don't understand hearing SM as fake but finding AM acceptable. It's definitely worse. Malleability after that point is cool but... can't get past that sound. But yes my god that one guy on Youtube has nailed SWAM strings! I replicated everything in his setup and it's honestly difficult like trying to play violin for the first time. Maybe I need more practice.

I would not say that it sound false, some manage to do wonders with SM.
But having other library of strings that I wanted to replace by SM Strings thinking it would easily do the job, I quickly realized that something was missing compared to these said libraries.
No matter how hard I tortured the plugin's parameters, I couldn't make it sound the way I wanted.

I don't have AM, but have seen a lot of videos on it that show this plugin is more of a rough-and-tumble violin simulation than the SM is.
After, having Friktion and seeing that in principle it is similar to AM Strings, I can confirm that it is more difficult to achieve a good result (and even more so for Friktion).
You will have to spend a lot of time tweaking the sound, but the purpose will be much closer to reality, because as said above, it is a simulation of the real parameters of the strings (position of the bow, pressure on the bow...).

Example of what SM is not able to do and which shows the difference with Friktion and SWAM Strings:

 
Last edited:
Hi @hag01

Interesting topic.

Physically Modelled Instruments are not very easy to play, and to have them sound very good right away, they need to be approached like a real instrument, take time to practice using them, with various controllers, and treat them with additional acoustic treatments, i.e. Reverb, EQ, Compression, ..etc. They are totally different experience compared to using Sample Libraries, but the results once you get the right formula, and enough time training to play them, is something Sample Libraries can't achieve.

I'm currently trying to decide if I should buy the Sample Modeling Brass Bundle. They are on sale.

The alternative is Audio Modeling Brass, but they are much more expensive.

I think I will end up getting the SM Brass Bundle.

I'm also impressed by how much the Leap-Motion controller can add to the realism as seen in some of the YouTube videos posted on this page, I have one, but never use it, now I feel that's a big mistake, I should put it to work, and see how much more expression, and realism it can add to both PM Instruments, and Sample Libraries as well.

Thanks,
Muziksculp
 
I recently got Sample Modeling strings, and quite happy with it. It has ensemble patches, for which I just enable vibrato snd lower vibrato delay, record in cc dynamics with my controller knob, and it's done. And it sounds great, to me. Runs, trills, whatever, just play it in. Way faster than key switch based libs.

When it comes to solo, well that's where you have to pay more attention and record vibrato depth and speed changes. A breath controller helps here.

Edit: it also has keyswitches, but you don't have to use them that much
 
Hi @hag01I'm currently trying to decide if I should buy the Sample Modeling Brass Bundle. They are on sale.

The alternative is Audio Modeling Brass, but they are much more expensive.

I think I will end up getting the SM Brass Bundle.


Unlike the strings, I would rather recommend the brass of SM (which I own) rather than those of AM (based on the demos that I have heard).
Maybe AM's demos weren't done with enough care, but their brass didn't convince me.

Still in the things that did not convince me and returning to the strings, Arché is also an element to consider.

I just saw a more convincing example in this thread.
 
Arché is definitely on the higher end of convincing modeling I've heard, as is Friktion. It's disappointing that the SM/AM guys nail brass and winds but not the sound of those. SM strings are definitely like bowling with bumpers... it'll never sound as terrible as default SWAM strings but the constraints are also a problem. However, SM strings have been amazing in certain musical contexts - emotionally restrained, classical or neo-classical settings where you want a controlled, mostly vibrato-free sound. That's why i got them, an unbelievably convincing Mozart mockup.

Glad to see someone else thinks Arché sounds good - I saw one bad review on here and stopped paying attention... maybe I'll get it if there's a sale.
 
@lychee ,

Thanks for the feedback.

The Sample Modeling Brass bundle is currently on sale for $360. which is an attractive price given it offers various Trumpets, Trombones, French Horns, and Tube. and with various models.

As far as Physically Modelled Strings, well.. I have the SM Solo & Ens. Strings, and the Audio Modeling Solo Strings, plus Reason's Friktion. So I'm quite set with strings, and don't see anything special that the Arche Strings will add.

Cheers,
Muziksculp
 
Here is a Reaktor based Physical Modeling Violin called 'Serenade', it's free if you have NI's Reaktor, being controlled using a Leap-Motion controller. I don't have it, but it is not a bad idea to get it since it is free.




Get Serenade Here :


Cheers,
Muziksculp
 
This is genius...



Yes, Amazing performance using SWAM Violin !

Thanks for posting it.

Someone asked the creator of the track/video how many controller lanes did he use, he answers :

Quote : I used 9 controller lanes to the following parameters: velocity, expression, bow pressure, bow position, modulation (vibrato depth), bow noise, sustain, accent, env attack speed
 
Top Bottom