Audio Modeling SWAM Brass 1.5.1 vs. Sample Modeling Trumpet V3: A Comparison 

@King Dubbist Thanks for the feedback. Have you downloaded the SWAM 1.6.1 update? Audio Modeling is really doing its best to improve the sound. Currently I'm experimenting once again with the settings. Not being dependent on Kontakt scripting is a real advantage for realtime control, I believe, and so I hope the timbre in the end will be even closer to the real instruments.

@Denys Quintero You're welcome. Thanks for this feedback, that mentions favourable aspects of the instruments. One of these competitors is under continuous development; I'm still experimenting with the fairly recent SWAM V1.6 update.

Thanks for the response. Will listen to Yamaha Electone sound demos (not familiar with this instrument). Just read the news that Audio Modeling will release an update after NAMM 2021.

@Richard F.W. Davis Thanks for the detailed comments on this AM vs SM brass instrument performance comparison. Since you use both I am curious to learn how to best perform either staccato or legato with an EWI. For me each instrument in the bundle requires a specific approach when playing; that also holds for the strings vs. brass, and I carefully adapt my playing style (mainly breath and tongue control). This is as close as I could get to a classical sound at the time (in the meantime I use the updated and improved bundle, and should redo the comparison).

@@FransAbsil I don’t use an EWI but I do use a TEControl breath controller. I think that if you slant the SM or the AM towards attack enhancement mapped to breath pressure you will get closer. The accent keyswitch on the SM instruments can also be superseded by a full-out breath message.

@dileitor Thanks for the message, sorry to hear about your problem connecting Akai EWI 500 to the iPad. I am afraid I can't help, since I use all SWAM instruments on Mac and PC only. If your question is about Midi mapping or instrument presets, I might help. The connection is made in SWAM>Settings (gearwheel icon)>MIDI; if that fails it's best to contact Audio Modeling support. They are always most willing to help you quickly.

@Selle Dudok van Heel you're welcome. Will do some RAM/CPU tests with current versions of the AM and SM trumpet and get back to you. These performance parameters have never been an issue on a MBP 2018 and PC 2021 system. I've been using up to 30 winds plus brass instruments in parallel in orchestral projects, without problems.

Here are my RAM/CPU test results for Audio Modeling (AM v1.6.5) and Sample Modeling (SM v3) brass. 1) RAM: AM 0 (physical modeling), SM Tpt 62 MB, Tbn 35 MB, Tu 20 MB, FHn 36 MB (in Kontakt, hybrid sample/modeling approach). I did a Logic Pro (LPX) test on I9, 6-core MBP 2018, MIDI region with 16th note scales, 120 BPM, 15 instrument tracks simultaneously. Using Activity Monitor, RAM data: LPX 740 MB, LPX+SM Brass (13 single instruments + 2 4-instrument Kontakt instances) 1.85 GB. 2) CPU: LPX idle 17%, AM: 170% (LPX CPU meter average 20% per core, uneven distribution), SM: 250% (LPX CPU meter around 40%, more even distribution). (NB for I9 processer CU load may go up to 1200%, AFAIK; so very limited load). Playback: no glitches/stuttering, no hiccup problems (make sure the active track is a separate audio track, otherwise there will be terribly distorted sound and CPU overload). The PC system is far more powerful so I did not repeat the test there in Cubase. Conclusion for this test scenario: these instruments are not too demanding RAM/CPU-wise, in a typical case, full brass tutti phrase. If there's an orchestral tutti with large (string) sample libraries, the RAM and disk I/O performance obviously will change. Hope this answers your question.

@Astrah Cat Thanks for sharing your opinion. I agree on the AM, SM and Pianoteq sound quality obtained through hybrid and physical modeling. I am not convinced that keyboards are the most suitable controllers for modeled winds, brass and strings.

@Céline et Dany GILBERT You're welome and thanks for asking. With MIDI breath and RH keyboard controller, my recommendation for live trumpet is twofold: use Kontakt Sample Modeling trumpet instrument for the most realistic timbre (in particular in the note attack phase). Choose the Audio Modeling SWAM trumpet for best realtime response and most versatile sound control (as standalone SW or within Camelot); this instrument is still being updated with improved timbre. Controllers: the combination of CC2 and CC1 indeed should be sufficient for realistic playiing in a live setting. No need to use a DAW in live situation. Listen to all available demos on company sites and RU-vid, then decide. Does this help?

@@FransAbsil Thank you very much for your answer. What does "best realtime response" mean to you? Does this mean that there is latency with sample modeling? My first concern is to have a realistic sound and a certain ease of playing because I will be using my 2 feet, my 2 hands, my breath and my teeth! My choice would rather go to sample modeling, plus it's much cheaper. Am I making the right choice? I use a DAW because during the celebration I use other instruments (flute, harp...).

@Céline et Dany Gilbert, In case of only 2 CC controls (breath and bite, CC2/11 and CC1) the realtime performance aspect is irrelevant in the comparison between Sample Modeling and SWAM trumpet; so go for what you consider the best sound/timbre. Only in case of multiple parallel CC parameters there is a theoretical difference between these two model-based instruments (the specifically designed binary code for SWAM plugins should be able to outperform the scripting approach within Kontakt). In case of the TEC Breath and Bite controller 2 there are 4 input signals (breath, bite, 2 attitude sensors), in case of a Lemur or TouchOSC controller there may be many more (I program these as linked 2-dimensional faders/sliders) and then you may start to have some latency. So your preferred 1st choice should be perfectly fine. Thanks for elucidating on the use of the DAW. Good luck with your church service performance; may the technology never be the limiting factor in a worthy celebration.

@@FransAbsil Thank you very much for your reply. Indeed, I use a TEC Breath and Bite controller 2, which has 4 signals. But I tell myself that playing with the hands, feet and a controller, it will be difficult to use the 4 signals! 2 is already not bad, isn';)? However I can choose which CC I assign each signal to. I hope the virtual trumpet will sound good to embellish the celebrations!

@Marius Rahhell Thanks for sharing your experience with and opinion about this trumpet instrument.

I prefer the raw tone of SWAM but Sample Modeling wins in realism.

@giroliro3, Thanks for the feedback. I guess it's a delicate balance between realistic timbre and realtime performance capabilities, that will change with each update of these products.

@Angelo Dopwell, Thanks for asking about the 'falls'. The focus of this video is on the classical trumpet sound and repertoire; therefore there are no falls in the examples. For completeness, both the Audio Modeling and Sample Modeling brass have a key switch for fall and doit. However, when playing the EWI controller it is far easier and more realistic to perform the fall. Hopefully this answers your question.

@yasu takeout, Thank you for this comment. The result from Google Translate suggests that you consider the sound of the Sample Modeling trumpet far more realistic than and outperforming the Audio Modeling equivalent. If my interpretation is wrong, please correct.