Jump to content

Does anyone else prefer hardware MIDI modules to software synths?


Notes_Norton

Recommended Posts

First of all, there is more than one right way to make music, and my way might not be best for anyone else. But I'm curious.

Does anyone else prefer hardware MIDI modules to software synths?

It seems like it's getting more and more difficult to find good hardware synth modules. Yet there are plenty of soft-synths available.

The way I see it is this:

  1.  Hardware MIDI synths last forever. I have a Yamaha TX81z, a Korg DS-8 and a Roland MT-32 that were manufactured when PCs still used DOS-5 or maybe Windows 3.1, Atari/ST computers had built in MIDI ports, and Mac computers used Motorola CPUs. Some of the sounds on these modules are stale, but others are stellar and have no contemporary equal. Plus how many software synths from that era still work today? Zero. How many software synths have been orphaned when computer OS systems evolved? Plenty
  2. Hardware synths store all their sounds in ROM. Software synths must 'do the math' creating every note, every nuance, and every articulation while the sequence is playing. This taxes the computer's CPU which is also running the DAW and perhaps other background jobs
  3. Since all the sounds in hardware synths are stored in ROM, the latency is negligible and they all run about 5ms plus or minus 1ms. I've seen some software synths that have a latency of almost a half second.
  4. Since there is no tax on the CPU and since all the hardware synths have the same latency I can mix sounds from my rack of 10 or more synths. I can pick the best snare drum for the song from one module, the best kick drum from another, the best bass from another the best guitar from another the best trumpet from another the best sax section from another and do on. They will all be in sync.
  5. With practically zero latency I can add track after track in real time while the old ones are playing and have the new track in sync.
  6. With zero load on the CPU for sounds, the hardware synths can be better sounding or better responding. I have yet to see any software synth that can do what the physical modeling Yamaha VL70m can do -- and it's been discontinued for years. The nuances it can reproduce are light years ahead of any sample or sample-modeling synth.

Through the years I have added new synth modules while my ancient stone-age sound modules still contribute to my mixes.

To be fair, there are also advantages to software synths.

With all the talk about software plug-ins, sometimes I feel like a dinosaur. Are there any other dinosaurs out there?

Notes

  • Like 2
Link to comment
Share on other sites

I loved hardware synths and ROMplers!  I could control them with my Korg N5, MS2000 or MIDI keyboard and save all of my CPU's processing power for real-time effects.

That said, the next DAW I create will be as beefy as I can make it so I can add more software synths.

 

10-Studio2010-02.jpg11-Studio2010-03.jpg

  • Like 3
Link to comment
Share on other sites

I would say I prefer the convenience of Software- vs Hardware Midi modules. However I still use HW MIDI modules as there is a lack of Software modules alternatives, or alternative that sound the same. A FM7 just doesn't sound the same as a DX7 etc. There is not even alternatives, say for a D-10 or a FB-01 that I am aware of. 

Just to comment on your points 2 & 3, unless it is a ROMpler ... these MIDI (digital) modules are small dedicated standalone computer which computes the sound ... similarly to what any VST synth is doing. 

Link to comment
Share on other sites

I've got a stupid amount of MIDI modules/synths:

Kawai K1R, Kawai K1
Yamaha TX81Z, TG500 x 3, MU10 x 3, MU100R, SY77
Roland D110 x 2, U220, CM64 x 3, MT32, D550, JV1010 x 3, JV2020, VariOS
Korg Wavestation SR, X5DR, NX5R x 2, X5D
Casio VZ-8M x 2, CZ3000, CZ1, VZ1
Evolution EVS-1
Alesis D5, SR16
SMPro V-Machine

I sampled all the sounds I liked from most of them (which took me literally months), and now I hardly ever switch them on. The only two that get any use is the X5D (a great live workhorse) and the SMPro V-Machine, which can play the samples of all the rest. Apart from that, I do everything in the box nowadays.

I still have some of them "wired up" in the studio though. so I can always go back to the hardware synth and edit a sound / re-sample if I really need to.

  • Like 1
Link to comment
Share on other sites

13 hours ago, Notes_Norton said:

Hardware MIDI synths last forever

Not necessarily.

My JP-8000 gave up the ghost last year. I could probably fix it but instead I decided to put it back into its carry case.

13 hours ago, Notes_Norton said:

Software synths must 'do the math' creating every note, every nuance, and every articulation while the sequence is playing

Since upgrading my workstation I don't really find this a problem.

13 hours ago, Notes_Norton said:

With practically zero latency I can add track after track in real time while the old ones are playing and have the new track in sync.

I don't understand this one... Are you tracking to your DAW or to a multi track recorder of some kind? If the former, then sure there is latency involved.

Back in the early 2000's I was entirely hardware with all my instruments going into a conventional mixer and I mixed and monitored from there and finally recorded a stereo mix into my workstation. So, yes, in that situation, zero latency.

Perhaps I am misunderstanding something here...

 

So, to answer the original question, at this moment in time I have feet in both the hardware and software camp. And you can't get more hardcore hardware synth than one of these:-

modulargrid_811967.jpg

No ROM... No presets :D

Link to comment
Share on other sites

Yes my TX81z and my VL70m might be 'doing the math' just as the VST does, but because the VST shares the computer's CPU, certain shortcuts are taken in the soft-synth to save processing time. So in the end it just doesn't sound the same.

I've never found tuned percussion on a software synth that sounds and reacts like my TX81z and my DS8

I did try sampling, but with the FM synths, the attack, sustain, brightness with different attack velocities isn't duplicated with the sampler, resulting in less expressiveness.

And nobody has made a physical modeling software synth, so the now discontinued VL70m is the only option. My VL can do things like add flutter tongue and throat growl to a brass instrument. It can change the vowel sound on sax patches as if I were changing the shape of my mouth to get slightly more oo and aah sounds. When the pitch-bend (reed emulation) on the sax is flexed, not only does the pitch change, but the tone changes as well, as it does on a real sax. I can do lip slurs on the brass patches. The timbre of the instruments change gradually with pitch realistically including the shift when woodwinds go from register to register. There are so many other nuances that PM can replicate. It's like playing a real instrument instead of controlling a synth.

To me, recreating the nuances of the instrument bring out the nature of the instrument much more than the tone.

But like I said, there is more than one correct way to make music.

It's looking so far like I'm the only dinosaur here ;)

Insights and incites by Notes

Link to comment
Share on other sites

20 hours ago, iRelevant said:

Just to comment on your points 2 & 3, unless it is a ROMpler ... these MIDI (digital) modules are small dedicated standalone computer which computes the sound ... similarly to what any VST synth is doing. 

A large chunk of the OP stems from this and the underlying CPU "issue" can be resolved by baking tracks discretely as one progresses when necessary. This is required less as machines get more powerful, but fear of commitment will challenge some.

Soft synths have a definite cost advantage over hardware (probably the biggest argument), and complex hardware synths can get even more complex with controls and menu drill-downs that make them unwieldy and slow. I prefer drilling with a mouse on a computer... Navigation with buttons on hardware I've outgrown.

Bottom line is having an adequate MIDI controller as one's weapon of choice. It boils down to a performance in MIDI either way, just which processor one chooses, and their workflow. Recording the audio from a synth is akin to baking a VSTi.

[Aside] The cost perspective and convenience alone allows more people to enjoy music production. For guitar, hardware requires setup, proper mic'ing, ability to play at proper levels, etc. to get it in the box. For less money, a new player has more options available already ITB. Back in the day with small kids running around that love playing with cables, I had to spend 30 minutes just to setup and break down hardware. No one mics a hardware synth... at least that I have ever seen.

Link to comment
Share on other sites

On 4/6/2019 at 2:35 PM, Notes_Norton said:

The way I see it is this:

 I'm a dinosaur too, and started out with MIDI keyboards and rack modules from Roland , Korg, and Alesis. But today I am comfortable with virtual instruments. If I played live as you do, I might have a different perspective that emphasize hardware. Still have a few Roland JV/XP series ROMpler synths sitting around. :D

Quote

1. Hardware MIDI synths last forever. I have a Yamaha TX81z, a Korg DS-8 and a Roland MT-32 that were manufactured when PCs still used DOS-5 or maybe Windows 3.1, Atari/ST computers had built in MIDI ports, and Mac computers used Motorola CPUs. Some of the sounds on these modules are stale, but others are stellar and have no contemporary equal. Plus how many software synths from that era still work today? Zero. How many software synths have been orphaned when computer OS systems evolved? Plenty

Not so sure about the forever part. Depending on wear and tear, something is likely to give out eventually. Then the search for parts and schematics will likely be underway.

 

Quote

2. Hardware synths store all their sounds in ROM. Software synths must 'do the math' creating every note, every nuance, and every articulation while the sequence is playing. This taxes the computer's CPU which is also running the DAW and perhaps other background jobs

The part about software synths sounds correct, but hardware synths don't store everything in ROM. In regards to a digital ROMpler, typically just the raw waveform is stored in ROM, but then the hardware synth  also processes that signal by combining oscillators,  using filters, modulations for various envelopes and LFOs, and effects in real time as you play the instrument. This real-time workload is also what is being emulated in a virtual instrument using the CPU in the computer host.

 

Quote

3. Since all the sounds in hardware synths are stored in ROM, the latency is negligible and they all run about 5ms plus or minus 1ms. I've seen some software synths that have a latency of almost a half second.

I don't believe the ROM theory is responsible for the reduced latency. Hardware is just dedicated to the  real-time audio task, and a general purpose computer is not. But I can still  easily get latency under 10ms with my virtual instruments.

 

Quote

4. Since there is no tax on the CPU and since all the hardware synths have the same latency I can mix sounds from my rack of 10 or more synths. I can pick the best snare drum for the song from one module, the best kick drum from another, the best bass from another the best guitar from another the best trumpet from another the best sax section from another and do on. They will all be in sync.

Doing that all on one computer is a balancing act for the operator. There are free solutions like freezing instrument tracks as you go, or expensive ones like a major computer upgrade.

 

Quote

5. With practically zero latency I can add track after track in real time while the old ones are playing and have the new track in sync.

Freezing tracks can accomplish that as well.

 

Quote

6. With zero load on the CPU for sounds, the hardware synths can be better sounding or better responding. I have yet to see any software synth that can do what the physical modeling Yamaha VL70m can do -- and it's been discontinued for years. The nuances it can reproduce are light years ahead of any sample or sample-modeling synth.

I'm not familiar with physical modeling hardware, but there are some interesting modeled virtual  instrument solutions such as Chromophone 2 from Applied Acoustic Systems (AAS),  Modartt's Pianoteq, Arturia's V Collection, and IK Multimedia's MODO Bass, for example.

I agree that you will never get zero load on a CPU running virtual instruments, but computers are getting more powerful every year. So you can throw money at the problem and get more tracks.

For reference to the path we have traveled, take a look at this comparison of CPUs from Intel 386 in 1985 to Intel Core i7 6700K 2015.

40 Years of Processing Power

"To put the difference the years have made into perspective: 1 second of processing by Intel’s (2011) i7 3960X would have taken the best 1985 personal computer over 4 hours 29 minutes!"

:D

 

 

 

Edited by abacab
  • Like 1
Link to comment
Share on other sites

1 minute ago, abacab said:

If I played live as you do, I might have a different perspective that emphasize hardware.

Absolutely.

1 minute ago, abacab said:

Depending on wear and tear, something is likely to give out eventually.

Definitely, in my experience.

And what about effects?

I have some VST effects that cannot be replaced with a hardware version...

Link to comment
Share on other sites

On 4/6/2019 at 7:35 PM, Notes_Norton said:
  1.  Hardware MIDI synths last forever. I have a Yamaha TX81z, a Korg DS-8 and a Roland MT-32 that were manufactured when PCs still used DOS-5 or maybe Windows 3.1, Atari/ST computers had built in MIDI ports, and Mac computers used Motorola CPUs. Some of the sounds on these modules are stale, but others are stellar and have no contemporary equal. Plus how many software synths from that era still work today? Zero. How many software synths have been orphaned when computer OS systems evolved? Plenty

Ironically, my TX81Z is the first hardware synth I've had die on me, well I say die, it's output is really distorted. I've just not got around to narrowing down the issue (I'm guessing it's the D/A converter(s)).

5 hours ago, Notes_Norton said:

I did try sampling, but with the FM synths, the attack, sustain, brightness with different attack velocities isn't duplicated with the sampler, resulting in less expressiveness.

You can do this, but it involves sampling far more velocities. I normally sample 4 velocities, but for some sounds I'll go up to 6 or 8. You can even sample different modulations (like mod wheel, breath control etc) but it takes forever to sample, and results in very large instruments.

5 hours ago, Notes_Norton said:

And nobody has made a physical modeling software synth, so the now discontinued VL70m is the only option. My VL can do things like add flutter tongue and throat growl to a brass instrument. It can change the vowel sound on sax patches as if I were changing the shape of my mouth to get slightly more oo and aah sounds. When the pitch-bend (reed emulation) on the sax is flexed, not only does the pitch change, but the tone changes as well, as it does on a real sax. I can do lip slurs on the brass patches. The timbre of the instruments change gradually with pitch realistically including the shift when woodwinds go from register to register. There are so many other nuances that PM can replicate. It's like playing a real instrument instead of controlling a synth.

For the reasons mentioned earlier, I wouldn't even bother trying to sample a synth that was reacting to so many different controller types. 

I use the VL card in my MU100R, or my VZ-8M for my wind synth or SY77 with breath controller. That's one of the reasons they're still in my rack... but to be honest I don't get much use out of them. Like I said, it is possible to sample them, but you have to sample the synth at every possible combination of controller setting. It's just not worth it IMO... especially since I've actually got the synths!

  • Like 1
Link to comment
Share on other sites

I used to rely on hardware MIDI modules. I don't any more. With the power in new processors and the huge number of good software instruments its very hard to justify hardware any more.  However there is still a place for it in live performance. I still use my Sound Canvas for a few things.  Its nice to have units but now I don't think its essential.  

Link to comment
Share on other sites

40 minutes ago, synkrotron said:

And what about effects?

I have some VST effects that cannot be replaced with a hardware version...

I have been wondering about this. Is it any possible way to sort of route the HW audio insert in the AI to an own audio out, route an audio track to the current midi track, and then add VST effects to that audio track? Or something like that... ?

I know it is easier to render (record) the audio directly, but maybe you need the ability to edit the notes when adding the VST effect.

Link to comment
Share on other sites

I have about a half dozen hardware synths and a Yamaha O2R  - about a half dozen processors / Effects. I rarely take the time to power them up - but when I do , nothing really comes close in software to stacking synths and applying a custom real time mix on them. Most of the keyboard work I do involves the basics Rhodes , Piano, Organ, and typically just a touch of synth. Those basics are well covered in VST's - typically better then standalone voices in hardware. But, the occasional super fat synth stack doesn't come even close to using hardware IMO. I went through a period a couple of years ago where I bought  / sold and refurbished some pretty valuable older analogue synths - it seems like the prices doubled about 6 months after I sold my last collectable.

Link to comment
Share on other sites

You might like to take a look at Tassman for physical modeling. Still works perfectly on Windows (and my guess probably will for a long time to come despite being discontinued). Keep an eye on AlphaForever too as that's new, maintained and updated. Hopefully it'll surpass Tassman's ability at some point.

As for the thread, nah i don't use hardware... although i wish the hardware manufacturers would release their stuff in software form more often! ? I'd buy Analog Heat in a heartbeat.

Link to comment
Share on other sites

Most of the time, when clients walk into my studio, they don't get impress by how many VSTi & plugins I have, instead they see the racks of stacking hardware all over my studio.  For them, that's positive indication that they go to right place to make music.... ?

  • Like 1
Link to comment
Share on other sites

Please sign in to comment

You will be able to leave a comment after signing in



Sign In Now
×
×
  • Create New...