RMT vs. Softsynth

[makary]

Hi,

 

thanks for opinions.

 

"Cracow..." is made using 4x. The main reason for using this mode is the drums. I simply need as quick as possible frequency shift to get really convincing bassdrum or snaredrum.

 

For instance if you design a bassdrum sound you need:

1. trigger / click -> usually made of a short noise;

2. the body -> usually made of a frequency shift;

3. the release -> usually made of a constant frequency with quickly fading volume / amplitude.

 

If everything happens too slowly (1x for instance) the sound tends to resemble a poor sample of bassdrum pitched down.

 

Usually the main reason and often the only reason for using 4x is related to the drums. Other instruments would work fine in 1x. Just small adjustments would be necessary.

Edited October 30, 2016 by makary

[ivop]

Perhaps the new tracker emkay is working on should also allow each channel instrument to run at a different rate

Edited October 30, 2016 by ivop

[makary]

Really great and unique pokey music can be written using 1x. Proofs are provided by Dhor, Miker or Triace for example I just mention these guys in particular because I listen to their music frequently. Of course that kind of music can be used as game music, demo music and so on. Extremely big advantage, that's clear.

 

Using 4x is just pushing Pokey to the limits to see what audio quality can be achieved. Beside better drums, one can get a lot of interesting results by using short and quick envelope loops with bass sounds (a sort of YM buzzer effect), quick and wide frequency movements using portamento (a sort of Frequency Modulation sounds and effects) etc. Those require 4x so it's a kind of l'art pour l'art

Edited October 30, 2016 by makary

[rdefabri]

 

I'm an old metalhead and Pokey sounds get on my nerves sometimes but... Today I had to check if this is real:

 

https://soundcloud.com/user-799776082/cracow-acid-city

 

And it really is. Kudos to you Makary!

 

pretty awesome!

[emkay]

You know what ? Even on Atarionline.PL I stated that it sounds great.

Hats will be off, after listening to a piece of music, using 3 octaves for a melodic voice and use such sounds .... as the "technique" uses similar CPU time as some Softsynth tune.

Edited October 31, 2016 by emkay

[rdefabri]

Thanks. Looks to be no digtized stuff, in any case it's RMT running at 3 iterations per frame.

 

For the dummy that is me, can you clarify what this means? Is it a mode in RMT or something different?

[Rybags]

(4 iterations per frame actually).

 

Makary explains the reasoning for more RMT calls in his earlier post up the page http://atariage.com/forums/topic/43419-rmt-vs-softsynth/page-2?do=findComment&comment=3621581

 

Normally you run RMT at once per frame, ie 50 fps for PAL, 60 fps for NTSC. There's a trick to make music run the same speed and that's to omit every 6th RMT call if it's an NTSC machine.

Of course music itself rarely runs anywhere near that fast but on the Atari we have to take care of ADSR and frequency shift effects in software.

As explained for better quality special instruments like accurate sounding drums even 50 fps isn't enough, higher fidelity comes by cranking up the rate even more.

 

So, the number 1 reasoning for going beyond 50 RMT calls/second is better sounding instruments. As the raster bars show it chews up a fair chunk of CPU time but if you were doing drums by digital sampling the CPU consumption for the effect as well as the tracker would leave just about nothing.

[makary]

As regards RMT vs. Softsynth... I know the former a bit, I'm learning to use the latter.

 

When using RMT (or other Pokey based tracker) a musician is limited to using a small range of waveforms. Basically noise + square and its variations. Of course I can mix some waveforms, add some HP filter and so on to add more variety to the timbres produced by Pokey. Still the domain of what can be produced is really limited.

 

When using Softsynth a musician can define almost any waveform he wishes to use. Consequently the amount of timbres that can be achieved is really incomparable to the capabilities of the regular Pokey trackers... Well, just listen to "Wodospadem Czasu" again

 

Beside Softsynth music being so demanding for CPU, it is also demanding for a musician "The composer" is based on a musical programming language. The language is simple but still you need to "switch" mentally to it's unusual representation of notes and their properties.

[rdefabri]

(4 iterations per frame actually).

 

Makary explains the reasoning for more RMT calls in his earlier post up the page http://atariage.com/forums/topic/43419-rmt-vs-softsynth/page-2?do=findComment&comment=3621581

 

Normally you run RMT at once per frame, ie 50 fps for PAL, 60 fps for NTSC. There's a trick to make music run the same speed and that's to omit every 6th RMT call if it's an NTSC machine.

Of course music itself rarely runs anywhere near that fast but on the Atari we have to take care of ADSR and frequency shift effects in software.

As explained for better quality special instruments like accurate sounding drums even 50 fps isn't enough, higher fidelity comes by cranking up the rate even more.

 

So, the number 1 reasoning for going beyond 50 RMT calls/second is better sounding instruments. As the raster bars show it chews up a fair chunk of CPU time but if you were doing drums by digital sampling the CPU consumption for the effect as well as the tracker would leave just about nothing.

 

Ah, yes - I understand and that makes sense. Emulating the ADSR envelope in software could be problematic for more dramatic changes, so more cycles would alleviate that issue. Thanks for the explanation - 2 more dumb questions:

 

1) consuming more cycles then means it's not possible to have music playing during another process (e.g., a game), correct? This would be similar to how Berzerk freezes when the speech is played....

2) Is there a setting in RMT that allows for the increased cycles / iterations? I'm guessing there is, I'm just not familiar enough yet.

 

Thanks again!

Edited November 1, 2016 by rdefabri

[rdefabri]

As regards RMT vs. Softsynth... I know the former a bit, I'm learning to use the latter.

 

When using RMT (or other Pokey based tracker) a musician is limited to using a small range of waveforms. Basically noise + square and its variations. Of course I can mix some waveforms, add some HP filter and so on to add more variety to the timbres produced by Pokey. Still the domain of what can be produced is really limited.

 

When using Softsynth a musician can define almost any waveform he wishes to use. Consequently the amount of timbres that can be achieved is really incomparable to the capabilities of the regular Pokey trackers... Well, just listen to "Wodospadem Czasu" again

 

Beside Softsynth music being so demanding for CPU, it is also demanding for a musician "The composer" is based on a musical programming language. The language is simple but still you need to "switch" mentally to it's unusual representation of notes and their properties.

 

I'm a tinkerer, but I really struggled with Softsynth. I'm absolutely getting RMT, even to the point where I can experiment (albeit, very simplistic as opposed to what you or others are doing). I like the RMT interface, it makes a lot of sense.

 

For me, I want the experience and the output to be something that would be achievable on a real Atari 8-bit machine. That's why I've avoided using stereo mode on RMT, even though it's cool. If memory serves, Softsynth is executable on an 8-bit machine, so I'd love to get into it more, but it seems a bit complicated and I don't think there's a manual in English (although I could be wrong).

[makary]

 

I'm a tinkerer, but I really struggled with Softsynth. I'm absolutely getting RMT, even to the point where I can experiment (albeit, very simplistic as opposed to what you or others are doing). I like the RMT interface, it makes a lot of sense.

 

For me, I want the experience and the output to be something that would be achievable on a real Atari 8-bit machine. That's why I've avoided using stereo mode on RMT, even though it's cool. If memory serves, Softsynth is executable on an 8-bit machine, so I'd love to get into it more, but it seems a bit complicated and I don't think there's a manual in English (although I could be wrong).

 

I understand your points very well. I use a stock 65XE. Everything I write is supposed to be working well with that hardware.

 

Right, Softsynth is not extremely friendly. But step by step... we're bulding our relationship Last but not least, there is a manual in Polish.

 

Iterations in RMT: left-hand upper corner, "MUSIC SPEED: 10/10/1". That's the default setting and the last digit indicates iterations.

[rdefabri]

 

I understand your points very well. I use a stock 65XE. Everything I write is supposed to be working well with that hardware.

 

Right, Softsynth is not extremely friendly. But step by step... we're bulding our relationship Last but not least, there is a manual in Polish.

 

Iterations in RMT: left-hand upper corner, "MUSIC SPEED: 10/10/1". That's the default setting and the last digit indicates iterations.

 

Oh right - I did notice that music speed, and I played with it to no apparent effect. I'll dig into that - thank you!

 

My Polish isn't so good I'm an English speaking American, I used to speak Italian and French, but that won't help me much with Softsynth!!! :)

Edited November 1, 2016 by rdefabri

[makary]

 

Oh right - I did notice that music speed, and I played with it to no apparent effect. I'll dig into that - thank you!

 

My Polish isn't so good I'm an English speaking American, I used to speak Italian and French, but that won't help me much with Softsynth!!! :)

 

I'll try to translate some essentials of the manual in my spare time.

[rdefabri]

 

I'll try to translate some essentials of the manual in my spare time.

 

You are too kind! If I'm ever on the other side of the pond, in Poland, I'll buy you a beer. Maybe even virtually!

[Heaven/TQA]

Now that's someone explaining me after 30 years softsynth.... if I would know that earlier I would have forced my musicians to use it even when it's only for end tune....

[rdefabri]

How are you guys converting the RMT output to .mp3? I'd love to capture both audio and video from an emulator, but Camstudio tried to put too much nonsensical SW on my work PC, so I won't install it.

[pseudografx]

How are you guys converting the RMT output to .mp3? I'd love to capture both audio and video from an emulator, but Camstudio tried to put too much nonsensical SW on my work PC, so I won't install it.

You can use WinAMP diskwriter feature together with the ASAP plug-in.

[emkay]

You can use WinAMP diskwriter feature together with the ASAP plug-in.

That doesn't work correctly.

[pseudografx]

That doesn't work correctly.

It does for me.

[+MrFish]

How are you guys converting the RMT output to .mp3?

 

foobar2000 with ASAP plugin.