SoloRack v1.0 is here

Modular Synth design and releases (Reaktor, SynthEdit, Tassman, etc.)
Post Reply New Topic
RELATED
PRODUCTS
SoloRack

Post

Having a lot of fun with SoloRack! Works nice in Reaper with the multiple DAW in/out modules running to another instance of SoloRack on another track...I can see four rows now. :) Also started playing around using Silent Way LFO to feed into SoloRack....oh yea this is going to be fun!
Dave

Post

des wrote:Works nice in Reaper with the multiple DAW in/out modules running to another instance of SoloRack on another track...I can see four rows now. :)
As they say, Necessity is the mother of invention :tu:
www.solostuff.net
Advice is heavy. So don’t send it like a mountain.

Post

Here is a little piece I recently created with the SoloRack. All sounds come from one patch (10 rows) and all patterns, chord changes, etc. are generated through sequencers and random S&H.

I only added a bit (well, a lot) of external reverb (smartelectronix ambience) and delay.

https://soundcloud.com/martin-lueders/solorack-study

Image

Have a nice weekend,
Martin

Post

Wow that sounds great! So were the sequences started automatically or did you have to trigger them manually?
Dave

Post

des wrote:Wow that sounds great! So were the sequences started automatically or did you have to trigger them manually?
The main tempo is taken from the DAW (via the Tempo from Daw module, 5th row on the left).
But, I also could have used one LFO instead, as I am not really sync'ing to anything else (apart from one external BPM sync'ed delay).

This "master beat" makes sure that all other changes of pitches happen on the beat and in sync with the various random melodies. For these random melodies, I usually use a sequencer module which is running at a higher frequency (possibly modulated) and then use the output of those sequencers through a sample and hold module, which again is in sync with the song.


Cheers,
Martin

Post

Very impressive! I see you have several lines coming out of the Midi Notes port. What are those being used for?
Dave

Post

des wrote:Very impressive! I see you have several lines coming out of the Midi Notes port. What are those being used for?
They go into oscillators to set the "master tune" for the patch. Again, they are currently not necessary. I did patch all the external CV2Midi and Tempo from DAW as I wanted to be able to include the patch into projects with other instruments as well. But then it developed into the direction that I wanted to see what I can do with one rack (plus effects) alone.


Martin

Post

Very nice sounding martin_l, and quite amazing that you achieved all that in one patch!! :o :tu:

I also wonder how do you feel about the scrolling up and down in such a large rack, while dragging cables and modules? Is it good or slippery? As the graphics have improved in beta 2, the scrolling may become too fast at times. Or may be you prefer using the mouse wheel?
www.solostuff.net
Advice is heavy. So don’t send it like a mountain.

Post

S0lo wrote:Very nice sounding martin_l, and quite amazing that you achieved all that in one patch!! :o :tu:

I also wonder how do you feel about the scrolling up and down in such a large rack, while dragging cables and modules? Is it good or slippery? As the graphics have improved in beta 2, the scrolling may become too fast at times. Or may be you prefer using the mouse wheel?
The scrolling works quite smoothly. Not too fast, as you can control the speed by how far you go beyond the current view.

One thing, I was wondering, though, is whether you could make the downsampling an optional feature for the additional outputs. In this patch, I use several of the additional outputs in order to be able to use different external effects for various components of the sound. To be honest, I did not notice much aliasing on the other outputs, but that is also because I used it either for the bass sound, or for quite harsh sounds anyway. I think the option to enable the downsampling filter for them would be a great feature. They even could come as separate modules with a pair of input and a pair of output connectors.

Cheers,
Martin

Post

martin_l wrote:One thing, I was wondering, though, is whether you could make the downsampling an optional feature for the additional outputs. In this patch, I use several of the additional outputs in order to be able to use different external effects for various components of the sound. To be honest, I did not notice much aliasing on the other outputs, but that is also because I used it either for the bass sound, or for quite harsh sounds anyway. I think the option to enable the downsampling filter for them would be a great feature. They even could come as separate modules with a pair of input and a pair of output connectors.
The downsample filter is a CPU heavy FIR. The decision to leave out oversampling (downsample filter) for the extra outputs was hard to make for me. The reason for it being this way is pure CPU usage, because oversampling is handled by SoloRack's engine not by the modules them selfs. Because of technical reasons, doing it for all outputs would either have to consume much more CPU, or at least unnecessarily complicate the code for finding out which outputs are being used and which are not, and do it only for the used ones. And this all had to be done in a critical section of the code where CPU is crucial.

I speculated that about 90% of the users are only going to use the main 2 outputs most of the time. And hence I decided to not oversample (downsample filter) the other outs for the time being. May be I'm wrong.

I could do it, it will affect the cpu more or less. Or may be change the scheme, and let the modules handle the oversampling them selves. This would solve the CPU issue, but it's a strategy change. It means that future 3rd party devs will have to handle their own oversampling, plus a few other complications regarding who does what. Needs more thinking :)
www.solostuff.net
Advice is heavy. So don’t send it like a mountain.

Post

S0lo wrote:
martin_l wrote:One thing, I was wondering, though, is whether you could make the downsampling an optional feature for the additional outputs. In this patch, I use several of the additional outputs in order to be able to use different external effects for various components of the sound. To be honest, I did not notice much aliasing on the other outputs, but that is also because I used it either for the bass sound, or for quite harsh sounds anyway. I think the option to enable the downsampling filter for them would be a great feature. They even could come as separate modules with a pair of input and a pair of output connectors.
The downsample filter is a CPU heavy FIR. The decision to leave out oversampling (downsample filter) for the extra outputs was hard to make for me. The reason for it being this way is pure CPU usage, because oversampling is handled by SoloRack's engine not by the modules them selfs. Because of technical reasons, doing it for all outputs would either have to consume much more CPU, or at least unnecessarily complicate the code for finding out which outputs are being used and which are not, and do it only for the used ones. And this all had to be done in a critical section of the code where CPU is crucial.

I speculated that about 90% of the users are only going to use the main 2 outputs most of the time. And hence I decided to not oversample (downsample filter) the other outs for the time being. May be I'm wrong.

I could do it, it will affect the cpu more or less. Or may be change the scheme, and let the modules handle the oversampling them selves. This would solve the CPU issue, but it's a strategy change. It means that future 3rd party devs will have to handle their own oversampling, plus a few other complications regarding who does what. Needs more thinking :)
I think I don't understand (or misunderstand) how you handle oversampling. My thinking is (and that's how my plugin works) that internally everything works at the over-sampled sample rate (not even aware of the external sample rate). Only at the very end, i.e. in case of the SoloRack I would presume this is the 2DAW module, the down-sampling should happen.

The down-sampling itself would consist of two parts: the down-sampling filter to make sure that no frequencies above the external Nyquist frequency survives (or at least they are strongly attenuated), and then the actual down-sampler, which either only picks every n'th sample, or does some further averaging. As I understood your previous posts, in the 2DAW module (2 standard outputs) you have both parts, while the additional outputs only have the down-sampler without the filter.

If this is correct, it should be rather simple to insert additional and optional filter for each of the additional outputs. I do know that they are very expensive, hence my idea to make them optional (either through a small switch next to the output sockets) or even as an additional independent module which performs that filtering (without the downsampling), which the user could then insert before the output, if needed.

I also do see the need for an unfiltered (and hence aliasing) output if you want to send control data. For instance, I tried to write an external CV2Midi plugin to be used with the Softube modular, and that does not work exactly because they have anti-alias filters on all outputs. If you get the trigger pulse on one output the voltage on the other is still wiggling around the value you want, because of the filter.

Obviously, you cant output the over-sampled signal directly, as the DAW is only working at the normal sample rate. That also limits the idea to split larger patches into several racks which are connected through these additional outs and ins.

In my understanding (as I said, I don't know whether I do understand how you implemented things) an additional anti-alias filter with two ins and two outs and possibly a single control for the order of the filter would be the simplest option, as it would not require any changes to your existing modules.

Cheers,
Martin

Post

Thanks for your suggestion about the anti aliasing filter. I will try to figure out the best way to do it. Without confusing the user, keeping a good standard for 3rd party developers and while keeping CPU usable.
www.solostuff.net
Advice is heavy. So don’t send it like a mountain.

Post

hi Solo,
recently, i have play a lot with solorack, and the engine is really good!
an Eta on futur modules?
( pwm at audiorate and delay time are sweet!)
in repear, the automation parameters works perfectly, but in studio one 3.5.1, the host didn't detect anything.

Post

kokotte wrote:hi Solo,
recently, i have play a lot with solorack, and the engine is really good!
an Eta on futur modules?
Thanks. Three new modules are already finished. More will be in the works. Eta,, hmmm have not decided yet. But if things go well. May be within a month. But again, I've been wrong before with my estimations :)
kokotte wrote: ( pwm at audiorate and delay time are sweet!)
I'm glad that you liked and noticed. Specially Audio rate pwm which is one of the tough things to do right digitally. specially when combined with sync.
kokotte wrote: in repear, the automation parameters works perfectly, but in studio one 3.5.1, the host didn't detect anything.
Thats a first regarding Studio One!!. I didn't test it in that DAW yet. thanks for reporting that.
www.solostuff.net
Advice is heavy. So don’t send it like a mountain.

Post

Hi SOlo,

I was experimenting a bit more, and think I found a bug in the Sample&Hold module.

When the trigger is not fully regular, e.g. when some trigger pulses are suppressed, it seems that the S&H module is taking an additional sample after the actual trigger pulse. As it is a bit difficult to explain, I include a picture of the patch and also a screen shot of an oscilloscope, showing the trigger pulses in the upper part and the S&H output in the lower. I marked the "wrong" samples with an error. As you can see, the output value of the S&H is changing even though the trigger signal does not have a pulse at these times.

Image
Image

When the trigger is regular, this effect is not obvious.

You can download the patch here: https://www.dropbox.com/s/g9mcjau7vfzvt ... g.srp?dl=0

I am using Reaper 5.40 (32bit) as host (Win64).

Please, let me know whether you can reproduce this behaviour.


Cheers,
Martin

PS. It seems the embedded images are not working. Please try to open the links directly:

https://goo.gl/photos/i3YMVJGcJEoKruNz9
https://goo.gl/photos/S6BGMjmZC9fkMLiU7
Last edited by martin_l on Fri Aug 04, 2017 9:29 pm, edited 1 time in total.

Post Reply

Return to “Modular Synthesis”