Percussa Home / Shop / Contact /

A long list of questions

#1

Hello. I’ve been considering picking up an SSP for years, now, and even though it’s been years I’m still unsure. Currently I am leaning even more towards yes on the fact alone that the SSP can be used as a multichannel audio/DC interface/recorder, however I’m still on the fence as to justify taking up that much rack space I need to consider all the modules that I won’t be able to fit in my case - particularly for live performance. I’m really hoping the SSP is not just a capable substitute, but a joy and inspiration to work with.

Here are a lot of questions which I think will really help me gain a better understanding of what creative possibilities are at hand as well as what limitations. I have read the manual, but there are a lot of modules with no links and some that do are just support issues or feature requests that tell me little. I have also watched videos at one point or another and dived deep into many forum posts, software update announcements, etc. - indeed I have spent hours trying to answer the following questions on my own, though it’s likely I missed a point here or there. So apologies for the long list of questions and thanks in advance for any time spent in response. None of these questions alone are remotely close to being deal breakers or anything, I just really want to know!

ADSR

  1. Do the ADSR segments support curves (log, exp, etc)? (exponential is particularly important for percussive sounds, for example)
  2. If yes to #1, are those curves modulatable?

Reverb

  1. What parameters are available?
  2. Can all parameters be modulated?
  3. Can more than one reverb be used? There is no entry for the reverb module in the forum manual index, though it is mentioned in the global effects section. There is also a forum post complaining about how unusable the module is and that nothing can be changed or modulated, but it’s really really old and I’d be surprised if that was still the case.

Delay

  1. Did this ever get sorted out? Modulating delay time is super important to me: Delay and Reverb Modules If that has been sorted out, what is the algorithm? Is it the cross fading type that results in smooth changes (like a 4MS or Disting) or is it the doppler effect?
  2. Is feedback patching possible (e.g. low filter in the feedback path)?
  3. Can more than one of these units be used?
  4. Is the delay clockable and if so can the clock be divided and/or multiplied?

MIDI

  1. Can MIDI CCs from a controller be used to control parameters? (The current MIDI documentation explicitly states that this is not possible, however I’ve seen forum posts suggest that this was added at some point)
  2. If yes to #1, am I able to choose what channel? For example Channel 1 CC 1 vs Channel 2 CC 1 and use them both without conflicts?
  3. Again if #1, is anyone using a MIDI Fighter Twister controller with the SSP? That would make me extremely happy (love that controller, I have two, but zero usefulness in Eurorack for me now - this could be its own forum post I guess)

Simple Delay

  1. What’s the unit precision? Milliseconds? Microseconds?

Of all of these questions this is the most important for me. I need something that will give me 100% wet delay down to an increment of at least 1ms on an input to compensate for various latencies throughout my system introduced from other digital modules. The way I deal with this now is not ideal and really hoping I can solve this with the SSP.

Network page

16 pages of 8 slots :thinking:

  1. Can things be moved after being created without losing connections for organizational purposes?
  2. Any other visual tools to aid organization?
  3. Does anyone ever run out of slots ie. is 128 the maximum module count? If so, why is there a limit?
  4. Are interconnections between modules constrained to those within those 8 slots, and is that what the BUS module was made to address?

General

  1. Is there a concept of abstraction or encapsulation? I see that there is a “macro” control, but has any progress been made on Meta Control Module (you REALLY want this)? Either more outs from the DCG module or similar?
  2. I see that there are 200 presets but I imagine these are complete snapshots of the entire state rather than “units” that can be mixed & composed. Is that true?
  3. If yes to #1, are things like “change all instances of the microtonal quantizer’s scale at once” possible?
  4. Is there a way to live loop / sample / “granularize” incoming audio?
  5. Is there a way to pitch and/or frequency shift “live” signals?

Thanks again

1 Like
#2

For question about midi

[1 & 2]
You can use MIDI CC from MIDI controller to control parameter of SSP’s module.
And you can choose midi channel and midi CC Number in MIDI module and the number of modules to insert is not limited. So, you can split Channel 1 CC 1 and Channel 2 CC 1.

[3]
You can use a MIDI Fighter Twister if it is Class Compliant USB midi devices.
I tested several midi controllers which are class-compliant, and those worked fine.

For me, the midi function is one of the reasons I bought SSP.
I love midi pad controller and I wanted to build a sampler which has a unique feature at eurorack controlled by pad.
But it will be a complicated setting if without SSP.
(Imagine, how many midi module and sampler is needed using 16 pad with midi velocity and CC!!)

If there was no SSP, I didn’t start Eurorack.

#3

Awesome. That’s a big plus for me and means I won’t have to sacrifice my sacred analog offsets.

The MIDI Fighter Twister is advertised as class compliant, however I have experienced issues across different OSs but there are so many variables who knows what the cause (could be banging them around in my backpack for years). Thanks for the reply.

#4

Many questions with a quick reply - Hope this helps you make your decision. In summary the SSP is a unique machine in Eurorack land that can be used in many many ways.

ADSR

  1. Attack: Exponential to linear. Decay and Release: Linear to Logarithmic
  2. Yes

Reverb

  1. Mix, Ambience, Speed, Image
  2. Yes
  3. As many as you like. Reverb is an independent module that can be patched anywhere in the signal chain

Delay

  1. With -this - I guess you mean modulating the delay time? Unfortunately not yet. At the moment the Delay is a set and forget as far as delay time is concerned
  2. Within the grid certainly. You can route the delay output to anything you like and ultimatley send it back to its inputs. Not the -every- patch connection has offset and scaling controls at its input and output. In this case this would allow you to very precisely control a feedback path
  3. As many as you like. delay is an independent module that can be patched anywhere in the signal chain
  4. This would be the case if the delay time modulation would work

MIDI

  1. Yes. 8 controllers can be defined per MIDI module. You can use as many MIDI module as you need
  2. Yes
  3. Don’t know about that one. Never heard of the MIDI Fighter. If it is class compliant, than most likely

Simple Delay

  1. Hard to tell. The range is 0.000 to 0.995 in 0.005 steps

Network Page

  1. Only Copy/Paste of individual modules within a Preset
  2. This display is fantastic. Navigation through the grid is easy. The display is used to great extent in some modules and sparingly (under)used in other ones.
  3. I have never used more than 8 pages and have never run out of space / CPU cycles either so havent really maxed out the SSP (although I only work in 48kHZ and switching some Presets to 96kHz immediately creates sonic mayhem in more extended Presets)
  4. You can either use multiples of the same module (e.g. a MIDI Input module for every couple of pages) or use BUS modules. The BUS module is unique to the SSP and is essentially a 512 * 512 signal routing matrix - in 64 groups of 8 inputs and 64 groups of 8 outputs

General

  1. Only Copy/Paste of individual modules within a Preset
  2. Yes. A Preset is a complete set of modules patched together. To get in the direction what you’re referring to is a matter of creating your own template Presets (Though this will give you a head start, it does not cover assembling from multiple templates)
  3. No. This is not always what you would want either.
  4. Yes. There is a dedicated module that does that
  5. Not sure. There is no dedicated Pitch Shifter module and I’m not sure whether the Grain or Sampler module allows for that. It wouldn’t surprise me if it did though. I havent got time to test it right now.
#5

a couple of things I’d add…

feedback patching - on any digital system will have some limits ‘inter’ module due to the use of sample buffers. (128 *44.1k on SSP) - but for many use cases it’ll be ok.

pitch shifting -
GRA should be a able to do it, but I suspect it wont be as clean as you might want.
CLDS (my VST) also has ability to pitch shift - in a number of ways , see the Mutable Instruments clouds manual … and dont forget the extended modes :slight_smile:

Reverb usablity…
yeah, I think the comments you refer to are due to it it not handling DC offsets well, which means you should put a HPF in front of it. However, my PMIX mixer removes this issue , you can put RVB on an AUX channel an it will remove DC offsets ‘automatically’.
(I find removing DC offsets makes a lot of the SSP modules sound better/more reliable)

tape modulation - again, perhaps play with my CLDS module… which has a looping delay mode (again in clouds manual)

Im pretty new to the SSP, and had some similar expectations to those you list…

I love the usb audio interface aspect, its really excellent - makes recording things very simple, and also bring a laptop into the mix if you want :slight_smile: e.g. playing with BWS grid or VCVRack

Recorder, very useful, again the fact it records multi channels is excellent.

this plus the number of physical inputs means Im using it as the centre of my rack and even bring in other synths and an octatrack… its really the hub.
this is why I developed PMIX so that i could easily control levels etc.

so something things that I love…
I love the screen, and really like the patching workflow.
support for VSTs for 3rd party dev (I’ll return to this ;))
INP/OUT show waveforms on all IO - very useful
scale/offset on every module inlet and outlet
GRA, STE, SAM, BUS modules :slight_smile:
(ARP is starting to grow on me too)

there are a few things, that I think are ‘functional’ that I don’t personally gel with.
in particular RVB and LFO just don’t do anything for me… but perhaps its just I cannot ‘dial them in’ - MATTS in his youtube video loved the RVB … so perhaps its ‘taste’

there are a few that I get mixed results with … i.e. sometimes great, sometimes not so great , WTO falls into this category for me.

but the SSP is modular, so mostly its about how you patch it and use it… its whole is greater than the sum of its parts.

which brings me on to my final point… which is quite personal…

the SSP is a big undertaking for a small company… not only has Bert produced lots of great modules - but also the hardware and the software foundation that these run on.
This is why Im really excited by the latest improvements to the VST support.

I think there is a tremendous opportunity now for us to help move the SSP forward, to bring some more great modules to the SSP, to let the community help with the development.

so… as I mentioned I didn’t gel with the RVB , and there were some limitations to DLY - thats why I created CLDS, as Clouds is a well respected in these areas - and it works great on the SSP.
I loved the PMM module (still do), but wanted more physical modelling so added RNGS (Rings)
and I wanted a bit more beyond the oscillators from LFO, so have PLTS (plaits).

so where I see ‘gaps’, Im looking to fill them… also with utilities like PMIX, or a new matrix switch I have in dev…

in fact, Ive so many ideas… its hard to know which to work on next :slight_smile:

2 Likes
#6

Thank you @titaanzink and @thetechnobear for the detailed replies. That’s a lot of good news as far as my concerns go. The SDK also looks quite promising. Hopefully I can try my hand at it some day.

1 Like
#7

So, my main reasons for considering an SSP right now are:

  1. 16 input interface for recording into DAW
  2. 16 input mixer for live performance without DAW

Ideally I’d love to have the only difference between 1 & 2 be if the SSP is plugged into the computer or not. I’m fairly certain this will be the case, but I have just one detail I’d like clarity on:

I absolutely need per-input delay offsets that are accurate to the millisecond. I enquired about the resolution of the Simple Delay unit, to which @titaanzink replied:

Hard to tell. The range is 0.000 to 0.995 in 0.005 steps

Does anyone know what is represented by that range and step resolution?

I need very specific, static offsets on all the inputs to account for latencies between modules in my rack. Besides compensating for latency, an ms accurate delay can do wonders for a mix of otherwise “on time” signals. Anyway I need accuracy of 1ms, and need to be able to set that anywhere between 1-10ms (hopefully more than 10ms in supported).

I assume that this is totally possible and am also hoping that having 16 of these is not noticeably taxing on CPU. Cool thing I’ve realized is that this alone (in addition to interface capabilities) is enough for me to justify purchasing an SSP, so everything else it can do is just a really, really big cherry on top.

#8

I’d need to check with a scope to be accurate (and I dont have time right now) …
but I’d suspect SDL is not down to ms ( im pretty sure SDL max > 1 sec , even it if was 1 sec, if it used linear time that would be 5ms, it is not linear time then not very useful for delay compensation).
however, as I said I’d need to check to be sure.

yeah, the SSP is great as a mixer, thats been one of its primary purposes for me… hence why I wrote PMIX. as you say, so useful, that you can just connect USB and you can easily multitrack into your daw etc.

delay compensation - yeah, this is not something I don’t think Ive seen in eurorack,
(anyone know of a eurorack module allow multi channel delay compensation?)

but for sure many of users are using an increasing number of digital modules, so it’s becoming more important. however, i don’t remember ever seeing a discussion about it on eurorack forums!
so, I guess many do the ‘final mix’ in the daw, so handle this there… so little demand?

that said, it be pretty easy to create a sample accurate delay ( * ) , or even add delay compensation to PMIX, just its a very limited use-case… and not very ‘exciting’ :wink:
(delay lines are not cpu intensive… so thats a non-issue)

( * ) for delay compensation, I think delay should be sample accurate, not just millisecond.

2 Likes
#9

I’d be really surprised that a digital delay wouldn’t be expressed in milliseconds or fractions thereof. I don’t think I’ve ever used a non-clocked digital delay that wasn’t. Of course microseconds would be better. Very helpful for getting gates/pitch CV to line up when using a quantizer (O_C and possibly Scales have microdelay tuning options to account for this).

OH! :laughing: what about the Delay module?

I think the creative use case for millisecond delays in electronic music speaks for itself. Or maybe this is just semantics, yes, sample accurate, as long as the value we are dialing in isn’t expressed in samples (or has options to control what unit is being is used, would be clutch).

Thanks for the reply. Anxiously hoping to learn what the Simple Delay, or Delay modules can do here.

#10

i think expressed in samples would be more universal (in line with other interfaces in the studio) and more general when using different sample frequencies on the SSP (48 vs 96 vs 192 khz)

#11

delay modules… often are in time units, but not always… e.g. Ive seen many eurorack modules where its just labelled as time with no units. but for sure with a digital display its easy enough to display time.

as for doing down to mSec delay times, sure its useful in a few use-cases, but not really that useful as a delay-fx (imho)

however, your main question seem to be about delay compensation, here samples tends to make more sense, since the time value can be pretty small.

anyway, no particular reason for it to be either/or a module could display both.

btw: can you talk about a solid use-case that you have for this?
what modules are you trying to use delay compensation for?

1 Like
#12

I have to respectfully disagree about millisecond delays being “useful in a few use-cases”. This is the main ingredient of Karplus Stong, chorus, flanging, phasers, resonators, the old trick of delaying only the left or right channel of an otherwise mono signal to give it stereo width. It is an immensely powerful tool for adding varying shuffle or “humanization” to a groove, especially when the time can be modulated, modulate the modulator and turn those hi hats into a robot every once in a while. It’s how you make a clap!

But yes, I’d love to be able to use a basic millisecond delay to push my non-latent signals forward to match my latent ones. This is how Ableton’s track delays work. They are expressed in milliseconds. Likewise Ableton’s MIDI sync offsets are expressed in milliseconds. Now, Ableton does express driver compensation in samples, but that’s a setting for the program - doesn’t address the issue of tracks needing to be nudged. If you have something that is latent, you adjust by ear and measurements of time are more natural to us.

The main offender in my rack is an ER-301 - a whopping 9ms. I work around this by routing a sub mix of non-latent signals into the 301. It works perfectly, but then I cannot multitrack in this scenario. In a studio setting with either a Percussa SSP or an Expert Sleepers ES-9 (the two options I’m considering) this will be no problem as I can use Ableton’s track offsets, and in a live situation I’ll just fallback to routing through my ER-301. I would much rather have 100% wet delays set to 9ms via a Delay unit, and then further adjust these to gain more coherent grooves (on time isn’t always “on time” :wink:) regardless of being in the studio or playing live without a computer.

To summarize, a millisecond delay is a generic utility for sound design, but can even solve my very specific latency problem.

1 Like
#13

Ableton allows track delays to also be specified in samples for more accuracy (click on ms, to switch)

yes, Im very familiar with building fx, and yes we use small delay lines all over the place.
in the DSP code this is in samples, only at the ‘user level’ is this rounded to mSec, just to simplify the display etc … but this inaccuracy (and audio buffers) means often building compound fx in user space has quite some limitations.
a solid example…
you mention karplus strong. milliseconds are no where near accurate enough, if you want to have it tuned… so a user-space delay used for this is a ‘rough approximation’ at best.
(if you know the maths, think of the size of the delay line of 440hz compared to 441hz)

so I guess… our view just differs by how much we value precision, and if we are able to accept the ‘issues’ lack of precision causes.

wow, 9ms on the ER 301…that knocks it off my wish list :wink:
but yeah, I’d want to some how get rid of that in a ‘live setup’ too… so I can see your ‘issue’
(I’ve a similar, but not so severe issue with the Qubit Nebulae)

ok, back to SDL - so I connected an oscilloscope to test it…

as expected, it’s no where near millisecond accurate.
0.0005 = 22.6 mSec, 0.010 = 42.6 mSec
so it looks like its a 4 second delay buffer…, with 20 mS resolution.

interestingly, it looks like a connection between modules is 1 sample buffer delay.
this actually makes sense, since in my test, the modules are likely running on different cores.

makes me wonder…
does the SSP enforce this 1 buffer delay on modules on same core, for consistency… or only impose the delay when necessary (across cores)?

anyway SDL as it stands is not going to help you resolve your issue.

thats not to say its technically infeasible… just not currently implemented.
either:
a) Percussa could update SDL for smaller delay times…
b) I (or someone else?) implement alternative delays lines or delay compensation.

I’ve no idea if Percussa are interested in (a) , obviously it has to compete with other priorities.

on my side… similar really…
pmix:
I’d like to add delay compensation to PMIX, it’s not ‘hard’, but I’d want to do it in such a way that it has zero impact when not used… so that needs careful consideration.

millisecond delay module :
generally, Im not going to write things thats are direct equivalents to existing modules (e.g. SDL).
however, I have been considering writing a more elaborate delay module, in particular Im interested in writing a multi-tap delay with heads that can be modulated.
But that’s a bit down the priority list… so I haven’t considered what the interface would look like yet.
but I like to have things ‘sample accurate’, so perhaps it will also cover this use-case :slight_smile:

swat module:
Ive started development of a utility module, that can easily have ‘algorithms’ added to it with simple UIs.
at the moment, Im using it as a development tool, to help me test modules (and my patches :wink: )
but I might release it at some point - haven’t decide for sure yet… but if I did, then a simple sample level delay line, would be the kind of thing id consider added to it. ( * )

( * ) its purpose, is simple algos with only a few simple controls.
yes… it’s inspired by disting… no it doesn’t/won’t have the same algos as disting.
the indecision about release, is i think the UI is a bit obtuse (as is disting), and could be painful to support… (albeit, unlike disting it has a help page for every algo :wink: )
its also possible that the ‘ideas’ and code Im developing for it, may make their way into another module idea I have (that is way too early to discuss :wink: )

1 Like
#14

I see what you mean about sample accuracy and to be clear I’m not advocating for less than sample accurate: my argument is that a Time parameter is better expressed in actual units of time in the UI/display, the more granular the better, with milliseconds being the minimum required for a delay to be useful. That’s my argument. Didn’t realize that about Ableton’s track offset or probably did but never needed it toggled. That’s pretty cool and looks a lot like some folks at Ableton had the same conversation you and I are having. You want samples, I want time. What to do? I really like how MaxMSP’s delay~ unit and many other units use max’s time format, so they can support anything from samples, time, hz, ticks, musical meter. To support something similar, SSP would need some sort of parameter configuration where you would select what unit the Time parameter uses. That would be super cool, but probably not in the cards.

It is a pity about the Simple Delay, but I do see a work around. I could use the SDL set to 20ms for my non-latent signals and put an 11ms delay on the ER-301 outputs via the ER-301’s micro delay (9 + 11 = 20). That’s a pretty thin solution but hey, it does seem like it could do the trick.

A note about the ER-301, it does have a low latency firmware, but at the cost of less CPU in user space. The firmware is also pre v1. I am a software engineer by trade and an electronic musician of 25 years and IMHO it is one of the most brilliant pieces of electronic music software ever designed. I think you’ll find users here who have both and I’m willing to bet that there is a certain warmth these users have towards the 301 that is not equal towards the SSP. So please, don’t let my words cross the ER-301 off your list anymore than I would let wavejockey’s expressions in this forum cross the SSP off mine :wink: There is a reason why I’m considering spending $2K just to interface and compensate for the ER-301. Ironic that I got into modular to in part get away from all of this software bullshit and hear I am, looking at yet more computers to solve issues caused by computers. Perhaps I’m approaching this all wrong :thinking:

Truly appreciate the help in understanding the SSP and hearing out my use cases and issues. It’s surely still on the table!

2 Likes
#15

yeah, the ER-301 is a very different beast to the SSP… they are not really comparable.

Ive done a lot of coding on Bela, which uses same chipset as ER-301, so I’m quite familiar with its pros/cons from a development/hardware perspective.
my impression from the outside (forums etc) - I think the ER-301 main strength is its a very solid implementation , also having customs units in lua, whilst not low level enough (for me), its helped build a community - which I think is a good thing.
(I know there’s been talk of a low level api, and replace SOM… but I don’t count me chickens :wink: )

so yeah, ER-301 is not totally off my list,
but honestly there’s so much potential on the SSP, Im going to be busy with it for a while yet…

2 Likes
#16

Another fine forum thread to catch up on!

@thetechnobear, did you see Brian open sourced everything (?) for the ER-301 in the latest release?

1 Like
#17

No, I don’t really follow the ER-301

definitely interesting, a few nice thing Ive noticed:

  • its primary purpose i think is to encourage module developers - being able to see the full source really does help debugging etc.
  • an emulator, helps when doing developing
  • brian, seems to be moving issue reporting to github, this will really help clarity.
  • its MIT license, so that makes it easier to re-use parts if you want. (Im surprised its not GPL)

definitely a really nice move forward, encouraging developes is of course a difficult step - but I think Brian has made it as easy as possible.

I had quick wander around the code, to see what might be interesting for the SSP.
honestly, on that quick look, I didn’t see much really - much of it is pretty familar (and available from other projects) .
that said, personally one aspect I enjoy from developers open sourcing code - is the ‘education’ aspect… I quite often look at others project - just to get some ideas on how they solved certain problems, to see if it might help me - so the repo looks good for that :slight_smile:

congratulations to Brian though, I think whilst its a bit scary making your code (and living!) open source - I do believe he will see some really postive upsides.

Im also sure the 301 community are happy - as he had promised the ability to write ‘full’ c++ modules from the beginning - so thats now a reality !

#18

Speaking of an emulator… On the Percussa site it talks about “A Mac running SYNTHOR software.” On further investigation, it looks like this was an old direction around the Audio Cubes hardware… I guess this is no longer relevant for us? Having an emulator would greatly shorten the feedback loop to see what changes look like.

#19

yeah, i think the synthor name has been ‘repurposed’, as you say previously it was used for the audio cubes which needed to be connected to a mac/pc… but that was not what we know and love.

an emulator is not really needed for the SSP, since we are developing VSTs , we can simply run and test our VSTs in any vst host.
I use the Juce ‘test host’ since its doesn’t make some assumption that daws do (like number of audio io), you can use it to control the encoder/buttons parameters, so that works ok (for testing).
you could create a vst host, that looked lilke the SSP with buttons/encoders… but Ive found this unnecessary.

as for the ER-301, well that doesnt have a choice, the modules are not vsts - so developers need an er-301 ‘host’ for development purposes.

btw: do all the modules work in that emulator? I looked at a few modules and they seem to have neon instructions hardcoded (some have it optional), Id guess these will not compile on an intel machine. (unless he is supply a conversion layer, which Ive done before… but i didnt see it)

of course, not saying an emulator for the SSP wouldn’t be fun… but it honestly would not add that much for me.
also… as above, an emulator not only needs to provide the ui (not that hard to be honest),
but also all modules would need to provide non-neon implementations given that many developers will run under intel. there are also SR issues e.g. many would want to run at 44.1k SR, which the SSP does not support natively.
I admit none of this is too hard, a desktop has plenty of cpu, so you can provide a non-optimised version, and use re-sampling for the SR issue, but its still extra complications, for what is not core product development.

1 Like