Digitone input noise floor

Elektron Gear Digitone/Digitone Keys
#1

So before I create a ticket I thought I’d ask everyone what their experiences was with this. I’m noticing with the left and right level at 100 for the inputs that there is a significant difference in the noise floor. How much do you ask?

there is roughly a 20dbfs difference in noise floor between the inputs being completely turned off, and them being at their default level (100). By comparison there is a difference of about 2dbfs between the inputs being turned off and being turned to full (127) on my Analog Four MKII.

Is this a hardware fault or are other people experiencing this? For the record, this is with no devices plugged into the inputs, though the results are the same if they are.

#2

i never have my inputs up past 40 no noise for me.

#3

I keep mine at 50. I have two. 50 keeps the output from each digitone identical when running one thru the other.

1 Like
#4

Thanks for your reply but it doesn’t really address what I’m asking. If you put it to 100, which is the default when you load a new project, are you seeing the same drastic jump in noise floor? It may not seem like you would ever need it that high, but I’m trying to see if this is the way the hardware is designed or a defect. I’m leaning towards defect since I don’t have this problem on my a4 but it’s possible the preamps on thw digitone aren’t the same.

On a track I’m working on i have it a 70 for my model d, I’m hearing quite a bit of hiss. I heard it in Cenk’s Digitone Experience recordings and thought the device was noisy but realized it was the inputs.

#6

Weird my default is 0 when I load a new project

2 Likes
#7

I understand your reply but what is your noise floor at 100 compared to 0. Could you measure that for me?

Edit: I’m pretty sure the level in Elektron devices is logarithmic so there would be a drastic increase between 50 and 70 and 70 and 100.

#8

What version of the firmware are you running?

Edit: it should also be noted that the preset folder, which is locked, has the inputs defaulted at 100. And my Dataline Experience project is at 100 but you are right a new project is defaulted to 0. But I don’t think that really matters since that level of noise should not exist at 100 when it doesn’t in other products.

Could someone please answer my original question and put their inputs to 0, measure the level (I did this in ableton) and then turn it to 100 and measure the level (nothing has to be plugged in) and let me know if they are seeing the same increase in noise floor? 20 dbfs?

#9

~20dB here (nothing connected to the inputs, which is by the way not too interesting). As the gain setting goes over a very huge range this looks completely fine.

2 Likes
#10

Thank you for checking. It just means I won’t use these inputs since that’s too noisy for me (I was wondering for weeks why my Digitone was the hissiest device in my signal chain). I’ll just go direct through my a4. Too bad cause my Model D really likes the Digitone pairing.

#11

If your model d really likes it then keep it and turn them down to fifty. Unless there’s still too much noise for you? I have my headphones plugged into one digitone and to match that exact volume on the second digitone I have to put the inputs to 50. If I put it to 100 it is double as loud. Way too loud. Which obviously will increase the noise floor as well. But they match perfectly at 50 so I just keep it there and I don’t hear too much noise I think it sounds good. But I also keep my main volume knob at about 1:00 and I end up turning each tracks amp down a little. Newest os

#12

For what it’s worth I ended up using balanced cables to connect the 2 digitones

#13

Technically this sounds very strange to me.
I checked the levels:
Output of Digitone via balanced connected directly in to RME Fireface 800.
Noisefloor with inputs off: ~ -104dB FS rms at a volume (3 o’clock) where I usually use the device without clipping.
Noisefloor with open input changes to:
-98 dB FS @ input gain = 50
-83 dB FS @ input gain = 127

This is absolutely fine and is what I would expect from a professional audio device. If there are problems with noise floor, than the levels are choosen very strange. Every device would introduce more hiss from its own output. Near to 100dB SNR is really huge. You should easily find a solution for this.

BTW gain differences:
127 “0”
100 - 4dB
50 -17dB
20 -33dB
10 -45dB

2 Likes
#14

I think part of it is that all my elektron devices are at full scale on the master level since i can turn them up with no noise whereas if i gain up my preamps you start to introduce preamp hiss (even the cleanest preamps will).

But that also means the input noise on the digitone is more pronounced.

It really all depends on what you’re putting into the inputs. The digitone, or even my model d can be really loud so you can (just checked) turn the inputs down to 50 and still get a good amount of level ( i ended up having to turn up my master gain quite a bit on the model d) some inputs are naturally quiet (like say, a deepmind 12) so more input level is required. That gain staging game we play.

Ugh curse my sound designer’s mind. Working in sound for a living has its drawbacks. It has made me very neurotic about things like device noise floors. I’ve sold / returned many products because they had what I considered an unacceptable level of self noise.

#15

So here are my exact measurements. This is with nothing plugged in.Using a Behringer X32 Producer.

Digitone.

-98 dbfs @ 0
-95 dbfs @ 50
-90 dbfs @ 70
-84 dbfs @ 100
-79 dbfs @ 127

Analog Four MKII

-98 dbfs @ 0
-98 dbfs @ 50
-98 dbfs @ 70
-97 dbfs @ 100
-97 dbfs @ 127

For my ears, anything higher than 90dbfs and I start to hear the noise. Does that matter? In the context of a completed song? Probably not, but for the metrics of the difference in input quality between the A4 and the Digitone? I think so. Higher noise floor means less dynamic range. And as previously stated, when you buy a Digitone today and load presets, or buy Cenk’s Digitone Experience pack, they both default at 100 for the inputs which gives an audible hiss. Which is really strange when a digital synth is louder in hiss (again at first I didn’t know why) than my modular or an analog synth.

At first I was like oh, maybe this is normal for Elektron? But thinking back, I didn’t have this much noise on the OT, A4MKII. Can’t remember with the Digitakt.

1 Like
Unity gain settings on the inputs for all Elektron gear
#16

Just checked. With unbalanced in, to get a 1 to 1 level from input to output through the Digitone, you have to set the input level to 28.30. It is a luxury (and not a fault of the device) to have the possibility of ~25dB additional gain for the (unbalanced) inputs available. That is not reason to complain about.

As shown above, there is NOT “much noise”. This is plenty of headroom and a low noise floor. Physics are limiting the analog gains.

I was interested in this question and quickly compared with the Octatrack MKII. Same behaviour, as expected. With full gain on the input, I also get ~20dB more noise. I don’t see a relevant difference to the Digitone. Digitone seems to have even more gain accessible which means the specs reagarding noise floor are even better. (would have to take more time to check in detail) Both devices fullfil the requirements of professional studio equipment and have a low noise floor.

Can’t compare to the A4 here at the moment. You have to make sure to compare gains and the same input levels. With the abstract Elektron numbers (0…127) it’s not possible to compare anything. It is possible that the gain is different. With a huge range of gain it’s the physics which amplifies existing hiss.

If you can clearly hear the noise at -90dB FS in your system you are listening at quite high sound pressure levels btw. Take care of your ears.

Unity gain settings on the inputs for all Elektron gear
#18

So, checking to see if the gain is the same on both devices, I’m using the a440 pilot tone on my model d to measure as a reference.

On the Analog Four MKII, with the input volume at 127 and the fx track level at 127 the pilot tone is at -35dbfs.

On the Digitone with the same settings, the pilot tone is at -14dbfs. So you’re right, there is quite a bit more gain coming out of the Digitone. When I bring it back down to -35dbfs to match the gain of the A4, the noise floor is at about -95dbfs. So I guess you’re right.

Though that poses the question of why the presets are set to 100 by default. Maybe something to change in an OS revision.

#19

20dB can be quite a lot of gain in the analog domain if you are looking at noise floor and SNR type of things.

On OS 1.02 I get in a new project all input levels set to 0.

I would prefer to set a default setting by my choice for a whole or a new project. For live performance I decided against using the Digitone inputs at the moment because I was annoyed to set the levels for each new pattern.

P.S.: If I press Func I get 0, 100 and 127 as steps for the levels (or is it that what you mean by “default” level?). This values could be changed to more meaningful values, e.g. 0, 30, 50, 100, 127. this would make things already a lot easier.

Edit2:
Even better would be to have the option if the input levels are globally set or pattern wise.

#20

What I meant by default is when you pull the Digitone out of the box, brand new, press play in the presets project everything has inputs set to 100. I purchased Cenk’s Digitone Experience album and patches/project, which is amazing btw, but it’s the same there too.

#22

Were you just listening for it, or were you measuring it in a daw? @Chiron echoes the experience I had using his RME interface, but as he pointed out, the gain at 127 is much higher than it is on the A4 MKII.

My mixer inputs with nothing in them idle at about -101dbfs. With something plugged in but powered off, about -99dfs. -98 for most stuff when it’s turned on. All of that is below hearing. But when I turned up the inputs at 127 i got up to -79dbfs, which is in the audible range. At least for me. If you’re not seeing that at all that makes me wonder if there is a bad batch.

I’m positive it’s not a mixer issue since the same doesn’t occur with other devices. Even my Behringer Model D direct into the mixer with everything cranked doesn’t break -92dbfs.

#24

What numbers were you getting?