Slice (unwanted) "tremolo" effect

Which means recording slots behave differently? :thinking:

Seem so. You can try it yourself, if you like. I don’t know why it is like this, but it is true. Maybe @Elektron has more technical details. ¯\(ツ)

That is logical but in reality you’re not “adding” attack, on the contrary you are introducing a rising slope at the beginning of the next slice’s amplifier envelope and the clicking you hear is the end of the previous sample cutting off abruptly (at the slice point). Which is why the ATK=1 solution isn’t a universal remedy against clicks: it’s only efficient if the click is due to an abrupt start point of the sample or slice

1 Like

You’re right, “adding” is the wrong term here. English is not my native language, so it isn’t easy for me to describe everything exactly. :slight_smile:

I just wanted to say that it playes seemlessly if no attack or release is introduced. In most cases you would introduce a bit of attack to “smoothen” the start of a slice/sample, of course if the slices/samples are independent from each other and the goal isn’t to play a “static” sound from slices.

“introduce a slope on the attack” would be more appropriate… because “introducing a bit of attack” means lowering the the “hold” (aka “sustain”) level of the ASR envellope (BTW English isn’t my native language either… even my everyday language isn’t my native language!) (sorry for the nitpicking :wink: )

Yes, I confirm. Can reproduce now anytime. See screenshot I posted earlier. It’s not the recording itself which is perfect, it’s the playback of the recording buffer

Geez that was a little defenisve. Just wondering why the sample wouldnt be cut at the trasients in this case.

Its like running a test on something youd never do to beginh with. Testing a non use case.
Dude it was not meant as a jab. Sorry if i set you off.
Dont let the machine get the better if you. We make machines do what we want them to.

Ill freakin try that non use case test today and see if i cant lend a hand vs setting you off.

(I dont recall that happening to me)

Perhaps your samples bpm doesnt match the project bpm when you added the sample to the project, so you are getting artifacts when you cut it?

Sorry dude.

Well I stumbled upon it because I wanted to do that very something maybe YOU would never do :wink: That’s the power of the Octatrack: we can go into so many directions …

Anyway I didn’t mean to start an argument. What matters to me doesn’t necessarily matter to someone else, and I can perfectly understand that. But I won’t explain in-depth where exactly this fits (or eventually might not fit) in my workflow, that would distract from the core-issue. I just want to figure out if I make a mistake somewhere or if I came across a hardware/firmware limitation.

1 Like

Im just saying you dont use a knife to eat soup.

But, best of luck man.
Maybe its something Elek can sort if no one can find a solution.

(Did you submit a ticket?)

A case for mono vs stereo samples. Stereo samples have offset L/R

Also, in any DAW, if you cut off a track mid sample, not at a transient, you WILL get a pop. I think since there is a microbleem of time between the end of the sample you chooped in a bad spot, and the start, that time given allows for the pop. The computer takes that micron to say stop slot 1 and start slot 2…POP

No problem :wink:

@pinup57, bravo for pointing that, I could test it.
I used the Isaac guitar chord sample from Elektron MKI demo. Bpm is 120.
In a regular Flex slot, no clicks. Atk = 1, clicks.
Same sample loaded in Recording1 slot, clicks.
Atk = 1, no clicks, slight clicks at the end of the chord.

I’m more interested in realtime recording without saving (autosaving is a dream), so I have to deal with clicks. Anyway I also use it to create rhythms, and it covers clicks. Of course I hate clicks.

I totally disagree. :slight_smile:
These settings DO matter, and can change clicks behavior.

2 Likes

Yes, that’s what all this was about!

Well, of course you’re correct, when I said that I was referring to the exact test I was performing.

1 Like

So I guess we have figured out that we either need to file a bug report (those 37 samples of silence between slices when played back from the Recording buffer) or/and a feature request (auto-save).

2 Likes

That period of silence at the end of each slice is really a strange thing. Especially since the values of the next slice continues exactly where the values stopped before the silence. It looks like that the slices gets somehow shortened (thus the silence).

When you compare the waveforms of the playback from flex and from record buffer is there a difference in wavelength?

When I record into the record buffer the resulting file is sample-accurate in lenght.

Hi, I edited a short clip out of rehearsal footage to demonstrate why those clicks are bothering me . This is only a very simple guitar riff: except for a kick on the beat and a hat plus a crash on the 1st beat of a 4 bar cycle so that I don’t get lost no samples or synths are used. All the rest (bass, arpegios etc) is generated in real time using one track of the Octatrack simply by pitchshifting slices in the record buffer while recording the inputs. Needs decent speaker or headphones to notice those clicks in all their glory. It was just a jam, all is rather random. Definitely not a “non-use case” AFAIC!

1 Like

Good groove and nice setup! Where do you live? Clicky indeed. You can reduce that.
Is OT Master?

I get rid (or reduce? :smile:) of clicks with ATK 1, HOLD min*, REL between 64 and 120, and I plock every 2 steps, 1/8th notes. I use rec trigs corresponding to trigs instead of slices.


*Maybe you can increase that below slices length, it’s in steps values.
1 Like

Tuff times.
Good luck.

That video of yours is super cool too. What a deep machine this is! I only have it since a week, so much to learn here… I live near Nîmes, France, btw. I’ll try your workflow, I don’t really get how that works but it sounds different and interesting. Yes, the OT is master, and it doesn’t do much really, just recording the incoming guitar, and shifting the pitch of each slice. No effects used on the OT, not even a LFO or a reverb. The kick and hat are from the Toraiz SP-16 down below, synced to the OT’s clock. I uses a iConnectivity iCM4+ midi interface to get everyone to talk to each other.

1 Like

Length is not what I meant. I meant the wavelength of the sin waves. Has one sin wave a shorter period in samples than the other? Because it looks like there is some unwanted negative timestretch going on (thus the periods of silence).