Umpf - does it sound worse than redrum in default?

[tobypearce]

Strange thing tonight.

I was searching for a new kick drum and noticed that when I load the same kick into redrum and umpf they sound noticeably different.
Both devices are initialised, the rack is otherwise completely empty.

To my ears Umpf sounds much duller, like with a lpf (which isn't enabled). Sending the kick into the compression route brings back some bite, but it still doesn't sound the same as the straight kick being played through Redrum.

Can anyone replicate this or let me know what's going on?

[eusti]

Strange thing tonight.

I was searching for a new kick drum and noticed that when I load the same kick into redrum and umpf they sound noticeably different.
Both devices are initialised, the rack is otherwise completely empty.

To my ears Umpf sounds much duller, like with a lpf (which isn't enabled). Sending the kick into the compression route brings back some bite, but it still doesn't sound the same as the straight kick being played through Redrum.

Can anyone replicate this or let me know what's going on?

Well, on my system the Redrum is - as a reset device - 1 db louder than the same sample played by Umpf (reset).
When I bring down the Redrum one db they sound the same here. But that is just checked with one sample.
Not sure if that's what you're experiencing.

D.

[tobypearce]

Can you get them to cancel out if you invert the phase of one of them and play at the same time?
When I do this all the low end goes but I'm left with a significant click!

[tobypearce]

Really odd.

I've uploaded the reason file - perhaps it's just my system, or something weird with the kick I used (from Native Instruments)

Now that I experiment further:

2 x Redrum pass the null test

Redrum and Kong - nearly but not quite, even when volume matched. Small difference though

Kong and Umpf - leaves a notable click on upper frequencies

Redrum and Umpd - leaves a notable click on upper frequencies

I'm now wondering why these 3 straightfoward drum sample players all sound a bit different. I'm not a puritan so I don't mind this, but the Umpf one sounds really out of whack and would discourage me from this RE.

[eusti]

Can you get them to cancel out if you invert the phase of one of them and play at the same time?
When I do this all the low end goes but I'm left with a significant click!

Well, yes. But have you tried the same with Kong vs Redrum?
Because two identical devices (one with reversed polarity and one normal) null.
But neither Umpf nor Redrum nor Kong nor the WAV in an audio track nulls to either of the others...
I'm sure we'd have the same result if we added NN19 or NNXT as well...
Not sure why... Maybe the timing is slightly off... Or the volume... As - at least I - the volume was only adjusted kind of roughly via Selig Gain.

D.

[dioxide]

You forgot to embed the sample.

[guitfnky]

if you want to test if they’ll null but they’re at different volumes, maybe bounce them and normalize before you see if they’ll null.

[dioxide]

Also: use the sequencer not Redrum. The sequencer has sample accurate timing, Redrum has some variation. If the samples still don't cancel out, then something else is going on.

[eusti]

So, I just normalized both parts and they still don't null, but got closer.

D.

[tobypearce]

This time with sample. Thanks!

[dioxide]

I did my own tests and yes, I couldn't get it to null. I set velocity to 0 on all devices. Set all 3 Kong volume parameters to neutral values. The same with Umph. And none of the three devices null with each other, even when they are triggered from the sequencer. I think it could be a timing issue?

[eusti]

I did my own tests and yes, I couldn't get it to null. I set velocity to 0 on all devices. Set all 3 Kong volume parameters to neutral values. The same with Umph. And none of the three devices null with each other, even when they are triggered from the sequencer. I think it could be a timing issue?

Even the same units didn't null?
When I created a Kong for instance, reset the device, loaded the sample, added a Selig Gain device, adjusted to what the first Umpf had shown (-7db), then duplicated the device/ track in the sequencer and reversed polarity in the copied device I got those identical devices to null.

D.

[dioxide]

Identical units null. I couldn't get different devices to null though.

[eusti]

Identical units null. I couldn't get different devices to null though.

Great! Same here! At least something consistent in this story!

D.

[dioxide]

Okay. More weirdness.

I exported loops from each. Kong and Redrum seem accurate. Umpf is doing something weird to the transient. Kong is different from Redrum. As far as I can tell this is because Redrum downsamples the 24bit original to 16bit (yes it does this).
EDIT: Actually I'm unsure if Redrum still does this. I couldn't see any mention of it in the manual. Perhaps this was only in early versions of Reason.

Here are the screen grabs. If you look at the original sample you will see that Kong and Redrum are more or less correct.

Kong.png (25.28 KiB) Viewed 4465 times

Redrum.png (29.51 KiB) Viewed 4465 times

Umpf.png (32.33 KiB) Viewed 4465 times

[dioxide]

Note that there's also a bit of DC offset at the start of the sample but the wave shape in Redrum and Kong is more or less correct. There must be some filtering going on in both though.

Original sample.png (40.8 KiB) Viewed 4463 times

[eusti]

It seems at least Umpf Club and Umpf Retro do null.

D.

[dioxide]

I think it could be a bug with the Attack in Umpf. The envelope can't be disabled.

[eusti]

I think it could be a bug with the Attack in Umpf. The envelope can't be disabled.

Ok. But we seem not even able to make anything null as soon as we use a different device... Even the wav files in the sequencer aren't nulling against anything... At least here... Maybe the volume needs to be matched much more closely... Not sure...

At least the issues that I was hearing Redrum (reset) vs Umpf (reset) were resolved when I dropped the Redrum volume 1 db.

D.

[Boombastix]

The dullness or a heard transient click can be from the very start of a sample if it starts at a zero crossing or not. Even the tiniest non-zero start can be heard.
You may wanna check if the sample start is truly at a zero and compare the behavior with a sample starting at zero vs one starting at non-zero to see if they behave the same or not. Just in case there is a difference in how the different units manage that (it is long shot but, you never know...)

I did some test myself and found a BUUUG in Redrum. In Gate mode it won't play the sample to the end, it chops it after the midi note ends. Kong plays it through. Also, to try null between Kong and Redrum, gave me sooo much inconsistency. Really weird.

This was my setup:
Put Kong and Redrum in a Combinator. Reset them both. Put single midi notes (C1, vel=100) in a midi clip in the sequencer to trigger both at the same time. Set Kong/Redrum vel to level=0.
Routed each via Selig Gain to gain adjust to the same peak value. Put these two Gains pan knobs to zero (i.e. 100% mono)
Routed Redrum to another Selig Gain for polarity invert only.
Merged the two audio streams with a Blamsoft Audio Merger.
Then another Selig Gain before it goes back to the Combinator to assure nothing was clipping.

Did a bounce in place and this is what came out!
You can see how the audio cancels out on the first hit. But only until the midi note ends, then Redrum goes silent and the audio jumps up as the Kong is still playing the sample.
Next hit, now they do not cancel out very well, third hit another but different again!

What the heck is this!? Pict shows Midi on top. Bounced audio that should have cancelled out.

Kong n Redrum.gif (8.36 KiB) Viewed 4395 times

[dioxide]

Even the wav files in the sequencer aren't nulling against anything... At least here...

I thought I'd have a go at this with Kong, as it looks like the best bet. I didn't get a 100% null but got fairly close. I needed to use Lectric Panda Amp instead of Selig Gain as it is more precise.

[dioxide]

I did some test myself and found a BUUUG in Redrum. In Gate mode it won't play the sample to the end, it chops it after the midi note ends. Kong plays it through.

I think this is correct, although it is frustrating when bouncing patterns to the sequencer. When triggered from a pattern, Gate mode is like a one-shot mode, like in Kong. But once you bounce it to the sequencer, it acts like a Gate on a synth or sampler, where it is only triggered for the length of the note. Bounce Pattern to track from Redrum isn't smart enough to bounce down the note lengths.

Have a go at the Gate Mode and Decay in the NN-Nano Amp Env section. There are some similarities to Redrum. The difference is that Redrum has a sequencer and the Drum Length in Gate Mode is actually shortening the note/gate length of the sequencer trigger. If you connect it to a Skope you'll see how you can alter the length of the CV Gate output. Having it set to Decay mode just produces a click/trigger signal.

[Boombastix]

I did some test myself and found a BUUUG in Redrum. In Gate mode it won't play the sample to the end, it chops it after the midi note ends. Kong plays it through.

I think this is correct, although it is frustrating when bouncing patterns to the sequencer. When triggered from a pattern, Gate mode is like a one-shot mode, like in Kong. But once you bounce it to the sequencer, it acts like a Gate on a synth or sampler, where it is only triggered for the length of the note. Bounce Pattern to track from Redrum isn't smart enough to bounce down the note lengths.

Have a go at the Gate Mode and Decay in the NN-Nano Amp Env section. There are some similarities to Redrum. The difference is that Redrum has a sequencer and the Drum Length in Gate Mode is actually shortening the note/gate length of the sequencer trigger. If you connect it to a Skope you'll see how you can alter the length of the CV Gate output. Having it set to Decay mode just produces a click/trigger signal.

Redrum was not triggered from its own Redrum pattern, it was triggered by a midi note in the sequencer. Both Kong and Redrum was set to gate mode (and triggered by the very same midi note). But they behave differently, so one has to be wrong, no?

[esselfortium]

Redrum has always behaved that way since dinosaur times, it's a feature.

[Boombastix]

Redrum has always behaved that way since dinosaur times, it's a feature.

Ahh, OK. That explains the odd gate behavior of Redrum. But it won't explain why the cancellation on some hits are good but on other hits it is clearly different. I can not see that I did anything wrong in the test, and I'm just triggering midi notes...