Why are samples are louder in Slate compared to Obsidian?
I have the same snare sample loaded in initialized instances of both Slate and Obsidian. For some reason the snare sample is significantly louder in slate than it is in Obsidian. I double checked both instruments to make sure all levels were zeroed out, turned off the filters in Obsidian, and maxed out the amp envelopes level in Obsidian as well. I even made sure that they both had the same amp envelope types “AD”, matched their settings and maxed out their velocities. I compared their levels in the mixer section and for some reason, the snare sample is louder in Slate compared to Obsidian.
After some trial and error I was able to match their levels, but only after having (OSC 1) in Obsidian bypass the Amp env completely and routed straight to the global filter.
So I can get the snare sample to the same level in Obsidian, but only if I bypass the Amp env? This has me a bit confused now since the same sample is going through the same type of Amp env in Slate without any issues. I’m not sure if i’m missing something, but it seems that the Amp env section of Obsidian is lowering the overall volume of the sample for some reason. The “level” knob in Obsidian’s Amp Env is maxed out at 100% so it shouldn’t be lowering the volume of the sample.
Can someone shed light on what might be going on? I was originally trying to use Obsidian as a sampler in order to create pitched snare rolls, but came across this issue along the way.
Comments
I asked the same question awhile back.
https://www.blipinteractive.co.uk/community/index.php?p=/discussion/363/obsidian-sampler-file-playback-is-quiet#latest
Thanks for the response. I read your thread and there seems to be a general consensus that Obsidian lowers the volume of samples at different stages in order to prevent the signal from being too loud. From my experience, however, the only place where Obsidian seems to lower sample volume is during the Amp Envelope section. If you bypass AMP in the filter section of Obsidian, overall sample volume will be unaffected and you get the exact same volume level as you would in Slate. This seems odd to me since samples in Slate go through their own Amp Envelope section without any volume adjustment whatsoever. I guess i’m just confused as to why the Amp Env section of Obsidian would arbitrarily lower the volume of your sample, while the same Amp Env section in Slate doesn’t seem to affect sample volume at all. Furthermore, the theory that Obsidian lowers sample volume at different stages to prevent overall loudness is a bit confusing since we have complete control over sample level & gain throughout each of these stages.
Interesting observation - to be honest never noticed that preAMP / postAMP difference, good job
That difference is not that big if you set amp from 80 to 100 and you disable velocity -> amp modulation (or you play notes with 127 velocity)
Anyway agree, there is difference in volume before / after AMP stage even when all other parameters affecting volume are set flat. Based on my observation (did not exact measurement, just checked it by eye) it's difference of 3dB single osc Obsidian vs. single sample Slate.
One of reasons is probably because in Obsidian you can play polyphonically - so lets say you can play accord of 3 or 4 notes - where in Slate pad is playing always just single voice. In case there would be not headroom in Obsidian, all non-monophonic patches would sound too loud compared to Slate ..
I don't think this will be changed anytime in future - because it would affect ALL existing patches and projects, especially in cases where is distortion used in global filter or distortion/compressor in track FX, so it's very unlikely this will be ever changed.
Surely you’re not saying Slate is monophonic?? You never played a kick and a snare on the same beat? I guess you mean it’s not usually playing the same sample twice over? Still, that line of reasoning just doesn’t seem right to me. Playing a chord might involve playing a sample multiple times, but they would be at different pitches, effectively making them different samples.
Moot point, since you’re right that it can’t be changed now without hosing up existing projects. But I think your reasoning is a bit off though.
Unless your sample playing the chord in Obsidian is a pure sine wave, it’s going to have a lot of the same harmonics on all notes which will add up and cause the volume increase. Drums landing on the same hits normally don’t have as many of the same harmonics because that would just create mud (like un-eqed low toms and kicks at the same time).
You can of course use the instruments in ways which make the volume drop on Obsidian unnecessary but it is based on a sound principle and empirical study.
I mean single PAD is monophonic, not whole Slate
This is about comparing playing same sample on single pad in Slate vs. single note in Obsidian, and reasons why in Obsidian it's quieter ...
As i remember this was discussed during developement a lot, Matt did some adjustments back then, but at the end it started to be "if i add there, in some case it is perfect but in other use cases it starts to be too loud" .. endless loop
But it always adds to overall volume
Just try it - play one note, two notes, 3, for - you will hear that perceived volume is increasing .. if you record it as sequence of coninuously adding notes starting with one and they you watch mixer VU meter you will see that i'm right 
One note plays quieter than 3 notes ..
btw. i never perceived this difference as issue - as i remember other beta team members are much more active in those slate vs. obsidian volume discussions than me
))
basically there are dramatically bigger volume differences between different HW synths i have, so somehow i accepted this as something which is not worth to try solve in any way, especially in digital domain with 32bit mixing where you can increase or decrease volume of any track as much as you want / need without any clipping ..
Btw. just tried to load same sample in Logic in EXS24 and in Ultrabeat and difference in volume is even bigger than Obsidian vs. Slate (without any tweaks, with default patch/bank)
did smal experiment - if both notes (in obsidian and in slate) are playing with velocity 127 AND AMP volume in Obsidian is on 100 - then lowering Slate's main output (fader in tab "OUTPUT") to -6dB gets same volume like Obsidian.
With 127 velocity but Obidian ENV set to 80 (default value in init patch), you need put main volume faer in slate to approx -11dB to match Slate volume vs. Obsidian volume.
Obe more important thing ! With drum samples, or generally short samples with sharp percussive attack - be sure that you lower Obsidian AMP "Attack" to zero - becasue in Slate default ADSR attack is 0 but in obsidian it is set to 5 ms (or something like that) which is usually good for lead / bass / pads sounds but not that good for drum/percussion sounds
Anyway, interesting discussion :-)
Wow, this actually explains a lot. Never noticed
Thanks!
I should write book about NS. Most probably also Matt would find there something he doesn't know :-)))))
How does one bypass the Amp Envelope? Awhile back I was trying to point different envelopes to different oscillators so each oscillator had control over its own adsr but I didnt have too much luck. I think I got it to where it kind of worked but there was an audible click at the start of each note.
You can bypass the Amp Env of (OSC 1) in the filter routing panel of Obsidian.
Make sure the blue dot next to (OSC 1) is highlighted in between the Amp section and the Global Filter. The only downside is i’m pretty sure it only applies to (OSC 1).
Pretty please
Good stuff!
This is whole discrepancy is probably my fault. I pressed hard during beta for Slate to 'bang' out of the gate the way the TRG did (as compared to the Eden default patch).
Thing is, Obsidian is just way more complicated than Eden. Beyond the static(ish) AMP envelope, there remains a myriad of ways for Obsidian to be over driven (or just spike in volume). Many of them dynamic via modulation sources and totally unpredictable.
Considering those possibilities and that it is far from an edge case to be playing 6-12 Obsidian notes (each of which can individually fire off differing modulation amounts...), Matt was understandably (to me) conservative with Obsidian internal levels.
No biggie. Just turn stuff up.
IMO it was a good decision, since I suspect it would cause more issues if the Slate samples were lower - they would sound too low in the default mix.
@itsinthemusiq, try creating and saving an Obsidian patch with a blank sample which you can then edit and save each time. There's all sorts of good reasons for using Obsidian for samples, e.g. modulation with LFOs (and as you say pitch effects).
I haven't used it much but this blank patch has helped me with, for example, long samples. Just load in whatever samples you need and remember that there's no "trigger" option - you have to hold down the key / draw in a long note. download patch here
I created this so I could record in longer audio samples, but I've also used it (with a track mixdown as the only sample) as a sort of mastering tool.... For when I know I won't be able to stop myself going in and tweaking the midi forever
If it helps you, all's good!
Perfect