KVR Audio

munchkin · Post by **munchkin** » Thu May 31, 2012 11:06 pm

I've bought the latest DCAM preset bundle during the into offer. The sound banks have some quality presets BUT...

With some of the banks the sound level of the presets is all over the place varying as much as +/- 3dB. This means I have to constantly ride the volume level when I'm auditioning presets not to mention worrying about my speakers and the neighbors.

I've noticed this discrepancy with other commercial sound banks and think it's just not good enough. I mean how difficult is it for the sound designer to do a final check on the sound levels of the presets to get them to match as much as possible. I'm talking about equity between similar sounds such as bass presets for example. I know this is not an exact science but come on +/- 3dB difference between bass presets in the same expensive sound bank is ridiculous.

Has anyone else noticed this?

RunBeerRun · Post by **RunBeerRun** » Fri Jun 01, 2012 12:10 am

I agree that a synth should set a certain level and stick with it. Also low volume is no good, I do well with default being just under clipping.

quayquay17 · Post by **quayquay17** » Fri Jun 01, 2012 12:22 am

When I demo-ed sugar bytes' "Unique" about half of the presets were literally clipping...

munchkin · Post by **munchkin** » Fri Jun 01, 2012 1:48 am

It seems like the developer/designer doesn't bother to test preset sound levels before release.

Xenos · Post by **Xenos** » Fri Jun 01, 2012 3:08 am

I've noticed this on occasion as well with lesser known synths by smaller devs. The bigger devs, though, are very adamant about keeping factory patch levels consistent and specific. I always thought that was a good habit to have as messing with the volume knob every time you switch patches is a royal pain in the a** lol.

quayquay17 · Post by **quayquay17** » Fri Jun 01, 2012 3:46 am

Especially when most people use the presets as a quick form of demo-ing the synth. For me, if it clips, that's a completely ruined first impression.

GeorgeZ · Post by **GeorgeZ** » Fri Jun 01, 2012 6:14 am

Xenos wrote:I've noticed this on occasion as well with lesser known synths by smaller devs. The bigger devs, though, are very adamant about keeping factory patch levels consistent and specific. I always thought that was a good habit to have as messing with the volume knob every time you switch patches is a royal pain in the a** lol.

I hear ya mate

Royal!

risome · Post by **risome** » Fri Jun 01, 2012 6:19 am

Its a must to check levels if there not matched its kind of a preset bug Lol

ouroboros · Post by **ouroboros** » Fri Jun 01, 2012 7:34 am

I agree, especially re: clipping. I find the problem is with, the difference between a mono lead and a pad or something that might be played as a chord.

BertKoor · Post by **BertKoor** » Fri Jun 01, 2012 10:50 am

munchkin wrote:With some of the banks the sound level of the presets is all over the place varying as much as +/- 3dB.

3dB off is already not that much

Sample-based instruments that use solely peak-normalised samples have far bigger differences in RMS level.

himalaya · Post by **himalaya** » Fri Jun 01, 2012 10:58 am

All my presets are volume tested. Not just for velocity, but key range volume, and modulation wheel volume. This last one can bring unexpected surprises if lots of fancy modulation is assigned to the mod wheel and the level can go up drastically, so very often this needs to be compensated with a slight volume drop on the mod wheel. Polyphonic patches are tested with chords (obviously).

In most of my work i use a velocity midi limiter when testing presets so that they play at max velocity across all octaves when testing - this is the best way to see if the preset is clipping, or if the presets need to be set to a specific dB threshold.

No doubt there may be occasions when a preset slips through the net or the synth developer makes further changes. If you catch it, report the issue directly to the developer and the problem has a chance of being corrected.

osiris · Post by **osiris** » Fri Jun 01, 2012 11:08 am

I had this problem with DCAM on FL Studio, and fixed it using the Processing menu, but I'm sorry I can't remember what was done.

Lotuzia · Post by **Lotuzia** » Fri Jun 01, 2012 11:30 am

himalaya wrote:All my presets are volume tested. Not just for velocity, but key range volume, and modulation wheel volume. This last one can bring unexpected surprises if lots of fancy modulation is assigned to the mod wheel and the level can go up drastically, so very often this needs to be compensated with a slight volume drop on the mod wheel. Polyphonic patches are tested with chords (obviously).

In most of my work i use a velocity midi limiter when testing presets so that they play at max velocity across all octaves when testing - this is the best way to see if the preset is clipping, or if the presets need to be set to a specific dB threshold.

No doubt there may be occasions when a preset slips through the net or the synth developer makes further changes. If you catch it, report the issue directly to the developer and the problem has a chance of being corrected.

+1 Same here, I have some automatic process to take care of all this, plus I always make a manual check before release.

I'll add that playing monophonic or polyphonic will make some differences, more or less noticeable for this ot that synth) and that some patches beeing likely to be played in both modes ( like keys ) this will reflect when browsing presets ( a single note will have a lower volume than one 6 notes chord )

osiris · Post by **osiris** » Fri Jun 01, 2012 11:50 am

I found it. Pretty sure it only applies to FL Studio but never hurts to check others.
In Processing on the wrapper make 'Ensure processor state in plugin callback' is unchecked.
That fixed my problem with volume spikes DCAM and RP Delay.
Maybe I hope that will help someone.

Sampleconstruct · Post by **Sampleconstruct** » Fri Jun 01, 2012 11:57 am

I also do level matching for all of the presets I release, sometimes with evolving sounds it's difficult though as the highest level of the sound might only be reached after 10 or more seconds or so and the beginning can be subtle/soft which can give the impression that the particular preset wasn't level matched/tested if the user only puts his fingers on the keys for 2 seconds and doesn't have the patience to wait until the sound has fully evolved. But that is usually mentioned in the patchlists I release with my Banks, where every sound has a short description and some playing tips, a lot of work but user-friendly.

Synth Preset Level Fluctuation