Login / Register  0 items | $0.00 NewWhat is KVR? Submit News Advertise
User avatar
Architeuthis
KVRAF
 
2873 posts since 27 Jan, 2006, from Phoenix, AZ

Postby Architeuthis; Tue Sep 26, 2017 9:26 pm Plugins that produce DC values dangerous for customers?

Is it bad to allow audio plugins to produce DC offset limited between -1 and +1?

Expert Sleepers Silent Way does this.

I need to allow this for my oscilloscope music synthesizers series especially, since they will be played back via an X/Y oscilloscope.
User avatar
BertKoor
KVRAF
 
10060 posts since 8 Mar, 2005, from Utrecht, Holland

Postby BertKoor; Tue Sep 26, 2017 11:22 pm Re: Plugins that produce DC values dangerous for customers?

Sure it is bad! Without any counteractions they'd produce clicks when their audio channel is switched on/off, and naieve level measurements will be skewed.

But technically there's no wrong or right. Even a DC offset outside +1/-1 could be found in the wild.
We are the KVR collective. Resistance is futile. You will be assimilated. Image
My MusicCalc is back online!!
User avatar
Architeuthis
KVRAF
 
2873 posts since 27 Jan, 2006, from Phoenix, AZ

Postby Architeuthis; Wed Sep 27, 2017 9:08 am Re: Plugins that produce DC values dangerous for customers?

BertKoor wrote:Sure it is bad!
oops, when I wrote "bad" I meant dangerous. DC signals are not inherently bad IMO, they are useful for voltage controllers and cross-plugin modulation. Just want to know if they are dangerous. Sounds like maybe I just need a disclaimer and warning and some options on the synth to allow DC (not by default).
Xenakios
KVRian
 
988 posts since 9 Sep, 2005, from Oulu, Finland

Postby Xenakios; Wed Sep 27, 2017 10:18 am Re: Plugins that produce DC values dangerous for customers?

Architeuthis wrote:Just want to know if they are dangerous. Sounds like maybe I just need a disclaimer and warning and some options on the synth to allow DC (not by default).

It would seem plausible that loudspeakers and headphones might not like extended periods of DC input because they'd actively need to keep the cone/membrane at a non-central position. However, I think modern audio interfaces will have DC-offset eliminating filters at the outputs, so it maybe isn't so likely the speakers/headphones will actually end up getting the DC signals.
JCJR
KVRAF
 
2147 posts since 17 Apr, 2005, from S.E. TN

Postby JCJR; Wed Sep 27, 2017 11:14 am Re: Plugins that produce DC values dangerous for customers?

Perhaps the biggest risk of frying something would be gear fancying itself "ultra hifi" capable of normal operation down to dc. For instance some of the old crown amps were dc capable and sometimes were used for such as servo motor controls for speed control on huge tape machine motors. Giant power opamps.

Most audio gear is ac coupled and would probably only click as Bert mentioned.

There are some fancy preamps and such that don't have caps or transformers in the signal path, but have fancier ways of rejecting dc and providing dc stability without putting coupling caps or transformers in the signal path.

But once in awhile I've read about devices (wrong headed in my ignorant opinion) considering themselves hifi audio because the freq response is as close to "DC to light" that can be built. ;) Such a signal path would smoke speakers fer sure if you intentionally or accidentally feed it dc.
User avatar
antto
KVRAF
 
2470 posts since 4 Sep, 2006, from 127.0.0.1

Postby antto; Wed Sep 27, 2017 12:29 pm Re: Plugins that produce DC values dangerous for customers?

JCJR: well, that would be too bad
sometimes a DAW plugin can "explode" (produce a NaN or +/-INF), other times an audio program might hang and the last audio buffer may repeat for a long time (if it's small enough it could contain a chunk of signal that's biased, essentially containing DC offset)
so people with such "hifi" systems are probably going to smell the magic smoke sooner or later
It doesn't matter how it sounds..
..as long as it has BASS and it's LOUD!

irc.freenode.net >>> #kvr
JCJR
KVRAF
 
2147 posts since 17 Apr, 2005, from S.E. TN

Postby JCJR; Wed Sep 27, 2017 8:55 pm Re: Plugins that produce DC values dangerous for customers?

antto wrote:JCJR: well, that would be too bad
sometimes a DAW plugin can "explode" (produce a NaN or +/-INF), other times an audio program might hang and the last audio buffer may repeat for a long time (if it's small enough it could contain a chunk of signal that's biased, essentially containing DC offset)
so people with such "hifi" systems are probably going to smell the magic smoke sooner or later

Agreed. At least a person would have to go to some trouble to acquire a rig capable of shooting themselves in the foot so thoroughly-- Needing DC capable audio interface, mixer, speaker processor, amps.

Maybe it makes some kind of sense, just seemed weird to me. I got similar twinges reading about "high quality" audio devices supposedly flat to 100 kHz and higher. I realize that HF rolloff "close to 20 kHz" can mess with audible high frequencies' phase, but amps capable of delivering high power, octaves higher than we can hear, seems to invite high frequency parasitic oscillations or partially-demodulated radio interference in the audio system, frying expensive HF drivers with no audible warning. The tweeter equivalent of silently DC-cooking a woofer voice coil.
camsr
KVRAF
 
6687 posts since 16 Feb, 2005

Postby camsr; Thu Sep 28, 2017 12:27 am Re: Plugins that produce DC values dangerous for customers?

As long as it doesn't make it's way to the audio output, it's fine.
Image
kippertoffee
KVRist
 
237 posts since 3 Mar, 2004, from Denmark

Postby kippertoffee; Thu Sep 28, 2017 1:01 am Re: Plugins that produce DC values dangerous for customers?

I would say, as long as the user understands what they are using, it's OK. As someone said, most soundcards are AC coupled so will not output DC. Motu are DC coupled iirc, and a small number of others. I would guess if this is for oscilliscope visualisation that your users will understand what they are working with.
User avatar
Tale
KVRist
 
478 posts since 12 Apr, 2010, from Lowlands of Holland

Postby Tale; Sun Oct 01, 2017 2:17 am Re: Plugins that produce DC values dangerous for customers?

Exactly what do you mean with "DC"? Because you could have a signal generator output a 25% duty cycle pulse in the [+1.0; -1.0] range at 440 Hz. This has a DC offset, but it likely wouldn't be dangerous at all (well, maybe you should turn down the volume a bit, hehe). But you could also have it output a 50% duty cycle square wave at 0.01 Hz. This won't have a DC offset, but it will output +1.0 during 50 s (!), and this could be potentially dangerous I guess.

But I assume you are talking about outputing a constant non-zero value, right? That is probably not a good idea in general, but I can see how this could be useful in a modular environment. I guess you could output such a signal to a channel less likely to be the master output (i.e. not channels 1 & 2).
Martinic Scanner Vibrato
JCJR
KVRAF
 
2147 posts since 17 Apr, 2005, from S.E. TN

Postby JCJR; Sun Oct 01, 2017 12:24 pm Re: Plugins that produce DC values dangerous for customers?

Thanks Tale for the interesting ideas.

I gather that a small number of audio interfaces are dc capable. Dunno much about it, but possibly most-needed by modular-synth people to computer-generate multiple conventional synth control voltages from multi-port interfaces.

Your mention of square waves sparks an idea perhaps somehow impractical or maybe already being done by somebody--

A naive square wave frequency of samplerate divisible by power of 2 might not have excessive aliasing nasties for this task-- At 44.1 k samplerate maybe a naive square wave at 5512.5 Hz or 2756.25 Hz. Low enough that maybe the interface antialias filter would not ring too badly when hit by the naive square wave. But high enough frequency to allow fast enough control rate (after rectifier lowpass filtering) for most analog synth purposes.

Feed the square wave to an ordinary not-dc-capable audio interface. Feed interface audio outputs to relatively simple "precision rectifier" opamp circuits. The kind of circuits with the diodes inside the opamp feedback loops. Then feed the rectified outputs into simple maybe 6th order analog lowpqss filters to supress residual high frequencies. Finally a trimmable, temp-stable output driver stage so that maybe a +/- 1.0 square wave in the computer can be adjusted to output a 10 volt dc control signal. Or whatever control voltage range that the modular system expects.

The final piece of the system would be a "CV plugin" which when inserted to a daw track, it would receive controller events and modulate the level of its generated square wave audio according to the daw automation info.

Maybe that would allow CV use of any decent multi-port interface, not have to search out special dc capable interfaces? The little maybe 8 or 16 channel rectify/filter/trim hardware box ought to work with about any decent audio interface. Unless the idea has fatal flaws.

Such a system ought be rather simple and inexpensive to build in the basement but maybe it would be expensive if someone tried to manufacture and market such a thing. Unless the rectifier box could be low-cost, it would present no advantage compared to paying extra for special dc-capable interfaces.
User avatar
Architeuthis
KVRAF
 
2873 posts since 27 Jan, 2006, from Phoenix, AZ

Postby Architeuthis; Sun Oct 01, 2017 12:47 pm Re: Plugins that produce DC values dangerous for customers?

JCJR, I think you describing what Silent Way offers. But my concern is not modular synth gear, although that is one benefit. My concern is that I don't want to stop the user, from say, modulating an oscillator down to 0 hz, or hearing the direct output of an envelope generator. (and SEEING it on an oscilloscope.)
JCJR
KVRAF
 
2147 posts since 17 Apr, 2005, from S.E. TN

Postby JCJR; Sun Oct 01, 2017 2:01 pm Re: Plugins that produce DC values dangerous for customers?

Architeuthis wrote:JCJR, I think you describing what Silent Way offers. But my concern is not modular synth gear, although that is one benefit. My concern is that I don't want to stop the user, from say, modulating an oscillator down to 0 hz, or hearing the direct output of an envelope generator. (and SEEING it on an oscilloscope.)


Thanks! Last time I looked at their stuff got the impression they were mainly doing dc-capable audio interfaces, but some of them do indeed appear to encode dc levels in audio as I described. The prices don't seem outrageous. I'm not into modular or analog synths, except decades ago.

I don't think it is a hazard so long as the user knows what he's getting. When debugging code I'll write very low freq control signals to audio tracks for debug, but usually avoid hitting the play button. Just for visual examination.

So far as I know my interface, speaker processor and amps are all AC coupled and it probably wouldn't hurt anything. The 18" sealed sub is flat to about 25Hz in the room but handles enough power (about 25mm xmax) at my usual listening levels about the worst could happen is SLOWLY moving the subwoofer cone maybe a little farther than average, for brief times when the dc is changing "fairly fast". At typical about 90 dB SPL nearfield listening level, real strong bass heavy music rarely moves the subwoofer cone past a few mm.

My amps (and most "pro" amps) have built-in output protection relays designed to open in presence of DC. So even if I somehow managed to get DC to the amp outputs, the relay circuits would notice and disconnect the speakers.

Moderator: Moderators (Main)

Return to DSP and Plug-in Development