Obs problems on Ubuntu

AaronD

Active Member
Just FYI, from the pro audio side of things:



XLR is:
Pin 1 - X - Shield / Ground
Pin 2 - L - Live / Hot
Pin 3 - R - Return / Cold

TRS is:
Tip - Hot / Live
Ring - Cold / Return
Sleeve - Ground / Shield

Same signals, different plugs, handled the same way, hence the combo jack. It's also valid to use a separate jack for each type, like on an analog mixing console. That's cheaper than a combo, and might be what you're used to, but when space is tight and only one is valid at a time anyway, the combo works well.



TS has a longer sleeve, effectively just removing the insulation between Ring and Sleeve. Thus, a TS plug keeps Cold grounded, so that the unbalanced Hot still has something to refer to. That's usually how a guitar cord works, at the receiving end.



TRS IS NOT STEREO!!! Nor is it anything else. It's just a 3-pin connector that can have anything on it. Some common incompatible standards are:

Balanced mono, as above:
Tip - Hot / Live
Ring - Cold / Return
Sleeve - Ground / Shield

Unbalanced stereo:
Tip - Left
Ring - Right
Sleeve - Ground
Ground can't be an effective shield here, because it's now the reference for both signals. Anything that it picks up will appear on the signals, again because it's the *reference* for those signals.

Unbalanced mono insert:
Tip - Send
Ring - Return
Sleeve - Ground
This is used on analog mixing consoles, to take a signal out of the board, through some external processing, and put it back into the board at the same point that it came out. Effectively *inserting* the outboard processing at that point in the board's internal chain. It has a switch built into the jack, so when there's nothing plugged into it, it connects the Send and Return together, so the internal chain "just works" on it's own.



If you get the standards mixed up, because they do all use the exact same connectors, you WILL be confused! They've been common forever, mostly for lack of anything else, but if you use them, you really need to know what you're doing, as the original designers intended you to know.

Unbalanced stereo (like a phone or computer) feeding a balanced mono input (like a mixing console or a pro-level USB interface), will result in the center channel dropping out. You're actually listening to the *difference* of the two speakers instead of their sum, or the Side channel of Mid+Side encoding. Lead vocals, bass, and whatever else is panned center, will go away, and you're left with a mix of everything else that depends on how hard the mixing engineer panned each one. (how far off to either side it is)
To fix this, use two input channels, not just one, and use a cord that splits the stereo signal into two mono's. Likely a TRS on one end and two TS's on the other: one TS connects its Tip to the Tip of the TRS; the other to the Ring of the TRS. That same cord may also be used for Inserts, and sold as such, with the TRS going in the board, one TS going to an Input jack of the outboard gear, and the other TS to an Output jack.

Unbalanced stereo (same as above) feeding an unbalanced insert, will result in only the Right channel getting in, without an input gain control because the Insert point comes after that (and often after the EQ too, in an analog mixing desk, but not always), and the Left channel fighting the board's output driver for what signal should be on that wire. You probably won't notice that fight, until you use one or the other device for something else and wonder why that part of it doesn't work anymore. (the fight fried it) And you don't get Mono either, only Right, even though it splits from there into both speakers. That's still only the Right channel of what the mixing engineer intended.
This is often the next thing people try after the previous one sounds weird or "broken", but it's the one that could *actually* break something despite sounding mostly as expected.
DON'T EVER DO THIS!!! If you have the previous problem, use that solution. Only if you've recorded something from the Insert Sends, are you safe to play it back unchanged to the Insert Returns. For example, a multitrack recorder in the Insert loops of a console.

Unbalanced mono in an Insert jack, depends on which direction you're going:
If you're taking a signal *from* the board, that works, and the grounded Ring will make the rest of the channel strip "dead". Some people leave the plug half-out so the built-in switch doesn't activate, but that's just inviting someone else to push it in for you while you're gone because "it was about to fall out", and leave you wondering why that channel suddenly stopped working.
If you really *must* do this, make your own "Insert splitter" from a bare TRS jack that you connect Tip and Ring together, and also bring that out to the wire. Push that all the way in, and the channel strip will still work because of the Tip-Ring connection, and you also get that signal out to do something else with. For example, you might put it into to the Ring of a different channel Insert, to double-patch that mic: this gets you separate controls after the Insert point, like mute, fader, and routing, and whatever else that means for your board, only sharing the controls before that point, like the input gain.
If you're putting a signal *into* the board, then you have the same fight as the unbalanced stereo Left channel above. DON'T DO THAT!!!

Balanced mono feeding Unbalanced stereo, likely for recording on cheap gear, will appear to work just fine in headphones, might be weird in live speakers, and will drop out completely when the stereo track is mixed electronically to mono. Or it could only end up in the Left channel, while the Right channel is silent. Or it could "wander" between Left and Right channels, with the live-speaker weirdness and electronic-mix dropout becoming worst when the wander happens to be centered between Left and Right. Which you get, depends on how the balanced output works. A balanced receiver won't care: it just takes the difference between Hot and Cold and passes that on, which is a massive clue towards understanding why the weirdness and dropouts when you preserve both.
To fix this, use balanced receivers.
If you already have a recording that does this, you can try to fix it by inverting the Right channel and *then* mixing to mono. This is a poor simulation of a balanced receiver, but it might be good enough to work.



Again, if you're going to use 'phone plugs - TS or TRS, regardless of size because adapters also exist - you really need to know what each one is doing and don't mix them up. Just because they fit together does NOT mean they work together.

This is especially true if you have an amplifier with TS outputs! If you do the math for some of them, it gets pretty close to an American wall socket! (120V RMS: just because it's an audio signal does NOT mean it's safe!) And you just casually have that exposed at the other end of an unmarked 'phone cord when it's not plugged into a speaker!
Not to mention using a thin-wire guitar cord for the amount of power that it takes to drive that speaker. Also not good. Big-wire speaker cords also exist, with the same connector, and those technically work, but then you REALLY need to keep straight which is what! (and they still have an electrical shock hazard)

Before Speakon connectors (made specifically for speakers, and touch-safe in both genders), some people used wall-socket connectors, which had the advantage of using plain old extension cords to go as far as they needed, and the hazard of plugging one into an actual power outlet. Speakon was (and still is) a welcome relief. USE IT!!!
 
Last edited:
Ok, so in your opinion the Focusrite is more of a commercial gimmick to refer to a certain commercial target, the fact that it is good for both voice and guitar depends on its design/quality.
I've seen the inputs on the Behringer, so they accept both Cannon and Jack cables! I would say excellent!
 
The situation as you describe it has a lot of details, I'll have to start by simply inserting the microphone into the sound card and connecting it to the PC to make Obs see it and record a decent voice, then slowly I'll do some tests to better understand the All.
 

AaronD

Active Member
Ok, so in your opinion the Focusrite is more of a commercial gimmick to refer to a certain commercial target, the fact that it is good for both voice and guitar depends on its design/quality.
No, it's good too. Not a gimmick. Designed and made well. Just not as flexible as the Behringer.

...then slowly I'll do some tests to better understand the All.
As you do that, understand what you're actually doing, so that you never connect two output drivers together. Most modern things will recognize that they're fighting and give up, but not everything will.
It's cheap enough - practically free - to add self-protection to a mass-produced chip, but it's still an extra add-on to its primary function.

Every once in a while, I see a discrete design get all hyped up again. That will probably be expensive and not have the extra self-protection, so it won't give up the fight, and it'll destroy itself instead. Tubes/valves are always discrete, so they're all in that category. You don't want to have a habit that ends up killing someone's expensive fragile thing.

...so they accept both Cannon and Jack cables!
Those are the "slang" terms for XLR and TRS. Not as much detail, and potentially confusing because "jack" can refer to *any* panel-mounted connector that accepts a cable-mounted "plug". An "XLR jack", for example.
And yes, there are XLR's with more than 3 pins, that are not compatible with other pin counts. But the default is 3.

Also, there TS, TRS, TRRS, etc., that differ only in the number of Rings. A phone or PC headset (not headphone), for example, is TRRS:
Tip - Left ear
Ring - Right ear
Ring - Ground
Sleeve - Mic with 5V / 2kohm phantom power
An external resistor on the Mic wire to ground is used to communicate what's plugged into it:
0ohm (short circuit, 0V) - headphone only, no mic, because a TRS headphone plug naturally does this.
2kohm (centers at 2.5V) - mic present
infinite (open circuit, 5V) - nothing plugged in
The volume and transport control buttons also mess with this resistance.

This also means that for guaranteed compatibility, you can't just put a generic external signal into a phone's mic input. The 0V-centered signal will make the phone think there's no mic, and not use it. You need a 2k resistor, as above, and a capacitor to allow the DC-offset, after turning it down as needed to not overwhelm the mic input. Then it works.

A separate TRS mic jack is:
Tip - Mic without phantom power, but it does have the DC-offset capacitor built-in
Ring - 5V power to the mic, self-protected against a short to ground (like a TS plug would do)
Sleeve - Ground
Now imagine what that would do if you put a naive XLR mic to TRS adapter in there. Because of the details of what each thing actually is, you probably won't break any of it, but do understand what you're actually doing.
 
Last edited:
Thanks for the info, I'll see what to do...
Fostex+SM57 I would need the optical cable which becomes USB but I only find optical cables (Toslink) which become RCA.
Or connect the Fostex to the PC via Jack outputs but a sound card would always be needed.
So definitely better to go with Behringer or Focusrite and use the SM57 with these.
Or opt for the economic choice of the microphone with the sound card, but with the risk of having a poor result.
 

AaronD

Active Member
Thanks for the info, I'll see what to do...
Fostex+SM57 I would need the optical cable which becomes USB but I only find optical cables (Toslink) which become RCA.
Or connect the Fostex to the PC via Jack outputs but a sound card would always be needed.
So definitely better to go with Behringer or Focusrite and use the SM57 with these.
Or opt for the economic choice of the microphone with the sound card, but with the risk of having a poor result.
I think you're getting it. Not just what to do, but also how to figure it out. Have fun!
 
Hi AaronD, the Behringer has arrived, I connected the 57 and I'm using Obs with Windows now, the signal arrives because by touching the microphone I see the volume bar on Obs advancing, but when I start recording the audio is not captured, I'm definitely doing something wrong in the settings, can you help me?
Thank you
 

AaronD

Active Member
What's in your Advanced Audio Settings? And Settings -> Output and Settings -> Audio? (you should be able to predict that question)
 
HI! Here are the sound settings!
 

Attachments

  • 1_.jpg
    1_.jpg
    83.9 KB · Views: 23
  • 2_.jpg
    2_.jpg
    84.3 KB · Views: 20

AaronD

Active Member
That's fine so far. That only gets you the mic input, which you already said you had. What about Advanced Audio Settings and Settings -> Output?
 

AaronD

Active Member
Hmm... You have the Monitor disabled (I think is what that says), which keeps the Output on. And you have the Output going to at least Track 1, which should be what the Simple recording mode picks up. That should work.

You won't have it in headphones though - that's what the Monitor is for, and you choose the Monitoring device on the Settings -> Audio page. According to your previous screenshots, you've left that as Default, or whatever the OS has selected.

You say it's not in the recording?



Also, if you want separate volume control of each input, you'll need to have two copies of that stereo source (two globals set to the same one), then in the mixer, right-click and Rename them to something that makes sense for each, then in the Advanced Audio Properties, pan them hard to either side and check the Mono box for both.

Now each one has a separate volume control, but you'll still want to experiment to see if the filters (EQ, Compressor, etc.) that you're going to use (if any) work sensibly. That is, do the filters on one signal only see that one signal because the Pan and Mono come first? Or do both sets of filters see both signals and respond to them because the Pan and Mono come last? You probably don't want the latter.

A DAW won't have that problem, but you then have to loopback or virtual audio cable the finished signal from there into OBS, to pass through unchanged. All of that is free too, after some googling.
 
I managed to get the audio, I did some tests, restarted Obs, but frankly I don't seem to have changed anything, I still hear the audio, but the volume is low, the volumes on the PC are at maximum, the input volume is set to maximum so as not to clip, to record the voice I would need the microphone to be 50 cm away, the 57 is probably not suitable for this, will it take a panoramic?
 

AaronD

Active Member
It'll take anything that has an XLR or 1/4" (6.5mm) plug on it. If the mic needs phantom power, it'll do that too. (the 57 doesn't) (phantom power doesn't get to a 1/4" plug, in case you're concerned about frying a guitar; it's a different set of contacts)

I have a Behringer XM8500, which is their clone of a 58, maybe a full meter away. Gain is not a problem. Though with that setting, it would clip if I came closer. Set the input gain for how it's actually going to be used: it's not absolute. Audacity might help with this, just to be a big, high-resolution meter. Or maybe OBS's meter is good enough.

I also have a "shotgun" mic from Thomann in Germany. (it was cheaper to buy from them and pay shipping to the U.S., than to buy from here; we have some stupid markups on media gear, and no decent "house brand") The idea was to be directional enough to reject a lot of the room reverb, which it does, but not completely. (magic still does not exist, unfortunately) It does need phantom power, which you can turn on with a switch on the back of the Behringer box.
I like the sound of the XM8500 better, so if I end up padding the room, I think I'll switch back to it.
 

AaronD

Active Member
But with all of that said, I wonder if the reason you think it's low is because you're comparing a good "live" level to what you're used to from a CD or streaming service.

Live needs to be lower because it doesn't know what's coming. Natural spikes need to be captured cleanly too, and so you need to leave enough headroom to do that.

A finished track knows what's coming, and has tamed most of it, so it can use all of its headroom to overcome the limits of the distribution medium. (or to "stand out" on the radio by being superficially loud...and boring / tiring)


You have two options here:

If you want to keep the "live" sound, then your audience will have to turn their volume up. No other way around that.

If you want to match their expectation of volume, then you'll have to slam the meter. Compress it hard enough that there aren't any more peaks, but tweak the compressor settings so that it doesn't make itself obvious, then use the compressor's makeup gain to fill the meter.
 
So is it normal that my 57 from a distance of 50/100 cm picks up the voice poorly and the recording comes at a low volume as it is a suitable microphone to keep close to the sound source?
 

AaronD

Active Member
So is it normal that my 57 from a distance of 50/100 cm picks up the voice poorly and the recording comes at a low volume as it is a suitable microphone to keep close to the sound source?
It *is* meant to be close, but that does not mean that it's poor from far away. Turn up the preamp gain (on the box) to where you get a decent signal anyway. That's just part of how you run a professional rig.

You don't calibrate it close and then put it where it's going to go. You put it where it's going to go, and then calibrate it in-place to do what you want. There's no set setting for a given mic; everything depends on the specific use.
 
So you mean the Behringer gain? I tried and it improves a bit, but you don't have to set it to the maximum otherwise it starts making noise, it has to be set almost to the maximum and in any case the microphone cannot stay at 1m, but I found a solution, I can hang the microphone from the ceiling towards the bass so that it is 20/30 cm from the face and at this distance it is possible to have an acceptable volume, then in a while I will buy a more suitable microphone to stay at a distance.
 

AaronD

Active Member
So you mean the Behringer gain? I tried and it improves a bit, but you don't have to set it to the maximum otherwise it starts making noise
Yes. And I do have it all the way up for one of mine, and it's fine. Turn the fridge and air conditioning off though.

For unavoidable noise, like computer fans, a Noise Suppressor can come in handy, but that's not magic either. Especially if the noise is not constant. You might need to change some settings in your computer to keep the fan on (counter-intuitive, I know), so that a noise suppressor doesn't have to chase that change. This only works for low-level noise though; a jet engine is still going to dominate regardless.

It also helps if you can set the noise suppressor to barely do enough and no more, but I don't think OBS's version has any controls.

I found a solution, I can hang the microphone from the ceiling towards the bass so that it is 20/30 cm from the face and at this distance it is possible to have an acceptable volume, then in a while I will buy a more suitable microphone to stay at a distance.
I tried that briefly: hanging overhead. Probably the same idea that you had - as close as possible but still out of frame - but it sounded weird. So I ended up with both the vocal mic and the shotgun mic next to the camera instead, gained up at the preamp to a decent level, and for a while I recorded both and chose the best one later. I like the sound of the vocal mic better, but the shotgun's room rejection makes it better overall. Since that depends on the room, your mileage will definitely vary.

It's worth spending effort to make the raw mic sound good, instead of trying to fix it electronically. The less you can do electronically, the better, but still do everything that it needs.

For what it's worth, I don't do audio processing in OBS unless I absolutely have to. Only one rig has to, and I really want to rebuild it differently for several reasons. That's just one of them. Another rig uses a physical mixing console, and another uses Ardour on the same PC as OBS, both of which take all of the inputs themselves and do all of the audio processing, and provide a finished mix for OBS to pass through unchanged.
So the video processing and audio processing are kept separate, and done by dedicated things that are designed to do exactly that. Only at the end are they combined together and sent out.
 
Top