Sound quality issues with Behringer UMC1820

[planetguru]

Hey folks

I'm new on this site, so apologies if I am in the wrong forum.

I have been using a Behringer UMC1820 audio interface together with my DAW and I have never really noticed that it has a sound quality issue, but something changed in the last few days.

I got my hands on the Waldorf Blofeld, which puts out an absolutely gorgeous sound (directly to headphones). however, I notice that when I pass it through my UMC1820 interface, the sound is noticeably degraded. I have also tried it with a different synth (Novation X-station) and found exactly the same issue, so I wouldn't blame this on the audio source. There is something going on with the interface itself.

I have stripped the interface out of the rack, so there are no other cables, or sources of interference. I have removed the USB connection to my DAW and I'm just running in monitoring mode. Still, the same issue.

I have experimented with different gain levels, line/inst settings etc, but nothing changes. It just sounds a bit grainy, loss of depth and purity and janky to the ear.

I have done a further test where I recorded the audio into Logic and listened to it direct from system headphones on the mac. The sound was the same - still not quite the same quality as the source audio. I did that to eliminate the output/monitoring circuit on the UMC1820 in case the sound was still perfect internally. This does not seem to be the case.

Is this expected with this grade of hardware, or am I doing something wrong, or is it time to upgrade my interface? I have had the device for about 3-4 years.

Thanks
Chris

[Hugh Robjohns]

It's hard to interpret your description of sound quality without a common frame of reference, and 'janky to the ear' means nothing to me.

However, 'a bit grainy' and 'loss of depth' could well be related to the interface which does sit near the bottom of the quality range of available interfaces.

[Drew Stephenson]

I wonder if the incoming signal is too hot and things are being overloaded?
What cables are you using to connect the synths?
Into which inputs?
How are the source selection switches set?
What are the gain settings for those inputs?
What signal levels are you seeing on the sig and clip lights?
What signal levels are you seeing on your DAW meters?

[The Elf]

Could well be the age-old balanced/unbalanced cable confusion. Please tell us in detail what cables you are using to connect all this stuff.

[James Perrett]

The differences in sound quality should be very slight unless something is wrong. First check that you are using the correct cables for the gear that you are using - different bits of gear have different types of output stages and you need to follow the manufacturer's recommendations exactly when connecting gear up.

I'd also check to see where the UMC1820 is getting its clock from - a poorly handled jitter clock can cause the symptoms you describe.

Also, when doing listening tests like this, you need to be scrupulous when setting levels. Even something as small as a 1dB level mismatch can cause a big difference in perceived sound quality.

[Hugh Robjohns]

I wonder if the incoming signal is too hot and things are being overloaded?

Would that be described as grainy and lacking depth, though?

I guess "Janky to the ear" might mean overload distortion, but I think most people know what it sounds like when things are overloaded.

[Hugh Robjohns]

Could well be the age-old balanced/unbalanced cable confusion.

Again, possible, but the description doesn't seem to fit with the actual sound of a stereo source feeding a mono balanced input which is normally incredibly quiet and lacking bass, often with excessive reverb.

[ajay_m]

I don't know if the umc1820 has any internal panning or routing but inadvertently panning a stereo signal to mono on a mixer is certainly a good way to provoke the "something's wrong" instinct.
You could try recording a sinewave through the interface then insert a spectrum analyser plugin into the daw and see if it's indicating any spurious harmonics.
You may be able to download a test wav file with a low distortion sonewave at 1KHz from somewhere. Then assuming you have a test wave recorded at the same frequency, time align it by zooming in and getting it 180 degrees out of phase with the reference tone. Then sum the two to a mono track and adjust gain on one track till you get a null. It should be possible to get at least a 70dB null, if not, and you're confident the two signals are precisely time aligned, then you have a problem with the interface. Note that depending on clock accuracy there could be some "beat" between the two tones but the periodic nulls should be pretty deep

[Hugh Robjohns]

You may be able to download a test wav file with a low distortion sonewave at 1KHz from somewhere.

Here, perhaps?

https://www.soundonsound.com/techniques ... test-files

....but I dont know that your null test will really prove anything useful in this case.

[planetguru]

Thank you all for the comments. I'll try and respond to the questions to the best of my (limited) ability:

Firstly 'janky to the ear' - you are right, doesn't mean anything. I would go so far as to say that if I called my wife in and asked her to listen, she would not notice anything wrong, but when you switch over to the headphone output of one of those synths, it's like chalk and cheese. The sound is simply way better from the source. Through the interface, it's somewhat grainy and lacks purity, depth, clarity and possilby even losing certain frequencies, certainly towards the lower end.

Now cables. I really had no appreciation that cables might cause the issue. Whenever I need a cable, I just hop on Andertons or Amazon and buy a reasonably priced 1/4" TRS Balanced Stereo Audio Cable of the right length (something like this usually: https://www.amazon.co.uk/dp/B07CKSG5VB . I don't go for the most expensive, but usually avoid the cheapest. That's really as far as my selection goes. I have cables from Roland and other manufacturers, but to be honest they all look exactly the same to me. If you know of an online resource that discusses this, I would be very keen to learn more.

Drew, to answer your questions:

What cables are you using to connect the synths?
- See above
Into which inputs?
- There is only one input for each channel on the UMC1820
How are the source selection switches set?
- I'm not sure which switches you mean.
What are the gain settings for those inputs?
There are no settings other than toggling line/instrument level and 'PAD'. I have fiddled with those, but no change.
- What signal levels are you seeing on the sig and clip lights?
There is no level indicator, just a single LED showing audio detected. The clip lights are not flashing. I tend to adjust to stay just below clip level.
- What signal levels are you seeing on your DAW meters?
I'm seeing -18dB to -21dB

I have tried different cables off the rack just to see if there is any difference, but nothing discernable. I have also tried different inputs on the UMC1820 just in case there was a defective circuit on one of them - but again, no difference.

Oops.. just found this still open in my browser.. thought I had submitted it earlier this afternoon

Chris

[N i g e l]

On my blofeld module....

The stereo/L output is TRS for stereo but if R is also plugged in, the stereo-ring is, I think, grounded via resistance.

The R is just TS.

Using two TRS to mixer balanced inputs would give silence on L –ve
& hum/noise on R -ve.

Does using TS cables make a difference ?

[Hugh Robjohns]

Now cables. I really had no appreciation that cables might cause the issue. Whenever I need a cable, I just hop on Andertons or Amazon and buy a reasonably priced 1/4" TRS Balanced Stereo Audio Cable of the right length (something like this usually:

This may be part of your issue... depending on how the Waldorfs outputs are wired.

If you haven't already tried a pair of 1/4inch TS UNBALANCED MONO instrument cables from your collection of cables, order some from amazon and see if they work better.

They might not make any difference (I suspect not from your earlier descriptions).... or they might make a world of difference. Mono TS cables are, technically, what you should be using, anyway.

I have cables from Roland and other manufacturers, but to be honest they all look exactly the same to me.

One type has two terminals (tip and sleeve, TS) and the other has three (tip, ring and sleeve, TRS). They are used for different applications and functions, and confusing things can happen if you use the wrong one for the wrong application.



But the differences you're hearing could just be down to the different headphone amps in the synth and interface...

...or it could be that your DAW is mono-ing the stereo synth signal...

... or it could be that your DAW or interface is mixing the direct synth signal with the (slightly delayed) DAW signal...

...or it could just be that the interface really is affecting the sound quality.

[planetguru]

Hi Both

Worth a try, but no difference. I have tried a twin TS to single TRS lead, but the sound difference is just the same. (similar to this, but not the same brand: https://www.andertons.co.uk/roland-5ft- ... -1-4-inch/)

I'm starting to wonder if this Behringer hardware is cranky :|

Chris

[Hugh Robjohns]

Worth a try, but no difference.

Not surprised.

I have tried a twin TS to single TRS lead, but the sound difference is just the same.

Details matter hear! I assume you mean a single TRS (from the synth stereo output) to twin TS (into the interfaces two inputs) — you'd get very different results going the other way!

I'm starting to wonder if this Behringer hardware is cranky :|

I don't think anyone would argue that it's the best there is... but there are forum members here who are quite happy with the way it sounds.

[Stuart79]

If you know of an online resource that discusses this, I would be very keen to learn more.

I mean...

[planetguru]

Details matter hear! I assume you mean a single TRS (from the synth stereo output) to twin TS (into the interfaces two inputs) — you'd get very different results going the other way!

Yes that is correct. I have also tried individual TS-to-TS leads out of the Blofeld L & R into oppositely panned channels on the Behringer.

[N i g e l]

I'm starting to wonder if this Behringer hardware is cranky :|

I don't think anyone would argue that it's the best there is... but there are forum members here who are quite happy with the way it sounds.

I think this is more about the specific unit rather than the model in general.
If the unit has been working for a couple of years & then suddenly its not, what has changed ? Different setup or unit gone faulty?

Ive compared my 1820 to the blofelds headphone output and dont get any "chalk and cheese" difference, although it took a while to get the levels matched.

as James said...

...when doing listening tests like this, you need to be scrupulous when setting levels. Even something as small as a 1dB level mismatch can cause a big difference in perceived sound quality.

[planetguru]

when doing listening tests like this, you need to be scrupulous when setting levels. Even something as small as a 1dB level mismatch can cause a big difference in perceived sound quality.

I have been wondering about this and to be honest, I don't completely understand. On the synth end (Blofeld, Moog, Novation or whatever), I don't know what the output level is. How can I determine that, and then what level should I be trying to match that to? Or does this refer to something else?

[N i g e l]

on the blofeld, i turned the'phones up to max, as the blo' doesn't drive studio 'phones very loud. I then adjusted the 1820 to match.

[James Perrett]

How can I determine that, and then what level should I be trying to match that to? Or does this refer to something else?

To do it properly you need some way of measuring the sound level of the headphones. Alternatively, you need a way of measuring the signal level feeding the headphones. You just need to ensure that both levels are within about 0.1dB of each other (given that I reckon I can hear 0.2 or 0.3dB changes in eq).

[ef37a]

I saw a YT test of UMC interface some months ago and the guy seemed to know his onions and said the interfaces was showing "noticeable distortion" at just -5dBFS which got worse as the level increased. I cannot remember which member of the UMC family it was.

Now! One has to be very careful. That was just one man's view of ONE sample of an interface and we must not tar them all with the same brush. I can say however that the range have a very low maximum output. Around +6dBu iirc. That could be a problem in some setups.

I had a UMC 204HD and was impressed with the mic pre amps but I did not keep it long enough to make any careful tests on it. Passed it onto my son but he now has a MOTU M4 (as do I) so the 204 is largely parked.

Dave.

[N i g e l]

One of the utubes I watched before purchasing the Ber 1820 was a drum shootout with berry, mackie & focusrite.

There were subtle differences but were they important & was that proportional to the difference in cost? No, especially for synth sources.

Over the past 2 or 3 years ive noticed inputs getting better from tascam, zoom, yamaha & indeed Behringer itself. I am perhaps more interested in the driver latency now.

[Arpangel]

I've been using an 1820 for a long time, with no desire to change it, it doesn’t have any obvious glaring sound quality issue to my ears.
However, I do find that setting the inputs to "instrument" level, not,"Line" provides a more usable signal range in my set-up, it allows me to adjust the input levels into to my DAW a lot easier, and it just suits the output levels of my mixer a bit better.
Also I'd like to mention possible overloads, I keep the input levels quite low into the 1820, if the green indicators light up you’re already getting a bit too hot I find.

[ef37a]

I've been using an 1820 for a long time, with no desire to change it, it doesn’t have any obvious glaring sound quality issue to my ears.
However, I do find that setting the inputs to "instrument" level, not,"Line" provides a more usable signal range in my set-up, it allows me to adjust the input levels into to my DAW a lot easier, and it just suits the output levels of my mixer a bit better.
Also I'd like to mention possible overloads, I keep the input levels quite low into the 1820, if the green indicators light up you’re already getting a bit too hot I find.

That last sentence aligns with the YT guy's finding and my caution about the low OUTPUT headroom. The interfaces do seem to have creditably low noise, just need to watch the output level you demand from them. For comparison my M4 has another 10dB over the UMC but about the same or better noise floor.

Noice to see you up and posting Tony!

Dave.