DAC output is low in voltage (if memory serves, like 0.7V or so), but if you can accommodate that, it performs just fine.
Thanks! Looks like I'll get a headphone amp to pair with the 204hd then, as that's the only place it is a bit lacking. (For my TYGRs)I have a UMC202HD (which has the same DAC circuit AFAIK) and it's perfectly adequate in that role. I eventually replaced it with a Topping E30 and honestly could hear very little difference -- perhaps a slightly more "open" sound at the higher frequencies, but not by much. I now use it primarily as the amp for my (very easy to drive Yamaha RH-5Ma) headphones with (IMO) very good results.
So is the dac out really bad on the 204hd? If i use the rca outputs, how bad is it?
It be noted that this should only be needed if you can't get high enough output otherwise (which might well happen), as to my understanding the distortion should also go away if the output level knob is turned down. That's one time where having monitors with an overly hot input may actually come in handy... a spare mixer could also be used to add some gain. Obviously you don't want to overdo it either, just dial in the levels to a point where neither maximum output not noise are a problem.1. The full-scale distortion is a real problem, that's why I always add one more limiter on the output channel when I'm playing with virtual instruments or band programs. The limiter can keep a safe headroom from clipping or near-clipping distortion, although a limiter is still not good for the sound, but still better than the clipping problem.
I can't believe it's that bad, you may just have set the input gain too high. The input level Amir measured at (which gave -1.752 dBFS at still-acceptable distortion) was 0.54 Vrms or about -3 dBu, which would be insanely loud on a dynamic mic - one rated -54 dBV/Pa and 300 ohms would have to be subjected to a whopping 139 dB SPL to produce -6 dBFS at minimum gain. Short of flat out screaming directly into the mic (or placing it in front of a guitar amp cranked to 11), you are unlikely to encounter such levels. Certainly not when miking up an acoustic guitar.2. The maximum input level is really low for sure. One time I was trying to record some guitar with dynamic mics and it easily clipped!
Agreed with both of your points.It be noted that this should only be needed if you can't get high enough output otherwise (which might well happen), as to my understanding the distortion should also go away if the output level knob is turned down. That's one time where having monitors with an overly hot input may actually come in handy... a spare mixer could also be used to add some gain. Obviously you don't want to overdo it either, just dial in the levels to a point where neither maximum output not noise are a problem.
I can't believe it's that bad, you may just have set the input gain too high. The input level Amir measured at (which gave -1.752 dBFS at still-acceptable distortion) was 0.54 Vrms or about -3 dBu, which would be insanely loud on a dynamic mic - one rated -54 dBV/Pa and 300 ohms would have to be subjected to a whopping 139 dB SPL to produce -6 dBFS at minimum gain. Short of flat out screaming directly into the mic (or placing it in front of a guitar amp cranked to 11), you are unlikely to encounter such levels. Certainly not when miking up an acoustic guitar.
It's not great or bad, everyone just need to fully understand the pros&cons of this unit and choose to fit suitable usage.
Which is the correct fix, of course.1. Despite the output gain, the distortion happens when the output level in DAW is reaching the maximum. So I add limiter to the DAW output master channel and crank the gain up, and ask the mix engineer to adjust the volume (more gain if needed for my channel) when I play live.
Well, yeah. The same absolute amount of preamp gain at a given mic level will give different digital levels if ADC clipping levels differ, and the Behringer tops out short of 0 dBu which is quite low. This is why @Julian Krause likes to refer to system gain, which is the inverse of input sensitivity (minimum analog input level that would result in 0 dBFS; for practical reasons, one would usually measure this at a lower level like -12 dBFS and then multiply the analog input level such that it would nominally hit 0 dBFS).2. My point is, with the same mic and less gain, the UMC204HD can reach a much higher digital level than a normal mid-range interface with the same or even higher pre-amp gain.
I'm aware of these, and totally agree.Which is the correct fix, of course.
Think of digital levels and analog output levels as two distinct things. You should not be hitting 0 dBFS in the DAW to begin with, ideally even without a limiter (it should not have to catch more than the occasional rogue peak). Since digital levels are invariably limited and maximum undistorted analog output level is rather wimpy, adding more gain afterwards is the only possibility if levels aren't high enough otherwise.
Strictly speaking, you could be hitting the limits in up to 3 places: digital, analog stage before volume, and analog stage after volume control. The first two are effectively lumped together.
The good news is that nowadays interfaces with substantially higher output and input levels no longer cost that much more than the Behringer, which has simultaneously crept up in price. Right now a 204HD is $109 at Sweetwater, while an EVO 4 is $129, a Minifuse 2 is $149 and a Tascam US-2x2HR or Focusrite Scarlett 2i2 gen3 are $179-180. If you don't need more than one mic input there's also the Scarlett Solo ($120) or the UAA Volt 1 ($139). There's a ton of options up to $200.
Well, yeah. The same absolute amount of preamp gain at a given mic level will give different digital levels if ADC clipping levels differ, and the Behringer tops out short of 0 dBu which is quite low. This is why @Julian Krause likes to refer to system gain, which is the inverse of input sensitivity (minimum analog input level that would result in 0 dBFS; for practical reasons, one would usually measure this at a lower level like -12 dBFS and then multiply the analog input level such that it would nominally hit 0 dBFS).
With system gain, you can compare apples to apples. For example, an Audient iD22 provides mic gain between -10 dB and +60 dB, with the lowest setting permitting a maximum input level of +22 dBu. That's an input sensitivity of +22 dBu - (60 dB - (-10 dB)) = -48 dBu, or a system gain of +48 dBFS/dBu (+50 dBFS/dBV).
Julian's system gain for the UMC202HD turned out to be +51 dBFS/dBu, absolute preamp gain itself being a tad short of 50 dB.
In effect, both are hitting a similar input sensitivity, the UMC202HD will even get a bit more digital level out of dynamic mics (podcasters / streamers rejoice), but the iD22 with its 70 dB gain range can be turned down a lot more to accommodate LDCs that are being sung / screamed into or even pro line-level audio.
Has anyone tried to mod this audio interface or similar behringer audio interface to add a on/off switch button?
The ADC part (inputs) is mediocre, though in my loopback testing, it was better than that. The issue is the voltage level of that output, which is not high. If you just want to know what the outputs are doing, you'd do better to have a separate ADC with known low distortion. Can you borrow one?Hi,
(I'm mainly interested in the distortion of the output channels).
Thanks, so in Armirs measurement at the beginning it was not fed into the UMC 204 input? The second harmonic is at around -100 dB while I'm at around -50 dB . I could feed it in the Zoom R8, but I need two PCs for that... would be doable.The ADC part (inputs) is mediocre, though in my loopback testing, it was better than that. The issue is the voltage level of that output, which is not high. If you just want to know what the outputs are doing, you'd do better to have a separate ADC with known low distortion. Can you borrow one?
Take a look at the first couple graphs. Amir noticed that turning down the gain by 3dB had a dramatic effect on the distortion.Thanks, so in Armirs measurement at the beginning it was not fed into the UMC 204 input? The second harmonic is at around -100 dB while I'm at around -50 dB . I could feed it in the Zoom R8, but I need two PCs for that... would be doable.
I also see a difference in THD when the Main Volume is at ~50% compared to 100%. THD increase by a factor of ~2 when I increase the volume of the output. But as it is weak anyway I can't work with half of the max. volume.
I think I don't understand it. So the signal is generated with an external generator and then fed into the UMC input. I wonder how the gain of the input is then set. Just below clipping, as low as possible, etc. I tried of course changing the input gain, but I didn't improve that much. Pressing the Pad button helps a bit though. But still THD at around 0.3 or so.Take a look at the first couple graphs. Amir noticed that turning down the gain by 3dB had a dramatic effect on the distortion.
It's fed in as a digital signal, i.e., via USB, not the analog inputs.I think I don't understand it. So the signal is generated with an external generator and then fed into the UMC input. I wonder how the gain of the input is then set. Just below clipping, as low as possible, etc. I tried of course changing the input gain, but I didn't improve that much. Pressing the Pad button helps a bit though. But still THD at around 0.3 or so.
Maybe related to this: https://www.audiosciencereview.com/...dio-interface-review.9856/page-6#post-1048970Hi,
I have a question concerning the distortion of the UMC 204 (I'm mainly interested in the distortion of the output channels). I measured it with REW and I get THD ~0.5% which is quite mediocre.
see here: https://www.audiosciencereview.com/...hringer-umc-204-distortion-on-main-out.36995/
Here's another measurement which looks similar: https://www.audiosciencereview.com/...nterface-for-rmaa-dac-test.8182/#post-1302848
Could it be that this is caused by the fact that I'm going from the output into the input, so the distortion adds up? How is Amirs measurement done? Is the output measured separately for instance by going into a high quality analyzer? (edit: Obviously output into input of the same device)
I wonder if the unit is broken or if I should get something better. Amirs measurement looks much better, but I question that it's comparable to what I did. I also measured a Zoom R8 the same way and there I get THD ~ 0.003% with the same method. So a factor of 100 better at least.
Yes, I knew this. I set the volume in alsamixer (Linux) to -3dB to prevent clipping, which is audible easily by playing a sine tone. I thought first that it's the volume of the Behringer, but with -3dB in the software and the volume control of the Behringer on maximum there's no audible clipping.