Probably most of the users of this device are engineers and are used to dB units so it might not be an issue, however I am an amateur and for me Vrms are more “user friendly” units. I am measuring DACs, DSPs, and car head units where typical users are used to see Volts as the output unit. I would like to request to add a selectable option to display Vrms on the X axis of the THD vs Level graph.
As well it would be great to have the attenuation buttons in Volts as well as the generator level. I mean as an amateur, my brain works much better with “normal” volts. Maybe there could be a selection in preferences to choose “global units” for everything? like if I select Vrms everything everywhere is in volts, the graphs would he generated using V on the X axis.
Hi @RAW-CAt, you can display measurements in Vrms, but most graphs need to display several orders of magnitude of information, and so the graph makes the most sense in dB of some kind.
When you are setting the generator and reading peaks, you can opt for readings in volts if you’d like.
dBV is quite easy to get comfortable with. 0 dBV is 1V, -20 dB is 100mV, -40 dBV is 10mV. And -50 dBV is 3.16mV. You can see the repeating pattern.
dBu is harder to remember, but if you think of it as an offset from dBV, it’s a bit more palatable.
Below you can see the readings in Vrms.
Hi RAW-Cat. I think you are right to note that decibels are complex and that people struggle with them. You have also had your answer from Matt.
I hope you don’t mind me mentioning that dB are not a different unit but a different measurement. Two reasons, first, decibels are power not voltage; so Watts not Volts. We don’t measure power directly, hence Volts are used as a surrogate and so the axis are labelled “V” in some way. Second, measuring power is itself a surrogate for volume, human perception of loudness. This is why dB are naturally a relative measure like percent (as we have no absolute sense of volume). So dB graphs and values are really volume graphs and values, even if what was measured was a voltage and the way it is expressed is absolute.
It does not help that so much of the signal chain uses voltages to carry the signal and all digital values are voltages. So power makes almost no appearance in our working. This makes the power connection seem irrelevant and our obsession with dB seem perverse. Indeed, the only place where power usually appears, is the end of the chain. However, the listener sits at the end of the chain. It is their perception that matters in the end. We do our calculations as if working with power so that we do our thinking as if working with volume.
A voltage measurement or a voltage graph makes complete sense and is a very valid thing to do, but it is distinct and separate from a volume measurement or a volume graph in decibels (even where measured and expressed in Volts). If you want to ensure that the input of one device will cope with the output swing of another, for example, voltages are fine. If you also want to know the impact of a filter, attenuator, amplifier, or other sound-changing feature, voltages are hard to understand. From experience then, and this is just something to be aware of, those who do most of their work in decibels make fewer mistakes than those that work in volts.
That is very understandable from engineers perspective. I guess that my use of this device is more amateurish. My followers are average people that are interested about head units output voltage via RCAs. Or a DSP output voltage, that kind of thing. If I will show hem a graph of a DSP that outputs +18dBV, they will have absolutely no idea what that jis and how that compares with a different brand DSP that advertises 2.5V output.
Hi @RAW-CAt, what graph specifically are you thinking about? Maybe I’m just misunderstanding your question
This is an example of a THD vs level graph I made using REW. In my video or written analysis I can say that this DSP is cleanest at 3V output and maximum output us 6V.
If I plot the same with QA software, I need to make conversions from dBV to Vrms in order to compare this Head unit with the Helix DSP above.
Ahhhh, thanks @RAW-CAt, now I better understand what you are after.
You can get pretty close to this out of the box using the PWR THD vs Power plug-in. And then, set the resistor to 1 ohm and change your X axis title to be volts. This works because if you set your load 1 ohm, then volts and watts are the same (which is in line with @Richard’s comment above–there are a lot of places in the discussion of noise where the measurements are referred to as “noise voltage” but there’s an implicit load assumed and it’s often 1 ohm.)
In any case, the plot will look like this:
I made the plot with the following options for the measurement:
Hi @RAW-CAt, for the next release, the AMP THD vs Input/Output plug-in will have a setting that will allow you graph the X-axis as dB on a linear scale (either dBV or dBu) OR as Vrms on a log scale.
For example, with ```Plot Vrms for X-axis":
And that yields:
And with the checkbox un-ticked:
That yields:
Nice! Now I will have almost no use for REW😂
That is truly awesome.
I was only sitting there a few days ago with my calculator making conversions (dBV to V) and doing spot measurements attempting to find a sweet-spot with an input switching IC.
Additionally, it’d be great if I could plot the THD (without +N) only vs level.
just remove the flag…
Hi. I would like to see a tile available that shows only the Distortion component, D. There are already tiles that show Distortion + Noise (N+D), Noise only (N-D) and for completeness, although easily calculated, it would be nice if there was only Distortion (D). Great, again in my opinion, would then be if there was also the graph 
Yes of course, but that is a value relative to the signal S. In other words : THD= D/S and using dB expressions THD = D - S. What I was asking for was a tile of just the value of D and not its value relative to S (THD). It is a little like THD+N and N+D. THD+N is the component relative to the S signal of N+D. Of course, this is just a requirement of mine that perhaps no one cares about!
Ignore my ignorance… but why would you need just the value of D?
Hi, simply because D is an absolute term of distortion and not a value relative to the signal. In other words, D does not depend on the signal, but measures the harmonic component of distortion in absolute terms. For example, with D being equal, THD decreases as S increases. It is a bit like N+D and THD+N. N+D (like D) is an absolute value and measures the effective component (in power, which can then be easily translated into voltage) of noise+distortion. The same thing happens with D: it measures the effective component of distortion alone, independent of the signal. Of course it is easily calculated, but if it were available, at least for me, it would be convenient.
You want distortion (component) shown as a voltage
Yes, that is correct. Since noise (N-D tile) and distortion+noise (N+D tile) are shown as voltage (expressed in dBV), for completeness I would like the distortion component D (D tile) to be as well. In a sense this would be a kind of “residual display” (without taking noise into account) expressed directly with a number. As I said, it is easily calculated, but if there is a tile that shows it directly, at least for me, it would be very convenient.
Noted, thanks @Claudio!
Great, thank you @Matt