Audio Interface Issue

[Majik]

@DavidP

I can’t explain in depth how the dB scale works (I’m sure @Majik could do that). Rather than saying 0dB is the threshold for distortion, think of it as the point beyond which you will get clipping of the signal.

That is, basically, correct.

What you care about, when recording, is “level” (signal level). This isn’t the same as “volume” or “loudness”, but is related. The actual volume will depend on how much you turn up your amp when you play it back.

Sound is measured using decibels (dB) which is, in itself, just a logarithmic scale. Logarithmic scales are used because the way we perceive audio is, basically, logarithmic in nature, so normal linear scales don’t work very well. Decibels also makes the maths relating to audio signals a lot easier.

Decibel scales are always relating one thing to another, so will always be done with some sort of reference level. There are different types of decibel-based scale used depending on what aspect of audio you are measuring.

For example, measuring the actual signal level in a cable could use the dBV scale (relative to 1 volt) or dBu scale (relative to 0.7746 volts).

How loud something sounds is measured using SPL (sound pressure level) where 0dB is defined as the lowest limit of human audibility. Note that SPL is complex subject but it is possible to map between, say, dBu and SPL if you know the gain of the amplifier and the efficiency of the speakers.

Back to recording:

For recording, the “level” in the DAW is an abstract number which represents how high the signal is. This is represented by the dBFS scale where FS is “full scale”. That means 0 dBFS is the largest possible signal level. Specifically, 0 dBFS is the loudest signal that can be represented by the digital PCM format. On digital systems, you will normally hard clip above 0 dBFS, which is a very bad thing.

Which is why recording well below 0dbFS is a good thing as you can always boost the signal level in the DAW.

Note that another thing to be concerned about is “dynamic range” which is the difference in level between the loudest and quietest sounds you are recording. I’ll post about that separately.

Cheers,

Keith

[Majik]

To expand a little on dbFS: as I said, the scale based on 0 dbFS being the loudest signal that can be represented by the digital PCM format. This is the reference point and, therefore, there are (strictly speaking) no value above 0dbFS, which is why all values are negative: -12 dBFS represents 12dB lower than 0 dBFS.

I say “strictly speaking” because DAWs normally do allow you to create signals within the DAW that are higher than 0 dbBS. But this is entirely a function of the DAW. You cannot capture signals with an audio interface above 0 dBFS, and you cannot render audio files for playback with signals greater than 0 dBFS. If you try, everything above 0 dBFS will be lost.

The reason DAWs allow this is because the DAW is “non destructive”. Everything inside the DAW is just numbers and some maths, so you can represent things that don’t exist in reality. The DAW does this to protect you and your data from corruption.

For example: if you record a signal with a peak level of -12 dBFS, and then apply a gain to it of 18 dB, you will end up with the whole signal level being increased, so that the peaks are now at +6 dBFS.

-12 dBFS + 18 dB = +6 dBFS

(as you can see, the maths is straightforward, which is why we use dB).

Now, if you did that by accident, then you later applied a gain reduction of -12dB, then you would bring the signal level down to something that could be rendered but, if the DAW chopped everything off above 0 dBFS then your recording would be ruined.

Now, you may ask why they start at the top with 0dbFS as the reference point and go down. Why don’t they start at the bottom and go up?

This is because the dbFS scale is independent of the resolution that you record at.

There are a number of digital recording formats which have different resolutions. Most of us will be recording at 16 bit or 24 bit. These are the practical formats which fit the limits of human hearing, and which also match with computer architectures, which generally work in groups of 8 bits.

16 bit recording represents the signal with 65,536 different levels (0-65,535)
24 bit recording represents the signal with 16,777,216 different levels (0-16,777,215)

In each case 0 means “no signal” and the highest value means “the highest signal”.

If we had started the dBFS scale with 0 dbFS = 0 in the PCM scale, then there would be a mismatch between recordings in 16 bit and 24 bit: you could make a 24 bit recording much louder than a 16 bit one. This would be confusing in the DAW where you, generally, don’t care about the resolution whilst working in the DAW itself.

By starting the scale with 0 dBFS being the loudest possible signal (65,535 for 16 bit, 16,777,215 for 24 bit) then we have a common reference point that makes sense, and is important for how we work within the DAW as, generally, when editing music we care about “how loud can I go” much more than “how quiet can I go”.

I’ll talk about that more when I post about dynamic range.

Cheers,

Keith

[Majik]

Dynamic Range

Dynamic range is the difference between the loudest sounds captured, or represented in the DAW, and the quietest

We have already noted that the loudest sound we can possible capture or play back is 0 dBFS. What is the quietest?

Well, with all recording media whether analogue or digital, the lowest level we can capture is normally dictated by the “noise floor” of the media itself. All media has a noise floor. Noise floor is commonly represented on a dB scale referencing the dBFS scale.

For example, analogue tape, as used in high-end recording studios has a dynamic range of up to 77dB. This means that, if the highest signal you can capture is at 0 dBFS, the quietest one is at -77 dBFS.

0 dbFS - 77 dB = -77 dBFS

For digital media, the lowest level you can record is the smallest non-zero value you can represent using the digital format. This is 1. What 1 means depends on the resolution you are using. Therefore, higher resolutions have a wider dynamic range (and an equivalent lower noise floor).

Roughly every bit of resolution equates to 6 dB of dynamic range. Thus a 16 bit resolution digital recording can have a dynamic range of 96 dB, and a 24 bit digital recording can have a dynamic range of 144 dB. Note that both of these are far higher than the “gold standard” of the analogue world.

Also, practically speaking, 24 bit recording/playback is limited by the noise floor of the analogue components in any audio interface/DAC to around 120 dB.

(There are actually some approaches, like “dithering”, which can extend the dynamic range of a 16-bit signal to 110-120 dB).

In fact, by doing the calculation in reverse, the equivalent bit depth for analogue media can be calculated.

So, for studio tape, the equivalent digital resolution is 12.8 bits (or less).
For cassette tapes or vinyl (70 db dynamic range) it is 11.6 bits (or less)

So, you can record 12 dB down from peak using a 16 bit audio interface and still have a better dynamic range than the best analogue recording studio that ever existed!

Of course that doesn’t mean you will equal the quality you get, because a lot depends on things like sensitivity and other characteristics of the mics, the linearity of the pre-amps, and the noise introduced in any analogue parts of the chain, like hum or background noise from heating systems, etc. And, of course, the general acoustics of any room you are recording with a mic in.

The question then is: how much dynamic range do you need?

Probably a lot less than you think.

A lot depends on the type of music. A dynamic classic piece will require a wider dynamic range than a heavy rock number. It also depends on the audience: as a listener, your experience is limited by the dynamic range of the your listening media. That is, the air between your speakers and your ears. Your listening environment has a “noise floor” too. As does your car, which is why you may find yourself turning up the volume to hear quiet sections of music whilst you are driving.

According to one of the “bibles” on the subject (“Mastering Audio, The Art and The Science”, by Bob Katz) about the best dynamic range you can hear will be in a good cinema, which has a dynamic range of around 63 dB at best.
A good home cinema (with acoustic treatment) will have a dynamic range of less than 50 dB.

Most homes have a dynamic range of around 35 dB

And, in a car, you typically have a dynamic range of less than 20 dB.

Note that these values are all still well within the range of 16-bit digital recording.

So, if you want your music to be heard in these environments, you need to keep the dynamic range within these limits. This is, typically, done using dynamic range compression (compression plugins in your DAW).

Note that, before you start worrying about the impact that compression has on your audio, it’s a standard part of the recording process. As long as it’s done with care and not “overcooked” then it will only enhance your recordings. Of course, if “overcooked” is what you want, then go for it. The Oasis album “What’s The Story, Morning Glory” is famously over-compressed, for artistic impact.

Every single commercial recording in the last 50 years and, probably, every single recording ever, has had compression applied to it.

If you are uploading to something like Youtube, there’s some guidelines here:

• Dialogue: -6db to -15db(Nb. Most YouTubers tend to stick at -12db max)
• Overall mix Level: -12db to -20db
• Music: -18db to -20db
• Sound Effects: -14db to -20db

Note that, here, they mean “dBFS”.

There’s also a useful reference at Mastering for Soundcloud, Spotify, iTunes and Youtube. – Mastering The Mix which includes a table for different platforms, including Youtube.

Also note that Youtube applies it’s own level “normalization” to any audio you upload to bring it into the range -12db to -20db.

(Apparently, more recently they’ve made this -14 LUFS: YouTube Changes Loudness Reference to -14 LUFS).

The dynamic range for Youtube is, generally, stated as >9dB.

Cheers,

Keith

[TheMadman_tobyjenner]

John

Just wanted to know how the UM2 is connected to the laptop ? Are you using the USB connection or via RCA cables. When I used mine for vocal recordings on the laptop on a couple of Roman’s projects, I only ever used USB and the vox on Born To Be Wild and Irish Pub Song were plenty loud !

I am wondering if that could be the problem ?

I have fished mine out the loft and dusted it down and later this evening, I’ll test the levels into Reaper in the PC.

Cheers

Toby

[TheMadman_tobyjenner]

Keith

With regard to overall mix levels you reference, having the Master Track RMS metering levels with a -14 offset (Reaper default with 6 threshold) should give you a reasonable if not optimum level for YouTube ?

Cheers

Toby

[Majik]

I’m not really an expert on this, mainly because I’ve not spent a lot of time trying it. But I reckon if you at least preview your content normalized down to -14 LUFS.

You don’t necessarily have to master it to that level, as Youtube will do that normalization for you.

It’s probably also worth listening to it on different systems, including a laptop speakers and a smartphone, which is where a lot of people will hear your video. If the dynamic range is too large, it’s not going to sound good on those devices. I would suggest compressing stuff down to 10-20 dB dynamic range to accommodate those devices.

It should be noted that Youtube (and many other platforms) have started using the “LUFS” (aka “LKFS”) scale, which is similar to dBFS, but is loudness weighted using the K metering. This gives a different result to pure dBFS.

LKFS

In fact K weighted scales are much better at representing what things sound like to us. If you can use a LUFS/K-Meter instead of dBFS, this is recommended. K metering is a more complex subject to understand the principles behind than dBFS, so I’m not going into detail here, but there’s plenty of information on the Internet about it if you want, for instance:

Metering Explained - VU LUFS LU K-Ref.... - Samplecraze

By the way, the aforementioned Bob Katz was also the inventor of the K meter system.

Cheers,

Keith

[Willsie]

Toby, I connect to my laptop via USB cable as provided with the AI. No other output cable were provided although there are L & R outputs that are described in the manual as being used for monitors. Reading the info for the AI it is stated that the mic input has a pre amp. No such statement is made for the INST 2 input. I wouldn’t know if this is a typical setup. I’m interested to hear how you get on with your AI. Your post suggests it is a UM2 as well?

[Willsie]

Great info Keith, that will take me some time to assimilate. I’ve bookmarked this thread but don’t know where these are accessed?

[TheMadman_tobyjenner]

John

NOTE this is not and has never been my GOTO audio interface. It was used on two projects for vocals only, in the garage and into a laptop, requiring a hell of a lot of shouting

I am now using a Behringer UMC1820 and prior to that a number of Xenyx Mixers.

A very frustrating experiment using my old UM2. Connected to PC via USB only.

Managed to record guitar via OBS after setting the UM2 as the default recording device in Windows Sound Control Panel. Upload to YouTube -6 db. Gain 2 in the 3 o’clock position and acoustic pre-amp set around 75% volume.

Then did separate mic test (Shure SM58) into UM2 XLR, declared as mic in Zoom, again with Gain in 3 o’clock position. Upload to YouTube -17.4 db. Not surprised as pretty much had to max the gain when used on previous projects.

Vox In Reaper
Setting up audio preferences as ASIO and then the UM2 no problems with mic. 3 o’clock gain about -16 db in Reaper 4 o’clock around -10 to -8 depending on voice volume.

Gtr In Reaper
OK for starters I only ever used the UM2 for vox projects in the past, so decided to use the T-Bucket with Fishman pre-amp.

• With Audio System declared as ASIO -zero sound from guitar.
• With Audio System declared Direct Sound massive latency - unusable
• With Audio System declared Wave Out moderate latency unusable
• With Audio System declared Dummy Audio no sound
• With Audio System declared WASPI HFR !! Sound with negligible latency but only around -20db

So pretty much useless as a Guitar input but OK for mic only.

Before writing this, did a quick google for Reaper/Guitar/UM2 and loads of post for zero sound in Reaper using the correct driver, in this case ASIO4ALL, as Behringer ditched own brand drivers a good few years back and they were most likely “white goods” anyway.

So after nearly 2 hours, Guitar and UM2 not working as designed but OKish with MIC but not particularly good levels using a dynamic mic - pretty much needed maxing out.

I will let you draw you own conclusions.

Cheers

Toby

[Willsie]

I take it your a big fan of the UM2!

[TheMadman_tobyjenner]

Ronseal it ain’t ! That’s for sure.

[DavidP]

@Majik Thanks for the extensive right up on the dB scale and dynamic range. Maybe it is something that should be cut from this topic and posted as it’s own topic in an appropriate category … what do you think @Richard_close2u

@Willsie

John, I’ve listened to Test4. Volume and audio quality more than good enough for making recordings for AVOYP just as is.

How did you achieve the panning?

The test recording from Zoom is good enough as is for OM events. I also recall that videos made via Zoom meeting recording were at a lower volume level given identical settings to a recording made from OBS.

Kenny’s tutorials are one of the reasons to pick Reaper IMHO. I know YT is full of educational videos, but his videos are so well done, extensive, organised that it makes learning to use Reaper a pleasure.

I assume that when time permits you will make some recordings into Reaper. As I said before, I am interested in what levels you achieve, both vocal and guitar, from UM2 into Reaper, what you see in Reaper as you adjust the gain on UM2.

Based on what you’ve said and now Toby has shared, it sounds to me that that the AI itself may be a part of your experience.

That said, as I said above, I think at this stage you are able to make both AVOYP videos using QT and participate in OM on Zoom, having resolved the distortion issue you had a couple of OMs back and the lack of sound on the Mac recently.

So where to next depends on your aspirations.

For example, you could introduce OBS as a tool for both making videos and streaming to Zoom. The one reason I can see to do that would be to be able to have more control on the audio output. OBS allows you to add in some effects that you could use to raise your levels, add some reverb onto the vocal (if you wanted to do so) etc. You can do this in OBS without adding Reaper into the chain. You can also play with extra cameras.

These days, I only use Reaper when I want to start producing songs that involve more than just me playing and singing. My Simple Blues Lead study project is a typical example of that. For just playing and singing, with a better quality audio than can be achieved by just using a mobile phone to record the video with the phone mic (though these days for AVOYP it is more than good enough), I use only OBS.

Glad you’ve made this progress, so look forward to both more videos in AVOYP and you performing at the next OM.

[TheMadman_tobyjenner]

John

Time to reflect.

I think if your primary objective is AOVYP and Open Mic via Zoom, like David I think the audio quality is now at an acceptable level. The main thing is you have now have a clear audio with no interference. Like David I would suggest using OBS for videos, as you can enhance the audio as it supports VST plugins. And you can stream from OBS to Zoom, which should give you better Open Mic levels (certainly from the figures I was getting last night). Also using OBS I don’t think you would have to drive the UM2 so hard but it seems you need to be fairly high on the gain settings to get a decent output. So you are good to go !!

MHO if you want to be using Reaper in the future, you may need a different AI in order to hear the Guitar. And it would seem it is not just Reaper, same issue was reported for the UM2 with other DAWs. As yet I have not found anything that suggests a solution but I’ll keep looking.
So as to Reaper, as I said :

@TheMadman_tobyjenner

Ronseal it ain’t ! That’s for sure.

Now get some recording done, so we can see all this has borne fruit

Cheers

Toby

[TheMadman_tobyjenner]

Any Behringer U-Phoria UM2 users out in the Community ??

So reaching out the the many thousands of members we now have in our Community, is anyone else using a UM2 for guitar with a DAW ??

Did you have issues getting it to work ?
Did you get it to work ?
If so how did you get it to work ?

Let see if we can collectively help @Willsie in order that he can use Reaper in the future.

Cheers

Toby

[DavidP]

@TheMadman_tobyjenner

Any Behringer U-Phoria UM2 users out in the Community ??

Might be more effective post the questions as a new topic with the quoted text as the title?

I think the volume of posting is up on what it used to be and I suspect a good chance that this may be missed by people scanning posts and only dipping in to a subset, perhaps not being drawn to read this Topic.

[TheMadman_tobyjenner]

Might have found something so hold that thought
Need to walk dog first

[TheMadman_tobyjenner]

OK we now have Reaper and sound via the UM2 !!

Not sure if I should put this down as a schoolboy error but I think the last time I did what I needed to fix it, was over 5 years ago and I’ve slept since then !!

Basically, the UM2 was not selected in the ASIO Configuration. Doh !!

When I have added new AIs in the past, I think I only once had to manually set the unit in the ASIO Config and as I say I have slept since then, What bothered me was not seeing the UM2 (USB Audio Codec) as an Input, when the Reaper Audio Preferences settings were Audio System = ASIO and ASIO Driver = ASIO4ALL. But it showed up when selecting the other "system " options.

So I selected ASIO Configuration and saw that “High Definition Audio Devicse” had been declared automatically from the device list. Selected the entry for the UM2, came out, picked up the acoustic and recorded at levels up to -7 Db. So not the AI it was the configuration

I guess when I used it on the laptop a few years back, it was the only AI that had ever been connected, so ASIO would have selected it by default. And all subsequent AIs used on the PC had been picked up by default automatically,

So question for @Willsie does this help you on the Reaper front ?

Cheers

Toby

[Majik]

@TheMadman_tobyjenner

So I selected ASIO Configuration and saw that “High Definition Audio Devicse” had been declared automatically from the device list. Selected the entry for the UM2, came out, picked up the acoustic and recorded at levels up to -7 Db. So not the AI it was the configuration

Nice one Tony. I was confused about that because, as far as the PC is concerned, an audio interface is an audio interface and if it works in a given way in one application, it should work the same in all other applications (there are exceptions to that, especially on Linux).

@DavidP

Maybe it is something that should be cut from this topic and posted as it’s own topic in an appropriate category

I’m happy to do that if people think it may be useful. It could be posted under a broader topic which could include Gain Staging, but it might be better if I review it and, maybe rewrite some bits and see if there’s any pictures I can include.

I did write this off the cuff and in a hurry, so it might not be the best guide as it is.

In particular, I would like to research and revise the section on Youtube uploading as I think that can be improved.

Cheers,

Keith

[Willsie]

@DavidP I agree @Majik 's post is great reference material and deserves to be easily accessible. I’ve bookmarked it although I don’t yet know how to access these yet.
Test 4: the way this recorded is down to how QT or my laptop works, nothing to do with any choices from me. All I did was select USB Audio CODEC as the sound input in System Preferences on my MacBook. This meant both my voice and Guitar were recorded on to the left side (or should the correct term be channel) through the mic, gain 1, and the guitar alone is recorded on the right side/channel, through the inst, gain 2. I don’t know why this works this way. Might explain why some recordings I have uploaded to YouTube (e.g. Fire & Rain) only output on the left side, which members have commented on. if I didn’t have the guitar going through gain 2 properly. So, no panning from me.
Yes, when time permits, difficulty in itself, I will be trying to use Reaper. I am thinking of finding something to replace the UM2 for recording the guitar after @TheMadman_tobyjenner 's and my experience with it.
I am looking forward to familiarising myself with both Reaper and OBS. Time permitting, because first and foremost I want to improve my playing and singing and after the time I put into that and my family chores time is very limited!

[Willsie]

@TheMadman_tobyjenner the primary aim is to improve musically and I’m recognising that all this recording and production is a big part of that. I very much appreciate the help you guys are giving me.