Need help with F8 gain staging

[pasternak]

Hi - I’m using the original F8 to record my jazz trio rehearsals (I’m not direct recording. I’ve got one dynamic mic in front of the bass cabinet. One dynamic mic in front of upright the bass cabinet. Three condenser mics on drums). And I’m having some difficulty gain staging the mics.

First issue: Turning the trim knob counterclockwise to zero reduces input from mic on a select channel but it does not reduce input on a channel to zero. To be clear, each VU meter continues to be active and I am able to record sound from each mic with trim knobs set at zero. Shouldn’t zero trim setting reduce input from each mic to zero ( I have the knobs set to trim rather than pan functionality)?

Second issue: My recordings require a great deal of upward normalization in Logic Pro X even though I’m trimming input on each channel between -25 and -12 db. It follows, I want to dial up input but I’m concerned with clipping. And I want to use the advanced limiter settings. But can’t find any material online describing the particulars of these settings. I want natural sounding compression. So I imagine I should use a soft knee. Can anyone here offer me some rough advanced limiter parameters to manage mics on guitar, bass, and drums, by chance?

I might mention I’m trying to eliminate as much room sound as possible, so mic placement is 6 to 15 inches from source. So input is pretty hot - hot enough to record with trim set to zero.

I’m a musician and, obviously, not a recording technician. And I’d really appreciate getting any live-music F8 recording advice you might have for me. Thanks so much for wading through this enormous post!

 

[TalismanRich]

On the F8, the gain control does not go from 0 to max. It covers a range from +10 to +75dB gain which should easily cover most microphone levels.

INPUT 1–8 Connectors XLR/TRS combo jacks (XLR: 2 hot, TRS: TIP hot)
XLR inputs (MIC) Input gain +10 – +75 dB
Input impedance 3.3 kΩ
Maximum input level +14 dBu (at 0 dBFS, limiter ON)

What type of mics are you using. A dynamic will have a much lower output (generally) than a condenser. Are you seeing indications of overloading? Make sure you are recording to 24bit, so you get the maximum amount of headroom. Then you should only have to worry about actually overloading the preamps themselves. That should be fairly obvious.

If you are seeing overloading, what instruments are involved? That would have a bearing on how I would set up the limiters.

 

[pasternak]

On the F8, the gain control does not go from 0 to max. It covers a range from +10 to +75dB gain which should easily cover most microphone levels.

INPUT 1–8 Connectors XLR/TRS combo jacks (XLR: 2 hot, TRS: TIP hot)
XLR inputs (MIC) Input gain +10 – +75 dB
Input impedance 3.3 kΩ
Maximum input level +14 dBu (at 0 dBFS, limiter ON)

What type of mics are you using. A dynamic will have a much lower output (generally) than a condenser. Are you seeing indications of overloading? Make sure you are recording to 24bit, so you get the maximum amount of headroom. Then you should only have to worry about actually overloading the preamps themselves. That should be fairly obvious.

If you are seeing overloading, what instruments are involved? That would have a bearing on how I would set up the limiters.

Thank you for the helpful response, TR. You cleared up my first and biggest concern - the range of the trim nobs.

Instruments & mics:

Electric Guitar & 10” cab : sm57 at 6” & off-center
Upright bass & via 12” cab : sm57 6” & off-center
Standard drum kit : matched Rode NT5s overhead
& Shure KSM27 3’up an out front. Mics are close to minimize sound of the room and to improve isolation. This is jazz. So the kick is not getting the royal treatment here and is adequately covered by the existing mics for the moment. May add a mic to the kick once I figure out how to manage the other mics.

I’ve been recording at 24bit.

As I recall, f8 meters via rodes / drums are maxing out at roughly -12 with trim nobs at zero. Guitar and bass rise to this level to match volume of the drums. All nobs off. I’ll check this again tomorrow at rehearsal.

As hot as this looks on the meters. The f8 WAV files always need need major normalization via Logic Pro X. I mean, I can barely hear anything at all when I sample a file with QuickTime before I import it to Logic. Does this surprise you at all? It’s not making sense to me. I apparently need to dial in some additional gain. But according to the current activity on f8 meter, I expect I’ll need to figure out how to use the limiter before I do.


We are largely recording for educational purposes. But we do like to circulate our best recordings a la demo. So I’m working on my recording skill. Might also mention that I am playing bass while recording. Which makes monitoring a challenge.

So, why the feeble wav files? Can you offer me a couple tips for setting the limiter? Thanks for your help.

 

[bouldersoundguy]

It's pretty normal to run the levels low during recording. It's good to stay well away from clipping at that stage. It's especially the case with percussive sources like drums where the peak-to-average ratio is high. The raw tracks will sound quiet compared to finished recordings. Typically, the finished volume is gotten during the mastering phase, after mixing.

 

[TalismanRich]

I agree. If you are getting peaks on the drums at -12, then you have a decent amount of headroom. Compared to an NT5, an SM57 will have a much lower output for an equivalent SPL, so you can run the gain higher on the input. With the F8, you have a noise floor that is about -120dB, so if you lose 12-15 dB to headroom, you still have over 100dB to work with. The KSM27 will have an output level similar to the NT5, however, you also have a 15dB pad that you can use if you want to use preamp gain. Unless you are getting overloaded peaks, I would leave the pad off and just lower the gain knob.

Most people say not to worry about seeing the waveforms, and as long as you can mix and adjust to get close(within about 5-10dB) to full scale for your output, they are probably right. I like to see the waveform for reference, especially when finding places to edit or fade. If you can't zoom the view in your DAW, you can normalize the tracks to something like -2 or 3dBFS and you'll find that any noise floor is going to be ambient sound in the room you're recording, not noise from the recorder.

As far as the limiter, on drums, I tend to go soft knee and a fairly quick attack IF needed. Going into the F8, if you are hitting -12, then you could set the limiter threshold to something like -4 or -5 with a soft knee and a fast attack just to take care of any abnormally loud hits. Probably 99.5% of the time, the limiter won't even kick in. Leave the compression and limiting to the mixing stage if you are trying to balance an instrument's overall level or achieving a particular LUFS value.

That said, I tend to like putting a compressor on bass guitar, something like 2 or 3:1 ratio with a relatively low threshold to smooth things out. That's probably because I suck at bass guitar and my levels vary more widely than they should.

 

[keith.rogers]

When you look at each single track, using the PFL button, what is the TRIM set at, and are you sure you have the input set to Mic? The TRIM is set in the PFL view or using the Menu with a track selected. There is no "0" setting for that. PFL also gives you the view of the actual track level.

Make sure you don't have your F8 set to record the second SD card at -10dB, and then trying to use that for import to the DAW. (And, don't be confused by the *Fader* control, which only affects the mix output to main LR or balance in headphones, though not overall headphone level.)

I used the F8 pretty much since they first came out and still have my F8n. You should have good audio that is not needing a lot, if any, gain applied in Logic, if you have your input type, phantom and Trim set for the channel. They are great devices.

Make sure the F8 has the latest firmware, and I'd use the advanced limiter settings, or at least look at them.

 

[pasternak]

On the F8, the gain control does not go from 0 to max. It covers a range from +10 to +75dB gain which should easily cover most microphone levels.

INPUT 1–8 Connectors XLR/TRS combo jacks (XLR: 2 hot, TRS: TIP hot)
XLR inputs (MIC) Input gain +10 – +75 dB
Input impedance 3.3 kΩ
Maximum input level +14 dBu (at 0 dBFS, limiter ON)

What type of mics are you using. A dynamic will have a much lower output (generally) than a condenser. Are you seeing indications of overloading? Make sure you are recording to 24bit, so you get the maximum amount of headroom. Then you should only have to worry about actually overloading the preamps themselves. That should be fairly obvious.

If you are seeing overloading, what instruments are involved? That would have a bearing on how I would set up the limiters.

Thank you for the helpful response, TR. You cleared up my first and biggest concern - the range of the trim nobs.

Instruments & mics:

Electric Guitar & 10” cab : sm57 at 6” & off-center
Upright bass & via 12” cab : sm57 6” & off-center
Standard drum kit : matched Rode NT5s overhead
& Shure KSM27 3’up an out front. Mics are close to minimize sound of the room and to improve isolation. This is jazz. So the kick is not getting the royal treatment here and is adequately covered by the existing mics for the moment. May add a mic to the kick once I figure out how to manage the other mics.

I’ve been recording at 24bit.

As I recall, f8 meters via rodes / drums are maxing out at roughly -12 with trim nobs at zero. Guitar and bass rise to this level to match volume of the drums. All nobs off. I’ll check this again tomorrow at rehearsal.

As hot as this looks on the meters. The f8 WAV files always need need major normalization via Logic Pro X. I mean, I can barely hear anything at all when I sample a file with QuickTime before I import it to Logic. Does this surprise you at all? It’s not making sense to me. I apparently need to dial in some additional gain. But according to the current activity on f8 meter, I expect I’ll need to figure out how to use the limiter before I do.


We are largely recording for educational purposes. But we do like to circulate our best recordings a la demo. So I’m working on my recording skill. Might also mention that I am playing bass while recording. Which makes monitoring a challenge.

So, why the feeble wav files? Can you offer me a couple tips for setting the limiter? Thanks for your help.

I agree. If you are getting peaks on the drums at -12, then you have a decent amount of headroom. Compared to an NT5, an SM57 will have a much lower output for an equivalent SPL, so you can run the gain higher on the input. With the F8, you have a noise floor that is about -120dB, so if you lose 12-15 dB to headroom, you still have over 100dB to work with. The KSM27 will have an output level similar to the NT5, however, you also have a 15dB pad that you can use if you want to use preamp gain. Unless you are getting overloaded peaks, I would leave the pad off and just lower the gain knob.

Most people say not to worry about seeing the waveforms, and as long as you can mix and adjust to get close(within about 5-10dB) to full scale for your output, they are probably right. I like to see the waveform for reference, especially when finding places to edit or fade. If you can't zoom the view in your DAW, you can normalize the tracks to something like -2 or 3dBFS and you'll find that any noise floor is going to be ambient sound in the room you're recording, not noise from the recorder.

As far as the limiter, on drums, I tend to go soft knee and a fairly quick attack IF needed. Going into the F8, if you are hitting -12, then you could set the limiter threshold to something like -4 or -5 with a soft knee and a fast attack just to take care of any abnormally loud hits. Probably 99.5% of the time, the limiter won't even kick in. Leave the compression and limiting to the mixing stage if you are trying to balance an instrument's overall level or achieving a particular LUFS value.

That said, I tend to like putting a compressor on bass guitar, something like 2 or 3:1 ratio with a relatively low threshold to smooth things out. That's probably because I suck at bass guitar and my levels vary more widely than they should.

Very, very helpful. Thanks so much. I expect you meant to write - not to worry about ‘not’ seeing the wave forms.… And, indeed, I do not see a wave form until normalization in DAW. I’ve been normalizing at -1. Rehearsal today. I’ll give -2 a shot tomorrow. I’ve been getting decent recordings according to my untrained ear. But everything looks so graphically wrong to me ( Strong meter but feeble wave forms) I lost confidence in my ears. Gonna give your rough limiter parameters a shot here in about 20 minutes. Thanks again for straightening me out.

 

[keith.rogers]

Post one of your raw WAV files in a shared Dropbox, Google Drive, whatever location so we can see (and *hear*) what you are talking about.

You should not be "normalizing" your raw tracks to -1 or even -2dB. And, how are you doing this? If you truly have a file that needs to be move up for mixing, in Logic you'd do that applying Gain in the Region (Info) window. You can also increase the visible size of the wave by clicking the Waveform Zoom icon in the upper right side. (Pics attached.)

 

[TalismanRich]

I don't have a problem with normalizing a track when necessary. I don't use Logic, my DAW is Reaper.

In Reaper, it can be easily done by right clicking on a track, choose Item properties and click on Normalize. If you want to boost several tracks with the same amount, you can choose to Normalize to common gain, or just have each track Normalized to boost to each track's full level. It really doesn't change the relative S/N but it does let me find points more easily when I'm editing or mixing.

You can use SHIFT Up or Down arrow to increase or decrease the waveform height, but it does all tracks equally. If you have one small track and others are ok, you can't just zoom the small track.

It's just quicker for me to adjust the level on tracks as needed.

 

[rob aylestone]

The thing you need to do is stablish how loud you can push the levels till distortion starts - digital devices often seem to have wierd meter ballistics which might mean dead accurate or a little 'averaging'. Difficult on the drums, but a clean source would let you look at the waveform in the DAW and see how pure they are. A piano, acoustic guitar or a pure tone from any acoustic instrumnt would do. That all said - your Zoom is pretty good at the top end and I suspect the meters actually over-read a bit. My H4 can go to the danger area by accident and rarely does it actually distort. I too find that the levels recorded are actually lower than expected. The important thing is to work yourself out a 'safe' rule and stick to it. DAWs make it look worse by their level scales - they look low, but the top section is always stretched compared to the bottom - log display making normal levels appear low, when they're not.

 

[pasternak]

Post one of your raw WAV files in a shared Dropbox, Google Drive, whatever location so we can see (and *hear*) what you are talking about.

You should not be "normalizing" your raw tracks to -1 or even -2dB. And, how are you doing this? If you truly have a file that needs to be move up for mixing, in Logic you'd do that applying Gain in the Region (Info) window. You can also increase the visible size of the wave by clicking the Waveform Zoom icon in the upper right side. (Pics attached.)

View attachment 123902 View attachment 123899

Hi Keith, I’m normalizing via Logic Pro X. I often apply gain manually in logic as you suggest. However, normalization strikes me as a quick way to maximize gain without fear of over doing it. I am just getting started. Perhaps I am mistaken. I’m not an audio tech. Just hoping to do an adequate job with the gear for rehearsal/demo purposes. So quite happy to rely on automated gain staging (normalization) if it gives me acceptable results. And I think it does. The contrast between the high meter activity on the f8 and the low wave forms in the polywav made me wonder what I was doing wrong. But the fact is, my recordings are noise-free after normalization. And others in forum are familiar with this phenomenon and tell me not to worry about it too much. So I expect I’m doing a better job than I thought I was. If you think I’m off track, then please set me straight. I’m wide open to suggestions. Otherwise, I think I’m set. Thanks, Kieth.

 

[bouldersoundguy]

One thing to consider is that levels accumulate. On average, when you double the track count you increase levels by 3 dB. So if you normalize everything to 0 dBFS or so, there's a good chance you'll run out of headroom. (Using 32 bit floating point processing mitigates this, but it's still good practice to manage levels thoughtfully.) Of course you can just pull faders down, but it's nice to have tracking levels that let you keep most of your faders in their most precise zone, near 0.

As for drum sounds, if I really know what the peaks will be, I might let them slide up toward -6 or even -3 dBFS. Also, if I plan on hitting them with a limiter later anyway, I might use a limiter just as a safety. I won't typically do that with other sources that aren't so peaky (though I do compress things when I'm familiar with what/who I'm recording).