Pretty sensible advice here Andi ... One little thing though - When you used the Auto Input level on the PRD Dallas plugin, the 2.5 seconds wasn't enough time and it missed a large Peak. Had you used a longer Time it would have caught it and set more of a cut on the Input. It shouldn't be hitting the Red after self-levelling... 😉
Indeed, the RMS level of the pickups is quite low. But peak is not 500mV. It's a couple of volts. And sometimes more than 10-15V if you hit it right. Even with weak singles-type pickups in a stratocaster. You can see this on an oscilloscope, or you can deduce it by comparison with a known signal source by just viewing recorded tracks in the DAW. A multimeter will not actually reveal this.
Finding reliable numbers for actual output is difficult - I was surprised that DiMarzio actually quotes numbers, and not surprised that I couldn't find the notes to explain what and how. The method works well enough because VU meters are so slow that they don't tell us much about peaks anyway, but you've got me intrigued so I've asked @maxquordalspleen7261 to fire up a scope and take some measurements for me. 10 - 15V could get a bit brutal on converters on the way in.
Yes, this may seem very surprising but, indeed, during the attack, spikes of amplitude many times the rms level appear in the signal. But they last a very short time. Most buffers powered by standard 9V clipping peaks of the signal. Most preamps built into basses can't handle the peaks cleanly when a bass player starts slapping or popping strings. The maudio 2626's line input has a fixed gain and is capable of transferring up to +19.6dBu, which is corresponding roughly to voltages over 10V peak. Plugging an instrument directly into it is rather unlikely to overdrive it, because the impedance of this type of input is generally very low(10-20kohm). But it is enough to plug between the instrument and this input a simple unity-gain buffer with an input impedance of 1M and a large headroom(e.g., made using an opamp and supplied symmetrically +/-15V) and the light will sometimes glow red. It doesn't take much: just solidly hit the powerchord with a pick(not a hammer :D ) directly above the bridge pickup and even the fact that it's 3-4mm away from the string won't prevent overdrive. The pickup model makes little difference. I've achieved this on L500 bridge pickup, bkp black hawk and even in split mode but even such SD SH-4 has enough punch(just hit harder :D ). Interestingly, emg81, which seems to have more output than all the others doesn't have this effect(power supply voltage limitation).
@@rzutyjenkin398 That 19 dBu is pretty much the same as RME quotes for their "Lo Gain" setting, it's significantly higher than the "standard" +4 dBu calibration at 13 dBu. I can get an over on my converters with a low E power chord hitting 3 or 4 strings, into the hiZ on my RME, but not in normal playing. Going back to the original idea, if the hardware can't handle the input then it's probably not worth worrying about trimming plugins for accuracy - guess we'll all just have to get the valve amps out again (or use a DI box with a pad - but that opens a whole other can of snakes).
@@AndiPicker Don't get me wrong, I think the idea of driving plug-ins keeping relatively realistic levels is absolutely right. My intention was to highlight how difficult pickups' signal can be to handle in real life. Enormous dynamic range and voltage spikes far beyond common expectations. And that's why it's difficult to talk about the transparency of any device, which on paper may seem completely flat and linear. But that doesn't mean you should give up or panic for this reason. I guess the truism ''use your ears'' is confirmed once again. Another thing is that recording is a completely different beast. Rehearsals or live performances are all about emotion and hard-to-tame expression. Punches with full force are not rare. During recording, the focus is on control and precision. The dynamic range is still large but we tend to stick to what sounds good, which means trying to play consistently(not necessarily like a machine but with accents placed in a musical way). One more thing seems important to me: recording using a sound close to the final one. Making the whole sound from scratch having only DI often gives strange results and can result a lack of dynamic consistency.
@@rzutyjenkin398 I'm with you - recording is a matter of fitting a bunch of parts into a box that's a bit too small for them and trying to make them sound massive. We filter what we hear in real life according to our situational awareness - when the sound is coming from a set of speakers we tend to hear what's actually there, and that means sculpting - sometimes drastically. I always prefer to record with an amp, even if I don't use the amp - it gives the player something to play against. Getting a sim "about right" seems to me to be good enough - give the code a chance to do what it was designed to do - but he notion that we can get it "exactly" right is a step too far for me - as you say, guitars are awkward beasts even when no one is recording, and until we get sims with leaky caps and 24 Ohm cabs with damp speaker cones and grill cloths with 30 years of smokers' tar and spilled beer clogging them, I figure it's good enough to be good enough and leave the obsession to someone else. 👍
Hi Andi, geek corner here. Ref. about 5 min in your vid where you mention multimeters and measuring frequency, If you/watchers happen to be trying to be randomly measuring an AC level with a basic multimeter for whatever reason then note to the wise: most budget multimeters have a rubbish frequency bandwidth so while you'll get the correct voltage value for 50Hz/60Hz mains and so forth, if you try and measure higher frequency for some reason you may (typically) read under / randomly wrong voltage. A posh meter like my professional life go-to (a Brymen BM869s, no affiliation, I just like it) is groovy, not all meters are bad in this respect, but the €20 eBay special with the incomprehensible instruction manual may not be so great. Certainly where any harmonics of significance are >1kHz unless you are sure of the meter spec. beware. Bottom E at 82Hz, OK, top E at 330Hz maybe, top E on 24th fret at 1319Hz rather questionable. Keep up the good work!
Hi Andi, sent you a screenshot from my oscilloscope abusing my Ibanez. Not the hottest pickups in the world but representative and hotter than my 1980's Fender although I think that may be on the cool side so to speak. The 200MHz 'scope had a x10 probe and a 1Meg input resistance so the pickups are basically unloaded. What I did was to set the 'scope to single trigger and slowly reduced the trigger threshold until by really whacking the strings I could get it to trigger. This was really hitting the strings stupid hard, not playing. I tested all the pickup settings and got the biggest signal I could. In my case maximum peak voltage was roughly 2V. Most pedals or indeed active pickups have a PP3 9V battery. That 9V can often be a bit less if you use a rechargeable NiMH battery BTW. Most analogue pedals use discrete components and/or op-amps which are not "rail-to-rail" so you don't get 9V = +/-4.5V of headroom. A NE5532 opamp (the mainstay of so many bits of recording gear) is about 2V drop on each rail so a 9V supply leaves you with a maximum signal of about 5V or +/-2.5V. Therefore much more than 2Vpk and you are in danger of clipping. All that would all be about typical I'm thinking. Perhaps less of an issue clipping into a Fuzz but for non-distorting pedal e.g. phaser or compressor more of an issue. You can get RRIO opamps of course where both their inputs and their outputs are designed to be able to go right up to the power rails, so you would get +/-4.5V but often they would not be your first choice for your pedal for various reasons. This makes me wonder... pickups, how hot is too hot? With regard to DAWs etc., for what its worth I remember having to fault find some kit where the complaint was it sounded like it was clipping. In the end I looked at the customers "reference" CD and found it's recording was clipped, the outboard units just faithfully reproduced the distortion. Lesson, avoid converter clipping, it sounds grim. As you say, once it is in digits you have no issues. Interesting can of worms opening I fear...
Hej Max, thanks for that. The thing about VU meters is that they're 1) brilliantly musical and 2) horrendously awful as precision test tools. This is all "rule of thumb" and I seriously reckon that contrary to the "we're all doing it catastrophically wrong" furror of a couple of weeks back most folks just set the trim to about the right place by ear and get on with it - if a Bluesbreaker combo sounds like a dimed 5150 red channel then the input is probably a bit warm - so trim it, unless you like the sound in which case record it! Obsessing over precise levels in a simulation of someting as imprecise as guitars, pedals and amps starts to seem like more of a hobby than a step to creating music - and getting it wrong is not the end of the world - as you will know better than most of us. 😘
Lots of discussion around this. I’m lazy and no one will ever hear my music. I go by if it sounds good to me it’s good. If it sounds trash then I’ll work on it. Seems to me as long as you don’t clip then no problem yeah?
Nail on head! I usually just trim the input until it sounds right or until the gain on the amp sim seems to go from clean enough to dirty enough - once it's on a record no one is ever going to be able to go back to a real amp and compare it - and how many times do we plug a boost pedal into a tube amp anyway? Once we get past our converters we have massive headroom in modern digital systems, and if the sim is clipping then it's part of the sound :-)
I am even more lazy 8^), I set my input level to 'don't clip' and expect the amp sim to sound good but it doesn't. I think that is what all the discussion is about. Still trying to understand why this is my problem and not the amp sim developer's one. With a correct 'no clipping' input, we should not have to worry about gain staging, adding or subtracting db's etc.
@@dekofschipper8412 The Waves PRS amps have that. Most others have an input control - basically set it to "not red" and it works. Othjer wise this stuff works.
@@AndiPickerProbably Waves PRS has a sensible solution there. Using amplitube, there really is no 'not red' input setting, so I would expect the default of 0dB shoud be correct. But no, seems you only get acceptible levels if you increase it to 12 dB, which somehow does not make a lot of sense to me.
@@dekofschipper8412 Amplitube 5 input meter is calibrated very close to a VU meter where 0 dBVU = -18 dBFS and the red light comes on at 0 on the VU meter, so if I gain a DI signal to peak aropund 0 VU it's just flashing the input red light on with the Amplitube input gain on 0
Pretty sensible advice here Andi ...
One little thing though -
When you used the Auto Input level on the PRD Dallas plugin, the 2.5 seconds wasn't enough time and it missed a large Peak.
Had you used a longer Time it would have caught it and set more of a cut on the Input.
It shouldn't be hitting the Red after self-levelling... 😉
I use my audio interface as a boost. I turn up the gain till the bass distorts
Right on... I'm a CH-47 guy too!
Indeed, the RMS level of the pickups is quite low. But peak is not 500mV. It's a couple of volts. And sometimes more than 10-15V if you hit it right. Even with weak singles-type pickups in a stratocaster. You can see this on an oscilloscope, or you can deduce it by comparison with a known signal source by just viewing recorded tracks in the DAW. A multimeter will not actually reveal this.
Finding reliable numbers for actual output is difficult - I was surprised that DiMarzio actually quotes numbers, and not surprised that I couldn't find the notes to explain what and how. The method works well enough because VU meters are so slow that they don't tell us much about peaks anyway, but you've got me intrigued so I've asked @maxquordalspleen7261 to fire up a scope and take some measurements for me. 10 - 15V could get a bit brutal on converters on the way in.
Yes, this may seem very surprising but, indeed, during the attack, spikes of amplitude many times the rms level appear in the signal. But they last a very short time. Most buffers powered by standard 9V clipping peaks of the signal. Most preamps built into basses can't handle the peaks cleanly when a bass player starts slapping or popping strings. The maudio 2626's line input has a fixed gain and is capable of transferring up to +19.6dBu, which is corresponding roughly to voltages over 10V peak. Plugging an instrument directly into it is rather unlikely to overdrive it, because the impedance of this type of input is generally very low(10-20kohm). But it is enough to plug between the instrument and this input a simple unity-gain buffer with an input impedance of 1M and a large headroom(e.g., made using an opamp and supplied symmetrically +/-15V) and the light will sometimes glow red. It doesn't take much: just solidly hit the powerchord with a pick(not a hammer :D ) directly above the bridge pickup and even the fact that it's 3-4mm away from the string won't prevent overdrive. The pickup model makes little difference. I've achieved this on L500 bridge pickup, bkp black hawk and even in split mode but even such SD SH-4 has enough punch(just hit harder :D ). Interestingly, emg81, which seems to have more output than all the others doesn't have this effect(power supply voltage limitation).
@@rzutyjenkin398 That 19 dBu is pretty much the same as RME quotes for their "Lo Gain" setting, it's significantly higher than the "standard" +4 dBu calibration at 13 dBu. I can get an over on my converters with a low E power chord hitting 3 or 4 strings, into the hiZ on my RME, but not in normal playing. Going back to the original idea, if the hardware can't handle the input then it's probably not worth worrying about trimming plugins for accuracy - guess we'll all just have to get the valve amps out again (or use a DI box with a pad - but that opens a whole other can of snakes).
@@AndiPicker Don't get me wrong, I think the idea of driving plug-ins keeping relatively realistic levels is absolutely right. My intention was to highlight how difficult pickups' signal can be to handle in real life. Enormous dynamic range and voltage spikes far beyond common expectations. And that's why it's difficult to talk about the transparency of any device, which on paper may seem completely flat and linear. But that doesn't mean you should give up or panic for this reason. I guess the truism ''use your ears'' is confirmed once again.
Another thing is that recording is a completely different beast. Rehearsals or live performances are all about emotion and hard-to-tame expression. Punches with full force are not rare. During recording, the focus is on control and precision. The dynamic range is still large but we tend to stick to what sounds good, which means trying to play consistently(not necessarily like a machine but with accents placed in a musical way).
One more thing seems important to me: recording using a sound close to the final one. Making the whole sound from scratch having only DI often gives strange results and can result a lack of dynamic consistency.
@@rzutyjenkin398 I'm with you - recording is a matter of fitting a bunch of parts into a box that's a bit too small for them and trying to make them sound massive. We filter what we hear in real life according to our situational awareness - when the sound is coming from a set of speakers we tend to hear what's actually there, and that means sculpting - sometimes drastically. I always prefer to record with an amp, even if I don't use the amp - it gives the player something to play against. Getting a sim "about right" seems to me to be good enough - give the code a chance to do what it was designed to do - but he notion that we can get it "exactly" right is a step too far for me - as you say, guitars are awkward beasts even when no one is recording, and until we get sims with leaky caps and 24 Ohm cabs with damp speaker cones and grill cloths with 30 years of smokers' tar and spilled beer clogging them, I figure it's good enough to be good enough and leave the obsession to someone else. 👍
Hi Andi, geek corner here. Ref. about 5 min in your vid where you mention multimeters and measuring frequency, If you/watchers happen to be trying to be randomly measuring an AC level with a basic multimeter for whatever reason then note to the wise: most budget multimeters have a rubbish frequency bandwidth so while you'll get the correct voltage value for 50Hz/60Hz mains and so forth, if you try and measure higher frequency for some reason you may (typically) read under / randomly wrong voltage. A posh meter like my professional life go-to (a Brymen BM869s, no affiliation, I just like it) is groovy, not all meters are bad in this respect, but the €20 eBay special with the incomprehensible instruction manual may not be so great. Certainly where any harmonics of significance are >1kHz unless you are sure of the meter spec. beware. Bottom E at 82Hz, OK, top E at 330Hz maybe, top E on 24th fret at 1319Hz rather questionable. Keep up the good work!
Hey Max, do me a set of measurements, and perhaps stick a scope across the output too?
Hi Andi, sent you a screenshot from my oscilloscope abusing my Ibanez. Not the hottest pickups in the world but representative and hotter than my 1980's Fender although I think that may be on the cool side so to speak. The 200MHz 'scope had a x10 probe and a 1Meg input resistance so the pickups are basically unloaded. What I did was to set the 'scope to single trigger and slowly reduced the trigger threshold until by really whacking the strings I could get it to trigger. This was really hitting the strings stupid hard, not playing. I tested all the pickup settings and got the biggest signal I could. In my case maximum peak voltage was roughly 2V. Most pedals or indeed active pickups have a PP3 9V battery. That 9V can often be a bit less if you use a rechargeable NiMH battery BTW. Most analogue pedals use discrete components and/or op-amps which are not "rail-to-rail" so you don't get 9V = +/-4.5V of headroom. A NE5532 opamp (the mainstay of so many bits of recording gear) is about 2V drop on each rail so a 9V supply leaves you with a maximum signal of about 5V or +/-2.5V. Therefore much more than 2Vpk and you are in danger of clipping. All that would all be about typical I'm thinking. Perhaps less of an issue clipping into a Fuzz but for non-distorting pedal e.g. phaser or compressor more of an issue. You can get RRIO opamps of course where both their inputs and their outputs are designed to be able to go right up to the power rails, so you would get +/-4.5V but often they would not be your first choice for your pedal for various reasons. This makes me wonder... pickups, how hot is too hot? With regard to DAWs etc., for what its worth I remember having to fault find some kit where the complaint was it sounded like it was clipping. In the end I looked at the customers "reference" CD and found it's recording was clipped, the outboard units just faithfully reproduced the distortion. Lesson, avoid converter clipping, it sounds grim. As you say, once it is in digits you have no issues. Interesting can of worms opening I fear...
Hej Max, thanks for that. The thing about VU meters is that they're 1) brilliantly musical and 2) horrendously awful as precision test tools. This is all "rule of thumb" and I seriously reckon that contrary to the "we're all doing it catastrophically wrong" furror of a couple of weeks back most folks just set the trim to about the right place by ear and get on with it - if a Bluesbreaker combo sounds like a dimed 5150 red channel then the input is probably a bit warm - so trim it, unless you like the sound in which case record it! Obsessing over precise levels in a simulation of someting as imprecise as guitars, pedals and amps starts to seem like more of a hobby than a step to creating music - and getting it wrong is not the end of the world - as you will know better than most of us. 😘
Lots of discussion around this. I’m lazy and no one will ever hear my music. I go by if it sounds good to me it’s good. If it sounds trash then I’ll work on it.
Seems to me as long as you don’t clip then no problem yeah?
Nail on head! I usually just trim the input until it sounds right or until the gain on the amp sim seems to go from clean enough to dirty enough - once it's on a record no one is ever going to be able to go back to a real amp and compare it - and how many times do we plug a boost pedal into a tube amp anyway? Once we get past our converters we have massive headroom in modern digital systems, and if the sim is clipping then it's part of the sound :-)
I am even more lazy 8^), I set my input level to 'don't clip' and expect the amp sim to sound good but it doesn't. I think that is what all the discussion is about. Still trying to understand why this is my problem and not the amp sim developer's one. With a correct 'no clipping' input, we should not have to worry about gain staging, adding or subtracting db's etc.
@@dekofschipper8412 The Waves PRS amps have that. Most others have an input control - basically set it to "not red" and it works. Othjer wise this stuff works.
@@AndiPickerProbably Waves PRS has a sensible solution there. Using amplitube, there really is no 'not red' input setting, so I would expect the default of 0dB shoud be correct. But no, seems you only get acceptible levels if you increase it to 12 dB, which somehow does not make a lot of sense to me.
@@dekofschipper8412 Amplitube 5 input meter is calibrated very close to a VU meter where 0 dBVU = -18 dBFS and the red light comes on at 0 on the VU meter, so if I gain a DI signal to peak aropund 0 VU it's just flashing the input red light on with the Amplitube input gain on 0