YouTube will reduce levels to their standard -14 LUFS, but they won't turn it up. They don't "level" or "normalize", in the traditional sense.
Ah yes, the eternal debate of what should be 0VU/+4dBu. krivium has pointed out one of the biggest stupid problems in audio in the last 45 years.
Over the years it's been everywhere from -22dBFS to -10dBFS depending on the organization, end user target, industry division, and on and on. On the original Sony PCM converters that were used for every CD in the beginning, there were built-in means of calibrating meters and a reference level. However, none of the calibration adjustments changed input gain, they just altered the total amount of headroom based on the user's choice of reference levels. The expanded reference meter on on the Sony PCM units was adjustable in 2dB steps from -10 to -20dBFS (though dBFS didn't exist as a term then). So, the 1kHz reference tone could be...well, anything. While analog tape machines already had VU meters, and didn't have a hard clipping point like 0dBFS, PCM was the very first recording system that clipped hard, and provided the user with (mostly) true peak meters...that in a world of VU meters on consoles. And that created a problem.
The way to look at it is in terms of headroom...the amount of high level "room" you had before you hit 0dBFS and ran out of digits, resulting in a very hard clip. Since a VU meter will read 8 to 10dB lower than a peak meter with music, and VU meters were and are still in use, there had to be some choices made. Where do we put the 0VU line-up tone in the digital world? And there were a lot of different levels chosen. Popular ones were -20dBFS, -18dBFS and -16/15dBFS, but there were also those who picked -12dBFS and -10dBFS. That's a 10dB spread, and that means there is no reliable universal standard. As I mentioned, within organizations standards were set. And within distribution chains, networks, etc.
Then the loudness war happend, and it got worse. Everybody wants to be loud, as if that makes people listen (and assumes they don't have a volume control). That pushed reference levels up, average levels up, peak to average ratios down, and we have today's mess.
I haven't done an exhuastive survey, but for the consoles I'm familiar with, analog and digital, the meter (if VU) is calibrated for 0=+4dBu, and there is 20dB of headroom above that point. Pretty standard. But the question is: what kind of meter are you looking at? If you have an audio interface/soundcard, it's probably a peak meter, and you're looking at total headroom up to 0dBFS. If it's a mixer, it might be a VU meter.
So the answer is: do not obsess about precision calibration, you can't ever achieve it because nobody's following the standards. For line-up, a 1KHz tone at -20dBFS and +4dBu is good enough. Then expect recordings to be hotter than that, and just turn them down if necessary.
For tone generation, every DAW will do this. Even Audacity can produce accurate tones. I would suggest that producing tones from software is more trustworthy than download tones from anywhere that probably went through who knows what compression before being posted.
Having learned more about gain structure, I now see how the Nelson Pass F4 amp could be a good solution. I'm currently using the F5m I had built, but the F4 does not have voltage gain. Since my goal is to preserve the analog signal as much as possible, I'm starting to think I should build the F4 instead. I had purchased everything to build a second F5m so already have the chassis/heatsinks/donut etc. @birdbox points out my mixer's output will not surpass the F4's max. so no pad needed, and no additional harmonics(?). ( I have not looked at the schematic/guide for it yet )
Yes F4 could be a good solution if you want to get rid of voltage amplification on signal path ( all amplification stages produces their set of own colorations but buffer are usually more 'transparent' imho).
The issue is about voltage driving them.
You are looking for an acoustic level at listening point. As such the efficiency and distance to loudspeakers will define the voltage needed to drive the F4.
From an acoustic pov, we know each time we double distance to source we loose 6db. Eg if located 2m away from loudspeaker able to ouput at max power 105dbspl/1m you'll have 99dbspl outcome at listening point. 4m away 93dbspl, etc,etc,...
Pro dacs can output +24dbu at most. Which means 12,27 V rms. This mean 17,34V peak ( Vx1,414). But i doubt your Scarllet can ouput such level. Check datasheet about max ouput level and come back to tell. Anyway if you use your desk to feed the amp you'll be limited to +21,5dbu out and SINGLE ENDED as it's balanced impedance but not differential drive ( which means you de facto loose 6db wrt a true differential output stage and the 'differential config shown in F4 manual ( you would need 2 x whole F4 to achieve this). There is way to make your output differential if needed but there is pro and cons about it... we will discuss that later. 😉
Check about your distance to loudspeakers too.
From there you can calculate the efficiency your loudspeakers need to achieve your target Spl at listening point. I would suggest 77dbspl with provision for 20db peaks over. Iow 97dbspl at listening position. This is typical level at which mixing/mastering occurs. And it's already kindof loud for prolongated listening sessions imho.
From there we can define the loss through acoustic travel ( distance to loudspeakers) and then power and voltage needed to achieve it.
It's a nice excercice to plan this anyway even if it doesn't work out as expected imho.
The issue is about voltage driving them.
You are looking for an acoustic level at listening point. As such the efficiency and distance to loudspeakers will define the voltage needed to drive the F4.
From an acoustic pov, we know each time we double distance to source we loose 6db. Eg if located 2m away from loudspeaker able to ouput at max power 105dbspl/1m you'll have 99dbspl outcome at listening point. 4m away 93dbspl, etc,etc,...
Pro dacs can output +24dbu at most. Which means 12,27 V rms. This mean 17,34V peak ( Vx1,414). But i doubt your Scarllet can ouput such level. Check datasheet about max ouput level and come back to tell. Anyway if you use your desk to feed the amp you'll be limited to +21,5dbu out and SINGLE ENDED as it's balanced impedance but not differential drive ( which means you de facto loose 6db wrt a true differential output stage and the 'differential config shown in F4 manual ( you would need 2 x whole F4 to achieve this). There is way to make your output differential if needed but there is pro and cons about it... we will discuss that later. 😉
Check about your distance to loudspeakers too.
From there you can calculate the efficiency your loudspeakers need to achieve your target Spl at listening point. I would suggest 77dbspl with provision for 20db peaks over. Iow 97dbspl at listening position. This is typical level at which mixing/mastering occurs. And it's already kindof loud for prolongated listening sessions imho.
From there we can define the loss through acoustic travel ( distance to loudspeakers) and then power and voltage needed to achieve it.
It's a nice excercice to plan this anyway even if it doesn't work out as expected imho.
Last edited:
^^ note I don't plan on using my focusrite scarllet with the F4, only the mixer which has 21.5dBu as you point out. I'll have to investigate what you mean about losing 6dB 'wrt'... not sure if that is a typo
I understand it's balanced impedance which means 1 wire just has a resistor and does not carry the differential signal like a 'real' balanced cable, so I guess that is what you mean by losing 1/2 ( 6dB )?
I don't get why Soundcraft would crap out on doing full differential balanced output.
I understand it's balanced impedance which means 1 wire just has a resistor and does not carry the differential signal like a 'real' balanced cable, so I guess that is what you mean by losing 1/2 ( 6dB )?
I don't get why Soundcraft would crap out on doing full differential balanced output.
It's not a typo.
It's the difference between single ended drive and differential drive.
Read F4 manual it's explained in there ( and Jaddie and i already talked about it in a previous post exchange in this thread). Take it for granted and measure distance to your loudspeakers. 😉
It's the difference between single ended drive and differential drive.
Read F4 manual it's explained in there ( and Jaddie and i already talked about it in a previous post exchange in this thread). Take it for granted and measure distance to your loudspeakers. 😉
Soundcraft have to be cheap, scale economy...
Yes single ended loose 6db wrt differential, because differential works on difference between pin2/pin3 ( hot/cold) and as it have 2 time the signal compared to a single ended drive: +6db ( 20log(2)).
Balanced (impedance to ground if you want to gives the complete name) have nothing to do with being differential drive, it's for different purpose.
Yes single ended loose 6db wrt differential, because differential works on difference between pin2/pin3 ( hot/cold) and as it have 2 time the signal compared to a single ended drive: +6db ( 20log(2)).
Balanced (impedance to ground if you want to gives the complete name) have nothing to do with being differential drive, it's for different purpose.
I don’t understand. You want to get/build an entire new power amp just so you don’t need a pad??
My numbers come out this way:
Soundcraft max output: +21dBu = 8.7Vrms
The F4 has no gain, so in = out.
8.7Vrms into 8 ohms is 9.4 watts.
Other than you might get the system to play loud enough if you aren’t too far from the speakers, how is this an improvement?
4 resistors and you accomplish the same thing with the amp you have.
A pad doesn’t color your audio. It preserves the analog signal better than any single element of your system apart from the wire.
You might want to take a step back and carefully analyze what your goal is and how to get there. You’ve already given up “unity gain” which doesn’t exist and doesn’t matter anyway. You over value things that don’t matter in favor of some vague thing that doesn’t exit, when your speakers and room are mangling your audio 100X more than anything in your system.
What about enjoying the music?
My numbers come out this way:
Soundcraft max output: +21dBu = 8.7Vrms
The F4 has no gain, so in = out.
8.7Vrms into 8 ohms is 9.4 watts.
Other than you might get the system to play loud enough if you aren’t too far from the speakers, how is this an improvement?
4 resistors and you accomplish the same thing with the amp you have.
A pad doesn’t color your audio. It preserves the analog signal better than any single element of your system apart from the wire.
You might want to take a step back and carefully analyze what your goal is and how to get there. You’ve already given up “unity gain” which doesn’t exist and doesn’t matter anyway. You over value things that don’t matter in favor of some vague thing that doesn’t exit, when your speakers and room are mangling your audio 100X more than anything in your system.
What about enjoying the music?
Thanks @jaddie, I think you are overlooking that I'm using this group to learn, and not pretend I know what I'm doing. I'll need to take various paths to learn, some of which may not make sense to you as someone with years of experience. For example, I had not started looking at the F4 yet, other than showing interest in it since IT WAS SUGGESTED. I have no ideas what the specs are yet since I have not had time to look at the guide but not sure why it was suggested if it's so crazy to consider it. How do I know what is is without first hearing about it, and then investigating it? Also, when you state "i've 'given up on unity gain, which does not exist or matter". I have not given up on anything, specially understanding what unity gain is which I know do thanks to this post. As far as I understand, it does 'exist and matters' so not sure what you mean by that. Gain staging and understanding it seems to be very critical and I don't plan on giving up on it... It may be obvious to you but I'm still learning about it. I"m not trying to convince anyone or myself something exists when it does not, and not sure where you get 100x more than anything.
P.S. I would DEFINITELY build another amp in the blink of an eye as that is the part I enjoy!
Thanks again for all the great info. I'll post back an updated when I've tested the solution I ended up with.
P.S. I would DEFINITELY build another amp in the blink of an eye as that is the part I enjoy!
Thanks again for all the great info. I'll post back an updated when I've tested the solution I ended up with.
Before jumping on the F4 could you tell the distance to your loudspeakers please. It could be nice for you to know the limitations/issues you could face and how solve them.
But without that distance it's impossible. 😉
But without that distance it's impossible. 😉
What about reducing the gain* of that mixers output stage with crazy 9V rms outputs? Solving the “problem” at the root. Solving by taking away a fault possibility not by adding yet another possible fault/imperfection.
*Just checked that service manual and it really is a question of changing 2 resistors (R506/R507 both 62 kOhm). It is a non issue of proportions but feel free to add compressors, limiters, computers, VU meters, pads, Klever Klippers and other possible new OCD factors 🙂
2 resistors......
*Just checked that service manual and it really is a question of changing 2 resistors (R506/R507 both 62 kOhm). It is a non issue of proportions but feel free to add compressors, limiters, computers, VU meters, pads, Klever Klippers and other possible new OCD factors 🙂
2 resistors......
Last edited:
Crazy output at +21,5dbu? Lol.
A Fairchild 670 was able to output +28dbu...
It's not consummer gear Jean-Paul but pro studio gear. Line are usually referenced to +4dbu with 20db headroom above. Pretty standard. At 21,5dbu it's not a beast regarding output voltage to say the least. Classic Neve designed modules ( Neve, Focusrite, Amek, Cadac,... ) are designed with +26db into 600ohms as target. Thats some serious line ouput stage.
A Fairchild 670 was able to output +28dbu...
It's not consummer gear Jean-Paul but pro studio gear. Line are usually referenced to +4dbu with 20db headroom above. Pretty standard. At 21,5dbu it's not a beast regarding output voltage to say the least. Classic Neve designed modules ( Neve, Focusrite, Amek, Cadac,... ) are designed with +26db into 600ohms as target. Thats some serious line ouput stage.
Last edited:
He uses a consumer grade power amplifier.
Connect it to +28dBu if you like. I would use normal signal levels and not worry about exceeding 3V by simply making it impossible to create 3V.
Connect it to +28dBu if you like. I would use normal signal levels and not worry about exceeding 3V by simply making it impossible to create 3V.
As we all do more or less. 😉
Your idea of changing the resistor is not bad but what will happen if he want to put something ( pro gear eg:an eq ) after the desk?
All being equal the pad is the universal answer we use in studio ( and you've got 2 studio tech from two different continent giving the same answer).
Your idea of changing the resistor is not bad but what will happen if he want to put something ( pro gear eg:an eq ) after the desk?
All being equal the pad is the universal answer we use in studio ( and you've got 2 studio tech from two different continent giving the same answer).
What if he marries 2 women tomorrow?
The simple issue is that studio stuff with crazy output levels is coupled to a consumer grade Pass F5m. Then he worries about damaging the Pass F5m by exceeding its 3V max. input level. 9V rms to a 3V max amplifier, yeah right that is error by design.
But feel free to marry yet another Quasimodo to that mixer to make it an odd but working couple.
2 resistors.....
The simple issue is that studio stuff with crazy output levels is coupled to a consumer grade Pass F5m. Then he worries about damaging the Pass F5m by exceeding its 3V max. input level. 9V rms to a 3V max amplifier, yeah right that is error by design.
But feel free to marry yet another Quasimodo to that mixer to make it an odd but working couple.
2 resistors.....
Last edited:
I think he worries to damage the loudspeakers rather than amplifier Jean Paul. 😉
Crazy ouput level from studio gear. It's funny to read that, it depend from the side you see it as i often think consummer gear is crazy without any clearly defined standard either... 🙂
That said whatever amplifier we use we often have the same kind of issues as even when from pro world ( P.A.) amplifiers don't match the standard line level in studio.
Whatever, the real issue comes from the console design by itself as it doesn't have a control room/monitoring section which usually deals with such issues: the ouput of a desk is not supposed to drive amplifier ( in studio) but a 2tracks. It's been already discussed.
Crazy ouput level from studio gear. It's funny to read that, it depend from the side you see it as i often think consummer gear is crazy without any clearly defined standard either... 🙂
That said whatever amplifier we use we often have the same kind of issues as even when from pro world ( P.A.) amplifiers don't match the standard line level in studio.
Whatever, the real issue comes from the console design by itself as it doesn't have a control room/monitoring section which usually deals with such issues: the ouput of a desk is not supposed to drive amplifier ( in studio) but a 2tracks. It's been already discussed.
Modifying the Soundcraft — difficult. Adding an external pad - easy.
This entire issue is caused by using the Soundcraft. It doesn’t have a monitor section, so you have to use the main output. It’s the problem, not the power amp.
In another thread he worried about the F5m. In this one too -> first post second sentence. It is even the subject.
The whole non issue is wanting to put a square peg in a round hole. Being used to building one offs I suggest to make the square peg round.
Last edited:
Maybe, i don't know i have only read this thread. That said if you plan to change 50 years of habits in a professional field by changing the standard used, good luck trying to convince people they have it wrong Jean-Paul. There might be reason beyong those choices of reference. 😉
Don't make my words BS. It is exactly how I described. I don't care what all those professionals may or may not do and how right they all are in their beautiful pro world where we don't understand anything of. It is totally irrelevant when using "normal" HiFi stuff.
It is a 3V max amplifier coupled to a 9V device so a device that may damage that first device. A wolf coupled to a lamb, (dramatic music playing) ready to bite. Take away the 9V possibility and it is SOLVED. What can not occur can not damage anything. Change the wolf to a sheep or the lamb to a wolf. Or my suggestion: make the wolf identify as a sheep. DIY too, soldering 2 nice resistors! 50 Eurocent solution, 100% effective.
It is a 3V max amplifier coupled to a 9V device so a device that may damage that first device. A wolf coupled to a lamb, (dramatic music playing) ready to bite. Take away the 9V possibility and it is SOLVED. What can not occur can not damage anything. Change the wolf to a sheep or the lamb to a wolf. Or my suggestion: make the wolf identify as a sheep. DIY too, soldering 2 nice resistors! 50 Eurocent solution, 100% effective.
Last edited: