Tested: Where Does The Tone Come From In A Microphone Preamp? — Transcript

Full transcript

1. 0:00This is a story about how I, my friends, my  heroes, and everyone else was completely,
2. 0:05embarrassingly wrong about one  very important thing about why
3. 0:09microphone preamps sound like they do.  Because nobody would do these tests.
4. 0:13(Turn off captions. Listen.)
5. 0:16My name is Jim. I'm this guy. You might know me  as this guy. And look at these huge consoles.
6. 0:21They're so cool and rare, and they  were used to make my favorite music.
7. 0:27My favorite engineers have told me how the  preamps and Neve consoles are warm and fat,
8. 0:31and the preamps and API consoles are fast and  snappy, and the preamps and SSL consoles are
9. 0:36tight and clear. And it always feels like an  honor to be in the room with one of these.
10. 0:40Preamps are something I didn't know anything about  when I moved to Nashville out of high school. They
11. 0:44were just kind of there and they all seemed the  same. You plug a mic in, turn it up, and whatever,
12. 0:48it does its job. So, I just used whatever was  built in to the cheap interface I had. But the
13. 0:53more I got into recording, the more I was taught  about their differences. And I started to realize
14. 0:57that if there are all of these tonal differences,  then my preamps that were just built in with my
15. 1:02interface would be holding me back. I wouldn't  have that sound. So to get those sounds, all
16. 1:08I had to do is buy a few of each type. And oh my  god, wait, really? 4,000? That's not even vintage.
17. 1:22Uh, okay. Plan B. My favorite album was  tracked through the Neve console at Ocean Way.
18. 1:26And I cared so much about understanding that  sound. I did my research and I learned that
19. 1:31the console at Ocean Way is a Neve 8078 from the  1970s. And I learned that Neve 8078 consoles have
20. 1:38Neve 31105 preamp and EQ modules. And that's  where the tone I wanted would be coming from.
21. 1:44And as fate would have it, one day I saw  a racked Neve 31105 preamp and EQ module
22. 1:50from a 1970s Neve 8078 console come up  for sale, ready to plug in and play.
23. 1:55And I made the most expensive purchase of  my life so far, trading 6 months worth of
24. 2:00rent for this preamp because I needed to  know what this tone was. And I loved it.
25. 2:51It was fat. It was warm. It was vintage.  It was oozing with vibe and cool factor.
26. 2:55And I was really happy.
27. 2:59Then after a few months, I got into a kick where  I'd test things. And I thought, hey, why not test
28. 3:03the Neve versus something silly like the built-in  preamps on that cheap interface I used to use?
29. 3:16And they sounded the same. So I tried  playing into every preamp I had.
30. 3:43This video will be very different.  With guitars and amps and microphones,
31. 3:47I knew they sounded different and I wanted to find  out why. With preamps, it kind of seems like they
32. 3:52all sound the same. And it is extremely important  to me that I find out when they don't because
33. 3:576 months of rent hangs in the balance. This is  the quest for the missing preamp tone. Okay. One
34. 4:05reason the internet says I might not have heard  the difference between real deal vintage Neve and
35. 4:08garbage is because electric guitars have limited  bandwidth. They do not cover the full frequency
36. 4:14spectrum that humans can hear. That's true. I do  have a drum kit. The kick drum is the lowest thing
37. 4:20I worry about and the cymbals are the highest  things I worry about. What if I mic those and
38. 4:23split it out to each preamp and flip back and  forth between them? What does that sound like?
39. 4:41Okay, that's not it. So, the internet says it  might be hard for me to hear the difference
40. 4:44on an individual track, but it's easy to hear the  difference in preamps when every track is recorded
41. 4:49through either something great or something  crappy because the character of the preamp builds
42. 4:53up across the mix. That sounds like it might be  right. It's just that nobody with the time to do
43. 4:58that has ever had the ability to do it well, and  nobody with the ability has ever had the time.
44. 5:04Until now, I'll take the bullet and  be that guy. I recorded an entire mix,
45. 5:08one mic at a time, with that mic split  out to both the Neve and the garbage.
46. 5:31Huh. Well, I didn't hear any  built-up preamp character,
47. 5:35but the internet says, "Oh, no, no,  this mix is too busy and commercial.
48. 5:38The preamp's character shines through on more  delicate music." Okay, I can give that a shot.
49. 6:02No, no, no. You hear the preamp character better  when you push the preamps with heavy music.
50. 6:07But I don't even listen to that kind of
51. 6:26I'm still not finding the missing  preamp sound. So, the internet says
52. 6:30maybe my examples just aren't the right  examples and I need to be more scientific.
53. 6:34Okay, I can do that. I can use Room EQ Wizard to  send a sweep of every frequency humans can hear
54. 6:39through each preamp and then show on the graph  exactly how different they are. For context,
55. 6:43everyone I know says Neves are warm. They also  say ribbon mics are warm. So, let's look at a
56. 6:48graph of a Royer 121 ribbon mic, which is warm,  compared to a Shure SM57, which is not warm.
57. 6:57Okay, cool. The Royer has about 9 dB less energy  up here and 9 dB more energy down here. So,
58. 7:03maybe the warm Neve has something that looks kind  of like this when compared to the garbage preamp.
59. 7:11And they're the same. They're within a  quarter of a dB from 20 Hz to 20 kHz.
60. 7:17But hey, maybe a dB is a lot. And that might mean  half a dB is something. And a quarter of a dB is
61. 7:23almost more than half of half a dB. So what if  that is making a difference? So I dialed in an
62. 7:28EQ to make a quarter of a dB dip here and here.  And then flipped it on and off and listen to it.
63. 7:52And I couldn't hear a difference.
64. 7:54But maybe I'm biased. Maybe I subconsciously don't  want there to be a difference, even though I'm one
65. 7:58of the few people who actually did get my hands on  the real thing. So, I took my subconscious out of
66. 8:03the equation and did a blind ABX listening  test for what's a good number? 100 rounds.
67. 8:14Ah,
68. 8:15I'm going to hit show results now.
69. 8:17And according to math, I would have  been better off flipping a coin.
70. 8:21But that's just my ears and I'm just a performer.  The internet still says that Neves are not only
71. 8:25warm, but they are punchy. What does punchy mean?  It means the preamps respond to the input and have
72. 8:31subtle compression. Subtle compression? Jeez,  maybe I'm just not as talented or accomplished
73. 8:36a listener and I simply can't hear the compression  that Neve preamps have that garbage preamps don't.
74. 8:41And luckily, compression is something I don't  have to listen for to see if it's there. Check
75. 8:44this out. This is a step test. It puts out higher  and higher levels into a piece of gear and plots
76. 8:50what it gets back. This perfectly diagonal line  is when nothing is in the path, and it gets back
77. 8:54exactly what it sent out. This hard horizontal  line is distortion, where the piece of gear
78. 8:59hits a ceiling and can't put out any higher  level, no matter the input. And this slightly
79. 9:03more diagonal line is compression. The higher  the ratio, the more horizontal it is. A subtle
80. 9:08compression might be a 2:1 ratio, like this. So  let's see how the Neve responds to a step test.
81. 9:14Oh, it's perfectly 1-to-1 linear until it hits a  ceiling and wigs out, which would be distortion.
82. 9:20So in other words, dynamically, it faithfully  reproduces any signal with zero compression
83. 9:25unless the signal is big enough to cause  hard distortion. So there's no compression,
84. 9:30not even subtle compression, just pristine signal  reproduction until the signal is louder than it
85. 9:35can handle and then it distorts. But now the  internet says "See, distortion! Neves distort! The
86. 9:41warmth and punchiness comes from the distortion!  The vintage British discrete analog distortion!"
87. 9:45Maybe it does. Preamps can distort. Anything can  distort. Like here's a delay pedal in bypass.
88. 9:52Its job isn't to distort, and nobody ever  hears it distort, because nobody ever puts
89. 9:55it in a situation where it would distort. But if  you're insane and you run a power amp into it,
90. 10:09So really when we hear distortion it could be the  Neve, or it could be the mic going into the Neve,
91. 10:14or it could be the next piece of gear the knee  goes into. And I want to figure that out. Let's
92. 10:18start with the mic. There are a lot of microphones  out there and I did a big 30 minute long video
93. 10:22all about them that's a lot like this video and  you should go watch it if you want to know why,
94. 10:26but because of all the work I did there. I'm  not worried about the mic itself distorting.
95. 10:30But if I'm thinking about preamps, it would  be good to know the biggest signals mics are
96. 10:34actually giving to the preamp when they're  recording music, which would mean putting
97. 10:38a mic in front of a kick drum, a snare drum, a  rack tom, a floor tom, a high hat, overheads,
98. 10:42an acoustic guitar, an electric guitar, amp  and cab, and a vocal. So I did all of that.
99. 11:09And this is my spreadsheet. So when  I'm reaming on all these instruments,
100. 11:12the highest thing that's going to hit the  preamp from any of them is 0.97 volts peak.
101. 11:16Let's remember that for later. Now let's think  about what the Neve goes into. The Neve is never
102. 11:21the last thing in the chain. It has to go into  something that records the audio. I'm just a
103. 11:25performer. I don't know anything about history,  but originally Neves were designed to send audio
104. 11:29to a tape machine. And tape machines had to be  set up so zero on their meters all meant the same
105. 11:33thing. And the better the tape machine, the more  headroom it had above that without distorting.
106. 11:37So a good tape machine might have had 14 dB of  headroom. When digital recorders came around,
107. 11:42they were replacing tape machines. And their specs  had to be at least as good as tape machines. So
108. 11:46they were set up at the factory so their maximum  level before distorting was +4 dBu with 18 dB of
109. 11:51headroom. And if you do the math, that means about  13 volts peak. So the biggest thing that can come
110. 11:56out of the preamp and hit the converters without  the converters distorting is 13 volts peak. Cool.
111. 12:01So in order for us to hear the Neve distorting,  this much coming in from the mic has to be enough.
112. 12:06But also the Neve has to be sending less than this  out. Let's figure out if that's what's happening.
113. 12:12I'm just a performer. I don't have a signal  generator and an oscilloscope, so I borrowed them.
114. 12:17This thing generates a signal and this thing  shows that signal. So if I make the signal go
115. 12:21all the way through the Neve, then I can  show it comes out of the Neve, like this.
116. 12:25The Neve on its lowest gain setting will distort  when it's receiving 4.75 volts from a microphone.
117. 12:30And when the Neve finally reaches distortion, it's  putting out 25.82 volts peak. Let's take a minute
118. 12:37and appreciate how mindblowing this is. First,  none of the mics are ever putting out a signal big
119. 12:42enough to make the Neve distort. Not even close.  You'd have to crank the gain knob up to get it
120. 12:47there. But even if you did, then the output level  of the Neve would be smashing the converters.
121. 12:52And long before you got any of the vintage  British warm discreet warm knee distortion,
122. 12:57you'd already be getting tons of cold cacophonous  bees-in-a-tinfoil-box digital distortion.
123. 13:03And yes, I double and triple checked this  using multiple methods, including making
124. 13:0780 different audio files to run through the Neve  and manually recording voltages for each of them.
125. 13:13In order to hear Neve distortion, you'd have to  crank the input gain past normal and then turn
126. 13:17down the fader on the console to at least -5.3 dB,  which isn't what happens. Every engineer I know
127. 13:23tracks with the faders at unity, or zero, and they  set the preamp gain so the converters don't clip.
128. 13:44And mathematically, objectively, every single  time in the past, present, or future that an
129. 13:49engineer has ever tracked through a Neve with  no converters clipping and faders at unity,
130. 13:54it is not possible for there to ever  be or have ever been Neve distortion.
131. 13:59That means distortion cannot be  an ingredient of the Neve sound.
132. 14:05Oh, okay. So, I have to crank the gain and turn  down the output so I can get the Neve sound!
133. 14:10What? No. How is it possible that you got that  from what I was saying? Because that's how you
134. 14:14get the Neve distortion! I want the special  distortion! So, I just found out that in all
135. 14:19normal cases, Neve distortion isn't actually  happening. But while I've been working on the
136. 14:23video, whenever I've talked with friends about  the idea that Neve distortion can't happen unless
137. 14:27you crank the gain and lower the fader, the  person I'm talking to, every time, has said,
138. 14:33"Oh, so I need to crank the gain and lower the  fader!" This is extremely frustrating because,
139. 14:38no, in the entire history of recorded music,  this behavior is an anomaly that amounts to
140. 14:43functionally 0% of cases. And by saying, "Oh, so  I need to crank the gain and lower the fader!"
141. 14:48you're prioritizing the special distortion of a  piece of gear over the reality that you've likely
142. 14:53never heard the special distortion of that piece  of gear in any music you've ever enjoyed. You have
143. 14:58been freed from needing that special distortion as  an ingredient in making good music. Why do people
144. 15:03still think they want that special distortion  when they don't even know what it sounds like?
145. 15:09Wait, what does that special distortion actually  sound like? This video isn't called Where Does the
146. 15:13Tone Come From In Distortion, but now it is, this  is Where Does The Tone Come From In Distortion.
147. 15:18I'm planning a dedicated video all about  distortion, but for now I'm only worried
148. 15:22about answering two questions. One, what does the  special vintage analog Neve distortion actually
149. 15:27sound like? And two, is it actually special?  Distortion happens when the music is too loud
150. 15:33for the gear to handle, or put a little more  accurately, when the amount of voltage is higher
151. 15:37than the gear can handle. Nobody uses the phrase  "distortion ceiling", but in the tests I've done,
152. 15:41that seems to be what's happening. Each piece of  gear has a ceiling where it can't put out any more
153. 15:45voltage and the wave gets clipped. The presumption  is that each different type of distortion behaves
154. 15:50differently around and above this ceiling, and  that's what makes them special. And maybe they do.
155. 15:56But I don't think I've ever heard a test where  anybody has tried to match the frequency response
156. 16:00so that isn't causing the difference, and match  the amount of distortion so that isn't causing the
157. 16:04difference, and then just compare two different  methods of achieving distortion and listen to
158. 16:08how big that difference is. Well, actually, now  that I think about it, I do remember one time
159. 16:13that happened. But anyway, if we take the Neve and  feed it signals like drums and vocals and guitars,
160. 16:17but keep track of how far above the distortion  ceiling we're going, what does that sound like?
161. 17:14Okay, that's real. That's vintage analog British  discrete Neve distortion. If you purposefully or
162. 17:20negligently set up your gear so that this happens,  that's what it sounds like. But is it special?
163. 17:26I could rent some preamps and do a big shootout,  but that wouldn't actually answer my question.
164. 17:30I want to test the absolute limits. And I'm  just a performer, I have no way of knowing if
165. 17:34these circuits or these plugins do something  special that might secretly make them kind
166. 17:38of like the Neve. I want something that  is unquestionably dumb and not special.
167. 17:44So, I wrote the distortion plugin. I wasn't  exaggerating. I don't know how to code,
168. 17:49but I know the crudest version of distortion  would be, "Hey, if it gets louder than this,
169. 17:53don't let it get any louder. Otherwise, don't  touch it." This would be hard distortion, the
170. 17:58hardest of distortion, or digital clipping. And  it's exactly what converters do when they clip,
171. 18:03which remember is universally regarded as very bad  distortion. But if I send the same audio through
172. 18:09it and match how far over the distortion ceiling  they are, what does that actually sound like?
173. 18:58To recap this whole distortion section of  the video. Typical mic choices for different
174. 19:03instruments don't produce enough voltage to  distort the Neve when recording music. And
175. 19:07if they did, the output voltage of the knee would  be so high it would clip the converters first. So
176. 19:12Neve distortion has never been an ingredient in  professional music, and you don't need to worry
177. 19:16about it, unless you turn down the output fader  so you could get Neve distortion without converter
178. 19:20distortion, and even if you did, it wouldn't sound  any different from the converter distortion you
179. 19:24could have gotten for free but purposefully  avoided. But Jim, says the internet, gear is
180. 19:30always distorting! Okay. Why do you think that?  Look at the total harmonic distortion percentage
181. 19:36in the manual! Okay, so total harmonic distortion,  or THD percent, is a real thing. It's a measure of
182. 19:42distortion, and some gear might be different.  Some might have .1%, some might have 1%, but I
183. 19:48don't know what that means. It'd be good to have  context for what that actually sounds like. So I
184. 19:52looked at the distortion page of my Neve sweep  tests. Since it sent out a perfect sine wave,
185. 19:56anything it gets back that deviates at all from  a perfect sine wave is distortion. Specifically,
186. 20:01when it's harmonics, or multiples of the frequency  being sent at any given moment. And I'm going to
187. 20:06be super generous here. As the gain knob is turned  up, the signal to noise ratio gets a little worse
188. 20:11and the distortion rises. And it looks like it's  highest here. So just to pick out a spot, with the
189. 20:16gain knob at +40, the highest the distortion gets  is -60.5 dB when the clean signal is at 15.4 dB.
190. 20:23That means the distortion is 45dB quieter. So  in normal use, the ambient distortion level is
191. 20:30at least 45dB quieter than the undistorted  signal. Can I hear that? Can you hear that?
192. 20:37Can anyone hear that? I tried a few different  ways of figuring this out. And my favorite,
193. 20:40easiest, fastest, and most unbiased way is to  take the Klippel listening test. It's a website,
194. 20:46and it gives you the chance to listen to either  a song or a couple of sine waves clean and then
195. 20:50with distortion. After that, you have to choose  whether you think you're hearing the original
196. 20:53undistorted version or the version with some  amount of distortion. It starts very easy and
197. 20:58then as the test goes on and you get some answers  right, the amount of distortion gets less and less
198. 21:02until eventually you're just guessing.  And then it gives you the results. So,
199. 21:06I did that in my acoustically treated studio with  really nice speakers that I tuned specifically for
200. 21:10the room and these are my results. Wait, Jim,  that's not fair! Not everybody has that! Okay,
201. 21:14so I also did the test in my kitchen on my  laptop with a pair of cheap earbuds and these
202. 21:19are my results. Wait, Jim, that's not what I'm  doing when I'm commenting on your videos! Okay,
203. 21:22so I also did the test in my bathroom sitting on  my toilet and listening through the phone speaker
204. 21:27just like you're probably doing right now. And  these are my results. In the studio, my threshold
205. 21:32for audibility of distortion on a pair of sine  waves is -71dB. That's actually insanely good.
206. 21:38It means the speaker system is clear and my ears  are capable. The sine wave test is sort of a best
207. 21:43case scenario for being able to hear distortion.  It makes it as easy as possible to hear a
208. 21:47difference. In the kitchen on the earbuds it was  -42dB and in the bathroom on the phone speaker
209. 21:52it was -51dB. So those situations are less ideal  for critical listening. But the real question is
210. 21:58how loud does distortion have to be to hear  it in music? And here are my results for that.
211. 22:08I'm just a performer. I don't know anything  about psychoacoustics. But every musical
212. 22:12instrument has some amount of overtones,  and harmonic distortion is overtones.
213. 22:17So very low levels of distortion are susceptible  to not being perceived in music due to auditory
214. 22:22masking, which is when low volume sounds at  a given frequency don't even make it to the
215. 22:26brain because that area of the cochlea is already  being vibrated by something else that's louder.
216. 22:31That's probably why when it came to music  in any of the three listening environments,
217. 22:35it got really really hard to hear the  difference between the totally clean example
218. 22:39and the distorted example when the distortion was  about 15dB lower than the clean signal, and then
219. 22:43literally impossible beyond about 21dB. Oh no.  Oh no. Oh no. [laughter] Oh, that's embarrassing.
220. 22:54This is going to suck. [laughter] Yeah, I think  that I got way lucky on that. But Jim, maybe
221. 23:01that's just your ears! Not everyone's... Before I  even had a chance to wonder if my ears were bad,
222. 23:05when I finished the Klippel test, it showed me  where the thousands of previous participants
223. 23:09ran out of hearing. And minus a few cheaters over  here, it's a pretty normal looking bell curve. So,
224. 23:15it's unrealistic to think that anyone can hear  distortion levels in music that are beyond 30dB
225. 23:19below the fundamental. What were the Neve numbers  again? Oh, yeah. -45dB. So, while analog gear,
226. 23:27or more likely all gear, has a nonzero amount of  distortion happening all the time, it is so small
227. 23:33that there is literally zero chance that you or  I are ever perceiving it. So, it cannot be part
228. 23:39of the Neve sound. And if you're still skeptical,  then cool. Take the test for yourself. It's super
229. 23:44easy. You can even do it on the toilet. But Jim,  you're clearly not a real engineer, because you
230. 23:49haven't mentioned slew rate! Every amplifier has  a slew rate! Alright, thanks for letting me know,
231. 23:54I'll look that up. Okay. Preamps are amplifiers.  They amplify everything that goes through them.
232. 23:59But they can't do it instantaneously.  It takes them some amount of time.
233. 24:04That amount of time is called the slew rate. And  some preamps have a faster slew rate than others.
234. 24:09And the worry would be that the preamp isn't going  to be able to catch up to the big sudden sounds
235. 24:13that it needs to amplify and it would cut off or  shape the beginning of those big sudden sounds.
236. 24:24Or wait, maybe the worry is  that the shaping is good and
237. 24:26the slowness of vintage preamps  makes them actually better than
238. 24:29boring affordable ones that are built  into whatever gear you already have.
239. 24:37But while this is the image the flowery  internet language conjures in my brain,
240. 24:41is that actually what's happening?  Slew rate is voltage rise over time.
241. 24:45So to know how it affects preamps, it seems  like it'd be extremely important to know,
242. 24:49how many volts are we talking about? And how much  time are we talking about? Well, I think I already
243. 24:55know the volts part from earlier. Just before hard  distorting, Neves can handle 5 volts peak on the
244. 25:00input and 25 volts peak on the output. So that  means when it comes to signal, 50 volts peak to
245. 25:05peak is the biggest swing any part of the preamp  will ever have to make. Now how much time does it
246. 25:10need? Well, these fades and Pro Tools that I made  as an example of how big loud drum hits might be
247. 25:15affected are 20 milliseconds. So that would for  sure be something. But the problems could start
248. 25:20way before that. Because if I zoom in on a wave,  it's going from its lowest point to its highest
249. 25:25point really really fast. Like 5 milliseconds  for 100 hertz. So if the slew rate is getting
250. 25:30in the way of that, it's going to be an issue.  In fact, as the frequency gets higher and higher,
251. 25:35the amount of time between the lowest and highest  points of a wave gets shorter and shorter,
252. 25:39and the slew rate would have to be faster and  faster to accurately reproduce it. So let's
253. 25:44shoot for the moon. What's the highest frequency  I care about? 20 kHz, the highest humans can hear.
254. 25:48What amount of time does it take a 20 kHz wave  to go from its lowest to its highest point? 25
255. 25:54microseconds, holy crap, that's fast. Okay, cool.  So, for the slew rate to ever affect what we hear,
256. 25:59a preamp would have to have a slew rate of less  than 50 volts per 25 microseconds or 2 volts per
257. 26:05microsecond. How does that compare to the actual  slew rate of different preamps? Well, maybe it's
258. 26:11because search engines suck now. Or maybe it's  because nobody has actually done this before.
259. 26:15But when I searched for slew rates of microphone  preamps, about half of it was stuff I wasn't
260. 26:19looking for, and half of it was people talking  without providing actual numbers or measurements,
261. 26:23which also isn't what I was looking for. And  buried deep in the results, I found one guy that
262. 26:29measured a Neve, an API, and a Millennia preamp.  And his results were 8 volts per microsecond for
263. 26:35the Neve, 10.2 volts per microsecond for the API,  and 24 volts per microsecond for the Millennia.
264. 26:41So, even the slowest of these was four times  faster than necessary to reproduce all audible
265. 26:47frequencies in all conditions. But maybe that  one guy is wrong. So, I'll be the guy. I'll do
266. 26:54it myself. Hey, remember that signal generator and  oscilloscope? Well, those can also show how fast
267. 26:59the slew rate is. I'm setting this graph so each  box is 2 volts tall and 1 microsecond wide. So, if
268. 27:04the slew rate of the Neve is less than 2 Volts per  microsecond, aka slow enough to ever affect the
269. 27:10sound of the audio in any situation, then the line  would be shallow, below 45°, like this. But if the
270. 27:16Neve is capable of reproducing a line that's  45° or steeper, then the slew rate is faster
271. 27:21than 2 Volts per microsecond, and therefore not a  problem. So what happens when I actually do that?
272. 27:29Oh, it's pretty steep. Let's do counting. It goes  to the right one two boxes and it goes up 1 2 3 4
273. 27:375 6 7 8 boxes. So the Neve is capable of at least  8 volts per microsecond. And amongst people who
274. 27:44talk about this stuff, the Neve is talked about as  a slow preamp. If the Neve is slow, and it's four
275. 27:51times overqualified, then everything else that's  fast would also be overqualified. And according to
276. 27:57math and anatomy, in these cases, slew rate never  is, never would be, and never has been audible.
277. 28:05But there's another way to look at it with  less math. We can just see if the bad thing
278. 28:09we're worried about actually happens. If the slew  rate is slow enough to be bad at amplifying high
279. 28:14frequencies at high voltages, then I can just  send sweeps of different voltages through the
280. 28:18Neve and see if the results are ever different.  So, I did. And there's literally zero difference
281. 28:25in high-end frequency response, even if the Neve  is working with the highest frequencies humans
282. 28:29can hear at maximum voltage. Okay, turns out the  cartoonish idea of the slew rate of a Neve shaping
283. 28:35the front end of drum hits like a crossfade  would only be possible if the slew rate was
284. 28:38literally 10 times, [clears throat] excuse me,  100 times, wait, 3,000 times slower than it is.
285. 28:45Jim, you're clearly not a real engineer.  You haven't mentioned impedance. The preamp
286. 28:49uniquely interacts with the mics because  of impedance. All right, I'll look that up.
287. 28:54Okay, impedance is an electrical term. It's like  resistance, but for when resistance is different
288. 28:58for different frequencies. Mics have an output  impedance and preamps have an input impedance. So,
289. 29:03this could make the same mic sound different  if it's going into a different preamp.
290. 29:07Okay. Does it? I got out my mic testing  rig and set up each of my mics. Then did
291. 29:12a sweep test with them plugged into the Neve and  a sweep test with them plugged into the garbage.
292. 29:23And these were the results. Okay, cool. This is  not nothing. It may even almost be something,
293. 29:30but I've got to check how big  these differences actually are.
294. 29:33And the biggest difference in any of these is  0.9 dB at 200 Hz. So, what does that sound like?
295. 29:58And everything else I got was less different  than that. There might be a mic out there with
296. 30:02a weird output impedance and a preamp out there  with a weird input impedance that makes this
297. 30:06a bigger deal. But at least with all of these,  impedance isn't an ingredient in the Neve sound.
298. 30:12But Neves have inductors! The sound of the Neve  is in its analog inductive EQ! You haven't talked
299. 30:17about the EQ! Whoa, hey, that's moving the  goalpost a bit. I thought we were talking
300. 30:21about preamps. This video isn't called Where Does  The Tone Come From In An EQ, but guess what? Now
301. 30:26it is. This is Where The Tone Comes From In An EQ.  Okay, so the point of an EQ is to change the sound
302. 30:31to make it brighter, darker, fatter, thinner, or  some other adjective. And it does this by changing
303. 30:36the frequency response of the piece of gear. In  other words, if you want something to be brighter,
304. 30:40the gear has to do something that makes it  better at passing through the treble. If you
305. 30:43want something to be darker, the gear has to do  something that makes it worse at passing through
306. 30:47the treble. I'm just a performer. I don't know  anything about circuits, but the inductors in
307. 30:52a Neve are part of an inductive circuit. That  is one way of accomplishing this. And engineers
308. 30:56love how musical analog Neve EQs are. And I  don't know what that means, but I know that
309. 31:01engineers are not so complimentary of the stock  Pro Tools EQ plug-in, which contains no inductors,
310. 31:06or any other analog parts. So, I did some tests  of my own. And the stock Pro Tools EQ makes clean,
311. 31:11cold, calculated frequency response curves  like this. And the Neve makes warm, creative,
312. 31:17musical analog frequency response curves  like this. Oh, they're pretty much the same.
313. 31:25But that isn't fair of me. I did have to dial  the Pro Tools EQ to match what the Neve does.
314. 31:29But the fact that I can even do that at all kind  of blows up the thing I used to think, which
315. 31:33was that digital EQs just can't do what analog  EQs do. But what happens if I change the gain?
316. 31:40Oh. Oh, wow. So that's the analog EQ sound. It  isn't that the curves are uncopyable. It's just
317. 31:47that the bandwidth of the boosts and cuts  might be different as the amount of boost
318. 31:51or cut changes. Like the Neve starts out wide  and gets more narrow, whereas the Pro Tools EQ
319. 31:56keeps the same width no matter the boost. So maybe  that's what engineers are calling musical. Sure,
320. 32:01every individual curve is matchable by digital  methods, but the analog EQs automatically have
321. 32:06this behavior and digital doesn't. Unless  you spend 250 bucks for this plugin, then
322. 32:12it matches almost perfectly. Or 100 bucks for this  plugin. Or 35 bucks for this plugin. But I'm just
323. 32:17a performer. I don't know anything about making EQ  plugins. And that brings up a really good point.
324. 32:22I'm just a performer. I'm just a guy who plays  guitar and owns a camera and spent 6 months of
325. 32:27rent on a preamp with the goal of figuring out how  much of an effect the legendary Neve console at
326. 32:31Ocean Way Nashville Studio A might have had on  the sound of my favorite music. Sure, I hunted
327. 32:36down the exact vintage module they have, but at  the end of the day, no matter how ironclad all my
328. 32:40tests and data are, it isn't the same as actually  getting to go to Ocean Way and test their board.
329. 32:51You know, I had to do it, right? This is Ocean  Way Nashville Studio A's 1970s Neve 8078 board
330. 32:59with 31105 preamp and EQ modules. This is the  board that my favorite music was tracked through,
331. 33:05and it is one of the most important things in the  world to me to find out what the sound of this
332. 33:10board is, and if I can somehow have it for myself.  So far, from the results of all of my tests,
333. 33:16it seems like... Neves are not warm.  They're flat from 20 Hz to 20 kHz,
334. 33:22just like this. Neves are not punchy.  They operate linearly until distortion,
335. 33:27just like this. Neves are not reaching  distortion under normal use, just like this.
336. 33:32Neves, when forced into distortion, don't sound  any different from converter distortion. Just like
337. 33:36this. Neves are not always distorting in a way  that anybody can hear in music. Just like this.
338. 33:42Neve's slew rates are more than four times  overqualified to perfectly reproduce all
339. 33:46frequencies in the human hearing range. Just like  this. Neve's input impedance is within a normal
340. 33:51range, so it doesn't cause major changes in common  microphones frequency responses. Just like this.
341. 33:56So theoretically, if I ran the same audio through  Ocean Way Nashville Studio A's 1970s Neve 8078
342. 34:03board with 31105 preamp and EQ modules,  and also the garbage built-in preamps on
343. 34:08my cheap interface, I can't think of any reason  they wouldn't sound the same. So I did that.
344. 34:30This was a huge moment for me. All of these years  I thought that my favorite music sounded the way
345. 34:34it does because of the Neve board, and its  preamps, and so did the people who made that
346. 34:38music. This would be true across the board on  all those records. First of all, I love Neves. I
347. 34:44didn't bring any outboard mic pres. I love the old  Neve in Studio A. It's still my favorite one in
348. 34:50Nashville. As it turns out, preamps just make the  mic signal louder. There's a bunch of different
349. 34:54ways to do that, but if they aren't changing the  frequency response, then they're just doing their
350. 34:58job. My friends, my teachers, and my heroes have  been misattributing the warm, punchy, musical
351. 35:05sound of the songs they've recorded, giving credit  to the preamps that actually should have gone to
352. 35:09them, and the decisions they made about mic  choices, mic placements, drum sizes, tuning,
353. 35:14muffling, cymbals, pickups, pickup positions, amp  circuit, amp settings, cab geometry, speakers,
354. 35:19acoustic guitar body size, string age, pan  positions, reverb send levels, volume fader
355. 35:24levels. Oh yeah, and I'm just a performer, so I  can't leave out all of the actual performances.
356. 35:31But I wasn't done. One thing I left off that  list that definitely belongs on it is EQ.
357. 35:36EQ can absolutely change the sound. If there's  one way that a preamp can sound like something,
358. 35:42it's because of the specific curves produced by  its EQ. So, I spent 4 hours in that control room
359. 35:48doing tests of every possible EQ change that  can be made on that board. Here are some of the
360. 35:54things that the Neve actually does. First, the mid  bells are wide at low gains and get narrower at
361. 36:00high gains. We saw that earlier. But the low bell  actually moves its center frequency higher with
362. 36:04higher gain. So, when the low bell knob says 100  hertz, it's only centered on 100 hertz when it's
363. 36:08on maximum gain. Otherwise, the center is actually  less than what's printed. Maximum gain for all
364. 36:13bands is about 18 dB with the shelves being  a little more up around 20 dB, which makes it
365. 36:18weird that the Universal Audio plug-in emulations  I tried only have the shelves go up to 12 dB and
366. 36:22the IK81 and Waves VEQ4 only have the bells go up  to 12 dB. And the maximum gain and gain travel for
367. 36:28a given knob is about the same for all frequencies  in a band, which makes it weird that the Kit Audio
368. 36:32Blackbird plugin is all over the place when it's  supposed to be modeling the same type of preamp
369. 36:36from the same type of board. All of the frequency  knobs snap to specific predetermined frequencies,
370. 36:41and they're all pretty close to what's printed  on the face plate, but they aren't always exact.
371. 36:44For example, on this channel, 390 is actually 406.  Also, there are 10 frequency positions in the mid
372. 36:50bands, and positions 1 3 5 7 and 9 have the same  bandwidth, and positions 2 4 6 8 and 10 have the
373. 36:55same bandwidth, but for some reason, the evens are  slightly narrower than the odds. And the real Neve
374. 37:00has small opposite resonances that make the curve  a little different from a standard digital EQ,
375. 37:05even if the bandwidths are matched. And  here's what the Neve actually sounds like.
376. 37:39I have a hunch that these curves are  what matters. And if they are matched,
377. 37:44even digitally, then it should sound the same.  There's no reason I found that it wouldn't.
378. 37:49But while all of the existing plugins I've  tried are close, none of them match it exactly.
379. 37:56So, I wrote the EQ plugin. You know what I  am. You know there's nothing special here.
380. 38:02There can't be. I don't know how to do that.  It's just an EQ. It just matches these curves.
381. 38:09But what makes it special is that someone cared  enough to put in the hours to measure the real
382. 38:13thing and then match the real thing, without  any AI at any point in the entire process.
383. 38:20AI cannot do this. I am the only one  with these data points. By the way,
384. 38:25this whole thing is brought to you by my PayPal  email address. I made 100% of this video by
385. 38:32myself because it needed to exist. And I've been  able to do that because when I write my PayPal
386. 38:37email address on something I've made to figure  out why music actually sounds the way it does,
387. 38:41enough people recognize how rare and pure this  is, and they chip in to keep the tests going.
388. 38:48So, if there's anything else special about the  sound of the vintage British discrete analog
389. 38:54inductive EQ on the 31105 preamp and EQ module  from a channel in the 1970s Neve 8078 board at
390. 39:02Ocean Way Nashville Studio A, then it shouldn't  be able to sound the same as a garbage preamp and
391. 39:07an EQ plugin written by a guitar player.  So, what does that actually sound like?
392. 39:44Which did you think was which?
393. 39:48So, usually in other videos, I'll just cut to  black and drop the mic and let the truth exist
394. 39:52for people to experience. I'm not in the business  of telling anyone what they hear, but I'm going to
395. 39:56do a little segment here called getting ahead of  it. One, is all of this actually real? Yes, 100%
396. 40:01of this video and every one of my videos is real.  The audio you hear is what I said it was. This is
397. 40:07not merely for clicks or entertainment. This is  rigorous. Two, are you saying that all preamps
398. 40:13sound the same? No, some preamps are broken, and  some preamps are made so cheaply or poorly that
399. 40:18their frequency response doesn't stay flat from  20 Hz to 20 kHz. I bought a little mixer for $30
400. 40:23on Amazon, and its treble falls off at around 15k.  But every musician everywhere can get their hands
401. 40:27on something like this, and it'll work great. But  Jim, my preamp sounds different. I know it does.
402. 40:33Okay, show me. In fact, to help you out, I made  a course that shows you exactly how to do a bunch
403. 40:37of the stuff I did in this video using tools you  already have and some free computer programs. I
404. 40:41walk you through it and then you can test not just  your preamps, but your pedals and your plugins and
405. 40:46other gear to see what it's actually doing. And  this has saved me and will save you thousands and
406. 40:51thousands of dollars. Then if you do that, you  can show me that your preamp sounds different.
407. 40:57Or maybe it doesn't. This course is actually a big  deal. It's called Find Your Gear's Hidden EQ. It's
408. 41:01brand new and it's on jimlillmusic.com. Available  instantly to you right now. Like a couple hours
409. 41:05from right now, you could know what all your gear  actually does. Four. Why should I trust you when
410. 41:10big famous guy says something different? I've  gotten texts from friends that are different
411. 41:14versions of this question where they watch an  interview or even in the room with a big producer
412. 41:18guy and he says Neves are warm or whatever.  So why trust me instead of big famous guy?
413. 41:25Because they never did this. Do you know what  it takes to make a useful comparison? It has
414. 41:29to be identical playing or it's useless from  an identical microphone or it's useless in the
415. 41:34identical spot or it's useless. Split out to the  two preamps you want to compare. Then the level
416. 41:38has to be exactly matched or it's useless.  And very short sections of the playing have
417. 41:42to be edited back to back with no gap or it's  useless. And maybe most importantly, the audio
418. 41:47file of the comparison has to exist so that you  or anyone else can actually listen to it right now
419. 41:54and revisit it later at any time to double check  it. These are the minimum requirements to begin
420. 42:00forming an opinion. If the story was, yeah, in  take three, the guitar was a little bright, so
421. 42:04he patched it out of an API and into a Neve. And  on take six, it sounded better. That doesn't do
422. 42:09any of that. [laughter] Big famous guys made huge,  great, awesome things. And those things were huge,
423. 42:15great, and awesome for a lot of reasons. And  the big famous guys that I've talked to have
424. 42:18never said that during their limited, expensive  time in big studios cutting major label music
425. 42:24that the appropriate thing to do was to question  whether the giant awesome console actually sounds
426. 42:28different from a Focusrite Scarlet. They're  already in the room with the big Neve! They just
427. 42:32use the big Neve! And they do whatever it takes  to make the richest person in the building invite
428. 42:36them back. So, nobody knew. And if you wanted to  know, you'd have to do everything I did in this
429. 42:40video. And nobody ever did that because it takes  a long time and it's hard and you risk a piece of
430. 42:45your religion and identity being chipped away.  And why do that when you have a bunch of other
431. 42:49stuff in life demanding your attention? Well, I  did it. And turns out when you actually do it,
432. 42:55this is what it actually sounds like. Now we know.  I gotta say, it's weird finding out that nobody's
433. 43:00listening. They're just repeating stuff they've  heard. So, that inspired a new project I'm doing
434. 43:03for anyone that likes this kind of honesty  and curiosity and wants to hang out with me
435. 43:07without waiting a year between my giant projects.  I just started a weekly show where I catch up on
436. 43:11important bands and artists that I should really  know about by actually listening, and you can do
437. 43:15that along with me. Too much stuff out there comes  from people who read the Wikipedia page and then
438. 43:19repeat it and pretend they're an expert without  ever showing you what it actually sounds like.
439. 43:23So, I'm not doing that. Instead, I'm telling you,  I might not know anything about this band. Let's
440. 43:26forget everything we've been told and just listen  to the songs and see what's in there and take home
441. 43:30some notes on how to get that vibe in case we're  ever asked to. I do that for fun anyway, so why
442. 43:34not post it so other people can learn, too. The  show is called Actually Sounds Like with Jim Lill.
443. 43:38If you had to tell someone about it, I'd describe  it as listening to music with a professional
444. 43:42musician and celebrating the details we notice.  The video version will be on the Actually Sounds
445. 43:46Like YouTube page and the audio version will  be on all the podcast apps. Rapid fire. Jim,
446. 43:50are you going to sell your super awesome,  Super Accurate EQ plugin? I'm glad you asked.
447. 43:54No, I'm not. I don't know how to do that. I  know that this works 100% perfectly on one
448. 44:00computer in this entire universe. I don't know  how to make it work on everyone else's computer.
449. 44:05And I don't know how to deal with anti-piracy  stuff. But are you Waves? Are you Universal Audio?
450. 44:11Do you think that people would buy this plugin?  Get ahold of me. It's a good sounding EQ.
451. 44:16Are you saying that Ocean Way and other expensive  studios are worthless? No! Plenty of reasons to
452. 44:20like big studios that have nothing to do with  their consoles sounding special. Are you saying
453. 44:24that expensive preamps are scams? No. They sound  great, but inexpensive ones can sound the same.
454. 44:29So hopefully people who are paying extra are doing  it for the extra features or workflow or because
455. 44:34it looks cool and feels cool and not because they  think just passing signal through it gives their
456. 44:38music the Neve sound. Because as it turns out,  that's not really a thing. Are you saying that
457. 44:42expensive preamp builders are scammers? No, they  don't know. They aren't asking the questions I'm
458. 44:47asking, so they haven't found the answers I  found. Specifically, the psychoacoustics and
459. 44:50blind listening stuff. I watched your video, and  I hear a difference and it's subtle. And subtle
460. 44:54changes are big difference makers. Check out that  psychological trick. The number of times I've seen
461. 44:59people attach their identity to a piece of gear,  talk about how it's a big difference maker, get
462. 45:03shown an actual comparison where it's not a big  difference maker, then walk it back by saying it's
463. 45:08subtle, and they get away with not admitting they  were wrong or changing their minds about anything
464. 45:12while also implying that it's the other person's  problem that they don't hear a difference because
465. 45:17it's their inexperience or lack of skill that's  keeping them from hearing what the pro hears.
466. 45:21Then they build the importance of the gear back  up by saying that it's all in the subtleties or
467. 45:26subtle changes make a big difference. Diabolical,  weak, subtle literally means not big. If anyone
468. 45:33says something is subtle, then that means doesn't  matter until every single other thing that is not
469. 45:39subtle is 100% accounted for. In other words, if  you don't have a preference for: mic choices, mic
470. 45:44placements, drum sizes, tuning, muffling, cymbals,  pickups, pickup positions, amp circuit, amp
471. 45:47settings, cab geometry, speakers, acoustic guitar  body size, string age, pan positions, reverb send
472. 45:51levels, volume fader levels, then you can't have a  preference for what preamp you use. All not subtle
473. 45:56things are more important than all subtle things.  Thank you. What about? What about what about? I
474. 46:02had an engineer once tell me that his boss told  him that phase differences in frequencies above
475. 46:0520 kHz can combine in ways that affect frequencies  under 20 kHz. And that's the sound of a console.
476. 46:12This is just made up. It's based on literally  nothing true. Humans can't hear relative phase
477. 46:18changes in different frequency ranges because of  the principle of wave superposition. Individual
478. 46:21waves combine to form the final wave that hits  your eardrum and they can all ride on top of
479. 46:25each other. And while relative phase changes  can make the final wave look different under an
480. 46:28oscilloscope, it doesn't make it sound different  when it reaches our brains. And it would take
481. 46:32me time and energy to load up a bunch of signal  generators with a bunch of sine waves and bounce
482. 46:36multiple copies of them with different flipped  phases and do a 100 rounds of an ABX test to show
483. 46:40that it can't be heard even if it was in the human  hearing range. And this guy was talking about what
484. 46:44happens outside of the human hearing range. But  I'm not going to do that for every sci-fi fairy
485. 46:49tale that gets conjured or repeated by someone  who isn't willing to bring any amount of evidence
486. 46:53themselves. It takes no effort to make something  up. You can use big words and try to say something
487. 46:57that feels kind of good and people who look up to  you will just go, "Huh?" and store it away and not
488. 47:02question it and then later repeat it themselves in  a social situation for social points. This video
489. 47:06could have ended 5 minutes in when I recorded  every instrument through the Neve and the Scarlett
490. 47:10and they sounded the same, but I went the extra  mile and had enough respect to give a bunch of
491. 47:15the other mechanistic theories a chance to show  themselves. And in the end, what did you hear?