Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Andy Simpson]

Chuck, further to your (old) post in another discussion;

On the subject of magnetic fields (and the eagles), in relation to valves/tape vs transistors/digital, I have begun to suspect something along these lines - 'more tubes, better sound' - for some time.

I have recently been experimenting with a quad set of TLA tube preamps that I've got ('4001' set, yes cheap but quite clean). To summarise, I've been chaining all four together in series and driving each to just before the point of clipping. The result is undeniably better in all cases. It just sounds more magical (I know it sounds stupid). How? I don't know. I can't believe that I'm getting significant drive from the 2nd,3rd and 4th pre-amp tubes, since the first pre-amp has already driven the signal to a shape that 'fits' the tube distortion profile.
Detail, presence, !*smoothness*! all improved markedly. In a word, silky. No, that doesn't even do it justice.

When I listen back to the sound of Sgt pepper (&rubber soul/revolver), and add up the number of valves it must have passed through, it comes to a large number: 1 in mic, 1 in pre-amp, 1 in compressor, 1 or 2 in line/buss-amp in desk, 2 in tape machine, 1 more in buss compressor, 2 in master tape machine, etc etc. The list goes on.
Then I think of the white album. Transistors. Not very magical or exciting (despite the fact that it was recorded and mastered to tape), especially compared with SGt pepper, revolver or rubber soul. On headphones there is no comparison at all. Not as much life.
Carol Kings 'Tapestry' also sounds magically musical to me. That vocal sound is beyond nice. How many tubes?

Also, I tried running the insert signal from my marshall DSL50 head through the set of 4 pre-amps, and fuck me if it doesn't sound just as magical as sgt pepper (especially with the spring reverb going through the pre-amp set as well)! I'm not kidding! I sat and played for hours. Everything I played came out magical!

Also, whilst on the subject, I think that with enough valve stages, one could concievably achieve massive 'compression' ratios in an extremely soft and musical way.

One day, I'm gonna get me a massive array of tube line-amps and see what happens. Maybe I'll get me one of those TLA all-tube desks, and just run it through every single channel and buss!

Btw, I have always thought that Abbey Road would've sounded _much_ better on tube gear! A thinner, more brittle beatles sound I can't find (except on the white album - but in my defense, I grew up listening to all the beatles albums on CASETTE!

Enough contention, I think!

Andy.

[Chuck]

Hi Andy,,

I think you have found your sound by running the signal through many tube stages.

Although many designers want to make their gear 100% transparent, it is always a cool test, when you  run a signal through it 5 or more times, because then you come to know the character of that piece of equipment, that hides beyond transparency...

So I consider your test very revealing. Please let me know of your recordings.

Charles

PS: I am also a big cassette fan, especially old BASF ferros. They just have something in their sound... and I wonder how long they will last...

[Andy Simpson]

Yeah, I just wish I could get my hands on some (cheap) line-level versions, 'cause feeding these pre-amps with line level is a waste of gain, and tends to get noisier than it should. This is where the line/buss amps in the TLA desk would be great.

It's a shame I can't just drive it through the tubes and back into the A<->D and repeat until the desired effect is achieved, but multiple A<->D conversions will not improve the sound. Infact, I have had better results when driving a mic source than a line-level D->A signal.

Anyway, I'll let you know when my next project is done, and send you a copy or whatever.

Cheers.

Andy

[Chuck]

Hi Andy,,

I don't know how much fun you have with the soldering iron,, but I think you are exactly on the right track of becoming famous for your own sound, and if you have the time and a little patience, what you would need maybe is this:

https://athena.safe-order.net/decware/zplates/zplatesp1.htm

You can build it after instruction, and then start to mod it and then are able to build exactly what you need, for very few bucks.

I have a Zen-amp-kit from this guy, and... this is really much fun, and he has much experience with this stuff.

His website is http://www.decware.com

Happy soldering,,

Charles

[Bob Olhsson]

Most '60s tube consoles had massive amounts of usable dynamic range compared to all but really high-end solid state gear. Much of the "euphoric tube distortion" mythology originated with early solid state manufacturers who were trying to defend the quality of their products to us. I learned years ago that there is remarkably little difference between the sounds of really well designed tube and transistor gear. It's just that there's always been a lot more poorly designed transistor gear available.

[Andy Simpson]

Point taken. However, you've only got to read a few Geoff Emerick interview articles (is his name dirty on this board?) to find out that they had great trouble getting decent sounds when EMI made the switch to transistor desks (though I'd love to know whether they kept the 1176 in it's resident 'buss compressor' position....I'd say not from the evidence).

Andy

[George Massenburg]

Your observations about tube sounds and the records that seemingly were made with tubes are interesting, and it's clear you're quite charmed with tube.

I, on the other hand, spent my first 10 years, give or take a few, trying to make tube consoles and tape recorders sound good.  I designed several boxes, and learned alot, But, honestly, I did (and, worse, was paid for) some of the WORST recordings in the universe.  Unlistenable, to say the least.

I couldn't wait to get my hands on better gear.

George

[Andy Simpson]

Charmed is probably an understatement.
I have long been of the opinion that it is the drive that makes tubes interesting to me. For example, there are probably quite a few overlooked mics that are naturally thin sounding (does the u47 come under this category?), especially when used at a distance, that would fatten up wonderfully when driven, and conversely modern large diaphram mics that are too 'fat' to drive much through a tube (without pre-eq). It is with tube drive that changing the mic/position REALLY changes the sound achieved.
I prefer to get an intimate vocal sound by drive rather than actual proximity - less problems with un-natural sibilance and more silkiness. For fun I demo'd an entire song today with just a 57 and my 'super-quad-tube' setup - a beautiful lack of top end! Gorgeous on vocals, accoustic and shakers - and what a delightfully firm but fluffy noise floor to sit on

In essence, I see tube-drive as a zoom-lens on the signal. Ie. you _could_ shoot the picture from right infront of the rabits face, but if you step back 20ft and get a good zoom lens on it you get lovely focal depth and all the detail is focussed on the rabits face and not 'wasted' on the whole picture - a nice classy zoom shot. You can see the light bouncing off it's whiskers - definately a nicer shot.
Does that make any sense?

Also, I enjoy it when (in a mix) each mic sits nicely in the sweet spot of the pre-amp tube - just before saturation - that when dynamics happen, things get wider instead of louder per se - I believe Carol Kings Tapestry is a good example of this, and a good example of headroom + saturation = very nice record when listenned to really loud.

I got my first taste of guitar amp (clean) tone when I realised that I could use more pre-amp tube gain on a thinner sounding pickup (the rear strat 'quack' position) and get just as fat a sound out, but with loads more harmonics and glorious bubbling - suddenly the guitar is right in the room with me. Whereas using a tone-bucking les-paul (std) I couldn't get enough drive because the output was so bassy and lacked the cancellation harmonics anyway.

Who would record with a transistor guitar amp?
Why should the same principal not be extended to all facets of the amplification process? - Obviously, a zoom-lens photo of a zoom-lens photo of a zoom-lens photo gets to look worse not better, but that's the art of it.

Also, as a foot-note, I truly believe that the Beatles made good records because they simply couldn't fail to get a musical (good) sound because of all the tubes, despite the amazingly short time taken. Ie. 10am - Get to the studio, plug in the u47s, turn up the gain until it almost distorts each channel (motown style). Same thing on tape. 11am - record about 5 songs before lunch. Then 'mix' all four channels while they pop out for a joint (also motown style).

[Andy Simpson]

Incidentally, as a reply to myself, does anybody know what kind of HF roll-off I could expect from an average u47 into 60's EMI tube desk onto 4-track 2" tape?

I have been experimenting with sm57 into tube preamps (as previously mentionned) and am really liking the (fairly steep) roll-off combined with the presence and detail added by the drive. I had never considered using a cheap dynamic for vocals until I tried this as an experiment, but the thing sounds incredibly musical when rammed through some valves, and I really don't miss the upper 8k+, infact, it always sounded somewhat unfriendly from my condensers (although by now you may have gathered that I don't really dig the whole clean scene - I'm just trying to guage whether there is anything significant above 8 or 9k on any beatles album?)

By the way, I just remembered the Miles Davis album 'Kind of blue' - what absolutely incredible tone! I challenge anyone to do that on SSL, and I'd like to see a freq. analysis of that one.

Andy

[davidc]

Many of my favourite albums were recorded through tube gear, but there have also been a lot of great albums recorded through solid state. I think its what you do with it that counts.

Anybody know how Joni Mitchell's "Blue" was recorded (1971)? This has always been a favourite of mine, and I assume it is early transistor from the sound of it, but I am willing to be surprised.

Best Regards

David C

[Andy Simpson]

What albums in particular? And how were they recorded? (And it doesn't count if the console was SS but they used valve mics, valve pre-amps and valve compressors!

My favourite reference albums for magic tone are:
'Pepper (reverb)/Revolver (dryness)
Tapestry (intimacy)
Kind of Blue (smooth-ass dry T O N E)

These albums all have strong presence/colour/flavour/character to my ears, and are unmistakably evocative of whatever flavour.

I do like Joni's Blue alot, but it could've been warmer and more intimate in my opinion....

I often wish Queen had been recorded by George Martin.....

....and I simply do not like Pet Sounds!

Andy

[Bob Olhsson]

andy_simpson wrote on Wed, 07 July 2004 16:23

I just remembered the Miles Davis album 'Kind of blue' - what absolutely incredible tone! I challenge anyone to do that on SSL, and I'd like to see a freq. analysis of that one.

I'm sure the fact that very few mikes were used compared with what became common ten years later combined with a wonderful sounding studio had far more to do with the tone than any of the gear did! They also probably didn't wear headphones which frequently screw up musicians' dynamics and blend.

[davidc]

Bob Olhsson wrote on Thu, 08 July 2004 04:31

They also probably didn't wear headphones which frequently screw up musicians' dynamics and blend.

Thats an interesting one Bob. Would you mind explaining it further. Every musician I know insists on wearing headphones, or else they reckon they cannot hear themselves.

Best Regards

David C

[punkest]

davidc wrote on Thu, 08 July 2004 09:05

Bob Olhsson wrote on Thu, 08 July 2004 04:31 They also probably didn't wear headphones which frequently screw up musicians' dynamics and blend. Thats an interesting one Bob. Would you mind explaining it further. Every musician I know insists on wearing headphones, or else they reckon they cannot hear themselves. Best Regards David C

Yes and specially singers, but not if you are playing with other musicians in the same room, without playing to a click track or any other cues from the control room.  All they needed to hear was their own ensemble.  Whenever they needed to use the talkback to say something to the musicians in the studio they used the studio speakers. Maybe also to play back a take for approval.


Hans Mues

[debuys]

If I can but in here I'll add some observations from a tube addict.

Tubes and solid state circuits can sound and scope identicly. Where the major diferences is when you push them.

Where you can hear the onset of cliping on a solid state circuit, a tube curcuit sounds far less clipped and distorts in a diferent manner. One that is actually pleasing to the ear.

When a tube gain stage begins to clip there is not an imediate sound of "dirt". Rather it compresses and it's frequency response changes. This is why a 30watt tube guitar amp sounds "louder" than a 30watt solid state.

Also, tubes excell in simple circuits free of any transistors or op amps. Add some solid state componets and the true character is easilly lost. Often times it's even worse sounding than if the circuit was all solid state. It depends on where the tube is in the circuit.

[ted nightshade]

Bob Olhsson wrote on Wed, 07 July 2004 20:31

andy_simpson wrote on Wed, 07 July 2004 16:23  I just remembered the Miles Davis album 'Kind of blue' - what absolutely incredible tone! I challenge anyone to do that on SSL, and I'd like to see a freq. analysis of that one. I'm sure the fact that very few mikes were used compared with what became common ten years later combined with a wonderful sounding studio had far more to do with the tone than any of the gear did! They also probably didn't wear headphones which frequently screw up musicians' dynamics and blend.

Plus, Miles insisted on going with first takes- in the case of an album like Bitches Brew, this is very evident- raw indeed! He believed that musicians would grow more conservative with every take, and he liked to capture them out on the edge, fresh. I know this became a big issue with Trane, who always wanted to try everything again, and again... as Miles noted he did that 3 times in most solos anyway! Ellington during the Money Jungle sessions with Max Roach and Trane took the engineer  or producer out into the hallway to plead with him just to go with Trane's first take- "he's just going to repeat himself".

Time has pretty much proved Miles right! From what I understand another copy of Kind of Blue leaps off the record store rack every few minutes or so, and probably will continue to do so as long as recordings are sold.

As for tubes- there is a dimensional quality with tubes that I have never heard with solid state. I will say that I admired a GML EQ in bypass for preserving that quality- most solid state things I have heard, at least with gain, tend to flatten it right out. I'm a great believer in the dimensional qualities of sound reproduction- very high on my list of what to listen for and what to strive for, and tubes can have a lot to do with that. Transformers can too.

The hitch is, the sweet spot thing! In other words linearity. Probably a bigger issue with transformers, which supply a lot of what tubes usually get the credit for. I was very struck by listening to the 3D Mic Pre CD, that the Manley 40dB pre, which I have worked with a lot, showed very poorly on the female vocal- man did I wish I could get in there and twiddle the gain structure and make it sound like it should! Other far more linear pres did far, far better in the arbitrary test gain structures they ended up in. But then on the acoustic guitar test, the same pre just exploded into color and dimension when the guitarist hit it just right. Basically at that point I felt that  the 40dB pre just blew away everything else for sheer life, drama, dimension... but without the gain structure just right and hitting it just right, what a dud!

Again I think it has more to do with the transformers than the tubes, but tubes are not especially linear either. However, a distressingly large amount of solid state gear just destroys the sound when pushed too hard- a good reason to go for solid state kit with sky high headroom and then a goodly margin beyond that. That extreme headroom, common to good quality tube gear, is very necessary for solid state to reach it's promise of linearity at a wide dynamic range.

[Andy Simpson]

You've hit the nail on the head regarding depth and dimension. Those are both big factors in valves sounding more 'life-like' (or should I say 'larger than life'?) than transistors.
I think that presence (the most commonly cited trait of valves) is very closely related to depth, only presence is the nearest end of the depth, if you see what I mean.

Also, this sense of depth tends to make the noise floor seem like an out-of-focus background in the picture, so you can get away with a much more present noise floor without any artistic discomfort.

Do you have any more info/ideas on how transformers induce or maintain the depth?

Btw, here's a quote from Geoff emerick re. tubes:
"
[Laughs] Well, that's because it was all tube equipment. All the albums up until Abbey Road were recorded through a tube desk. Abbey Road was the first album that was recorded through an EMI transistorized desk, and I couldn't get the same sounds at all. There was presence and depth that the transistors just wouldn't give me that the tubes did.

That must have been frustrating.

Oh, it was. But, of course, it gave a texture to the Abbey Road album after all, which is quite pleasant. But at first, being used to the tube desk and being confronted with the transistorized desk, it was like chalk and cheese. It was hard. And there was nothing I could do about it except craft the music around it; it was a much softer sort of texture.
"
Taken from:
http://www.findarticles.com/p/articles/mi_m0HEO/is_2002_Oct_ 1/ai_92280874/pg_2

I always find that a mix needs much less compression with valves, and creates alot less hearing fatigue after listening to a whole album (especially on headphones).

Also, I am guilty of the cardinal sin of engineering: I always expect the recorded signal to sound BETTER than the damn thing did in the liveroom! And with tubes it does!

Andy

[Chuck]

ted

Again I think it has more to do with the transformers than the tubes, but tubes are not especially linear either. However, a distressingly large amount of solid state gear just destroys the sound when pushed too hard- a good reason to go for solid state kit with sky high headroom and then a goodly margin beyond that. That extreme headroom, common to good quality tube gear, is very necessary for solid state to reach it's promise of linearity at a wide dynamic range.

Yes, the transformers add up to the sound, as they can smooth things out, especially when they don't measure perfect.

As for linearity: a single preamp triode is much more linear than a single transistor.

What makes a transistor circuit linear, is the use of multiple gain stages (at very least two transistors), achieving high gain, and then the use feedback to get back down again.

A tube circuit don't need that feedback to be linear.

You can consider a sonic character that is free of feedback, just like a unhindered free decay of a note, whereas in a feedback desing, the decay gets some grip, as if someone is always trying to grip and control the note...

Did you know that on a guitar amp, the presence knob is in fact a feedback knob ?

With presence full on, feedback is off.

With presence less than full on, feedback is activated from the output of the amp to the input.

To get an idea this way:

Presence is the exact opposite of feedback.

Of course without feedback more overtones are produced, also with transistors.

If you listen to the output of a single transistor (with no feedback applied), you also get that free flowing decay, but the overtones that are produced are not likely what you would want (except perhaps as an effect). This also has to do with the materials involved and their resonant frequencies.

Some years ago I have introduced a transistor power amp circuit, that had a switch to put feedback off.

With that one, it is very easy to experience the difference...

http://www.altmann.haan.de/splif_page/

Charles

[bobkatz]

andy_simpson wrote on Sat, 10 July 2004 06:52

You've hit the nail on the head regarding depth and dimension. Those are both big factors in valves sounding more 'life-like' (or should I say 'larger than life'?) than transistors.

I truly believe that harmonic proportions has 99% of the explanation of what's going on here. I've written about it extensively in my book and demonstrated it as well. The Cranesong HEDD does a fantastic job of emulating various tube and analog tape characteristics without the down sides.

As you add harmonics within the Cranesong, the depth image and other "tubelike" wonders, including that "livelike" quality starts to occur. But MORE is not necessarily better. Adding harmonics reduces the peak to average ratio of a recording and it is DEFINITELY a form of compression. A little bit goes a long way and too much sugar in the sound becomes diabetes.

I've carefully compared using the HEDD-192 digitally versus various tube processors (or a dub to analog tape), and in most cases going digitally and adding the harmonics yields far more transparent audio than going through the analog device.

We need to start putting as much scientific study as possible into our art, as effective emulation of "revered" analog processors will become more and more necessary as the analog versions become scarcer and scarcer.

I'm not saying it's all coming up roses. Many other digital devices and plugins I have tested which perform harmonic synthesis or even convolution have come up short in my estimation. Many of them have poor resolution and add inharmonic distortion. One of the most expensive convolution processors comes up very short to my ears (it sounds quite "digital"), though others have reported positive results with it.

I've tried several of these simulation/emulation/harmonic generator boxes, and the only one that does it transparently and with no downside is the HEDD-192. And only at 96 kHz, by the way. At lower sample rates,  the HEDD produces some fold-down alias distortion which shows up as "digititus". And in fact, that "digititus" resembles the problems I've heard with the competing units.

If you wish to test such units, be sure to test them at higher sample rates. I've got some explanations for this in my book.

Quote:

I think that presence (the most commonly cited trait of valves) is very closely related to depth, only presence is the nearest end of the depth, if you see what I mean.

This presence is absolutely related to the device's harmonic signature. I am convinced. Other factors include dynamic response. But the amount and types of harmonics and the device's dynamic response cannot be divorced. For example, an analog tape machine saturates as the level goes up, which is a form of compression, and at the same time harmonics (predominately 3rd harmonic) are added. It can be shown that the very process of adding 3rd harmonic distortion, and doing it in a level and high frequency-dependent manner, yields a dynamic response very similar sonically to that of an analog tape machine.

Quote:

Also, I am guilty of the cardinal sin of engineering: I always expect the recorded signal to sound BETTER than the damn thing did in the liveroom! And with tubes it does!

Amazingly, it can. It's a form of aural exciter, by the way. Creative use of distortion is largely how rock and roll is made (in a nutshell)!

[ted nightshade]

bobkatz wrote on Sat, 10 July 2004 07:31

andy_simpson wrote on Sat, 10 July 2004 06:52 You've hit the nail on the head regarding depth and dimension. Those are both big factors in valves sounding more 'life-like' (or should I say 'larger than life'?) than transistors. I truly believe that harmonic proportions has 99% of the explanation of what's going on here. I've written about it extensively in my book and demonstrated it as well. The Cranesong HEDD does a fantastic job of emulating various tube and analog tape characteristics without the down sides. As you add harmonics within the Cranesong, the depth image and other "tubelike" wonders, including that "livelike" quality starts to occur. But MORE is not necessarily better. Adding harmonics reduces the peak to average ratio of a recording and it is DEFINITELY a form of compression. A little bit goes a long way and too much sugar in the sound becomes diabetes. I've carefully compared using the HEDD-192 digitally versus various tube processors (or a dub to analog tape), and in most cases going digitally and adding the harmonics yields far more transparent audio than going through the analog device.

I believe that the HEDD-192 works for you that way in mastering, or in mixing, where you are adding a D/A/D conversion by going to analog and back, but what I found in tracking, was that the "resolution" (dimension, detail, etc.) of a recording made through the HEDD A/D was reduced significantly even by putting the process out of "bypass" into "process".

Now some "extra texture" could be invoked with the use of the remarkable HEDD process... but getting that extra texture during tracking via fine tube mics and pres (including the SLAM! hybrid tube pre), and most important- NOT LOSING IT through some routine analog degradation (weak cables, connections, unnecessary circuitry of any kind really) on the way to A/D, achieved a result in depth, dimension, harmonic subtleties, that the HEDD process could not achieve.

----
Negative feedback is an interesting thing indeed- I have a couple Manley devices, mics and amps, that let you adjust the feedback to some extent. All these are choices between very small amounts of feedback compared to solid state amps, but I definitely have a real taste for a little bit of negative feedback on a tube circuit, unless I'm going for "brawny" rather than "most dimensional, cleanest, most detailed". Brawny is a presence thing, but not just in the usual EQ range associated with presence- I really prefer the 440 monoblock tube amp as a bass amp with minimum negative feedback. Brawn!

On the other hand the SLAM! is a no-negative-feedback design and has brawn, dimension, detail, the works. Very appealing way of mixing the best of FET, tube, transformer, and digital. Clearly they all have strengths to offer the designer and user.
[/quote]

[bobkatz]

ted nightshade wrote on Sat, 10 July 2004 12:46

Now some "extra texture" could be invoked with the use of the remarkable HEDD process... but getting that extra texture during tracking via fine tube mics and pres (including the SLAM! hybrid tube pre), and most important- NOT LOSING IT through some routine analog degradation (weak cables, connections, unnecessary circuitry of any kind really) on the way to A/D, achieved a result in depth, dimension, harmonic subtleties, that the HEDD process could not achieve.

Certainly you're right... even the world's best D/A/D in the chain creates a transparency loss. So you're faced with a tradeoff if you're mixing via analog and all the wonderful analog tools we have at our disposal are likely to sound more transparent (better) than going D/A/D, for a long time to come. Someone mentioned the sound of a transformer; mixing on an API console does add a certain kind of magic to the sound. Some of it is supposed to be the distortion of the transformers. Wouldn't it be nice if we knew exactly what it is that's doing it? As I said, as we enter more and more into the "mix all digital" world, it's time that we find out. I believe that eventually all the "magic" will be encapsulated into transfer functions; it's only a matter of time; 5 or 10 years and there will be enough power to transparently emulate what we have.

Please don't call me a "faithful guileless defender of digital". I've got ears and I haven't heard much (yet) in the emulation department that meets the above promise---all I'm stating is the obvious extension of where we're going. In another forum here they've been depating the sound of the real Rhodes piano versus the simulator. Bottom line: even if it's now only at 95%, sooner or later it will be a 100. Only a matter of time.

[ted nightshade]

Well I'm all for it. Grateful to those on the bleeding edge!

[Andy Simpson]

Yeah, the tube is definately a form of compression, but it's so much more.

I think that when viewed as a soft-knee'd compressor, tubes have an INFINITELY undefined threshold and therefore are infinitely soft on the attack (of the attack) when they reign in the signal, which is much softer on the ear compared to the relatively on/off of a compressor.

Imagine catching an egg in a frying pan, and catching an egg in a pair of nylon womens tights (is that what they're called in US?). The tights slow the egg down gradually, acting over a long threshold, but the egg stops very quickly on the frying pan when it reaches the threshold and breaks.
An old chinese philosophy

I believe that it is this 'rubber-band' attack compression that makes it sound much better than transistor+compressor/exciter.
The soft pull only rounds the signal and adds the odd harmonics - whereas the compressor squares the signal to an extent and adds even harmonics too (even if they're relatively HF).

I also believe that the valves depth comes from the 'stretching' of the compression 'threshold', ie. the dynamic of the signal around the sweet-spot - the dynamic that is INSIDE the actual signal, NOT INCLUDING the entire dynamic of the _gain structure_, is much increased in perception, whereas the dynamic in the noise floor and headroom is reduced to make room.

To sum up my theory, a valve, by 'soft-threshold' compression, focusses more of it's dynamic range in the actual signal, rather than the entire gain picture, deepening the dynamic range of the actual signal so that the signal and it's harmonics appear farther 'apart', but moving the harmonics closer to the listener and the fundamental further away towards the noise floor (infinitely?).

To illustrate, imagine a vocal sound, peaking at -5dB. The signal and it's harmonics can be said to be between -5 and -10dB, with the noise floor at -80dB and a gap of 70dB between the signal and the noise floor.
After the valve, the signal is still peaking at -5dB, but it's fundamental is now at -20dB (and is now 10dB closer to the noise floor). The harmonics are stretched by 10dB between -20dB and -5dB.

I'm not sure if I'm talking complete shit here guys.

You may have noticed that I've used the word 'believe' several times in this post, so please arm yourself with a pinch of salt before pondering this too deeply!

Btw, I'm using dynamic mic->tubes->A/D, it's the best sound I've ever got. Period. Transparent to say the least.

Also, as a side question, is it possible that dynamic mics deliver more dynamic _inside_ the signal than a condenser?

Andy

[Andy Simpson]

Further to my last post;
I think that the valve widens the dynamic inside the signal by 'catching' the fundamental and allowing the harmonics to 'slip through'. In a nutshell.

Sort of a fundamental-compressing dynamic-expander?

This is screwing my head up.

Andy

[Bob Olhsson]

A lot of '60s era pro tube mike preamps clipped at +35 to +40 while many of the first solid state mike pres clipped at +18-+24. Meanwhile the tape was saturating around 14 dB above operating level. In many cases the old tube preamps weren't being pushed all that hard while the solid state stuff was often clipping most of the time.

This made recording with tube gear feel like walking on solid ground compared to walking on thin ice with a solid state board. At Motown in Detroit we had both a tube studio and a solid state room so I got to go back and forth with the same tapes several times a week.

[Andy Simpson]

This is fascinating.
How did this affect the overall picture of the gain structure?

We're not just talking about pre-amps.....presumeably the mix buss amp, line amps/tape returns, etc were all tube as well?
Not to mention the tape machines themselves.

Did you ever mix tube-desk recordings on the solid state desk? (Or vice versa?).

Cheers.

As a side note, to a rare person (who recorded with tubes in the 60's and can talk about it!), did you find you had to roll-off the bottom end all the time because the tube saturation brought it up so much? Or did the pre's do it as part of their circuitry? OR was this done on the tape bias curves perhaps?

My experience of tubes (when driven nice and hard) is that I ALWAYS have to roll about 10/15dB of the bottom off for clarity. No exceptions.

Andy