I have just done that..spending most of a whole day listening to an entire system which has been done with exactly that dedicated philosophy, costing a fortune and all singing all dancing, so the idea I might be OT is utter nonsense.
The speaker drivers were supposed to fill that exact goal, batch specially made drivers, frames made of pure brass, ss amp using exactly what you described
Had exactly the same speel as you just put out..."can be done bla bla" we will show you how to do it, nothing can beat it, even everyone else is stupid or mediocre...your microphones are crap, + you are really dumb to use a valve based amp system....
conclusion?
It sounded dreadful and hurt my ears.
Totally underwhelmed, incapable of reproducing St Sulpice Organ, really lacking in bass response, and violently shouty mid range.
For our own interests,-
In the next few days we will at least attempt a shoot out in our opera house (behind the scenes) using DPA mics.
Modern 21st century DSP based SS amp specially matched to the drivers it is supposed to use:-
v
1955 vintage valve amp I just rebuilt, - golden age transformers (known to be one of the best ever made), not matched really at all to the drivers, but with a Baxandall type tone stack to introduce small corrections....
Output into hard disk recorder, then spectral and FFT analysis, and of course some reasonably well trained ears, because that's what we use for recording every day.
My take on all this, is the unique non linear reponse generated by magnetic saturation characteristics of the valve transformers of the so called "golden age".
It makes sense.
This is exactly the way magnetic tape heads work, again with a marked tendency for soft clipping.
I never saw people trying to drive tape heads with current....
I looked for something other than Nelson’s specified damping factor of 3 which is an output impedance of about 2.5Ω. You have the kit, can you measure one?
dave
Any device has its own peculiarity and usually each design offers something unique to the sound.
There are many systems. I’ll counter with a frugal-phile™ system that follows the philosophy and sounds really good.
dave
A modulated current source is the best way IMO but the downside is it must be class A. I haven't come across a better way to do it though.
This can be emulated with transistors for the same affect with the same transfer curves.
What are the specifics of designing a current driven driver vs a voltage driven driver?
One thing is sure,
is a classic oxymoron case. Sound like what?
All valves sound different, larger or smaller IMD components as Jack will tell you.
I would hesitate a long time, let you chew a lot of gum, and eat masses of popcorn while you attempt to convince him of all the spurious arguments here.
I'll counter by adding, above was designed by you, therefore your idea
is about as objective as saying "fish and chips" is the best meal in the world.
Whether or not other people like it, becomes irrelevant to grandstanders, only your opinion appears to matter.
( I just had a triple dose of all that last week, so I got severe opinion allergy.
)Eg.
Our microphone manufacturers much prefer our approach, which is to test everything, compare like with like, measure stuff, & come up with VALIDATION.
Validation strategies work, because they rely on proven industry standards.
They do tend to be somewhat objective.Ie.
I don't believe a word you say.
My initial argument was that tubes have no unique sound in a well designed circuit compared to an equally well designed solid state circuit.
Therefore, there is no point in using tubes in a super triode configuration.
If anyone disagrees let me know. But the arguments against it so far have not been convincing at all.
Therefore, there is no point in using tubes in a super triode configuration.
If anyone disagrees let me know. But the arguments against it so far have not been convincing at all.
I really hate the entire idea that an amplifier design must be considered as part of a system. What if it isn't and it is a damn fine amplifier? That line of thought is incompatible with real life and only serves the manufacturer or dealer of a "system" complete. We all know that isn't how the world works, and some of the worst equipment I have seen was designed as a "system".
This thread is all about if tubes have a sound and that's it. It does not take into account an amplifier designed to run a set of speakers, and that cancels out when the consideration is changing tubes or circuit design.
Talking about a speaker - amplifier system is distinctly off topic and belongs in it's own thread. Too bad this wasn't caught earlier so that the OP could actually discuss what they intended to discuss instead of the hijacked side discussions.
How about we return to the actual intended topic. Speakers and how they interact are not a rational consideration to the talk about the sound of tubes.
-Chris
This thread is all about if tubes have a sound and that's it. It does not take into account an amplifier designed to run a set of speakers, and that cancels out when the consideration is changing tubes or circuit design.
Talking about a speaker - amplifier system is distinctly off topic and belongs in it's own thread. Too bad this wasn't caught earlier so that the OP could actually discuss what they intended to discuss instead of the hijacked side discussions.
How about we return to the actual intended topic. Speakers and how they interact are not a rational consideration to the talk about the sound of tubes.
-Chris
Validation strategy?
Measurement data v blind testing.
Hearsay 1,-
Btw, it's well known, amplifiers measuring very low distortion can sound evil, while amplifiers, particularly certain topologies which don't neccessarily measure so well, (I am not a DHT freak!), can sound lively and clean. Prove it!
Hearsay 2,-
It's possible to design an algorithm for DSP that imitates "valve sound", (I don't think so,-what valve sound??!). Some guitar amp design nuts have gone this route without noteable success.
Hearsay 3,-
Damping factor arguments in all their worn out clothes.
Hearsay 4,-
Who's say so on what constitutes a "well designed circuit". I don't buy that, cos it's usually on the back of either "snake oil" or some other marketing ploy, or Chinese copies of same.
Last edited:
Sorry that's utter nonsense.
You only have to see what Peter Walker's QUAD did as a manufacturer of amplifiers AND speakers.
The Quad ESL is one of the very few speakers that can actually reproduce & resolve a square wave,- or make a very good crack at it.
Quad initially sold it with a valve amp, and the systems were typically used in ABBEY ROAD and the BBC.
Hey guys do better!
It's well known a typical ESL like that, resembles a short circuit at HF.
One of the best lessons you can learn from systems put together there for over 5 decades, is precisely the question relevant to the discussion, whether or not you claim hating or liking amps as part of systems.
Does a HIGH resolution loudspeaker (QUAD ESL exceeds 16bit resolution), sound better with a valve amp driven from that manufacturer or the ss version?
We are about to do a similar but even more demanding test today.
When the flag drops the BS stops.
My "validation" is stated in my posts on the first page. Mostly listening comparisons through experiments and such. I use HD800 headphones for all my tests in order to put the amp under near ideal conditions and to avoid room variables.
During a variety of tests I was unable to get tubes to sound euphonic or special in a "good" way when using them in an amplification stage or an OTL output stage.
Only varying degrees of good or bad in the objective sense.
My solid state circuits have unmeasurable distortion but sound the same as the tubes when the tubes are put in a good circuit.
I would like to get tubes to produce "euphonic" sound but when making them distort they just sound crappier, not more euphonic.
In my definition a "well designed circuit" is one that solves as many conceptual problems as possible in order to make it as perfect as possible within the constraints of the topology. I don't mess around in this regard.
So in context of tube vs solid state, a "well designed" circuit is one that is designed as linearly and ideally as possible.
During a variety of tests I was unable to get tubes to sound euphonic or special in a "good" way when using them in an amplification stage or an OTL output stage.
Only varying degrees of good or bad in the objective sense.
My solid state circuits have unmeasurable distortion but sound the same as the tubes when the tubes are put in a good circuit.
I would like to get tubes to produce "euphonic" sound but when making them distort they just sound crappier, not more euphonic.
In my definition a "well designed circuit" is one that solves as many conceptual problems as possible in order to make it as perfect as possible within the constraints of the topology. I don't mess around in this regard.
So in context of tube vs solid state, a "well designed" circuit is one that is designed as linearly and ideally as possible.
Last edited:
And you measured the headphones??
Have curves for them?
In my experience (knowing the Senns intimately) that is about the weakest validation argument I have come across, adding which I am constantly moaning of inbalance between H-P drivers, throwing the mix left or right.
Not only are headphones exceedingly non linear, but they have mountains of faults specific to the Sennheiser range, which certain people have spent considerable time correcting.
It is highly possible in view of the curves and transient response they are actually incapable of the resolution required to make a critical judgment, - an essential reason why studio work is regularly invalidated using headphones.
Our professor was very very much against their use in live recording environments.
We were doing exactly that sort of comparison last night recording a Beethoven & Brahms concert with 10 or more microphones involved, and a shift of the orchestra physical placing between pieces. (ouch!)
We can't even validate a concert set up properly without resorting to multiple blind testing & switching between monitors and multiple headphones.
How are you supposed to be able to quantify amp performance at below 24 bit resolution?
It basically doesn't make sense and it's not VALID.
Last edited:
This is the FR curve of the HD800 but the absolute linearity of the headphones is irrelevant when they are being used as a standard reference for relative comparisons. Headphones are an audio microscope when compared to loudspeakers and don't have room reflections to interfere with assessment, that is the only point that matters. Even if that wasn't true it would be irrelevant in relative comparisons.
Last edited:
We are talking theoretically here. Neither of the 2 designs will (yet) sound the same. The 2 designs are coming toward sthe same goal but from different directions.
dave
- Status
- Not open for further replies.