Re Miniman and 6vheater. MiniMan is correct on 'ageing'. I can tell you that as an absolute fact. When working at AWA Rydalmere decades ago one of my jobs was ageing valves inside a great magnetic coil. It groaned very loudly like an armature growler plus 12 dB or so. The valve 'flashed'.
Valves went out into the field ready to operate interchangeably. There were not the fantasised/actual, more recently, significant differences in tube output....Quality assurance was very strict from whoa to go.
It may be that the event or quality of ageing has 'gone'...I can't say however Tektronix placed their valves under stress for ...was it 50 days?...I can't exactly recall...and then flicked the ones not making spec.
Sure, valves will deteriorate during storage, even leak however running them a while then checking them in service can be done using recent Japanese 'matching' devices used for bias adjustment.
The 'audiophile' obsession with old valves...and I have hundreds of them...may see deterioration but then even matched valves will suffer performance 'de-matching' through all the other aspects of the components. The fantasising on valve performance per matching/per manufacturer with older valves has long reached the point of super- saturation.
I have a tester modified to read individual triodes and my EICO offers that ability plus some more. When I test valves I work on comparisons rather than just 'numbers'. One might argue that's a form of 'matching' and I suppose it is but my intention is to separate the sheep from the goats....and whether the goats actually perform 'worse' or 'better' than the other, is moot. You get two valves which go equally 'off' the gM scale for example....does that make them the top performers?..it may.. but it's putting valves
into their specific, functioning circuitry which gives the opportunity to measure performance . The old military service manuals, for example gave tables of correct measurements of all their valves, after controls properly set and using specific test equipment but even when voltages were within 'spec' ...was it the valve causing the problem, maybe an intermittent inside it....which may show up only with the equipment being checked, in a certain position or under vibration?... or some intermittent in a component or.....My NOS 807's and 811's and 12AX7's and others which will all be tested where possible and pretty-soon sold will not fix equipment performance where components are the causes. MyART-13 Transmitter is probably the ideal tester for some of that gear (smiles0 but as we age, as our moods change, environments change things we are used-to sound 'different' . Was it the valve change, or the better speakers?...or the repositioning of audio gear or furniture in that room or another, for example, which makes my gear sound better....or worse.
The proof of the pudding is always in the eating but rejecting (audio-line-up) valves even on 'hearing' may simply be you go on until finding a valve which temporarily makes-up for circuit differences. If using matched valves, everything else has to be matched but even then values change and not consistently...one might go up 15% another down 12%..or 5%...A fully metered amplifier gives some hope even if the argument is 'what I'm hearing', the objective and the subjective juxtapose. My regards
Valves went out into the field ready to operate interchangeably. There were not the fantasised/actual, more recently, significant differences in tube output....Quality assurance was very strict from whoa to go.
It may be that the event or quality of ageing has 'gone'...I can't say however Tektronix placed their valves under stress for ...was it 50 days?...I can't exactly recall...and then flicked the ones not making spec.
Sure, valves will deteriorate during storage, even leak however running them a while then checking them in service can be done using recent Japanese 'matching' devices used for bias adjustment.
The 'audiophile' obsession with old valves...and I have hundreds of them...may see deterioration but then even matched valves will suffer performance 'de-matching' through all the other aspects of the components. The fantasising on valve performance per matching/per manufacturer with older valves has long reached the point of super- saturation.
I have a tester modified to read individual triodes and my EICO offers that ability plus some more. When I test valves I work on comparisons rather than just 'numbers'. One might argue that's a form of 'matching' and I suppose it is but my intention is to separate the sheep from the goats....and whether the goats actually perform 'worse' or 'better' than the other, is moot. You get two valves which go equally 'off' the gM scale for example....does that make them the top performers?..it may.. but it's putting valves
into their specific, functioning circuitry which gives the opportunity to measure performance . The old military service manuals, for example gave tables of correct measurements of all their valves, after controls properly set and using specific test equipment but even when voltages were within 'spec' ...was it the valve causing the problem, maybe an intermittent inside it....which may show up only with the equipment being checked, in a certain position or under vibration?... or some intermittent in a component or.....My NOS 807's and 811's and 12AX7's and others which will all be tested where possible and pretty-soon sold will not fix equipment performance where components are the causes. MyART-13 Transmitter is probably the ideal tester for some of that gear (smiles0 but as we age, as our moods change, environments change things we are used-to sound 'different' . Was it the valve change, or the better speakers?...or the repositioning of audio gear or furniture in that room or another, for example, which makes my gear sound better....or worse.
The proof of the pudding is always in the eating but rejecting (audio-line-up) valves even on 'hearing' may simply be you go on until finding a valve which temporarily makes-up for circuit differences. If using matched valves, everything else has to be matched but even then values change and not consistently...one might go up 15% another down 12%..or 5%...A fully metered amplifier gives some hope even if the argument is 'what I'm hearing', the objective and the subjective juxtapose. My regards
I put it in my car, or in my ear, both even smaller than my toilet. With appropriate transducer, I hear 30Hz just fine. In fact it is easier to make big bass in a small space than a large space.
"Bass gain" in cars is a well known effect. A couple 6x9s is wimpy in the livingroom but can be good boom in a car. Yes, a 23' wave in a 7' interior and 4' trunk.
Bass in headphones or ear-plugs is mostly limited by leakage. At the audiologist the tympanometer applies a very low-frequency wave as essentially "static pressure" to see if your ear-drum is stuck.
We do not have to "make the wave fit in the room". We only need a variation of pressure.
Last edited:
Sorry this is just plain wrong.
I have tested this problem extensively and measured the point where reflexions finally add up to a distance long enough actually to hear the true bass notes.
It's nowhere inside the listening space of a (my typical) room.
What you think you are hearing, and what you are actually hearing are 2 totally different things.
As speakers (and headphones btw) produce masses of distortion at ultra low frequencies, they become inherently ultra rich in harmonics.
The more cone movement, the higher the THD, so the smaller the cone area the worse it gets.
Try actually MEASURING the waveforms from speakers at those freqs with proper measurement mics. You will be in shock to see what's actually going.
You are hearing harmonics not (at min doubling), nothing evenly close to fundamentals.
That also stands for headphones btw.
Have you looked at the impulse and sq wave response of so called "hi end" headphones costing 1000s.
They are just dreadful, the only exception being extremely rare true ESL headphones.
In fact I never fail to underwhelmed by the current headphone offerings.
A cruel world,- once you start looking at stuff properly up close.
Anyway lets not spoil filament aging with a few arguments about what you can and can't hear.
s
ANSWER nobody knows, but the speaker designed by Kellogg has great difficulty reaching the resolution of even 16 bit audio, whereas ESL can reproduce sq wave pretty cleanly, as well as excellent transient response.
So, it's unlikely you have ever been presented with a square wave test that's any good anyhow.
The human ear can detect the difference between 16 and 24 bit audio, so I tend to trust it a lot more than marketing BS, and PR speel.
+
It's proven the human ear is very good at handling even sharp loud transients without going into overload or temporary deafness.
orchestral musicians have to handle that daily.
The day your sound reproduction systems can reproduce an uncompressed cymbal crash (which has content going up to 60khz) means it actually can be termed hifi.
The quad ESL of 50-60yrs ago could actually do that trick, as could some KLH equipped with Jantzen ESL tweeters.
Some people like to kid themselves, physics laws have changed in 50 yrs.
ANSWER nobody knows, but the speaker designed by Kellogg has great difficulty reaching the resolution of even 16 bit audio, whereas ESL can reproduce sq wave pretty cleanly, as well as excellent transient response.
So, it's unlikely you have ever been presented with a square wave test that's any good anyhow.
The human ear can detect the difference between 16 and 24 bit audio, so I tend to trust it a lot more than marketing BS, and PR speel.
+
It's proven the human ear is very good at handling even sharp loud transients without going into overload or temporary deafness.
orchestral musicians have to handle that daily.
The day your sound reproduction systems can reproduce an uncompressed cymbal crash (which has content going up to 60khz) means it actually can be termed hifi.
The quad ESL of 50-60yrs ago could actually do that trick, as could some KLH equipped with Jantzen ESL tweeters.
Some people like to kid themselves, physics laws have changed in 50 yrs.
I'm pretty sure you'll find the laws of physics are constant...
16 bit audio means, you can have 65536 speaker positions from -32768 to +32768. 24 bit makes that number jump to 16.7 million so it's understandable you'd hear the difference. I read somewhere that in comparison, a vinyl record approaches 750kbit. Sampling rate is more important IMHO. I'd take 16 bit @ 96 kHz over 24 bit @ 44.1 kHz any day.
16 bit audio means, you can have 65536 speaker positions from -32768 to +32768. 24 bit makes that number jump to 16.7 million so it's understandable you'd hear the difference. I read somewhere that in comparison, a vinyl record approaches 750kbit. Sampling rate is more important IMHO. I'd take 16 bit @ 96 kHz over 24 bit @ 44.1 kHz any day.
Surprisingly enough, this isn't really true, although you'll read it endlessly on the Internet. Bandlimited and dithered A/D/A conversion is (with idealized converters) mathematically an exact copy of the input plus a delay, plus a small noise (mostly the dither itself). There are no "stairsteps" despite what many people still say three decades down the road. They're wrong.
All good fortune,
Chris
It's an interesting comment Jazid. Sometimes seen with valve/tube testing., by coincidence last time was a 6VgGT...With more than reasonable warm up testing might show (forget gM) say a constant 60%. Get called away to the phone, leave tube alive in the socket... and comeback 10 minutes (e.g.) later test again ...now it's 70%'.
'
Perhaps feeling useful again gets the old tube ripping along..."I'm here, I'm here...please don't bin me...I've had my tin of spinach while you were away"....and the Tarzan jungle call issues from its pin 3 area....Ok brother...you're on!!...just leave Jane and the chimp at home...ok?
'
Perhaps feeling useful again gets the old tube ripping along..."I'm here, I'm here...please don't bin me...I've had my tin of spinach while you were away"....and the Tarzan jungle call issues from its pin 3 area....Ok brother...you're on!!...just leave Jane and the chimp at home...ok?
Somehow Jazid my observation on yours went to the wrong area....I'll try again. It's an interesting comment Jazid. Sometimes seen with valve/tube testing., by coincidence last time was a 6V6GT...With more than reasonable warm up testing might show (forget gM) say a constant 60%, test aftertest. Get called away to the phone, leave tube alive in the socket... and come back 10 minutes (e.g.) later test again ...now it's 70%'.
'
Perhaps feeling useful again gets the old tube ripping along..."I'm here, I'm here... You want audio...I'll give you audio...please don't bin me...I've had my tin of spinach while you were away"....and the Tarzan jungle call issues from its pin 3 area....Ok brother...you're on!!...just leave Jane and the chimp at home...ok?
'
Perhaps feeling useful again gets the old tube ripping along..."I'm here, I'm here... You want audio...I'll give you audio...please don't bin me...I've had my tin of spinach while you were away"....and the Tarzan jungle call issues from its pin 3 area....Ok brother...you're on!!...just leave Jane and the chimp at home...ok?
No, this is true for all A/D/A conversions. The only difference between dithered 16 bit and dithered 24 bit is noise floor level. Dither operates at the LSB, so both are small enough for delivered home audio. Audio production benefits from a comfortable professional margin, so "24 bit" - not actually attainable - is preferred.
Sampling rate comes into play because signal must be bandlimited to half of the sampling rate and dithered before any bit depth reduction including the original A/D and all mathematical operations, and the bandlimiting itself (a steep low-pass filter) causes artifacts near the band limit.
Sorry to be so off topic, but this is a common error that needs constant correction.
All good fortune,
Chris
Sampling rate comes into play because signal must be bandlimited to half of the sampling rate and dithered before any bit depth reduction including the original A/D and all mathematical operations, and the bandlimiting itself (a steep low-pass filter) causes artifacts near the band limit.
Sorry to be so off topic, but this is a common error that needs constant correction.
All good fortune,
Chris
Last edited:
Thanks for the information. I assume everything is dithered now? Was 8 bit dithered? I started out with 8 bit, and you can certainly tell the difference between that and 16 bit Still, much like I like the sound of big band music from a 78 better than live, I still love 8 bit sound
Still, much like I like the sound of big band music from a 78 better than live, I still love 8 bit sound
Last edited:
I agree with kodabmx's statement: "I'd take 16 bit @ 96 kHz over 24 bit @ 44.1 kHz any day."
A flute plays a note, C7, 2,096 Hz.
At 44.1k samples/second there are:
21 samples of the fundamental (2,096 Hz)
10.5 samples of the second harmonic (4,192 Hz)
7 samples of the 3rd harmonic (6,288 Hz)
and 5 samples of the 4th harmonic (8,384 Hz).
Tell me how many bits of resolution you get with only 21 samples of the fundamental note of 2,096 Hz,
(not very many).
Well, they say that the digital brick wall filter at 22.05 kHz smooths all of that out, wonderful.
Yes, it is so smooth, but a square wave has Pre ring and Post ring at the up transition, and Pre ring and Post ring at the down transition.
That may not be anything to worry about, unless you care about group delay (different relative phase shift angles versus different frequencies).
A flute plays a note, C7, 2,096 Hz.
At 44.1k samples/second there are:
21 samples of the fundamental (2,096 Hz)
10.5 samples of the second harmonic (4,192 Hz)
7 samples of the 3rd harmonic (6,288 Hz)
and 5 samples of the 4th harmonic (8,384 Hz).
Tell me how many bits of resolution you get with only 21 samples of the fundamental note of 2,096 Hz,
(not very many).
Well, they say that the digital brick wall filter at 22.05 kHz smooths all of that out, wonderful.
Yes, it is so smooth, but a square wave has Pre ring and Post ring at the up transition, and Pre ring and Post ring at the down transition.
That may not be anything to worry about, unless you care about group delay (different relative phase shift angles versus different frequencies).
As to hearing a 30Hz note, I feel "pressured" to say that it is not necessary to have a standing wave that relates to the wavelength of that frequency (no "reflection" on me).
But there is a law in physics.
Start with a specific container size (i.e. a listening room), gas temperature, and gas (air) pressure.
Now, vary any one of those, and at least one of the other quantities will vary.
When an acoustic suspension woofer cone moves, the space for the gas changes.
I like to think that the pressure in the room changes. Great, we have a 30 Hz pressure wave, and we hear (and feel it).
I'm just sayin'
But there is a law in physics.
Start with a specific container size (i.e. a listening room), gas temperature, and gas (air) pressure.
Now, vary any one of those, and at least one of the other quantities will vary.
When an acoustic suspension woofer cone moves, the space for the gas changes.
I like to think that the pressure in the room changes. Great, we have a 30 Hz pressure wave, and we hear (and feel it).
I'm just sayin'
Surprisingly, it only takes 2 samples per cycle (of bandlimited and dithered signal) to perfectly reproduce the original (idealized converters, yada, yada).
The artifacts of the bandlimiting are, of course, a whole different critter.
All good fortune,
Chris
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.