Nico,
I understand that many people do not listen to amplified music in a live situation, though many times there can be some sound reinforcement in large venues. Since you are recording your own source perhaps a kettle drum would be a better test of the worst case scenario for you personally? It doesn't sound like you will be listening to any pipe organs or train sounds so that would push your amplifier to your limits better than an acoustic guitar I would imagine. I do get where you are coming from, it is a different perspective than Tsiros who obviously listens to more pop music. On another note, Tsiros, I have seen direct feeds used for electric guitar and they weren't going through a traditional amplifier before being recorded. It is still in the active electronic realm, but not after going through a guitar head and speaker......
Steven
I understand that many people do not listen to amplified music in a live situation, though many times there can be some sound reinforcement in large venues. Since you are recording your own source perhaps a kettle drum would be a better test of the worst case scenario for you personally? It doesn't sound like you will be listening to any pipe organs or train sounds so that would push your amplifier to your limits better than an acoustic guitar I would imagine. I do get where you are coming from, it is a different perspective than Tsiros who obviously listens to more pop music. On another note, Tsiros, I have seen direct feeds used for electric guitar and they weren't going through a traditional amplifier before being recorded. It is still in the active electronic realm, but not after going through a guitar head and speaker......
Steven
Surprisingly, pop is the only music that i do not listen to, nor practice at all.
yes i've seen direct injection, and done recordings like that, but it's never kept like that till the end result. It's always either going through an amp/sim or reamped later. Reamping especially is something like a god's gift. The solid body electric guitar needs a cabinet simulation at least because it doesn't have any natural resonance to write home about.
Actaully you are waffling a load of crap which has nothing to do with the discussion. People like yourself wants to be heard but has nothing constructive to offer. You have contributed absolutely zip.
This is my thread and I suggest you start a thread of your own choosing a topic that interest you and have a debate.
This is my thread and I suggest you start a thread of your own choosing a topic that interest you and have a debate.
you attack me but you don't explain what i did wrong. an ad hominem is not a valid argument.
On topic,
Not really. headphones do not have the characteristic impedance curve of loudspeakers, so the load that the amp sees is very different. Their impedance is almost perfectly ohmic. No inductance.
Guys, let's keep it civil and to the topic. No-one "owns" a thread in the noncommercial parts of the forum. All technical points of view are welcome and encouraged.
This is the truly marvellous thing about sound reproduction, that it is capable of being superior to what is heard live, for the simple reason that the mic's, etc are usually set up in the absolutely optimum position and environment to pick up all the subtleties of the sound. Something we humans can rarely do in real life at a performance, no matter how much money is paid for the ticket. I first realised this over 20 years when I listened to a piano recital at the Sydney Opera House, and thought the sound was a bit lacklustre, not firing on all cylinders.
But sorry, Nico and Andreas, this was in comparison to a highly tweaked, pure digital setup. Analogue can do it too of course, but there is nothing instrinsic in CDs that stops it happening there. The simple answer is that convincing digital playback requires a lot of fussiness, being anal about everything. Analogue is quite tolerant of plenty of sloppiness in the playback chain, pleasurable sound emerges even if things are not quite right.
An analogy would be that analogue is like the steering of a 60's car. You could rock the top of the steering wheel a good 6 inches or more side to side and not much happens, the car still goes pretty well straight down the road. Try that on a modern, high performance vehicle and you'll go straight into the ditch! That's what digital is about ...
Frank
SY,
I profusely apologize to anyone who objects listening to a 50s recording of Dave Brubeck on a 78 rpm record with no certification or reference made that it was recorded using a dummy head and absolutely compatible with stereo equipment sixty plus years in the future.
That was a total un-audiophile thing to do and I realize that I should be banned indefinitely from every audio forum in the universe until such time that I can prove beyond any doubt that I have been rehabilitated.
I also apologize for insisting that a few notes played on an acoustic instrument and recorded onto analogue tape could qualify as a recording - it actually is nothing and it should be impossible to hear anyway.
And finally for mentioning that the recording was using no mic, in stead of no pick-up which would have invalidated the proposal for some another reason I suppose.
The idea was purely to compare the change in characteristic of a particular sound before and after incrementing reservoir capacitance as has been discussed and simulated on this thread.
I did not know that recording an instrument directly without an interfacing speaker onto tape was invalid and that hearing something afterwards was a only a figment of my imagination.
Tsiros thank you for pointing out that my PC and sound-card operating at a max of 48 kHz sample rate and 96kbps would far out perform an analogue recorder using an infinitely high sampling frequency and infinitely high bit rate. Heck, I cannot not even strip the Studer for spares as they are pre-digital era and probably rubbish.
Tsiros thank you for pointing out my erroneous ways I vow to convert.
I profusely apologize to anyone who objects listening to a 50s recording of Dave Brubeck on a 78 rpm record with no certification or reference made that it was recorded using a dummy head and absolutely compatible with stereo equipment sixty plus years in the future.
That was a total un-audiophile thing to do and I realize that I should be banned indefinitely from every audio forum in the universe until such time that I can prove beyond any doubt that I have been rehabilitated.
I also apologize for insisting that a few notes played on an acoustic instrument and recorded onto analogue tape could qualify as a recording - it actually is nothing and it should be impossible to hear anyway.
And finally for mentioning that the recording was using no mic, in stead of no pick-up which would have invalidated the proposal for some another reason I suppose.
The idea was purely to compare the change in characteristic of a particular sound before and after incrementing reservoir capacitance as has been discussed and simulated on this thread.
I did not know that recording an instrument directly without an interfacing speaker onto tape was invalid and that hearing something afterwards was a only a figment of my imagination.
Tsiros thank you for pointing out that my PC and sound-card operating at a max of 48 kHz sample rate and 96kbps would far out perform an analogue recorder using an infinitely high sampling frequency and infinitely high bit rate. Heck, I cannot not even strip the Studer for spares as they are pre-digital era and probably rubbish.
Tsiros thank you for pointing out my erroneous ways I vow to convert.
Last edited:
Frank, no disrespect meant but if you are listening to anything and I gate it do you think that the result is the same as the original, not a chance.
This argument has done its rounds and and will never go away because it was economics that forced everything to be digitized simply because humans have become tolerant to what has gone missing between black and white.
In a digital car as you mention, depending how closely you can mimic analogue depends how accurately you can steer.
If you only have left and right (1 & 0), going straight is a tri-state which does not exist in digital and it will depend on whether you want to be in the trees on the left or that on the right side of the road.
However if you can oscillate the steering between left and right the average might approximate a straight line. But if you stop for any length of time obviously you end up in the trees again.
This argument has done its rounds and and will never go away because it was economics that forced everything to be digitized simply because humans have become tolerant to what has gone missing between black and white.
In a digital car as you mention, depending how closely you can mimic analogue depends how accurately you can steer.
If you only have left and right (1 & 0), going straight is a tri-state which does not exist in digital and it will depend on whether you want to be in the trees on the left or that on the right side of the road.
However if you can oscillate the steering between left and right the average might approximate a straight line. But if you stop for any length of time obviously you end up in the trees again.
On the surface, the idea of telling if there's enough capacitance seems complex.
It certainly is not. Gorgeous bass H1/H2 balance = Their caps are big enough (aka the transformer is big enough).
Call Me Maybe - Carly Rae Jepsen (Disco Cover by Stacks of Wax f/ Nicholas Wells) - YouTube
After the fact, when its right, that's obvious. Therefore, the best thing this thread could do is predict necessary transformer va for avoiding negative consequences such as these:
needing transformer replacement
needing excessive capacitance
requiring capacitive multiplier
Which end of the horse is the cart?
So, predicting the required capacitance per given transformer and watts, does look like it serves the purpose of accurate transformer purchases.
Finding the best capacitance is so very simple, but finding the best transformer size to have the capacitance work right, is the hard part.
It certainly is not. Gorgeous bass H1/H2 balance = Their caps are big enough (aka the transformer is big enough).
Call Me Maybe - Carly Rae Jepsen (Disco Cover by Stacks of Wax f/ Nicholas Wells) - YouTube
After the fact, when its right, that's obvious. Therefore, the best thing this thread could do is predict necessary transformer va for avoiding negative consequences such as these:
needing transformer replacement
needing excessive capacitance
requiring capacitive multiplier
Which end of the horse is the cart?
So, predicting the required capacitance per given transformer and watts, does look like it serves the purpose of accurate transformer purchases.
Finding the best capacitance is so very simple, but finding the best transformer size to have the capacitance work right, is the hard part.
Last edited:
Daniel that is the core of this thread, what is the minimum requirement, if someone feels the need to double it then so be it at least it removes some marketing hype of amp X having one mega farad rail capacitors and that is why it is the best sounding amp, if we can show that 1000 uF is adequate to reproduce sound accurately and consistently then who cares about 1 mega farad.
Yes. Agreed. I just want to get a transformer big enough that I don't need 50,000uF or other huge figures. <other way to say the same thing.
P.S.
I had an idea for your headphone amp test: Temporarily power the amp decoupling caps (the elecro closest to the amp) from regulators (not using the board's power supply) until you determine the ideal amp decoupling caps (size at that location determines forward versus laid back). And then remove the regs.
Next, populate the power supply reservoir with caps. After that, insert a schottky in series (tiny voltage drop) between power supply reservoir caps to amp decoupling caps, so that the smaller faster amp decoupling caps can't fight with the bigger slower power supply reservoir. This looks odd, but isn't different than an automotive diode-based battery isolator. You can optionally expand this idea to individualized left/right channels isolators if you desire the slight stereo broadening of virtual dual mono, which could be entertaining with headphones.
You can determine power supply reservoir size without the noisy influence of the amp caps competing with power caps if you put a series schottky from power caps to amp caps (if split rail, both V+ and V- would need a series schottky). Now we don't have to observe which esr caps fight the least, since that audible effect is blocked. The simple isolator trick should make the determination of power supply reservoir size a bit more direct.
What do you think of it? It can look like CRCDC if you wish.
After that, you could try a variety of transformers with same voltage, different VA and observe the effect on power supply reservoir size.
The result could be plotted on a graph or chart.
Last edited:
Nico, I would say that it is almost impossible to know what the original was, unless you happened to have your head right next to where the microphone(s) were located at the time of the recording. Of course the recording won't be the "same", but it's close enough for me. Not sure what you mean by "gating" but I presume you're referring to the effect where detail below a certain level appears missing in much digital replay. This is faulty playback, pure and simple, a form of distortion, in the same sense that mistracking and inner groove deterioration are forms of distortion for vinyl.
I'm not referring to digital steering, but rather precision steering, where the suspension and tyres very sensitively and accurately react to the slightest input of the driver's hands on the wheel. As in an expensive sports car. What I'm saying is that digital is extremely precise; and if the system setup doesn't match up everywhere to that level of accuracy then you end up with unpleasant, or boring sound.
Which brings us back to the topic of this thread: part of that precision required for competent digital sound is bound up in how well the power supply does its job -- lack of rigour here will be penalised with unsavoury sound ...
Frank
Last edited:
Your car does that too? Oh no! I though I had the only one.
The middle is missing? For audio, the middle is DC/Silence. Digital is never silent? Oh no! Will a mild noise gate fix it?
I have seen DI to the mixing desk looping through an effects chain numerous times in demo or cheap-quick productions, for ''creative'' purposes in serious productions even, but more rarely. Not in seasoned rock bands studio sessions though. There much care is invested in miking up the cab just right until the guitarist is happy recognizing his trademark sound. But other genres can be altering it much with effects and DI boxes. An electric guitar is defined as an instrument-amp-cab combo in its original form no doubt, but its not strict in the wider recording musicians world.
I know I am repeating myself here, but there is no minimum. Given a particular peak output power into a particular load resistance, and given detailed knowledge of the output stage (e.g. emitter resistor value, how many Vbe drops), and given transformer off-load voltage and effective secondary resistance, and rectifier voltage drop we can arrive at an estimate of the minimum capacitance. Change any of those and the estimate changes. It could change by a lot if the transformer output voltage is only just big enough. Also, what if mains voltage drops by 5% - could mean little change or a big change in cap value.Nico Ras said:
So there is no magic value (so many uF per A or W). The best we can do is estimate a value, given all the above information, and then double it. You can estimate a minimum energy storage (see my earlier post - I can't be bothered to find it) but this is of little practical help as it requires exactly the right transformer voltage and most circuits will need to use significantly more.
Keep searching for a minimum cap value if you enjoy the thrill of the chase but, like unicorns, it doesn't exist.
Lol for unicorns,
Please, let me know if you see any potential threats and/or caveats in distributing the capacity along the circuit, and as per Tom suggestions, in particular placing more capacitance near the output stages? I've been told that, if the power should oscillate then output stage would have a low impedance current source and would probably fry, but is it likely that a stable amp would oscillate because of such a placement of capacitors?
Thank you
Please, let me know if you see any potential threats and/or caveats in distributing the capacity along the circuit, and as per Tom suggestions, in particular placing more capacitance near the output stages? I've been told that, if the power should oscillate then output stage would have a low impedance current source and would probably fry, but is it likely that a stable amp would oscillate because of such a placement of capacitors?
Thank you
Last edited:
the analog recorder does not have an "infinitely" high sampling rate, nor an "infinitely" high bit rate. The fact that it is analog and that there isn't a expressed nyquist rate, is irrelevant: There still is a limit. What? You think you can record a 1MHz signal on the tape? "96kbps" next to "48kHz", means a 2bits per sample quantization depth, so i guess you meant to say something else. No, your Studer, being electromechanical, isn't superior to a high quality digital system. Keep your sarcasm to yourself.
if you want to measure the impact of capacitors or capacitance in audioplayback you do measurements.
have you heard of any recording done with an electric guitar without the signal going through at least a cabinet simulator? If you plan on replying "yeah for pickups with midi output" then it's not an electric guitar anymore, it's a midi controller.
Terry,
I would like to try to per-unitize the transformer parameters and be able to scale them for different VA. It obviously wouldn't be as simple as multiplying all of the L and R figures by newVA / originalVA. So how? Or can that not work well?
Tom
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.