Hi Johnego from Indonesia.
No, when you desolder & upgrade op-amps to the flagship op-amps, they sound better.
It's not possibility.
For example Asus (the laptop company) released a desktop DAC with 6 x NE5532 in rollable sockets, so then you can just pull them out & insert modern chips which are superior to the NE5532.
Asus also released a fancier version which included the MUSES01 x6, instead of the NE5532 x6. If you read the datasheet of MUSES01 it's not low in THD & not an ultra performer in anything really, so I think Muses is like "we are going to colour the sound & don't care about numbers". Actually I think the numbers are the best sounds I have heard i.e. LME49990 & AD828. I bought THS4032 & never got around to solder them.
No, when you desolder & upgrade op-amps to the flagship op-amps, they sound better.
It's not possibility.
For example Asus (the laptop company) released a desktop DAC with 6 x NE5532 in rollable sockets, so then you can just pull them out & insert modern chips which are superior to the NE5532.
Asus also released a fancier version which included the MUSES01 x6, instead of the NE5532 x6. If you read the datasheet of MUSES01 it's not low in THD & not an ultra performer in anything really, so I think Muses is like "we are going to colour the sound & don't care about numbers". Actually I think the numbers are the best sounds I have heard i.e. LME49990 & AD828. I bought THS4032 & never got around to solder them.
What I mean is that you can downgrade the performance of a better amp so that they are similar, if the performances are not so different. But if you compare 741 with 797, may be it will be hard to make them sound similar (in a different circuit).
When I compared op-amps, I always use optimum/best circuit for each of them. So OPAMP1 in its best circuit will be compared to OPAMP2 in its own best circuit.
Most people don't have sensitive ears. But if you have, you will know that amplifiers (sound) are sensitive to stability. Some opamps just need the correct compensation cap to sound good. Sound differences are easy to detect if they sound wrong. They sound wrong because the implementation is not proper.
The surrounding resistors that work with the input capacitance may create a pole that depends on the bandwidth may create peak in the MHz region.
If you hear differences between opamps, you should be able to describe the differences. From your descriptions it can be predicted which variable that has to be changed so you will no longer hear it. So theoretically it is possible.
Last edited:
You can manipulate the whole signal to some degree as you mentioned with harmonic distortion for example, I suspect this would also blur the individual locations to some degree but I don't know. I haven't heard of mixers using this as a method of image location, but they may. It all seems a bit hit and miss to me.
No doubt all important when it comes to localisation, so making sure they are reproduced accurately is important. Again, I don't see the issue or need to measure them individually in the composite signal, not that you can anyway.
Hi Johnego from Indonesia.
I used mostly iBasso portable amplifiers which had op-amp sockets & included extra op-amps for rolling, so you could try every op-amp in the included kit to get closer to your favorite sound.
Every desktop DAC I had included rollable line-out sections as well. DAC + line-out op-amp is enough to drive an IEM, this is a short chain.
I know that not every single op-amp is rollable, a lot of them are incompatible & will overheat, or cause aberrant spikes in the ultrasonic inaudible frequencies etc like you said. However most of the time, the popular chips such as AD797, AD8620, LM4562 are rollable... then if you are just a normal citizen & can't read datasheets and insert a THS4062 into your speaker amp or DAC line-out, there is a risk it will overheat & not be compatible at all, however the ultimate amplifier is rollable & the ultimate DAC is rollable, that's all, that's why there's the little sockets & plastic boxes with extra amplifiers.
If I just wanted to enjoy the sound then I listen to my favorite artists with ES9018 & LME49990 / LM4562 into speaker or IEM & never sign up here.
I used mostly iBasso portable amplifiers which had op-amp sockets & included extra op-amps for rolling, so you could try every op-amp in the included kit to get closer to your favorite sound.
Every desktop DAC I had included rollable line-out sections as well. DAC + line-out op-amp is enough to drive an IEM, this is a short chain.
I know that not every single op-amp is rollable, a lot of them are incompatible & will overheat, or cause aberrant spikes in the ultrasonic inaudible frequencies etc like you said. However most of the time, the popular chips such as AD797, AD8620, LM4562 are rollable... then if you are just a normal citizen & can't read datasheets and insert a THS4062 into your speaker amp or DAC line-out, there is a risk it will overheat & not be compatible at all, however the ultimate amplifier is rollable & the ultimate DAC is rollable, that's all, that's why there's the little sockets & plastic boxes with extra amplifiers.
If I just wanted to enjoy the sound then I listen to my favorite artists with ES9018 & LME49990 / LM4562 into speaker or IEM & never sign up here.
It seems to be consensus on that the most faithful reproduction of a sound (all sounds?... close, far away, large, small??) can be binaural recording played back via headphones. So this tells us that perhaps it is the room the biggest obstacle to perfect reproduction. And its quite obvious that we have a system architecture problem: we add the playback room to the recording room. So how can the ever come out right? No, it cant. Just simple basic math. 1+1 is not 1.
The next big problem is mic positions and what it does to the perspective (angels), levels and tonal balance of what is recorded compared to a good listening position in a hall.
You see, it just cant be done by stereo and how we record and replay today.
0,1% or 0,001% isn't the biggest problem.
In a serious attempt I would not like to bother about distorsion profiles - it just should be so close to 0 that it is not a factor at all.
//
The next big problem is mic positions and what it does to the perspective (angels), levels and tonal balance of what is recorded compared to a good listening position in a hall.
You see, it just cant be done by stereo and how we record and replay today.
0,1% or 0,001% isn't the biggest problem.
In a serious attempt I would not like to bother about distorsion profiles - it just should be so close to 0 that it is not a factor at all.
//
Interesting.
In this case to separate a difference in opamps from a difference in how the opamp interacts with the circuit, one would have to check:
- layout & decoupling
- transient response into load for ringing or stability (also probe power supplies for ringing)
- frequency response up to several times GBW to check for suspicious peaks indicating borderline instability
'cause very often people comment on the sound of opamps, and it turns out to be a 55MHz opamp on a single sided board with decoupling caps two miles away...
Hi Chris,
just for curiosity, what about the measurements before and after the tweaking?
And also, if it's not a secret, what were the changes you made?
Andrea
To drive most headphones, important requirement usually is an opamp that has a strong drive. Many good opamps will shine as long as you don't overload them. Standard minimum load for most opamp is usually 1k, below that most of them suffer. Certain opamp will sound very good only when the load is 2k so you need to buffer them.
Isn't it better to create a circuit that is optimum/dedicated for each of the opamp, especially for the one with superior overall performance?
Thank you Matt. I do not mean being able to separate the composite and measure the properties of individual image. If something can be deduced about soundstage property from the signal then how a system respond wrt soundstage can be quantified more readily. Most probably we have nothing in that direction yet.
Hi,
Indra, step response and impulse response are interesting in particular for your loudspeakers.
Those are electro mechanical device so there is issue about cinetic energy (stop and start of motion), stored energy, etc,etc,...
In that case step and impulse will tell you many interesting things.
Most people agonize about freq response. In a way the concerns are valid but this is only one side of a coin: time domain is often second rank (or no rank at all).
This is a gross error: our auditory system/brain is a 'transient analyzer'.
You have most information about the family of instrument ( striked or bowed string, membrane hited, etc,etc,..) in the first ms of sound.
If this part of sound is modified you loose a lot of precious infos about 'quality' of instruments and overall about all the clues embeded* in the signal about the initial acoustic environement.
So for loudspeakers (in my view) you have to know how they behave regarding tracking of the 'envelope' of a sound.
There is other things at play of course like 'instant' compression ( thermal related but not only: how the system behave regarding short duration peaks- which you will likely found in the first ms of a sound), the behavior of phase, directivity behavior,etc,etc,...
For an electronic device theorically you shouldn't be limited about this enveloppe tracking if the circuit topology is well choosen and this whatever the kind of components used: tubes, opamp, transistors,...
People keep on talking about 'tube' sound but.... there is no tube sound! A tube if used in SOA is as clean as anything else. It even have some advantage regarding biasing as voltage rail is usually high which gives headroom so theorically better immunity to transient smearing.
It is the choice of operating points of circuit which gives the 'tubey' sound ( the EAR i talked about is an example in amateur world, in pro world Avallon tubes processor are an example too) and the transformers used.
If you still doubt then take a look at radio transmitters from 60's, 70's or 80's and ask for yourself if this tubes were allowed to saturate...
Iow the aesthetic choices made by designers of a tube amp is that only: an aesthetic choice on the rendering.
* the spaciousness or feeling of it is mainly related to the ratio of direct sound/early reflection and the nature of the later ( duration, freq balance, direction) and i do think that other acoustic artefacts comes into play too (SBIR).
So you can have a feeling of space even with a mono recording.
Indra, step response and impulse response are interesting in particular for your loudspeakers.
Those are electro mechanical device so there is issue about cinetic energy (stop and start of motion), stored energy, etc,etc,...
In that case step and impulse will tell you many interesting things.
Most people agonize about freq response. In a way the concerns are valid but this is only one side of a coin: time domain is often second rank (or no rank at all).
This is a gross error: our auditory system/brain is a 'transient analyzer'.
You have most information about the family of instrument ( striked or bowed string, membrane hited, etc,etc,..) in the first ms of sound.
If this part of sound is modified you loose a lot of precious infos about 'quality' of instruments and overall about all the clues embeded* in the signal about the initial acoustic environement.
So for loudspeakers (in my view) you have to know how they behave regarding tracking of the 'envelope' of a sound.
There is other things at play of course like 'instant' compression ( thermal related but not only: how the system behave regarding short duration peaks- which you will likely found in the first ms of a sound), the behavior of phase, directivity behavior,etc,etc,...
For an electronic device theorically you shouldn't be limited about this enveloppe tracking if the circuit topology is well choosen and this whatever the kind of components used: tubes, opamp, transistors,...
People keep on talking about 'tube' sound but.... there is no tube sound! A tube if used in SOA is as clean as anything else. It even have some advantage regarding biasing as voltage rail is usually high which gives headroom so theorically better immunity to transient smearing.
It is the choice of operating points of circuit which gives the 'tubey' sound ( the EAR i talked about is an example in amateur world, in pro world Avallon tubes processor are an example too) and the transformers used.
If you still doubt then take a look at radio transmitters from 60's, 70's or 80's and ask for yourself if this tubes were allowed to saturate...
Iow the aesthetic choices made by designers of a tube amp is that only: an aesthetic choice on the rendering.
* the spaciousness or feeling of it is mainly related to the ratio of direct sound/early reflection and the nature of the later ( duration, freq balance, direction) and i do think that other acoustic artefacts comes into play too (SBIR).
So you can have a feeling of space even with a mono recording.
Well, as I said previously, I think all you could do is make some kind of comparison, like between mixes that have spatial cues and those that don't, then you'll see a difference, but that's all, I doubt you'd be able to quantify it in a useful way, but I don't see how just looking at one lot of measurements would help. I think trying to take that statement to the, apparently logical, conclusion you are, is a mistake.
It doesn't add up because we are still technology limited: 25 years ago no one thoughts you could 'de-mix' a song ( from a track have the bassline separated from voice,etc,etc,...).
And now we have Celemony's Melodyne polyphonic! IRCAM have some very advanced algorythm for the same things too ( a friend demonstrated some to me some years ago).
With AI and processing power it may happen quickly to have acces to something which could easily help in analysing what interest you.
Indra you should take a look at David Griesinger work:
http://www.davidgriesinger.com
Indra,
You told you are frustrated by your set up. Have you the possibility to listen to different kind of loudspeakers easily ( like dipoles, panels, omni, coax,...).
This is imho the only way to know what YOU like and will be a step in the direction. Ime loudspeaker/room are the weakest link in reproduction and once you found the one you like everything is simpler.
You talked about air behavior regarding rarefaction and pressure: what you described is true and it gives you an idea of how it'll behave: distortion only on one side of a waveform: second harmonic mainly ( same behavior as tube class A regarding this point).
But before you'll suffer from this you'll be long time deaf!
Don't agonize over details before a more complete picture take shape in your mind.
And now we have Celemony's Melodyne polyphonic! IRCAM have some very advanced algorythm for the same things too ( a friend demonstrated some to me some years ago).
With AI and processing power it may happen quickly to have acces to something which could easily help in analysing what interest you.
Indra you should take a look at David Griesinger work:
http://www.davidgriesinger.com
Indra,
You told you are frustrated by your set up. Have you the possibility to listen to different kind of loudspeakers easily ( like dipoles, panels, omni, coax,...).
This is imho the only way to know what YOU like and will be a step in the direction. Ime loudspeaker/room are the weakest link in reproduction and once you found the one you like everything is simpler.
You talked about air behavior regarding rarefaction and pressure: what you described is true and it gives you an idea of how it'll behave: distortion only on one side of a waveform: second harmonic mainly ( same behavior as tube class A regarding this point).
But before you'll suffer from this you'll be long time deaf!
Don't agonize over details before a more complete picture take shape in your mind.
Last edited:
^Probably a low pass and a resistor in serie with the output.
Indra as you want, as long as we are not moderated i think it is ok.
For loudspeaker measurements i relink this:
https://www.diyaudio.com/forums/the...-dr-earl-geddes-gedlee-audio.html#post6636818
Listen to the podcast and read the D'Appolito approach, make a spinorama too and i think you won't be far away of the important factors ( other details count too but those are the important one in my view).
I would like to give the same thing to Andrea but i can't...
Indra as you want, as long as we are not moderated i think it is ok.
For loudspeaker measurements i relink this:
https://www.diyaudio.com/forums/the...-dr-earl-geddes-gedlee-audio.html#post6636818
Listen to the podcast and read the D'Appolito approach, make a spinorama too and i think you won't be far away of the important factors ( other details count too but those are the important one in my view).
I would like to give the same thing to Andrea but i can't...
Last edited:
Yeah, thanks. I should have taken blanket statements like that with a big chunk of salt.
Yet the statement if we can hear it then it is measurable is considered to be true. I really don't know what to think. It simply does not add up.
The statement is presumably more meant as "if a perceptable difference exist, than we are able to measure a difference" , not necessarily that we are always able to point to an explicit cause and effect relationship.
- Status
- Not open for further replies.