Regarding "headroom": The same circuit with different transistors, e.g. TO-220 and TO-3P, both with comparable "measured values", ALWAYS have a different sound. The TO-3P transistor sounds like more headroom. Why? The current is less accurate. Low frequency, mids, highs sound separated, the tones do not find each other, it seems bigger, more dignified, more relaxed. But also more unclear, slower, flatter, milder, the contours are washed out.
Looking and hearing measurements are not identical.
Looking and hearing measurements are not identical.
Headroom in music is useless. Maybe symphony orchestra are exceptions. But artists tend to remain in rythm and in controll of their loudness. Even when went high notes they never increase db. Thats what singing is all about.
In your world maybe, in mine it's not . And i would like you to show me how to record a singer with high dynamic range without overloading either mic or preamp... please show me! Because in the genre/style i'am active as recording engineer, we often need two mic to track singers: one for 'soft' parts, one for screamed ones.
Of course one the first is overloaded we switchnto the other.
But hey... you've got the answer we all seek!
So you listen to sinewave? Interesting... because it is what i get from your description, a stable signal without life. Should be boring.
This kind of definitive comment ( without anything technical as back up) i find useless.
Last edited:
Why? Because 'singers' use a very wide dynamic range (high crest factor): from whispers to full blast screamin' in very short times.
Once you setted up gain for the 'whisper' mic once you start talking with regular level microphone/preamp start to clip...
When they scream you are already in the 20% distortion (or more!) range with this mic. You need a second one with way lower gain to record the high level parts...
And no, those are not unskilled singers ( most of the guy/ girls are very good technical singers and they know how to regulate their distance to mic in 'regular' parts of songs) it's just the genre/style which ask for that.
This isn't anything new, check for the 'vocal' section (there is even a nice picture of a typical vocal recording setup):
https://www.soundonsound.com/techniques/extreme-metal
I know, so much headache for something this ....ugly. 😉
Once you setted up gain for the 'whisper' mic once you start talking with regular level microphone/preamp start to clip...
When they scream you are already in the 20% distortion (or more!) range with this mic. You need a second one with way lower gain to record the high level parts...
And no, those are not unskilled singers ( most of the guy/ girls are very good technical singers and they know how to regulate their distance to mic in 'regular' parts of songs) it's just the genre/style which ask for that.
This isn't anything new, check for the 'vocal' section (there is even a nice picture of a typical vocal recording setup):
https://www.soundonsound.com/techniques/extreme-metal
I know, so much headache for something this ....ugly. 😉
Last edited:
Yes but you seems to suggest a low power amplifier is not able to be dynamic, clean and produce headroom.
This i disagree.
I can be the case some have bad sound but it's a misapplication/misuse most of the time.
Some of the most beautiful sounding amp i've heard were triode or mosfet based low power amps. But used with drivers able to produce spl required by application ( high efficiency mid or compression drivers on horns).
If your spl requirements are lower than mine low efficiency drivers with low watts can do it ( lower because you live in a place where background noise is lower than the 55dba i've got, or you listen closer to the loudspeaker than i do, or you prefer other kind of music, whatever...).
No. I balance low volume and high volume not tones: mic i use are usually a pair, they have same frequency characteristics.
Compression is used of course, it is part of the sound artists looks for ( and you won't hear voice if it wasn't used: have you ever been located next to a metal drumer? Say whaaat?). Like 99% of voice recording ever done ( when we don't use hardware ones... we used tape which is the best compressor ever).
Compression is used of course, it is part of the sound artists looks for ( and you won't hear voice if it wasn't used: have you ever been located next to a metal drumer? Say whaaat?). Like 99% of voice recording ever done ( when we don't use hardware ones... we used tape which is the best compressor ever).
Last edited:
I would suggest only a low power amplifier is able to be dynamic, clean and produce headroom.
The reason: a minimum of parts, which modulate the signal with their noise;-)
The reason: a minimum of parts, which modulate the signal with their noise;-)
Well idk.
Some Smps driven class D are dynamic clean and have headroom too. And you can find them in kwatt range which makes them a good choice for low way of multiamped systems.
Some Smps driven class D are dynamic clean and have headroom too. And you can find them in kwatt range which makes them a good choice for low way of multiamped systems.
Assuming we are still discussing power amplifiers rather than amplifiers in general, it seems illogical to suggest that low power amplifiers could have more headroom than higher power models. They simply cannot because more power is essential to the larger dynamic range which, due to a noise floor and our hearing capability, can only be increased to widen it.
Practical considerations for ambient sounds, system noise, loudspeaker sensitivity, linearity and personal hearing loss will limit how low you can go with power and hence listening levels. Otherwise, we would likely need to consider headphones and modern hearing aid technologies to make use of really low power (say <5W) amps or perhaps just become super-rich and look for ultra-sensitive speakers.
Practical considerations for ambient sounds, system noise, loudspeaker sensitivity, linearity and personal hearing loss will limit how low you can go with power and hence listening levels. Otherwise, we would likely need to consider headphones and modern hearing aid technologies to make use of really low power (say <5W) amps or perhaps just become super-rich and look for ultra-sensitive speakers.
My experience: The higher the "power", the more complex circuits, the more components, the more modulation with its noise, the more noise: the less quiet-loud difference.
Cleanliness brings the dynamics. I do not confuse with maximum sound levels.
Cleanliness brings the dynamics. I do not confuse with maximum sound levels.
Cummb,
I tend to agree but at same time it seems to broad generalisation:
I'm currently switching to classD and the Hypex oem modules i use ( without input buffer implemented) push 180w for an unbelivable small size and low components count. The ucd 400w boards i'll use too are not much more bigger.
Another counter example is the 1kw 2 amplification stage Pass Labs amp i heard ( i forgot the reference of it, i think it's the one Nelson talk about into the cascode article ( or maybe SuSy one i can't remember).
None of them are plagued by issues related to noise floor ( ucd have some other drawback though. Nelson amp is a Nelson amp. ).
I tend to agree but at same time it seems to broad generalisation:
I'm currently switching to classD and the Hypex oem modules i use ( without input buffer implemented) push 180w for an unbelivable small size and low components count. The ucd 400w boards i'll use too are not much more bigger.
Another counter example is the 1kw 2 amplification stage Pass Labs amp i heard ( i forgot the reference of it, i think it's the one Nelson talk about into the cascode article ( or maybe SuSy one i can't remember).
None of them are plagued by issues related to noise floor ( ucd have some other drawback though. Nelson amp is a Nelson amp. ).
It wasn't intended like an insult (although I guess that that would make it even worse). Some consider audio to be more of an 'art' anyway. I didn't say that we 'only' need a scope, although now that you mention it, if you have access to a linear representation of the signal, you can always use it to create whatever SPL averaging you want, but you can't go in the other direction to recover a non-averaged signal that has already been smoothed.
For me, a useful measure that I haven't seen anywhere, would be statistical analysis of a lengthy musical signal in a similar vein as normalisation, but instead of only looking for the tallest sample, it organises the DC offsets of all the samples on a bell curve. That way, you know exactly how much time the amplifier spends at every voltage.
Not knowing the maximum output of a singer or instrument while recording is a valid concern and a big argument in favour of high bit depths, so you can have that 20dB of overhead, just in case. But once it's all recorded and you're interested in playback levels, you simply adjust the volume to suit, and measure the voltage. I think you're over-complicating things by converting precise voltages into vague SPL levels, and then guesstimating what the voltages might have been in the first place!
Hi Abstract,
Vague db value?
Please go back in message i posted and you'll observe that except for direct comparison between levels ( what db have been invented for at first) i always use a letter or a group of letters after the db mention.
Those are not arbitrary things, each one correspond to a clear reference ( db SPL are referenced to 1Pa @ 94db, a/c correspond to a certain kind of filtering, dbfs is referenced to clipping point,...). It's not because something is vague to you it is in absolute. Db are not anything vague in proaudio and this is my background.
In no way this is vague. Check with Wikipedia definition of each one.
Really i won't bother you with the article i linked but it's all about what you suggest... So you either haven't read it' understood what it is about or another motivation?
I'm not trying to convince you of anything, but it seems obvious we have very different approach to all this and you are locked on your pov ( as i am too).
This i can understand but dismissing approach used by pro to produce what you ( we) as end user use doesn't make it for me.
At least not with the arguments you raised ( i've got my own set of complain about the solution i talk about but they are probably not what you would expect and doesn't change the fact this is in my view the least offendable approach and why i choosed to implement it even at home).
Art. Ok i work with artists to help them accomplish the artistic vision they have. In no way i want my tools to be 'art' related as i'm a technician. This is all science to me. Of course ymmv, as your motivations. But art in technical matter too often equal to conspiracy to me.
You want to keep some thing 'mysterious' or 'magical' it's fine. I don't, this is technical subject, if technical mysterie happen then we have to investigate it to understand and make it repeatable. My pov.
Overthinking things, yes maybe ( no, often!) But not in this case i fear. Maybe you are oversimplyfing things by putting aside important matters ( in my view)? 😉
I've got nothing to sell you, no interest in spreading this view either ( as it always end up more or less at where we are now from my pov).
How fortunate the one which doesn't see there is an issue.
Vague db value?
Please go back in message i posted and you'll observe that except for direct comparison between levels ( what db have been invented for at first) i always use a letter or a group of letters after the db mention.
Those are not arbitrary things, each one correspond to a clear reference ( db SPL are referenced to 1Pa @ 94db, a/c correspond to a certain kind of filtering, dbfs is referenced to clipping point,...). It's not because something is vague to you it is in absolute. Db are not anything vague in proaudio and this is my background.
In no way this is vague. Check with Wikipedia definition of each one.
Really i won't bother you with the article i linked but it's all about what you suggest... So you either haven't read it' understood what it is about or another motivation?
I'm not trying to convince you of anything, but it seems obvious we have very different approach to all this and you are locked on your pov ( as i am too).
This i can understand but dismissing approach used by pro to produce what you ( we) as end user use doesn't make it for me.
At least not with the arguments you raised ( i've got my own set of complain about the solution i talk about but they are probably not what you would expect and doesn't change the fact this is in my view the least offendable approach and why i choosed to implement it even at home).
Art. Ok i work with artists to help them accomplish the artistic vision they have. In no way i want my tools to be 'art' related as i'm a technician. This is all science to me. Of course ymmv, as your motivations. But art in technical matter too often equal to conspiracy to me.
You want to keep some thing 'mysterious' or 'magical' it's fine. I don't, this is technical subject, if technical mysterie happen then we have to investigate it to understand and make it repeatable. My pov.
Overthinking things, yes maybe ( no, often!) But not in this case i fear. Maybe you are oversimplyfing things by putting aside important matters ( in my view)? 😉
I've got nothing to sell you, no interest in spreading this view either ( as it always end up more or less at where we are now from my pov).
How fortunate the one which doesn't see there is an issue.
Last edited:
I guess we also have to separate the circuits a little bit. The advantage of Classe D is the lower modulation of the signal, due to the signal, by the parts. The comparison should be intra-category: for example, little and big Classe D. And a revision of the amplifiers brought further advantages for the little one, also Classe D.
But what should be clear: a little one moves not as much mass as a big one;-)-; My recommendation therefore about: according to the circumstances (living room, cowshed, stadium - sleep-in level, room volume, garage punk, open air, and speakers) always choose minimal circuitry.
I would recommend something like this.
But what should be clear: a little one moves not as much mass as a big one;-)-; My recommendation therefore about: according to the circumstances (living room, cowshed, stadium - sleep-in level, room volume, garage punk, open air, and speakers) always choose minimal circuitry.
I would recommend something like this.
Well I've had time to try out different speakers with the 3 watt amp, they all worked well apart from one with polypropylene drivers, they seemed to need more volume, and that tipped the amp into distortion. I have rarely reached the point of distortion, and with the PAM amp, it's night and day, either it's perfectly fine, or just a bit louder and there's a knocking noise, nothing in between. I'm going to try some LM4871 class b amps, they seem to have much lower distortion figures. I'd be interested to try the LM4871 amps with some 94db efficient speakers that I've nearly finished.