A/D Converter High Res No high pass

While I'm no expert on circuit design, most (all?) of the audio circuits I've delved into have DC blocking capacitors at various points out of necessity. I suspect those can't be arbitrarily low.

What's the point of going below 20Hz?
 
While I'm no expert on circuit design, most (all?) of the audio circuits I've delved into have DC blocking capacitors at various points out of necessity. I suspect those can't be arbitrarily low.

What's the point of going below 20Hz?

Some 'High End' designs eschew capacitors and use DC servo circuits to keep outputs at 0 volts. I find the claims for better fidelity dubious...the MORE so since these same designers often extol the sonic virtures of transformers! The latter will always introduce more distortion than a well specified electrolytic capacitor.

Maybe it's a 'snob' thing? Capacitors are cheap, = BAD! Transformers are expensive, especially really good ones, = GOOD.

I always wonder? Say you have a totally DC coupled path from an AI to a DC servo'ed, DC coupled 200W active monitor and just ONE servo in the chain goes 18 volts ape shit? !!

I cannot remember the last time I had an electrolytic capacitor short?

Oh! BTW BSG, there is a church organ with a 16Hz pedal pipe! In Oz somewhere I think.
Dave.
 
I found this: "There are two instruments that have a full-length 64’stop. The first one is the Midmer-Losh organ at the Atlantic City Convention Center. The second one is the Pogson organ in the Organ Sydney Town Hall."
 
I found this: "There are two instruments that have a full-length 64’stop. The first one is the Midmer-Losh organ at the Atlantic City Convention Center. The second one is the Pogson organ in the Organ Sydney Town Hall."

Obliged Gekks. Now all OP needs is a mic that can get down there, a sub woofer to do same (would need at least a 1kW amp I think) and a FURKIN big room!

If instead he wants to record infrasonics, whales say then I suggest the inverse of a Bat Detector. Instrumentation recorders, strain gauge data say, used to use a DC to frequency converter to record to conventional tape. You STILL have to find a suitable primary transducer!

Dave.
 
While I'm no expert on circuit design, most (all?) of the audio circuits I've delved into have DC blocking capacitors at various points out of necessity. I suspect those can't be arbitrarily low.

What's the point of going below 20Hz?

I remember when the movie Earthquake first came out. They had a huge bank of CerwinVega subwoofers that were supposed to put out subsonic sound at very high volume. The idea was to feel the quake, not just hear some rumble. There were stories about damage to some places when it first came out.

It worked. I remember the feeling in my gut when things began to shake.
 
Some 'High End' designs eschew capacitors and use DC servo circuits to keep outputs at 0 volts.

Right, but the fact that I could fool with dozens of audio circuits and not have encountered one of those means that not passing below 20Hz is generally considered completely acceptable.

Oh! BTW BSG, there is a church organ with a 16Hz pedal pipe! In Oz somewhere I think.
Dave.

Yep. And my question remains. What's the point? Will reproducing the effect of a semi (sorry, articulated lorry) driving by improve people's chances of getting into heaven?
 
Right, but the fact that I could fool with dozens of audio circuits and not have encountered one of those means that not passing below 20Hz is generally considered completely acceptable.



Yep. And my question remains. What's the point? Will reproducing the effect of a semi (sorry, articulated lorry) driving by improve people's chances of getting into heaven?

Oh! Sorry chap I wasn't arguing with you. You asked "what about the capacitors" and I replied that they can be done without.

I have to say I do not agree with DC coupling. The idea is to do away with electro' caps so maybe a slight diversion into why?
There is no doubt that e caps produce harmonic distortion and on test it can be seen that it gets worse as frequency falls. But this is one of those 'obvious innit? Capacitor distortion is frequency dependent" Not so. What happens is that AS frequency falls the reactance (sort of resistance) of the capacitor increases and so the voltage dropped across it increases. Once that voltage reaches about 100mV measurable (0.001%!) is observed. The solution is simple! Make the fekkin cap bigger! A capacitor 4 or five times the value necessary for the desired LF cut off point is still a lot cheaper than a servo!

B.U.T. People, especially MARKETING people just seize on Capacitors are BAAAAAD! Lets fill the kit with vastly more complex electronics. Then again, "I" am NO expert! I have a certain level of technical understanding and that level tells me caps are fine. I could be wrong!

I hold the same view at the other end of the audio spectrum. There are very well regarded, very high end mic pre amps that go up to 200kHz and beyond. I cannot understand why? Yes, there are specialist microphones that go past 50kHz but nothing in any studio is likely to reach 20k let alone go beyond.

So, my view is cut your spectral cloth to the real world. 20-20k is good enough.

Dave.
 
I’m not sure that DC coupling is any more complex. Quite the opposite actually. At the very least, you’re eliminating caps from the circuit. There isn’t much else that even needs to change in a dual supply circuit. Course, if the circuit is designed for single supply with a virtual ground, the whole thing needs to be rewritten.

The one argument I can think of for DC coupling is if you wanted to be able to pass or capture control voltages using the same interface. There have been times when such a thing could have been handy.
 
I’m not sure that DC coupling is any more complex. Quite the opposite actually. At the very least, you’re eliminating caps from the circuit. There isn’t much else that even needs to change in a dual supply circuit. Course, if the circuit is designed for single supply with a virtual ground, the whole thing needs to be rewritten.

The one argument I can think of for DC coupling is if you wanted to be able to pass or capture control voltages using the same interface. There have been times when such a thing could have been handy.

Well, no so bad in an AI perhaps but when you look at all the buffer and summing stages' switches and pots in even a modest studio mixer that is a lot of servos!

My beef really is an extra complication for no sonic benefit. There is no downside to coupling capacitors if properly implemented.

One are where large time constants should be avoided in the signal path is in power amplifiers because a very short lived entry into clipping can cause many mSecs of latchup. Of course it is bad practice to clip PAs but ***t'appen. Should never happen with an AI or mixer.

Yes, there is now a 'fashion' for AIs to pass DC control voltages. I have no experience of that situation so I shall not comment further about it.

Dave.
 
This is all interesting and well and good for me to hear. I'm just doing Jazz things so harmonic distortion isn't that bad. I play out of a tube amp that can handle anything. It's not the biggest thing in the world, but right now I am just using the MOBO mic input and recording at 384,000hz 24bit. It cuts alot of the bass out of the recording and I have to seriously bump it up like 25db digitally.

I don't want to spend 1000's of dollars and don't care about transistors. I play out of a tube amp. With 2 8ohm stereo monitor cabinets and a 16 ohm ampeg in the middle. It's not the loudest set up, but I can't play that loud anyway, but that amp is capable of recreating sound much better than most I have heard. It can take a lot of punishment. And it never cracks. It sounds thinner than modern digital transistor equipment, but is much much clearer.

I saw this...
AK5538VN | Audio A/D Converters | PRODUCTS | Asahi Kasei Microdevices (AKM)

Which is not this...

RME ADI-2 Pro FS AD/DA Converter | Sweetwater

, but if it has a HPF? I don't necessarily want it.

THE 786khz resolution would help becasue I could send tiny signals and blow them up a lot better depending on noise. But I would rather not do that to bring up to normal, but instead enhance....



The digital recording is hardly a recreation of how warm and soft the original bass from the keyboard sounds. 30+db of signal gain in digital recording environment. It is still to quiet. + transcode in video to 96khz Vorbis audio.

I have one of these, but it's firewire and I have a bitch of a time getting it to connect. Right now I haven't been able to make it connect in six months.

edirol-fa-101-fa101-firewire-audio_360_456182867ba3d0b1c24943bf2272e74f.jpg
 
Ryan mate, you have some very strange ideas about interfaces, their LF responses and sampling rates. If you can't 'do it' at 44.1kHz you ain't gonna.

That interface really needs an XP computer. Roland say there is a W7 driver but I wouldn't trust it. Firewire AIs were a PITA for a lot of people. Most only ran with a Texas I chip as FW controller and many FW equipped laptops would not work at all.

Oh! And that amp chassis looks to have a blown pre amp valve.

Dave.
 
Ryan mate, you have some very strange ideas about interfaces, their LF responses and sampling rates. If you can't 'do it' at 44.1kHz you ain't gonna.

That interface really needs an XP computer. Roland say there is a W7 driver but I wouldn't trust it. Firewire AIs were a PITA for a lot of people. Most only ran with a Texas I chip as FW controller and many FW equipped laptops would not work at all.

Oh! And that amp chassis looks to have a blown pre amp valve.

Dave.

How can you tell?
The Video above is not with the amp, but straight into the mobo. I only posted that so you could have a picture of the amp I was talking about and it's giant cap. So your saying the mobo has a blown cap? I had to blow that bass up like a lot to to hear it on the recording 30+db. Nowhere near as good as in real life.
The last recording on bottom is through the amp.

Anyway I know that. 24 bit audio wil be able to record bass the same despite the resolution. High resolution audio just sounds better in my opinion. That is why I am on here trying to find out what is and isn't good. I want to do stuff with higher res audio becasue it can be manipulated more w/o getting all distorted. I've had that thing work before. I'd have to go down stream to an LTS version. It'll work better, but firewire audio is integrated in with Linux Kernel. It powers up and syncs , but ALSA plug dose not recognize it on this upstream OS for some reason.
[
1908 Colerain JAM.flac - Google Drive

Life in Lower East Tardiea.flac - Google Drive
]two edirol recordings on Linux
This strait to mobo and some bass boost.
testprojectmain out.flac - Google Drive

and this a few days ago using the tape out and mobo mic in of amp pictured above
a bit of a 40 min long guitar sesh I did a few days ago.
snippet2.flac - Google Drive
see I want the clarity of the high res digital recording, but I want that bass to come through also... So do you guys think that RME is the best bet here?


i can stream this stuff to any Bluetooth 5.0 device no problems. It now supports lossless audio.
 
Last edited:
Um, I get the feeling you're misunderstanding a few things about recording in general and digital recording in particular.

Typical digital recording is sampled at rates between 44.1kHz and 192kHz, and none of that will impact the low frequency cutoff. Going higher will either capture frequencies far above what a human can hear (which actually causes problems with reproduction), or just make the files much bigger than they need to be.

If a device has a high pass filter as a feature, it's almost certainly switchable. You can simply turn it off.

If what you hear when playing back your recording doesn't sound right to you, it could be your monitors or it could be something about how you're recording. You need to verify that your monitoring setup is accurate before you use it to judge your recording process. Once you can confirm that you are hearing things accurately, then we can start to address the recording process. What does other music sound like on your system?
 
Thanks BSG, I am glad I am not the only one who thinks that way!

Ryan, if you want an interface that definitely works in Linux check out the Native Instrument Komplete Audio 6. That will easily cover 20Hz to 20kHz at 44.1kHz sample rate, the noise floor is better than -100dBfs and distortion buggerall below 0dBfs. About the lowest latency you will find unless you go RME or thunderfart.

Dave.
 
Um, I get the feeling you're misunderstanding a few things about recording in general and digital recording in particular.

Typical digital recording is sampled at rates between 44.1kHz and 192kHz, and none of that will impact the low frequency cutoff. Going higher will either capture frequencies far above what a human can hear (which actually causes problems with reproduction), or just make the files much bigger than they need to be.

If a device has a high pass filter as a feature, it's almost certainly switchable. You can simply turn it off.

If what you hear when playing back your recording doesn't sound right to you, it could be your monitors or it could be something about how you're recording. You need to verify that your monitoring setup is accurate before you use it to judge your recording process. Once you can confirm that you are hearing things accurately, then we can start to address the recording process. What does other music sound like on your system?

Actually it is you that do not understand at all what i am saying. I hear you you don't need 786khz resolution for bass. But what you seem to be failing to understand is that I am not just recording bass here. I want that high resolution digital clarity, and I want my bass to come through also. I don't make CD's, so I don't need 44.1khz.

I'm pretty sure they have these HPFs hardwired in so as to not blow the equipment, or muddy the signal.

Almost everything works on Linux as dose Windows or Mac. This is not the issue here. I can go back to an LTS linux like UBuntu Studio and that Edirol will work fine.
 
Last edited:
Actually it is you that do not understand at all what i am saying. I hear you you don't need 786khz resolution for bass. But what you seem to be failing to understand is that I am not just recording bass here. I want that high resolution digital clarity, and I want my bass to come through also. I don't make CD's, so I don't need 44.1khz.

I'm pretty sure they have these HPFs hardwired in so as to not blow the equipment, or muddy the signal.

Almost everything works on Linux as dose Windows or Mac. This is not the issue here. I can go back to an LTS linux like UBuntu Studio and that Edirol will work fine.

I am far from any expert in digital matters but find a proper one and they will tell you that the "resolution" of digital recording is 6dB per bit and that decides the noise floor...Nothing else. The sampling rate determines the highest frequency you can record which will be a little under half said rate.

That is all. I shall leave you to it.


Dave.
 
Back
Top