Over 40db Less Harmonic Distortion than a FET: The OPA Alice Serena Mic

Рет қаралды 3,223

The Microphone Assassin

Күн бұрын

Пікірлер: 58

@mikeh2520 Ай бұрын

The OPA1642 is a JFET input op amp meaning that the FET is inside the chip.

@MarkYoshimotoNemcoff Ай бұрын

True, but here the exercise is to replace the FET circuit with something more modern, especially with the most popular quiet mic-use fets like the 305 and 5102 having been "end of life'd", and the big benefit is also that with the OpAmp, you don't have to manually bias each fet.

@astralaudio101 Ай бұрын

Hey many thanks for sharing your thoughts and experience. I’m interested in building but have become disillusioned by the wall of information. So it’s very appreciated..

@MarkYoshimotoNemcoff Ай бұрын

You’re welcome! The OPA Alice is a great way to get your feet wet because the board is populated for you and all you have to do is assemble and solder a few points.

@Blinkerd00d Ай бұрын

Ya gotta love the OPA2134 as well... I use the 2134 and 1642 as opamps for guitar fx pedals. I use them in EQ and modulation types of circuits.

@MarkYoshimotoNemcoff Ай бұрын

That's pretty cool. I love that. There seems to be a whole world of op amp projects out there that are waiting to be discovered.

@Blinkerd00d Ай бұрын

@MarkYoshimotoNemcoff I built a custom delay pedal for Brad Paisley and used the OPA2134. It sounded extremely clean, which can be hard to achieve with analog based delay circuits.

@TheRecordist Ай бұрын

No pro audio recording talk is EVER too nerdy from me. I eat, sleep, and breathe nerd. Gimme it all. But yeah. That sounds divine. Where is that gorgeous very subtle saturation coming from, I wonder? Your own voice? Is it the transformers in the Hairball Gold? Is it the capsule? It sounds so good, man! Very clean!

@MarkYoshimotoNemcoff Ай бұрын

It’s not the capsule. Put it in your veins, baby! In your veins. The OPA Alice is mic nerd heaven.

@TheRecordist Ай бұрын

@ woo hoo!

@joebloggs4369 Ай бұрын

The OPA appears to have a noticable roll off from 5k upwards. The FET circuit doesn't start until 10k and appears to be a gentler slope. SO there appears to be a trade off.

@MarkYoshimotoNemcoff Ай бұрын

Don't mistake the THD graph for a frequency response graph. A microphone frequency graph shows how a microphone responds to different frequencies across the audible spectrum, while a THD (Total Harmonic Distortion) graph illustrates the amount of distortion a microphone produces at different frequencies, essentially showing how much additional unwanted harmonics are added to the original signal at each frequency point; essentially, a frequency graph tells you how well a mic picks up sound across different pitches, while a THD graph tells you how much distortion is introduced.

@joebloggs4369 Ай бұрын

@@MarkYoshimotoNemcoff I'm not. What is shown by that graph is the spectral response of each amplifier when a single tone of 1 Khz is injected. Essentially what it shows is the noise floor spectral response of each amp below the injected frequency and the harmonics plus noise floor above the injected frequency and, of course the tone injected. What's not shown is the spectral response of the source signal being injected, which is surprising as no sinusoidal generator is free of distortion and I find it odd that the noise floor spectral response for each amp is nearly identical, something I find surprising and unlikely given than noise is random in nature. However the fall off of the high frequency noise at different rates gives an indication of the response of each circuit.

@maudiojunky Ай бұрын

@@joebloggs4369 If you are seeing a difference it must be the noise spectrum of the opamp (OPA1642) or an artifact of the test setup. Looking at the circuit, there is no HF compensation, so the opamp should be linear and unity gain until the MHz range for small signals.

@thestupendouschumley Ай бұрын

This mic is next on my list after I complete Duffy's Lei87 build in the TLM103 body...hot-rod complete with 3 switch options!

@MarkYoshimotoNemcoff Ай бұрын

It would be a great companion to Duffy’s awesome Lei87.

@miroslawkaras7710 Ай бұрын

I am not audio engineer, but i read abot PDM driven micophone that do not add in pass band harmonics. The main driver you can buy as cheap as dirt. Most of cell phone use the PDM mic.

@MarkYoshimotoNemcoff Ай бұрын

Yup, that tech is definitely advancing fast, but PDM (Pulse Density Modulation) microphones, while innovative in their direct digital output and compact design, aren't ready for the demands of professional voice-over work. Their audio quality, dynamic range, and sensitivity fall short of what’s required to capture the nuance and tonal depth voice-over demands. Additionally, their raw digital bitstream output requires specialized processing, adding unnecessary complexity to workflows where simplicity and efficiency are crucial. In contrast, condenser microphones, like the Neumann U87 or Sennheiser MKH 416, have long been the industry standard for their proven ability to deliver natural, broadcast-quality sound. Until PDM technology catches up, they remain more suited to consumer electronics and casual use than serious voice-over recording.

@miroslawkaras7710 Ай бұрын

@MarkYoshimotoNemcoff With current PDM mic the issue is that they try to put all the eggs in one bag, as this is monolithic design. I am not in audio, I am in RF test I may not have sufficient background. I just wonder where the harmonicas come in microphones. Is that the main copret is the transducer or the driving amplifier. If that is transducer, that is a pretty difficult task, however if that is an amplifier that is a different story. If I build a microphone I will use a conventional transducer and then get that immediately to PDM. Converting analog signals to PDM is very simple. Signal this form is also very immune to line noise.Converting PDM to PCM, with today CPUs or just embedded processors do not look like difficult or expensive end ever any more. Suppressing harmonics at that level that you show in the example mic , with such (relative) wide bands is not a simple task in totaly analog circuit.

@MarkYoshimotoNemcoff Ай бұрын

@@miroslawkaras7710 You raise some excellent points about PDM microphones and the challenges inherent in their monolithic design. In audio applications, particularly for professional voice-over or music recording, the issues often stem from the transducer rather than the amplifier. The transducer’s ability to capture nuanced audio with accuracy and minimal distortion is critical, and even slight imperfections can introduce harmonic distortion or color the sound in undesirable ways. The idea of using a conventional transducer with immediate conversion to PDM is intriguing and could address some of the limitations. However, while converting analog signals to PDM is straightforward, the resulting digital bitstream still requires precision in processing, especially in suppressing harmonics and maintaining audio fidelity. This is particularly challenging in professional audio, where the demands for tonal accuracy and clarity are much higher than in RF testing or consumer audio. Suppression of harmonics in a wideband analog circuit is indeed complex, but professional condenser microphones have refined their analog paths over decades to achieve excellent results. For now, PDM microphones, while resistant to line noise and cost-effective for embedded systems, struggle to compete with the nuanced sound and proven reliability of traditional designs in high-fidelity audio environments.

@billkaroly 8 күн бұрын

Have you built a micparts v57 kit? I wonder how it compares to OPA Alice with DC-DC Hex board. I'll probably build the OPA Hex for a comparison.

@MarkYoshimotoNemcoff 7 күн бұрын

I have and it was fun. What I like about the v57 is that it’s a great kit with great documentation and it’s more customizable than the Alice, but overall, the Alice pre built boards at JLI are a much easier build. The Alice will give you a cleaner sound and less noise, but they are both just different flavors of vanilla.

@billkaroly Күн бұрын

@@MarkYoshimotoNemcoff I bought my second V57 kit in 2018 and its sat on my desk ever since. I bought a Mackie EM-91C and installed a 2555B and the Alice board but I haven't tested it yet. Oh and I ordered another Alice and hex board for a 34mm build.

@BoydTimothy Ай бұрын

It definitely sounds clear! I love your passion about microphones. :) I wonder how this mic would do for singing vocals & instruments? Theoretically, would it be better than, say, the nicest U87? I’d love to hear your thoughts process about what, to you, makes a great microphone. When you evaluate microphones, say, a mic you’ve never used before, what are the things you look/listen for to know a mic is incredibly good? In this case (OPA Alice) you’re impressed with the nearly no THD and the ability to shape the sound with a preamp or in post. Is this a mic you’d use for anything and everything or are you going to reach for a different one for certain applications … and why?

@MarkYoshimotoNemcoff Ай бұрын

Those are great questions and they deserve their own future video to answer. Thank you for the suggestion. As far as if it would be better than a premium mic like a very nice 87, that's really some nebulous territory since I think mic choice is always situation-dependent. While you can really shape the OPA Alice to your preferred sonics, doing so is really going to depend upon your post production abilities while a mic like an 87 gives you a defined, and widely adored sound right from the get go. You can do quite a lot with the Serena here, but if you are asking if this could be the one mic to rule them all, I'd have to lean towards the belief that it's a very valuable mic to have in a collection of tools.

@poindextertunes Ай бұрын

@@MarkYoshimotoNemcoffsounds like it could be a daily driver for recording different sources everyday without having to scour your signal. I record rap vocals and deductive eq always seems to ruin the vocals. especially if you want the “in your face” sound that is rudimentary to the genre. I’ve definitely learned less is more in that arena

@BoydTimothy Ай бұрын

@ Roger that. Yeah, I tried searching through all your videos before commenting to make sure you hadn’t already made a video like that, but it would be extremely interesting to hear all your thoughts regarding selection of mics, what makes them good, what to listen for, what mic characteristics to look for, etc. I face the same thing trying to learn more about mics like the revered Neumann KM-84 (vs a lot of clones or the newer KM-184). So yeah, let’s say you walk into a studio, there are 10 similar-looking mics to choose from, and you aren’t able to see what model/brand they are, how do you choose which one(s) to use? Looking forward to that video. :)

@SoundSleuth Ай бұрын

I own a U87 1980's model that is very clean. My Serena mic is lower noise and lower distortion in side by sides. I have these in a couple studios and they have become their go to vocal mics.

@RegebroRepairs Ай бұрын

Wait, no one else put an Opamp into a Mic before? That's funny.

@MarkYoshimotoNemcoff Ай бұрын

CAD put an op amp into the E350 which had to be run off two internal 9v batteries, but what Jules did was identify an op amp that had lower noise, distortion and could run off phantom power and integrated it into the Alice circuit. The OPA Alice is his contribution to the evolution of mic circuits and that’s what his AES paper is about. Henry Ford didn’t invent the car, he just figured out how to do it for the modern world.

@maudiojunky Ай бұрын

Putting a high-performance opamp into something phantom-powered is less trivial than it sounds. Assuming a 5kOhm source impedance from a 48V source, you've only got about 5-10mA to work with. With its zener-regulated 12V supply we can assume the single-channel OPA Alice is designed to draw at most 7mA from phantom power. The classic low-noise NE5532 draws up to 16mA just idling, and it also has abysmally low input impedance. Older FET-input opamps have great input impedance and improved idle current, but bad voltage noise performance. It's pretty clear this has only become feasible with more recent chips. The OPA1642 in the OPA Alice, a FET-input chip from 2009, brings the idle current figure down to 4.6mA per channel with very high input impedance and comparable voltage noise performance to NE5532.

@jazzmastert Ай бұрын

Have you ever experienced a high noise floor on these specific designs Mark?

@maudiojunky Ай бұрын

@@jazzmastert I can't say from experience, but simulating the circuit gives me a total noise of 1.54µV average, peaking at 2.3µV at 20kHz and dropping below 1µV by 3.75kHz. This can be improved by reducing the size of the feedback resistors on the inverting stage, but further analysis would be needed to see if the increased loading causes other undesirable effects.

@RegebroRepairs Ай бұрын

@@maudiojunky That said, external power supplies are NOT unheard of for high performance microphones...

@GiovannideBari Ай бұрын

This microphone would be dope if used as source mic in a microphone modeling system. Would love to hear a k49 capsule on that circuit... Too dreamy 😍?

@MarkYoshimotoNemcoff Ай бұрын

I actually have an OPA Alice mic with a k49, so I promise I will put that on camera at some point in the future. Also, if you did down into the channel you'll find the Sphere clones I made and they do have OPA Alice circuits in them.

@GiovannideBari Ай бұрын

@@MarkYoshimotoNemcoff Found it! Thank you ☺... Keep going! Really love your videos 🤩

@MarkYoshimotoNemcoff Ай бұрын

@@GiovannideBari You're welcome! And thank you for reminding me to pull out the OPA with the k49 for a future video.

@joefrisco Ай бұрын

What do you think is in the OpAmp?

@MarkYoshimotoNemcoff Ай бұрын

Magic dust. Oh, and probably some kind of gain stage, differential circuit and biasing to mitigate the distortion, but mostly magic dust.

@joebloggs4369 Ай бұрын

It's a Burr Brown OPA 1642 dual op-amp. It's been around for a while.

@fx-studio Ай бұрын

@@joebloggs4369 The OPA 1642 is a good opamp for filtering but not amping as its slew is only 20v/uS, it simply can move fast enough for high frequency sounds.

@maudiojunky Ай бұрын

@@fx-studio That's complete nonsense. You don't need to worry about slew rate for audio work unless you're using something ancient like LM741 or doing something fancy like class D amplification. 20V/µs is fast enough to reproduce frequencies in the megahertz range just fine - audio is no problem.

@KingKong-mp6gj Ай бұрын

While the performance of the OPA Alice is impressive, the nearly -60dB k2 of the simple FET circuit is unhearable. I think that is reason no one really saw a need for improvement.

@MarkYoshimotoNemcoff Ай бұрын

I think the impetus toward moving to an op amp has a lot more to do with how many of the lowest noise FETs like the J305 are obsolete and in limited, dwindling supply and the commercial consideration of automated manufacturing moving away from through hole components. So, it’s less of an indictment of the FET, but more about how it can be replaced in the design process with even higher fidelity.

@rillloudmother Ай бұрын

xiya mi beltalowda: xidawang mic sounds lik wa bosmang.

@badmiddens Ай бұрын

First. Score. Neat.

@MarkYoshimotoNemcoff Ай бұрын

🥇🏆🥳🎉🍾🍻

@wesleybrehm9386 Ай бұрын

Why are you so close to the mic? The proximity effect is distracting, and the mouth noise without a pop filter is pretty distracting too. Is it an artistic choice to add more bass to your voice?

@MarkYoshimotoNemcoff Ай бұрын

I always work 4-5 inches off the mic.

@area51audio Ай бұрын

Because a lot of people love that sound for broadcast?

@wesleybrehm9386 Ай бұрын

@@area51audio You can just as easily add a bass boost to the EQ. It sounds better overall because you aren't sacrificing the high end information, and has a similar effect on the low end. Added bonus: no lip smacking sounds or heavy plosives. Having bad audio on a channel made to showcase audio... isn't a good look.