I think my new Rode NT1-A can be fake.

[rogs]

I have a challenge. Here are the files yet again, but this time they are slightly different..........

I can understand that there is little (no?) advantage in keeping the files as 'dual mono' tracks, but why did you decide to half the sampling rate for your new files?
Did you conclude that the original 96KHz sampling rate was simply unnecessary for this comparison?

I was wondering if any 'foldback' of ultrasonic harmonics that might occur with the higher sampling rate might affect how we perceive some of the higher frequency 'unvoiced' parts of the recording?

 

[kingkorg]

I can understand that there is little (no?) advantage in keeping the files as 'dual mono' tracks, but why did you decide to half the sampling rate for your new files?
Did you conclude that the original 96KHz sampling rate was simply unnecessary for this comparison?

I was wondering if any 'foldback' of ultrasonic harmonics that might occur with the higher sampling rate might affect how we perceive some of the higher frequency 'unvoiced' parts of the recording?

I really don't have opinion on this. It's never ending sample rate battle. I just ended up using 48k as a standard for all of my recordings, as everything ends up online these days at that sample rate. Since i "re-clocked" the takes in order to prevent null testing, i just downsampled to 48k.

The only reason i occasionally use up to 192k is when i know audio files are going to be stretched. Funny no one ever mentions this advantage of higher sample rates. If i have extremely flimsy drummer i go for 192k, which gives me room to stretch files a lot in order to quantize takes without loss in quality.

 

[rogs]

I really don't have opinion on this. It's never ending sample rate battle. I just ended up using 48k as a standard for all of my recordings, as everything ends up online these days at that sample rate. Since i "re-clocked" the takes in order to prevent null testing, i just downsampled to 48k.

The only reason i occasionally use up to 192k is when i know audio files are going to be stretched. Funny no one ever mentions this advantage of higher sample rates. If i have extremely flimsy drummer i go for 192k, which gives me room to stretch files a lot in order to quantize takes without loss in quality.

It was just a thought.... I have trouble in hearing any significant differences between the 2 microphones in the original recordings ... certainly not any differences worth spending an extra £9500 for !

I'm afraid my eyes tend to glaze over whenever I read words like ' breathe' - 'airy' - 'harsh' - 'shrill' - 'forward' - 'congested' - 'poky' - 'dull'- 'bright' etc, etc to replace engineering terms.... but I do accept that some folk genuinely get sucked into the belief that some of these really expensive mics are worth the silly prices charged for them.
All good business for those who manage to do the 'convincing'...

 

[kingkorg]

Made the same guess but I don't think the NT-1 esses are that much worse, U87 esses just sound different, more 'vintage', not necessary as 'true' as with NT-1. I think if we had the waterwall plots for both mics with some plosive inputs it would give us much more information than a frequency response plot.

There's the obvious difference in capsule construction. K47 being single backplate hypercardioid with through holes and two diaphragms. Rode's capsule is whole nother story. Then the headbasket. Obviously u47 is a tube circuit which will introduce some i guess audible harmonics? Compared to Rode which is basically distortion free at this level. So yes i bet waterfall would reveal interesting stuff.

It was just a thought.... I have trouble in hearing any significant differences between the 2 microphones in the original recordings ... certainly not any differences worth spending an extra £9500 for !

I'm afraid my eyes tend to glaze over whenever I read words like ' breathe' - 'airy' - 'harsh' - 'shrill' - 'forward' - 'congested' - 'poky' - 'dull'- 'bright' etc, etc to replace engineering terms.... but I do accept that some folk genuinely get sucked into the belief that some of these really expensive mics are worth the silly prices charged for them.
All good business for those who manage to do the 'convincing'...

I agree 100%. I can't hear the difference in the EQ matched takes. I've taken multiple blind tests, and i just cant do it. But i am glad Sammas chimed in and nailed which is which, proving his point. There has to be a difference between the two, since even non re-clocked EQed takes won't null completely. But how significant the difference is... I guess for people who clearly hear the difference, purchase of a 47 would be perfectly justifiable.

 

[Ricardus]

But i am glad Sammas chimed in and nailed which is which, proving his point. There has to be a difference between the two, since even non re-clocked EQed takes won't null completely. But how significant the difference is... I guess for people who clearly hear the difference, purchase of a 47 would be perfectly justifiable.

It's not statistically significant unless it can be done many times, and on different samples.

 

[Icantthinkofaname]

I don't doubt it's easy to measure, but mfgrs never publish such data. There's sometimes not even fully informative response and polar graphics. Used to be fairly common to see both 90 degree and 180 degree responses for cardioids; hardly ever these days.

It's one of the things I really appreciate about Oktava.

I wonder how much Rode paid that guy to say that.
Schoeps and Rode shouldn't even be in the same sentence

I think Rode could make great mics like Sennheiser and Schoeps if they chose to. The only Rode I've used that I really like is the NTG3, though I haven't used the new NT1. Mainly just the shotgun mics and "shotgun mics" like the VideoMics without interference tubes.

Frankly, I'm a bit surprised Rode doesn't have a supercardioid instrument SDC/small capsule yet. I get that you CAN use a shotgun mic indoors, and it works until it doesn't. I thought about picking up some of the VideoMic NTGs because they're small enough and don't sound too bad when you have the high shelf boost on, but I think the MK-012 hypercardioid (provided you have a quality shock mount and wind screen) just beats them for crash or plant mics or low/no budget location recording (I really think a max SPL of just 120 dB holds the Rode back compared to similar mics).

I really don't have opinion on this. It's never ending sample rate battle. I just ended up using 48k as a standard for all of my recordings, as everything ends up online these days at that sample rate. Since i "re-clocked" the takes in order to prevent null testing, i just downsampled to 48k.

The only reason i occasionally use up to 192k is when i know audio files are going to be stretched. Funny no one ever mentions this advantage of higher sample rates. If i have extremely flimsy drummer i go for 192k, which gives me room to stretch files a lot in order to quantize takes without loss in quality.

That's actually one of the things I like about the extended response of my MKH 8050 and recording in whatever the highest supported resolution of hardware is (right now I'm just using a Zoom F6 so it's 96K as an interface or 192 as just a recorder).

 

[kingkorg]

It's one of the things I really appreciate about Oktava.

I think Rode could make great mics like Sennheiser and Schoeps if they chose to. The only Rode I've used that I really like is the NTG3, though I haven't used the new NT1. Mainly just the shotgun mics and "shotgun mics" like the VideoMics without interference tubes.

Frankly, I'm a bit surprised Rode doesn't have a supercardioid instrument SDC/small capsule yet. I get that you CAN use a shotgun mic indoors, and it works until it doesn't. I thought about picking up some of the VideoMic NTGs because they're small enough and don't sound too bad when you have the high shelf boost on, but I think the MK-012 hypercardioid (provided you have a quality shock mount and wind screen) just beats them for crash or plant mics or low/no budget location recording (I really think a max SPL of just 120 dB holds the Rode back compared to similar mics).

That's actually one of the things I like about the extended response of my MKH 8050 and recording in whatever the highest supported resolution of hardware is (right now I'm just using a Zoom F6 so it's 96K as an interface or 192 as just a recorder).

There's the Rode NT3 which is hypercardioid. Super useful and great sounding mic. There's a mod for lowering the HF boost i came up with if that's an issue. No high spl limit problems, i used it as a main vocal mic for 10 years, spitting, bumping, falling, plosive bombardment 4 days a week on average, still as good as new but with even better low end response compared to new ones. I guess the diaphragm relaxed a bit over time. Which is a good thing.

 

[Sammas]

I am very glad you chimed in, and you are right. B is the NT1. However, the mics are matched for the frequency response, and the matching EQ is applied to the NT1. When played back to back, analyzed with several software analyzers they give the exact same results, frequency wise they are IDENTICAL with the eq applied to NT1!

I cant for the life of me hear any difference between the two even when i loop short, critical sections. As you obviously hear the difference there is no doubt there's something to it. And now we can conclude it's not frequency response related.

Some clock drift is introduced to both files as a measure to prevent people do the null test to find out which is which. So they won't null with the original files.

Nice monitoring and room certainly help hear more detail, but I learnt along the way that it often is just a case of knowing what to listen for rather than not being able to hear it. If you let me know what DAW you use, I can probably send you something to listen to at some point that might make it a bit clearer. Semi-regularly something comes across my desk that is such a pure representation of "a recording-related artifact" for lack of a better term, that it becomes really hard to un-hear it once you realise what it is and how it is reacting and unfolding over time against the source sound.

Speaking of which... time is the aspect that really matters. If you are writing a piece of music, you don't just tell a musician what note to strike and how loud it should be played. You also tell them how long to play it for. Music & sound is inherently time-based. Yet frequency response graphs have no axis for time at all. Pluck a guitar string and it will ring. Pluck a guitar string harder and it will ring for longer. The more energy, the longer the ringing. The less energy, the shorter the ringing. That is the *very* basic premise of resonance as the relationship between frequency and amplitude over time.

An EQ alters the amount of energy at any given frequency. Depending on whether you are cutting or boosting, it might also mean decreasing or emphasising the ringing in the time domain. That is why it is generally recommended to cut with EQs rather than boost, because adding energy increases resonance. It is also why EQ as a means of room correction offers limited success. It can be beneficial to a degree, but the more you chase a perfectly flat frequency response, the more it inevitably results in EQing a two-dimensional frequency response with little consideration of what is happening in a room in the time domain. The more expensive tools like Trinnov room correction use EQ and also do specific processing in the time domain to address ringing.

 

[Sammas]

It was just a thought.... I have trouble in hearing any significant differences between the 2 microphones in the original recordings ... certainly not any differences worth spending an extra £9500 for !

I'm afraid my eyes tend to glaze over whenever I read words like ' breathe' - 'airy' - 'harsh' - 'shrill' - 'forward' - 'congested' - 'poky' - 'dull'- 'bright' etc, etc to replace engineering terms.... but I do accept that some folk genuinely get sucked into the belief that some of these really expensive mics are worth the silly prices charged for them.
All good business for those who manage to do the 'convincing'...

I agree 100%. I can't hear the difference in the EQ matched takes. I've taken multiple blind tests, and i just cant do it. But i am glad Sammas chimed in and nailed which is which, proving his point. There has to be a difference between the two, since even non re-clocked EQed takes won't null completely. But how significant the difference is... I guess for people who clearly hear the difference, purchase of a 47 would be perfectly justifiable.

I just want to be clear that my point has never been pro U47. I totally think the prices are insane...

But I also think that Rode's video comparison is insane. As far as I am concerned, it doesn't matter if the conclusion is drawn from two audio files on the internet, or through the price tag a microphone has... it is one and the same belief to me. It is all the pursuit of a simplistic answer to judge quality, irrespective of the complexity of the production processes these tools exist in within the real world. Or in the case of the U47, the complexity of capital gains and appreciating value of an asset.

Rode is doing nothing but seeking to blur the lines by capitalising on a human tendency to need quick and easy answers.

Subjective listening tests absolutely have value. The work Floyd Toole has done with speaker design is nothing short of breathtaking, and the audio world is all the better for it. Listening tests for AD/DA converters that are stuck in audio equipment racks. Sure, it absolutely simplifies things. With microphones though? They are probably the most complex tool we use, and that rode video really just seeks to diminish their application and use within complex production processes.

As though they are nothing more than a cookie cutter slicing dough.

 

[rogs]

Subjective listening tests absolutely have value.....

We can see that is true from the prices asked - and apparently received - for certain items of audio equipment.
At the end of the day any - and I mean any - deviation from linearity in the audio chain will introduce distortion.
Now that distortion may be perceived as 'attractive' by some, and offensive to others..... But it is all still distortion, from an engineering point of view.
People have made fortunes - for example - from the non linearity introduced into valve (tube) based guitar amplifiers. Often by mistake - or simply bad design, but considered desirable by many folk.

I suspect the lowest distortion figures in microphones probably come from the 'push-pull' capsules fitted to the Sennheiser RF mics .
That should make them the most desirable, technically, but I'm not sure that's true?
The non linearity (distortion) introduced by various - notably tube based - microphones from suppliers like Neumann is often deemed highly desirable.
And of course Neumann will obviously capitalise on that reputation...... Why wouldn't they?

When 'upstart' companies like Rode came along and start introducing much cheaper microphones that actually perform quite well, the professional microphone 'establishment' are inclined to automatically ridicule the performance of these imposters.
It's the same story in many other aspects of audio electronics.... Certain model Hammond organs, certain Fender and Gibson guitars, certain Vox and Marshall amplifiers all have almost religious followings.
Objectively, many of their supposed 'superior' attributes are often nonsense. Subjectively, it's a totally different story.

There are some very capable marketing folk around. You just have to try and make sense of what is valid, and what is BS.
And it's not easy....I suspect we've all made the odd expensive mistake by getting that wrong, along the way!

 

[soliloqueen]

I am very glad you chimed in, and you are right. B is the NT1. However, the mics are matched for the frequency response, and the matching EQ is applied to the NT1. When played back to back, analyzed with several software analyzers they give the exact same results, frequency wise they are IDENTICAL with the eq applied to NT1!

I cant for the life of me hear any difference between the two even when i loop short, critical sections. As you obviously hear the difference there is no doubt there's something to it. And now we can conclude it's not frequency response related.

Some clock drift is introduced to both files as a measure to prevent people do the null test to find out which is which. So they won't null with the original files.

average FR over time != short term FR. if you look on a spectrogram, the frequency response is clearly different at any one point, it's just not different on average.


In the first phrase in this spectrogram, there is a difference of 6dB at 10k, and it certainly sounds that big to me. FR is not a static coloration on the sound of a microphone. FR does not affect the sound, the FR is a measure of how the sound is affected. the capsule folds the sound from the space into an audio signal depending on how it's designed, and every change in position, direction, probably even pressure and intensity etc changes the FR, and how the FR is changed by those things is different for every design. This is not something a static EQ can replicate. It's not "not FR related." It's intimately FR related. It's just that the FR changes on a second-by-second basis under normal use, you know?

i personally can't stand the nt1. i feel like the really high frequencies (12k and up) are way, way too loud. it's all i can hear. it's louder than the rest of the sound to me. it's like being in a room with a tube TV... the nt1 picks up these really offputting sounds that sound like rats scratching in the walls but really high pitch.

 

[kingkorg]

Great post Solilo.

Just to be clear, i'm not saying EQ matching mics like in this case is turning one mic into another. We all know it works to an extent. But like in with this example, i like to try to take the frequency response out of the equation by matching them FR wise, and see how much difference comes from other factors.

All that being said, i still can't hear the difference between the matched files even though i know it exists. That comes down to my hearing and perception i guess.

 

[mhelin]

Maybe there is a way to grab the IR (impulse response) of the mics and then you could generate (calculate) another IR which the when applied to one mic makes it sound like the other (maybe that cannot work 100% as the capsule with the microphone is not a pure minimum phase source but includes non-linearities which cannot be simply undone). Maybe that's already the way EQ matching is implemented anyway, simulating guitar amps and speakers is much more easier than microphones using IR's.

 

[MicMaven]

I have a challenge. Here are the files yet again, but this time they are slightly different. Who wants to guess which is u47, and which NT1? I would like to hear how did you spot the difference, and which part is the giveaway for any of the mics? With timestamps.

Here are GDrive links if attachments don't work.
https://drive.google.com/drive/folders/1R95-TU1wokh6tEJXAAMAnNzfQOO90TYa?usp=sharing

My wife and I both listened to the samples. We agree B is more mellow, thick and rich - so it is the Neumann - A is crisper and more articulate, with poppier ps and crispper consonants. She likes B while I like A - exactly for why they are different.

REGARDLESS, we like both microhones and it would not matter to our needs which one we use. James

 

[MicMaven]

I'm afraid my eyes tend to glaze over whenever I read words like ' breathe' - 'airy' - 'harsh' - 'shrill' - 'forward' - 'congested' - 'poky' - 'dull'- 'bright' etc, etc to replace engineering terms....

Sidebar Comment - Yes, and ... ahem ... then, again, these words sort of make sense to non-engineers. My wife follows my various electronics hobbies, but she is hardly an expert thereon - words like these listed seem to make sense to her and others who listen, but lack the technical background to indicate what they hear with more precisely or more "technically correct" locutions.

I suppose one must adjust to his audience in any field of endeavor. My good friend is an electrical engineer specializing in radio and aerospace product design. I was a trial atty and college professor in a former life. We get along well, but sometimes we must stop and translate what we are saying so we remain on the same page, as it were. Sometimes it is easy, other times not, as he thinks in technical engineering terms, and I have a different lexicon. But we manage well, if we take our time. I wonder whether you are addressing a similar situation with the above commentary. It can be difficult, at times, to get everyone on the same page using the same terminology. I think. James

 

[kingkorg]

My wife and I both listened to the samples. We agree B is more mellow, thick and rich - so it is the Neumann - A is crisper and more articulate, with poppier ps and crispper consonants. She likes B while I like A - exactly for why they are different.

REGARDLESS, we like both microhones and it would not matter to our needs which one we use. James

Sorry, B is actually NT1. Tricky, isn't it?

 

[soliloqueen]

Maybe there is a way to grab the IR (impulse response) of the mics and then you could generate (calculate) another IR which the when applied to one mic makes it sound like the other (maybe that cannot work 100% as the capsule with the microphone is not a pure minimum phase source but includes non-linearities which cannot be simply undone). Maybe that's already the way EQ matching is implemented anyway, simulating guitar amps and speakers is much more easier than microphones using IR's.

Unfortunately an impulse response does not include nonlinearities, by nature. An impulse response includes linear changes only.

 

[kingkorg]

I'll just quote myself from a PM I sent to MicMaven:

"This is something i do regularly. I take two mics, totally different ones and i use quite sofisticated measurement process to match their FR. Once i do, and i'm sure they match, i start the shootout process between the two in different scenarios. This way i eliminate obvious FR difference as a variable and i'm able to concentrate on tiniest of details.

What i've done here is way too simplified version where i use matching eq to bring two files closer over which i had no control during the recording. So the result is not ideal, but close enough to prove the point. Vocals are not ideal source for this, but oh well..."


Differences tend to show best when recording drum room mics facing away from drums, but not too far away from them. This shows best how rear rejection works, and sounds, while front facing the room feeds the front of the capsule with complex off axis dominated reflections of wide frequency range, high spl... Sharp transients involved as well. And it's a beautiful sound as well.

 

[kingkorg]

Unfortunately an impulse response does not include nonlinearities, by nature. An impulse response includes linear changes only.

I am obviously just guessing as i had no control over the recording, but i bet a lot of nonlinearities in this example come simply from the talent moving her head diferently in respect to the mics.

 

[MicMaven]

Sorry, B is actually NT1. Tricky, isn't it?

YEP -- ROOKIE MISTAKE. I appreciate you are not rubbing it in. But, at least I stuck my neck out and took the chance.

Serious Sidebar Comment -- I have performed similar blind A/B/C demonstrations of various dynamic microphones popular among ham radio operators. I made sample recordings of several microphones at multiple price points - from $15 to $99 - and a certain popular brand microphone costing $150 - all recorded at the same level, on the same device, with no EQ - to demonstrate a great many operators "want to believe" spending more money will make their signals them sound better, or (worse) travel farther, than the lower cost models. They sound sufficiently close to each other that NO ONE has ever consistently guessed which mic is which, either recorded on a computer audio interface or, especially, on a transceiver (which are typically rated as high as 10% THD and utilize a mere 2800 Hz transmit audio and RF bandwidth). Obviously, we are not talking Hi Fi here. But they fool themselves into thinking more money will somehow turn a purse into a sows ear.

Of course - YOU GUYS would probably score a lot higher than most ham operators on my test pool, because the very subtle differences between them just might matter to serious studio engineers. But, in MY world, NUTS, they all sound sufficiently alike it really does not matter which one we use. And herein lies the rub - it matters in some cases, and not in others.

Similarly, young podcasters demonstrating and recommending low cost microphones for other podcasters to use may find the NT1a and other low cost microphones well suited to their needs, whereas you guys, recording a symphony or other professional outfit for commercial production may want something else - and for very subtle reasons which matter to you, but not them.

So, I suppose it kinda all depends on whose ox is gored, by whom, and for what purpose. Eh?

James