I don't know what rode is talking about, but the essence of 32-bit "not clipping" is that with the recording you made at 13:10 you should be able to turn the amplification down in audacity and you will see some of the peaks restored where it "clips", which it really doesn't do immediately, the information in the file itself just goes above the 0 db level on the VU meter IF you exported 32 bit and a .wave file... not sure what the mic itself has to do with that... but I feel like that's what they are getting at... .Wave at even 24 bit... maybe 16 bit even has details in the file above 0, so by reducing the amplification you can brings those back.
Once a microphone signal is pre-amped and quantized there is no "amplification", there are only numbers. -8,388,608 to +8,388,608. Nothing more, nothing less. Just one of those numbers every 48,000th of a second if your sample is 48K. One can scale those numbers to look like clipping, or re-scale them to "remove" clipping if they were improperly scaled in the first place. I look at the data, not at "charts".
@@MaxoticsTV I'm literally trying to explain you the truth about what they are talking about, the grain on truth in their marketing BS. If you TAKE that recording you made... in Audacity, highlight it... do to the effect called "amplify" and turn it down -6 db or -12db, the loudest part will not just be a flat line like it is where you see it hitting zero... there IS information there that you cannot see... THAT is what they are talking about when they claim their mic doesn't clip... it's not the mic... it's the 32 bit .wav file that has information beyond 0 DB that exists without being hard clipped and flattened. Try it and you will see for yourself that THAT is the only true part of what they are claiming, it's something anyone who produces music digitally knows.
@@RooftopKoreansMusic I did that and it is flat, clipped. Here are the files if you want to look at yourself: drive.google.com/file/d/1w1CqIbgPhZq4SQfL7QTu-PxELvBPL_ol/view?usp=drive_link
@@MaxoticsTV Something else is going on with your recordings... there must be a limiter also... I did my own tests just now... the 32 bit recording that is over modulated dose retain SOME details when reducing the gain of the recorded clip as I thought it would, although the obvious over modulation is still audible in the audio... that's something they don't address, even if the clipping doesn't occur the over-modulated sound does and is not desirable. I'm using Ableton 12 to do my tests, I'm thinking Audacity has a limiter at zero built in. Maybe the rode software doesn't limit the audio when you record so saving a 32 bit file with information above 0 is possible
@@RooftopKoreansMusic I recorded that with Reaper using Rode's ASIO driver. Again, the results are exactly what I expect. Clipping happens in the preamp long before 24-bits are reached but it does write some garbage in 32-bit float past that--I think--not sure yet. I plan to do some tests comparing both Rode's ASIO and WASAPI.
I think you may be missing the point here. The Rode 'schematic' is only a simple representation of the concept of using more than one A/D converter to increase dynamic range. The highly commerically sensitive processes involved in gain matching, .overlapping and 'meshing' 2 x 24 bit converters to provide a 32 bit float stream are not going to appear in a data sheet! Just because the data sheet says 'mic 1' and 'mic 2' etc doesn't mean that's how they HAVE to be used.... There will be sophisicated proprietory processing involved here. They are unlikely to publish how it's being done in detail!
I get this all the time. "you don't understand the proprietary technique that makes 'x' possible.' They don't need to publish how it's done. All they need to do is provide reproducible tests and, well, hearing will be believing! I've explained in various videos why it isn't possible, though I'm sure it's not for trying. Again 1) no explanation in chips that the manufacturer would love to sell everyone and anyone 2) no empirical proof that 32-bit float recovers more microphone data than 24-bit. Thanks for comment!
@@MaxoticsTV As I mentioned above, I think you've missed the point. The 32 bit float format has a dynamic range which exceeds 1500dB .. which is astonishing ( and unclippable! ) 24 bit PCM is limited to 144dB (max). The electronics interfaced to create audio files in the float format may well clip -- depending on how things are set up. The 32 float files themselves will exist intact, unclipped. Where they really excel is at the lower signal level settings. 24 bit fixed point recordings will start to degrade, noise wise, at the lowest levels. Not true for 32 float files. This video (from 5.38 onwards) illustrates some of the dramatic advantages of 32 bit float: ruclips.net/video/y4oNd1RgGL0/видео.html I suspect you've got your Rode set up wrongly. You say you can't see the reason for the gain control. It's just to allow sensible monitor levels . Any clipping will be a function of the Rode mic electronics, not the 32 bit float file format itself.
@@arjay1949 Everyone missing the bus here ;) I understand the range of 32-bit float. There are some issues with 32-bit float but yes, practically speaking, it is unclippable. My guess is you don't work with data. Forget dbs, they are a scale that allows us to treat the data in a way that we hear (exponentially). Curtis was fooled by the software, but I won't get into that. Suffice it to say, all ADCs output data in linear format. In 24-bit thats 000000000000000000000000 to 111111111111111111111111, or −8,388,608 to +8,388,607 (it's split in half for negative/positive air pressure voltage) You can draw your x/y axis anyway you want, scale it in dbs or whatever, but THE FACT IS AN ADC can only output BITS! Wherever you set your gain you'll even end up with 4 bits (if not 8 bits or even 10) of noise or "clipped" data. I didn't create the world, this is how we found it. This information is found in every mic specification. When you read 60db of dr think 10bits. When you read 120db, think 20bits, but seldom is such a range used for a variety of reasons. You have it right in your last sentence. Clipping is a function of what the pre-amp sends the input of the ADC. You simply CANNOT get 24-bit of usable data into the pre-amp. Your argument is is you have 32-bits (float) that it changes that. How? Whether a ruler is 12 inches or 24 inches it will never make 10 inches more than 12 inches ;) Cheers!
@@MaxoticsTV Probably best we agree to differ? The 'internet' seems very divided on the usefullness of the 32 bit float format. Some folk (like me) find it useful .... Others (like you) don't. Problem is with the internet is that it's almost impossible to know which technical comments are valid, and which are not. The only results you can really trust are the ones you deduce for yourself! ...... (I imagine you'll be selling or returning your Rode NT1 Gen5 ?)
@@arjay1949 I haven't finished my analysis yet! So who knows! I might be wrong about the usefulness of 32-bit float recording. I'm working on it some more right this second. Keep an eye on my channel if you're interesting. My favorite mic is the Electrovoice RE20, so I can't see using the NT1, but again, I try to keep an open mind!
This is so interesting (Thanks!) So generally I've always thought my money should go into buying the best mics and pre-amps I can afford, because of this. The ADCs you'd think either work - or they don't ? as its just 0 to 1 - and any recently manufactured ADCs should be fine. I'm using an old Gen 1 Focusrite because precisely because it has 6 line level inputs into which I can use some decent pre-amps, and ignore the two built in ones. It seems great at 48/24. So Is there any point I guess ? in buying expensive ADCs if you have your own decent pre-amps ? I ask because it's getting hard to find budget ADCs with a decent number of line inputs as they all seem to want to rave about their built-in pre-amps and charge accordingly ..Do they all use much the same chips or are some 0 to 1 conversions "better". I instinctively think its not a good idea to chain pre-amps (esp good ones with cheaper ones) .
I speculate a bit in this video. The manufacturers test to show some benefit but no one is paid to test whether the benefit is ever noticed. I believe the only reason some old Audio Interfaces begin to fail is discrete components like capacitors surrounding the chips, not the chips themselves. Again, I'm not an electronics engineer. I think it's easy to digitize any line voltage. What's difficult is interfacing with PC, Apple, Linux, Android, iPad, etc. Audio remains a second class citizen on most operating system--I mean for the audio engineer. But with the Behringer UMC1820 for $190 what more does a home studio really need in brining in 8 channels of analog from mics? Want another 8? For $150 Behringer has the ADA8200? There's so much of this equipment out there. Lost of people raving about their gear, but none saying theirs converts to digital better than the other. Again, I'm just looking at it as an outsider.
@@MaxoticsTV thanks - that's all good to know. 👌There does seem a bit of a gap in the market for budget ADCs with say 8-16 line level inputs/outputs without any preamps ? Ferrofish and Motu seem to have them at I guess mid level price.. but not really budget - Then there's the pro level converters with that weird plug interface. The Behringer stuff would do but its full of preamps, whereas I'd be happy with just mono jacks line in/out without a pad, etc. Guess they're about - I'll keep looking and hope the Gen 1 keeps going ..😬
@ well for me, it’s convenience . Set up a couple of mics, mic an acoustic & a cabinet and stereo lines from a keyboard and some sort of audio in and bang - there’s 8. I really like to have everything ready to go before that fleeting flash of inspiration flies past. I know that’s common - sure not all projects are the same but being ready to go for most things (for me) helps a lot👀🔊
If no one notices the bottleneck does it exist? ;) It wouldn't bother me. To me, the worst part of the mic is putting the USB port within the XLR port. The general engineer literacy of the buying public keeps sinking to new lows haha.
![Blox Fruits Dragon Rework Update [Full Stream]](http://i.ytimg.com/vi/EqDAp8udhm0/mqdefault.jpg)
I don't know what rode is talking about, but the essence of 32-bit "not clipping" is that with the recording you made at 13:10 you should be able to turn the amplification down in audacity and you will see some of the peaks restored where it "clips", which it really doesn't do immediately, the information in the file itself just goes above the 0 db level on the VU meter IF you exported 32 bit and a .wave file... not sure what the mic itself has to do with that... but I feel like that's what they are getting at... .Wave at even 24 bit... maybe 16 bit even has details in the file above 0, so by reducing the amplification you can brings those back.
Once a microphone signal is pre-amped and quantized there is no "amplification", there are only numbers. -8,388,608 to +8,388,608. Nothing more, nothing less. Just one of those numbers every 48,000th of a second if your sample is 48K. One can scale those numbers to look like clipping, or re-scale them to "remove" clipping if they were improperly scaled in the first place. I look at the data, not at "charts".
@@MaxoticsTV I'm literally trying to explain you the truth about what they are talking about, the grain on truth in their marketing BS. If you TAKE that recording you made... in Audacity, highlight it... do to the effect called "amplify" and turn it down -6 db or -12db, the loudest part will not just be a flat line like it is where you see it hitting zero... there IS information there that you cannot see... THAT is what they are talking about when they claim their mic doesn't clip... it's not the mic... it's the 32 bit .wav file that has information beyond 0 DB that exists without being hard clipped and flattened. Try it and you will see for yourself that THAT is the only true part of what they are claiming, it's something anyone who produces music digitally knows.
@@RooftopKoreansMusic I did that and it is flat, clipped. Here are the files if you want to look at yourself: drive.google.com/file/d/1w1CqIbgPhZq4SQfL7QTu-PxELvBPL_ol/view?usp=drive_link
@@MaxoticsTV Something else is going on with your recordings... there must be a limiter also... I did my own tests just now... the 32 bit recording that is over modulated dose retain SOME details when reducing the gain of the recorded clip as I thought it would, although the obvious over modulation is still audible in the audio... that's something they don't address, even if the clipping doesn't occur the over-modulated sound does and is not desirable. I'm using Ableton 12 to do my tests, I'm thinking Audacity has a limiter at zero built in. Maybe the rode software doesn't limit the audio when you record so saving a 32 bit file with information above 0 is possible
@@RooftopKoreansMusic I recorded that with Reaper using Rode's ASIO driver. Again, the results are exactly what I expect. Clipping happens in the preamp long before 24-bits are reached but it does write some garbage in 32-bit float past that--I think--not sure yet. I plan to do some tests comparing both Rode's ASIO and WASAPI.
I think you may be missing the point here. The Rode 'schematic' is only a simple representation of the concept of using more than one A/D converter to increase dynamic range.
The highly commerically sensitive processes involved in gain matching, .overlapping and 'meshing' 2 x 24 bit converters to provide a 32 bit float stream are not going to appear in a data sheet!
Just because the data sheet says 'mic 1' and 'mic 2' etc doesn't mean that's how they HAVE to be used.... There will be sophisicated proprietory processing involved here. They are unlikely to publish how it's being done in detail!
I get this all the time. "you don't understand the proprietary technique that makes 'x' possible.' They don't need to publish how it's done. All they need to do is provide reproducible tests and, well, hearing will be believing! I've explained in various videos why it isn't possible, though I'm sure it's not for trying. Again 1) no explanation in chips that the manufacturer would love to sell everyone and anyone 2) no empirical proof that 32-bit float recovers more microphone data than 24-bit. Thanks for comment!
@@MaxoticsTV As I mentioned above, I think you've missed the point. The 32 bit float format has a dynamic range which exceeds 1500dB .. which is astonishing ( and unclippable! ) 24 bit PCM is limited to 144dB (max). The electronics interfaced to create audio files in the float format may well clip -- depending on how things are set up. The 32 float files themselves will exist intact, unclipped.
Where they really excel is at the lower signal level settings. 24 bit fixed point recordings will start to degrade, noise wise, at the lowest levels. Not true for 32 float files. This video (from 5.38 onwards) illustrates some of the dramatic advantages of 32 bit float: ruclips.net/video/y4oNd1RgGL0/видео.html
I suspect you've got your Rode set up wrongly. You say you can't see the reason for the gain control. It's just to allow sensible monitor levels .
Any clipping will be a function of the Rode mic electronics, not the 32 bit float file format itself.
@@arjay1949 Everyone missing the bus here ;) I understand the range of 32-bit float. There are some issues with 32-bit float but yes, practically speaking, it is unclippable. My guess is you don't work with data. Forget dbs, they are a scale that allows us to treat the data in a way that we hear (exponentially). Curtis was fooled by the software, but I won't get into that. Suffice it to say, all ADCs output data in linear format. In 24-bit thats 000000000000000000000000 to 111111111111111111111111, or −8,388,608 to +8,388,607 (it's split in half for negative/positive air pressure voltage) You can draw your x/y axis anyway you want, scale it in dbs or whatever, but THE FACT IS AN ADC can only output BITS!
Wherever you set your gain you'll even end up with 4 bits (if not 8 bits or even 10) of noise or "clipped" data. I didn't create the world, this is how we found it. This information is found in every mic specification. When you read 60db of dr think 10bits.
When you read 120db, think 20bits, but seldom is such a range used for a variety of reasons.
You have it right in your last sentence. Clipping is a function of what the pre-amp sends the input of the ADC. You simply CANNOT get 24-bit of usable data into the pre-amp. Your argument is is you have 32-bits (float) that it changes that. How?
Whether a ruler is 12 inches or 24 inches it will never make 10 inches more than 12 inches ;) Cheers!
@@MaxoticsTV Probably best we agree to differ? The 'internet' seems very divided on the usefullness of the 32 bit float format.
Some folk (like me) find it useful .... Others (like you) don't.
Problem is with the internet is that it's almost impossible to know which technical comments are valid, and which are not. The only results you can really trust are the ones you deduce for yourself! ...... (I imagine you'll be selling or returning your Rode NT1 Gen5 ?)
@@arjay1949 I haven't finished my analysis yet! So who knows! I might be wrong about the usefulness of 32-bit float recording. I'm working on it some more right this second. Keep an eye on my channel if you're interesting. My favorite mic is the Electrovoice RE20, so I can't see using the NT1, but again, I try to keep an open mind!
This is so interesting (Thanks!) So generally I've always thought my money should go into buying the best mics and pre-amps I can afford, because of this. The ADCs you'd think either work - or they don't ? as its just 0 to 1 - and any recently manufactured ADCs should be fine. I'm using an old Gen 1 Focusrite because precisely because it has 6 line level inputs into which I can use some decent pre-amps, and ignore the two built in ones. It seems great at 48/24.
So Is there any point I guess ? in buying expensive ADCs if you have your own decent pre-amps ? I ask because it's getting hard to find budget ADCs with a decent number of line inputs as they all seem to want to rave about their built-in pre-amps and charge accordingly ..Do they all use much the same chips or are some 0 to 1 conversions "better". I instinctively think its not a good idea to chain pre-amps (esp good ones with cheaper ones) .
I speculate a bit in this video. The manufacturers test to show some benefit but no one is paid to test whether the benefit is ever noticed. I believe the only reason some old Audio Interfaces begin to fail is discrete components like capacitors surrounding the chips, not the chips themselves. Again, I'm not an electronics engineer.
I think it's easy to digitize any line voltage. What's difficult is interfacing with PC, Apple, Linux, Android, iPad, etc. Audio remains a second class citizen on most operating system--I mean for the audio engineer. But with the Behringer UMC1820 for $190 what more does a home studio really need in brining in 8 channels of analog from mics? Want another 8? For $150 Behringer has the ADA8200? There's so much of this equipment out there. Lost of people raving about their gear, but none saying theirs converts to digital better than the other. Again, I'm just looking at it as an outsider.
@@MaxoticsTV thanks - that's all good to know. 👌There does seem a bit of a gap in the market for budget ADCs with say 8-16 line level inputs/outputs without any preamps ? Ferrofish and Motu seem to have them at I guess mid level price.. but not really budget - Then there's the pro level converters with that weird plug interface. The Behringer stuff would do but its full of preamps, whereas I'd be happy with just mono jacks line in/out without a pad, etc. Guess they're about - I'll keep looking and hope the Gen 1 keeps going ..😬
@@pch554 Who in the home market has 8 people over playing at the same time? ;)
@ well for me, it’s convenience . Set up a couple of mics, mic an acoustic & a cabinet and stereo lines from a keyboard and some sort of audio in and bang - there’s 8. I really like to have everything ready to go before that fleeting flash of inspiration flies past. I know that’s common - sure not all projects are the same but being ready to go for most things (for me) helps a lot👀🔊
@@pch554 For me it's 8 blinky lights make me happier than 2 ;)
Interesting 🤨 This chip story looks like a bottleneck of the product 🤔
If no one notices the bottleneck does it exist? ;) It wouldn't bother me. To me, the worst part of the mic is putting the USB port within the XLR port. The general engineer literacy of the buying public keeps sinking to new lows haha.