M-Audio AIR 192x4 bit and sample rate

[MattBass87]

So I bought a M-Audio AIR 192x4 on sale and it all works nicely but it says it does

24-bit/192 kHz


The properties in windows sound options for the speakers/headphones and input recoridng you can only choose between

16-bit / 48000 Hz dvd quality

24-bit / 48000 Hz studio quality

32-bit / 48000 Hz studio quality


But isn't 192khz 192000hz? those are all 48000hz..

Is 32-bit 48000hz equal or better than 24-bit 192000hz?


Thanks for any help.

 

[Wood and Wire]

Don't use it as your Windows sound device.

Use it (with its dedicated ASIO driver) with a DAW, to utilise its full capabilities.

32-bit floating point is "better" than 24-bit fixed point, in that you effectively have infinite headroom regarding "in the box" summing (that is to say, if the signal clips internally, you can just reduce the level, after clipping occurs, and the clean signal is completely restored).

However, your converters won't take advantage of 32-bit processing, and you'll use significantly more processing power, and hard drive space - so for the most part, you're better of learning how to gain stage properly, and sticking to 24-bit.

I can't think of a single good reason why anyone would choose to record at 192kHz, unless they're recording movie sound effects, that are intended to be pitch shifted down by several octaves.

 

[Bonafide]

But isn't 192khz 192000hz? those are all 48000hz..

Is 32-bit 48000hz equal or better than 24-bit 192000hz?

They are both 'better' for different reasons. But even 32 bit floating is rarely practical unless your entire system can support it and you have the experience to hear the differences from converters to your monitor chain.

Are you asking for theoretical reasons or pragmatic reason? Theoretically, all things being equal, 32 bit has more headroom (above 0db). But not pragmatic.

48k vs 192k = sample rate (how many samples are captured per second) If we ignore the Nyquist frequency involved in pro audio then yes 192k is better. But since the point of diminishing returns happens much earlier (even when the entire system can support it) you are left with nearly unusable large files and little or no audible difference.

For the sake of professionalism, many studio record at 24-96k, some at 32-48k but most still record at 24-48k or 24-44.1K and some mastering houses do record at 192 for archival reasons.

24 bit can capture all the needed headroom you'll need, granted you understand digital headroom. And 48k after conversion is still above the range of most human detection and therefore above perceivable distortion or artifacts.

 

[MattBass87]

I'm using reaper free trial daw with ampeg svt suite.. .. have no real intention to record yet just wondered why i couldn't select the 24-bit 192000hz

 

[silky smoove]

16/44.1 is perfectly fine for 99% or uses.

Additional bits of fixed point integers just means more headroom above the theoretical noise floor. Assume 6db per bit which gives you 144db of headroom above the point of noise introduction at 24-bit. That is exceedingly far beyond the needs of any practical music recording application.

Higher sample rates give you more room between your system’s antialiasing filter, if operable, and Nyquist. It also gives you less latency. Lastly it gives you increased demands on your computer both in terms of operation and storage. Use plugins that don’t alias, track with a chain that yields 8ms or less of latency and you’ll never need to use a higher sample rate for most cases. 44.1 is the standard deliverable format for music. 48 is the standard deliverable format for audio used in video. Either are fine for most applications as up or down sampling is sonically invisible for most situations.

 

[Krisfield]

HDTracks.com offers 192khz/24bit. I think this is preferable for classical music. Audiophile quality. 44.1khz/16bit is CD quality.

 

[Bonafide]

I'm using reaper free trial daw with ampeg svt suite.. .. have no real intention to record yet just wondered why i couldn't select the 24-bit 192000hz

As is the case with so many forum posts > there is benefit from educated responses laid out in simple terms to a question that wasn't meant to be asked, so the effort isn't wasted. The right question wasn't asked to begin with or was not clear enough to get the simple answer from the start, although @Wood and Wire answered it for you inadvertently.

HDTracks.com offers 192khz/24bit. I think this is preferable for classical music. Audiophile quality. 44.1khz/16bit is CD quality.

Recording and playback are different subjects. YouTube has '24 bit' audiophile videos where listeners swoon and praise...even though it is being played back to them at 128kbps AAC or 256kbps AAC at best.

A very small group of audiophiles subscribe to the 192k playback theory. But it has been proven to death it has little relevance and even detrimental effects. Audio artifacts increase quickly and steadily at the ultrasonic end of the spectrum. While you can’t actually hear the ultrasonic frequencies themselves, you can discern the artifacts they create in the frequencies you are able hear. On the opposite end, sometimes what we call ‘musical’ is little more than the absence of certain frequencies or the increase of the ones we find pleasing. Even cables can do this.

Anything over 48k (listening/playback) has no value. It’s supportive in pro audio applications for oversampling and for creating a nice headroom buffer for the analog filters. 25 years ago this was not the case, but it is such the norm today that it simply goes unnoticed because the problems that it is solving do not arise.

In the end, we can only hear what our playback system/environment can support. That said, the greatest piece of audio gear is our ears. The majority simply do not practice listening for the greater detail that is already there. Generalities about sound, thrown into the collective mix as gospel are usually little more than limited experience.

 

[Krisfield]

My current project is being recorded at 44.1/24. I bounce the MP3 file format at 320kbps, and a WAV file format at 44.1/24. Spotify excepts FLAC or WAV file formats. MP3 for Soundcloud that can be played on everything. AAC if it’s going to a radio station.

As is the case with so many forum posts > there is benefit from educated responses laid out in simple terms to a question that wasn't meant to be asked, so the effort isn't wasted. The right question wasn't asked to begin with or was not clear enough to get the simple answer from the start, although @Wood and Wire answered it for you inadvertently.

Recording and playback are different subjects. YouTube has '24 bit' audiophile videos where listeners swoon and praise...even though it is being played back to them at 128kbps AAC or 256kbps AAC at best.

A very small group of audiophiles subscribe to the 192k playback theory. But it has been proven to death it has little relevance and even detrimental effects. Audio artifacts increase quickly and steadily at the ultrasonic end of the spectrum. While you can’t actually hear the ultrasonic frequencies themselves, you can discern the artifacts they create in the frequencies you are able hear. On the opposite end, sometimes what we call ‘musical’ is little more than the absence of certain frequencies or the increase of the ones we find pleasing. Even cables can do this.

Anything over 48k (listening/playback) has no value. It’s supportive in pro audio applications for oversampling and for creating a nice headroom buffer for the analog filters. 25 years ago this was not the case, but it is such the norm today that it simply goes unnoticed because the problems that it is solving do not arise.

In the end, we can only hear what our playback system/environment can support. That said, the greatest piece of audio gear is our ears. The majority simply do not practice listening for the greater detail that is already there. Generalities about sound, thrown into the collective mix as gospel are usually little more than limited experience.

 

[And I]

Use the m-audio control panel to adjust the sample rate.

Recording at 192khz is going to eat up disk space fast and in a home studio you literally won't notice the difference vs 44.1khz.

 

[Aloe]

there's a popular opinion that 24 bits in a modern studio is a must, but you don't really need anything above 48k. I proudly disagree with it: I need and use 24/96, albeit 24/192 would be even nicer.

if you don't hear the difference in 96k and 44k -- well, you can use 44k, but I do hear three things (even if the final output will be 16/44):
1. I like how my ADCs work better at higher sample rate. they're simply recording better sound
2. plugins (virtual instrumens, reverbs) running at 96k sound superior to the 44.1k. it's notable even if your source recordings are 16/44
3. higher sample rates give you more headroom if you're pitch-shifting or time-stretching stuff, i.e. the result will sound better

I'd say that 24/96 is a studio standard now and been for years, but you'll always find people who claim, they don't hear the difference. try to hear it yourself, if they cannot it doesn't mean, you won't.

 

[silky smoove]

Avoid speculation of sample rate performance. Get analytical:

Review that and understand what aliasing is and you’ll have all the answers. Bit depth becomes irrelevant as even 16-bit provides 96db of headroom over the noise floor.

 

[Aloe]

Review that and understand what aliasing is and you’ll have all the answers. Bit depth becomes irrelevant as even 16-bit provides 96db of headroom over the noise floor.

and that's okay, if you're not going to dramastically boost some signals. there's nothing wrong of having better headroom, so you can turn quiet sounds loud and they'll sound better, at the expense of 1/3 larger files.

if you're mastering for 16bit, i.e. 96dB of headroom in final audio, you might wish more headroom for editing. I do. YMMV.

 

[Krisfield]

Not my image, but I keep this in my study/resources folder as a visual representation of sample rates and file formats. What it doesn't show is track file size. MP3 can be be as little as 3.8MB, and a WAV file as much as 63MB.

 

[silky smoove]

and that's okay, if you're not going to dramastically boost some signals.

If you’re boosting signals so greatly that a bit depth of 16 (96db) or the more common 24 (144db) isn’t sufficient, the problem isn’t the file format, it’s the engineer.

 

[Aloe]

If you’re boosting signals so greatly that a bit depth of 16 (96db) or the more common 24 (144db) isn’t sufficient, the problem isn’t the file format, it’s the engineer.

I'm advocating for 24bit, right, it's sufficient and even a bit more that's needed in real life.

 

[silky smoove]

I'm advocating for 24bit, right, it's sufficient and even a bit more that's needed in real life.

Absolutely. My point is that even 16-bit is more than sufficient for any purpose that does not stem from poor engineering practices.

 

[Aloe]

Absolutely. My point is that even 16-bit is more than sufficient for any purpose that does not stem from poor engineering practices.

well, 16bit inside my DAW sometimes makes me wish for a compressor/limiter before the audio interface.
with 24bit I can just record and apply compression/limiting later in software.

 

[silky smoove]

well, 16bit inside my DAW sometimes makes me wish for a compressor/limiter before the audio interface.
with 24bit I can just record and apply compression/limiting later in software.

How does the bit depth have any effect on that determination? Bit depth just dictates signal level above the noise floor.

 

[Bonafide]

The digital math is the digital math – the numbers aren’t going to change, so the analytical part of the conversation is never the full picture. Professional studios and experienced engineers use 24 bit as a capture standard. Despite any skill or ability to manage 16 bit dynamic range, there is no reason for recording with it at this stage. There are plenty of sources that have a wider dynamic range than 96dB and gear with a noise floor above 96db. Regardless of our ability to hear the detail (even the acoustic environment and ambient noise for that matter) capturing that information is both relevant and important. Speaking from personal experience, if you use analog gear or vintage analog particularly, higher dynamic range allows for lower levels to be captured reducing the noise floor of the recorded signal. It adds up.

All that aside and the fact that this topic goes in tiresome circles…the way one's entire signal chain (gear) manages the math is the relevant portion, and/or what it does sonically (perceptibly) to your signal. Same with conversion. Converter chip ‘specs’ mean little outside of the analog implementation. That is why high end converters such as Crane Song, Burl, RME, Lynx, Prism etc all sound so different. It is a constant give and take. If you are dedicated to good audio, it can take considerable time to refine and develop a process with ultra stable clocking and keeping noticeable quantization errors, dither noise etc to a personal sonic satisfaction. Does it matter? Nope. For most, not in the least. Most home studios aren't equipped to hear it with enough detail to matter.

Truth is plenty of people just don’t care or don’t notice. If it sounds good based on the reference chain they have, they go with it. In the end, it is always a balance of the two. 16 bit may be sufficient, but 24 bit is capable of capturing considerable more and has zero downside in a proper chain.

 

[Krisfield]

I’m enjoying this thread. I’m in the process of setting up an independent record label, so I’m having to research and learn about all aspects of recording and distribution. When you master then send your WAV file for distribution to Spotify, they will downsample the file to their streaming specs which is 44.1/16. That’s because if someone is streaming over 4G at 1.4Mbps, you’ll get uninterrupted streaming. So if you distribute a file at 96/24 to Spotify, it will be downsampled. YouTube’s specs are different, as are Tidal. It’s recommended that if you’re recording at 96/24, that you downsample the master yourself. Here’s the interesting part. Some recording studios are setting up two DAW’s synced together, with one specifically for Spotify specs, 44.1/16. The other for other platforms. If you’re licensing your recording, you’ll need a DAW setup at 96/24.