Does turning down the "preamp" slider in the itunes equaliser destroy dynamic range?
Hi all,
Does turning down the "preamp" slider in the itunes equaliser destroy dynamic range, or does it just simply move the entire range up or down?
Your advice is appreciated!
Greg.
Hi all,
Does turning down the "preamp" slider in the itunes equaliser destroy dynamic range, or does it just simply move the entire range up or down?
Your advice is appreciated!
Greg.
Turning it *down* prevents clipping/limiting if you use too much EQ. Ideally, the preamp should be turned down by the amount that your highest EQ band is turned up.
So turning the preamp down preserves dynamic range. I think this answers your question. Turning it up will either hard clip or limit, depending on the player. Winamp has had a limiter (a fairly poor one too) since version 5, i don't know about iTunes but i'm fairly sure it has a built-in limiter too.
So turning the preamp down preserves dynamic range. I think this answers your question. Turning it up will either hard clip or limit, depending on the player. Winamp has had a limiter (a fairly poor one too) since version 5, i don't know about iTunes but i'm fairly sure it has a built-in limiter too.
Thanks. I _think_ you have answered my question. I have my EQ flat, but I have turned down the preamp as the output of my DAC (benchmark dac1) is far too high and requires the volume pot to be turned right down for normal listening levels.
Any digital volume control (turning down) reduces dynamic range - it is done by dividing each sample by a fixed int number > 1 (or multiplying by float < 1). Thus the maximum possible value in the signal is reduced, while the minimum cannot be further lowered - the least significant bit.
Analog volume control reduces the dynamic range too if you take into account the fixed noise level of your chain. It is perfectly possible you will not tell a difference between reducing volume in the digital domain (via your PC) or the analog one (turning the volume knob on your amp).
Analog volume control reduces the dynamic range too if you take into account the fixed noise level of your chain. It is perfectly possible you will not tell a difference between reducing volume in the digital domain (via your PC) or the analog one (turning the volume knob on your amp).
Hi
As far as I understand it, any digital volume control will decrease the bit depth.
-so if you're playing a 16bit file out of a 16bit sound-card, the highest digital value (given to the top of the waveform of the loudest piece of audio) is 65535 (decimal equivalent). When you have a digital volume control set at 50%, the highest used value would be 32767 (effectively using 15 of the 16 bits).
As mentioned before, any volume control (digital or analog) will reduce dynamic range. Digital no more than analog though.
In conclusion:
-a digital volume control is theoretically less accurate than an analog one, but only very slightly, and it is perfectly acceptable (especially if you've 'run out' of analog control).
-I use it, and I recommend you do too if it is convenient.
It sounds like your amp is 'too powerful' for you application.
-you might benefit from some kind of output-based volume control (power soak), like a transformer-based thing. It's difficult to find good quality ones though (most would degrade the audio more than the digital control).
-alternatively, you could cheaply and easily build a bypass-able fixed voltage divider/attenuator on the line in to the amp. This would give the amp volume more 'resolution' for lack of a better word. -some of the old NAD amps have a "low power" button, which does this. You could have 3 or 4 levels. Just use good quality 1% resistors so you preserve your noise floor and stereo accuracy.
IN --------+------- Bypass (full level)
|
>
> 2k2 1%
>
|
+------- Loud
|
>
> 2k2 1%
>
|
+------- Mid
|
>
> 2k2 1%
>
|
+------- Soft
|
>
> 2k2 1%
>
|
_
GND
As far as I understand it, any digital volume control will decrease the bit depth.
-so if you're playing a 16bit file out of a 16bit sound-card, the highest digital value (given to the top of the waveform of the loudest piece of audio) is 65535 (decimal equivalent). When you have a digital volume control set at 50%, the highest used value would be 32767 (effectively using 15 of the 16 bits).
As mentioned before, any volume control (digital or analog) will reduce dynamic range. Digital no more than analog though.
In conclusion:
-a digital volume control is theoretically less accurate than an analog one, but only very slightly, and it is perfectly acceptable (especially if you've 'run out' of analog control).
-I use it, and I recommend you do too if it is convenient.
It sounds like your amp is 'too powerful' for you application.
-you might benefit from some kind of output-based volume control (power soak), like a transformer-based thing. It's difficult to find good quality ones though (most would degrade the audio more than the digital control).
-alternatively, you could cheaply and easily build a bypass-able fixed voltage divider/attenuator on the line in to the amp. This would give the amp volume more 'resolution' for lack of a better word. -some of the old NAD amps have a "low power" button, which does this. You could have 3 or 4 levels. Just use good quality 1% resistors so you preserve your noise floor and stereo accuracy.
IN --------+------- Bypass (full level)
|
>
> 2k2 1%
>
|
+------- Loud
|
>
> 2k2 1%
>
|
+------- Mid
|
>
> 2k2 1%
>
|
+------- Soft
|
>
> 2k2 1%
>
|
_
GND
Pedantic expansion on the subject coming up..
But first a disclaimer... *As you get to know me you'll discover that it's never my intent to belittle others but instead to help pass on COMPLETE knowledge. The Interwebs are great for finding half the story. Also, since I'm new around here I figures it helps to be explicit about my motives.
Anyway...
As mentioned, reducing level digitally reduces the bit-depth in the signal, which reduces dynamic range, increases signal:noise ratio (digital systems have a minimum floor of 1 bit, but frequently the first two bits are pretty close to meaningless random noise due to a whole host of factors).
What's interesting about this is if you turn UP (shift the bits upwards) of a digital waveform you will NOT INCREASE the dynamic range of the original signal either. You do increase level but could wind up reducing the range if you go to far and clip the signal and lose data as a result. Ideally by increasing level digitally you simply have the least significant bits filled with zeroes when you increase level. These bits could be then be filled with additional information which will possibly be audible -Although that's a discussion for another day.*
The same theory roughly applies to analog audio, but with one mjor difference. In the analog realm, increasing the level of a signal moves the noise floor upwards by the same amount.
The difference here is subtle. Both the digital noise floor and analog noise floors move up in level, but with the digital system you now have spare bits at the bottom of your range that are doing nothing, and therefor can now hold useful information. The analog noise floor on the other hand prevents additional low level information being added.
This effect is a major reason why studio gear usually runs at 24bit (fixed) precision. It allows there to be enough "Raw bits" in the source to allow multiple changes in gain and still be able to output full range 16 bit audio. Also it allows for signals to be recorded at much lower levels and retain enough resolution to be usable.*
As a side note here, it's a standard rookie mistake to think everything must be recorded so hot that every peak is near 0dBFS because more bits in the recorded signal must be better. Unfortunately that's dead wrong, because when you start to mix all these super hot signals together you're quickly going to run out of bits in your master bus and have to turn things down to prevent clipping. The practical result being that you may as well have recorded everything at a lower level in the first place and not run the risk of accidentally clipping your recording or overdriving the analog front end into distortion. Yike.
If someone has time to kill here's a nice practical experiment to illustrate the effect of changing levels digitally:-
Equipment needed:*
A multitrack track editor, pretty much anything will do including a whole slew of free ones.
Some music
Some time
Techniques used:-
Inverted null summing
1) Pick a favorite track with a very wide dynamic range, so not utterly compressed dance music, pop or numetal. A favorite of mine is simple minds Belfast Child, but something like O Fortuna (Old Spice theme in the UK) works just as well.
2) Copy from CD (DONOT use an mp3 as your source) into in the editor and save as "reference 16". *
3) change the bit depth of the file to 24 bit (maybe as simple as saving as 24 bit wav) name that file "reference 24". Technically you just increased the level of your track by moving it 8 bits to the right, and filling the lower bits with zeroes. Since you now have 24 bits to play with instead of 16 you won't have any impact on the qualit*
4) For the both reference tracks ONLY "invert polarity" aka "Phase reverse" aka "invert" each of them. Save each as inverted16/24 depending on your source
4) Reduce the level of boh "Invert" reference files by a significant amount. Let's say 75% or if you're feeling drastic try taking them to practically zero. Even a 25% reduction to the original level should show a pretty significant effect later.*
Save this files as "Reduced"16 or 24 depending on which of the reference files you used to start from
4) Open up the original versions of the "inverted" reference versions again and this time increase the level by 50%, and then immediately reduce it by 75% and save as "Clipped" 16/ 24. It would also be worth saving the 24 bit file as 16 bit too and call it "Clipped 16b
5) open up all the files and "Normalize" them ( automatically increases gain so there is at least 1 sample at maximum level.
Now the fun begins.
A) Start a new session/ project
B) Import all the tracks mentioned above
Mute all tracks but the original 16 bit reference (which is inverted).*
Unmute the "inverted 16" track. Whoa?! Where did all the audio go? The answer is because both tracks are prefect mirrors of each other there is zero difference between them an so you are hearing nothing. Mute either of them and you'll be right back where you started in terms of levels.
Ok, reset the mutes so JUST the reference 16 is playing.
Un mute the *"reduced 16" track. The sound will change drastically. What you are hearing is the audio that was IRREVOCABLY lost when you reduced the level. Pay particular attention to what happens in the "quieter" passages. Ouchy.
Mute "Reduced16" again and unmute "clipped16", this time it's playing you the audio information that was lost when the track was distorted due to digital clipping. Pretty unpleasant.
If you made a "clipped 16b file" mute the reference track, and unmute the 16b file. *If the team who wrote your application didn't screw up, you should hear NOTHING! This one proves that moving from 16 to 24 bits gains nothing in terms of extra information and both will lose exactly the same amount of information once you clip the signal.
Ok, if you're not bored yet rinse and repeat with the 24 bit files. The biggest difference should be apparent when the reference 24 and reduced 24 are played together. That combination will be very much quieter (less difference) than the 16 bit versions doing the same thing. This will be more pronounced in quieter passages.*
So after all this what have we learned, other than I can prattle on about this stuff until any sane person either falls asleep and runs screaming? Hopefully something useful for those new at this who bothered reading this far (thanks! 🙂, the main thing being is by knowing your tools and the impact that choices now have on the end result.
Oh, and that wherever possible you should be running 24 bit audio, not 16 bit.
*I hope it was worthwhile wading through this "digital audio 101" posting. Apologies if I come across as condescending, but I figured if you knew this stuff you would have given up reading log ago..
I'll leave sampling rate and a little nyquist theorem for another day (possibly never if I've overstepped my mark with this).
Rob*
But first a disclaimer... *As you get to know me you'll discover that it's never my intent to belittle others but instead to help pass on COMPLETE knowledge. The Interwebs are great for finding half the story. Also, since I'm new around here I figures it helps to be explicit about my motives.
Anyway...
As mentioned, reducing level digitally reduces the bit-depth in the signal, which reduces dynamic range, increases signal:noise ratio (digital systems have a minimum floor of 1 bit, but frequently the first two bits are pretty close to meaningless random noise due to a whole host of factors).
What's interesting about this is if you turn UP (shift the bits upwards) of a digital waveform you will NOT INCREASE the dynamic range of the original signal either. You do increase level but could wind up reducing the range if you go to far and clip the signal and lose data as a result. Ideally by increasing level digitally you simply have the least significant bits filled with zeroes when you increase level. These bits could be then be filled with additional information which will possibly be audible -Although that's a discussion for another day.*
The same theory roughly applies to analog audio, but with one mjor difference. In the analog realm, increasing the level of a signal moves the noise floor upwards by the same amount.
The difference here is subtle. Both the digital noise floor and analog noise floors move up in level, but with the digital system you now have spare bits at the bottom of your range that are doing nothing, and therefor can now hold useful information. The analog noise floor on the other hand prevents additional low level information being added.
This effect is a major reason why studio gear usually runs at 24bit (fixed) precision. It allows there to be enough "Raw bits" in the source to allow multiple changes in gain and still be able to output full range 16 bit audio. Also it allows for signals to be recorded at much lower levels and retain enough resolution to be usable.*
As a side note here, it's a standard rookie mistake to think everything must be recorded so hot that every peak is near 0dBFS because more bits in the recorded signal must be better. Unfortunately that's dead wrong, because when you start to mix all these super hot signals together you're quickly going to run out of bits in your master bus and have to turn things down to prevent clipping. The practical result being that you may as well have recorded everything at a lower level in the first place and not run the risk of accidentally clipping your recording or overdriving the analog front end into distortion. Yike.
If someone has time to kill here's a nice practical experiment to illustrate the effect of changing levels digitally:-
Equipment needed:*
A multitrack track editor, pretty much anything will do including a whole slew of free ones.
Some music
Some time
Techniques used:-
Inverted null summing
1) Pick a favorite track with a very wide dynamic range, so not utterly compressed dance music, pop or numetal. A favorite of mine is simple minds Belfast Child, but something like O Fortuna (Old Spice theme in the UK) works just as well.
2) Copy from CD (DONOT use an mp3 as your source) into in the editor and save as "reference 16". *
3) change the bit depth of the file to 24 bit (maybe as simple as saving as 24 bit wav) name that file "reference 24". Technically you just increased the level of your track by moving it 8 bits to the right, and filling the lower bits with zeroes. Since you now have 24 bits to play with instead of 16 you won't have any impact on the qualit*
4) For the both reference tracks ONLY "invert polarity" aka "Phase reverse" aka "invert" each of them. Save each as inverted16/24 depending on your source
4) Reduce the level of boh "Invert" reference files by a significant amount. Let's say 75% or if you're feeling drastic try taking them to practically zero. Even a 25% reduction to the original level should show a pretty significant effect later.*
Save this files as "Reduced"16 or 24 depending on which of the reference files you used to start from
4) Open up the original versions of the "inverted" reference versions again and this time increase the level by 50%, and then immediately reduce it by 75% and save as "Clipped" 16/ 24. It would also be worth saving the 24 bit file as 16 bit too and call it "Clipped 16b
5) open up all the files and "Normalize" them ( automatically increases gain so there is at least 1 sample at maximum level.
Now the fun begins.
A) Start a new session/ project
B) Import all the tracks mentioned above
Mute all tracks but the original 16 bit reference (which is inverted).*
Unmute the "inverted 16" track. Whoa?! Where did all the audio go? The answer is because both tracks are prefect mirrors of each other there is zero difference between them an so you are hearing nothing. Mute either of them and you'll be right back where you started in terms of levels.
Ok, reset the mutes so JUST the reference 16 is playing.
Un mute the *"reduced 16" track. The sound will change drastically. What you are hearing is the audio that was IRREVOCABLY lost when you reduced the level. Pay particular attention to what happens in the "quieter" passages. Ouchy.
Mute "Reduced16" again and unmute "clipped16", this time it's playing you the audio information that was lost when the track was distorted due to digital clipping. Pretty unpleasant.
If you made a "clipped 16b file" mute the reference track, and unmute the 16b file. *If the team who wrote your application didn't screw up, you should hear NOTHING! This one proves that moving from 16 to 24 bits gains nothing in terms of extra information and both will lose exactly the same amount of information once you clip the signal.
Ok, if you're not bored yet rinse and repeat with the 24 bit files. The biggest difference should be apparent when the reference 24 and reduced 24 are played together. That combination will be very much quieter (less difference) than the 16 bit versions doing the same thing. This will be more pronounced in quieter passages.*
So after all this what have we learned, other than I can prattle on about this stuff until any sane person either falls asleep and runs screaming? Hopefully something useful for those new at this who bothered reading this far (thanks! 🙂, the main thing being is by knowing your tools and the impact that choices now have on the end result.
Oh, and that wherever possible you should be running 24 bit audio, not 16 bit.
*I hope it was worthwhile wading through this "digital audio 101" posting. Apologies if I come across as condescending, but I figured if you knew this stuff you would have given up reading log ago..
I'll leave sampling rate and a little nyquist theorem for another day (possibly never if I've overstepped my mark with this).
Rob*
diginerd, I have no practical experience in audio studio, but why don't you run your digital chain at e.g. 32 bits? Even with 24bit sources you will have 8bits (i.e. 8 x (-6dB) = -48 dB) of overhead for "lossless" mixing. The way I see it - if I import my 24bit sources to 32 bits and lower volume by 4 bits, I will have +24/-24dB absolutely "lossless" range for volume control.
Or if your chain was 64bits and you imported to occupy the first i.e. 40 bits, you would not have to worry about clipping/loosing LSBs at all.
I know this is not reality of studio work but technically perfectly possible.
Or if your chain was 64bits and you imported to occupy the first i.e. 40 bits, you would not have to worry about clipping/loosing LSBs at all.
I know this is not reality of studio work but technically perfectly possible.
Actually the mix bus of several systems is greater than *24 linear bits or the equivalent 32 bit floating point used in native processing. That instantly creates a host of other issues though, all of which are navigable if you're aware of them.
Here are some of the biggest ones.
1) To get off that bus at 24 or 16 bits to go through converters, out to files or through plugins you need to either truncate your bit depth by lopping off the lowest order bits which is "A Very Bad Thing"(tm), or use dithering (which adds noise to make the rounding errors less audible).
2) there's a point of diminishing returns with bit depth and audible difference. There's also a trade off in terms of the amount of compute required to process higher bit depths which leads to less being available for additional processing.
it's easy to lose sight of the fact that 24 bit audio doesn't have 50% more resolution than 16 bit, but 256x the resolution.*
2^(24-16)= 2^8= 256
Here are some of the biggest ones.
1) To get off that bus at 24 or 16 bits to go through converters, out to files or through plugins you need to either truncate your bit depth by lopping off the lowest order bits which is "A Very Bad Thing"(tm), or use dithering (which adds noise to make the rounding errors less audible).
2) there's a point of diminishing returns with bit depth and audible difference. There's also a trade off in terms of the amount of compute required to process higher bit depths which leads to less being available for additional processing.
it's easy to lose sight of the fact that 24 bit audio doesn't have 50% more resolution than 16 bit, but 256x the resolution.*
2^(24-16)= 2^8= 256
32 bit is the way to go in digital audio. Modern 32bit dacs an work as a preamp too. An example is wyred for sound dac, the latest version.
another solution, for home theater, is the onkyo 5507.
Most professional soundcard have a dsp of at least 32 bit. But having both dsp and dac, i think is better than having a 32 bit dsp and a 24 bit dac.
Personally i'd buy the onky 5507 for stereo listening with subwoofer, but you spend much of the money for additional non need channels, and video components, and it wayy to big 🙂.
another solution, for home theater, is the onkyo 5507.
Most professional soundcard have a dsp of at least 32 bit. But having both dsp and dac, i think is better than having a 32 bit dsp and a 24 bit dac.
Personally i'd buy the onky 5507 for stereo listening with subwoofer, but you spend much of the money for additional non need channels, and video components, and it wayy to big 🙂.
Why is 32 bit the "way to go"? Do you mean fixed point 32 bit? Or floating point 32 bit? Float 32 is what most "Native" applications use today, which is functionally equivalent to 24 bits fixed point (aka linear), which is used in traditional DSP systems.
Major downsides to 32 bit linear audio are the increasd resource consumption, including disc space and compute, the need to dither back down to sensible bit depths for everyone else and you're dealing with a market that all but rejected high quality audio in favor of convenience. I'm looking at you Apple...
Even more importantly, THERE'S ZERO AUDIBLE BENEFIT for playback.*
In return what does 32 bit linear get you? Unless you're the bionic man there's no point having anything greater than 16 bit linear audio anywhere near consumer space. Certainly there are applications in the Pro Sphere. Examples include Pro Tools HD having a 48 bit mix bus for mixing multiple 24 bit streams with headroom, double or even triple precision plug-in calculations, but examples like that are exceptions and not the norm for good reason. They also need to be well understood to use correctly, and avoid causing more harm than good.
More bits and higher sampling rates are more marketing and an excuse to have sloppy design in your converters than being required to improve audio fidelity.
Here's a controversial, yet absolutely defensible position:-*
There are many more things that will yield audible results than wasting resources on capturing signal at resolutions that provide no benefit.
If a human can't perceive it, there's no point wasting resources capturing it and playing it back. Movies wouldn't look better if they captured the ultraviolet and/or the infrared spectrums as well as visible light you see in the cinema, so why do people insist that it is necessary to do the equivalent when dealing with audio.
Unless I'm missing something?
Nyquist theorem is very explicit when it comes to the sampling rates required to capture full-bandwidth audio. 44.1KHz contains enough data to reconstruct the entire audible range, 192k contains exactly the same data plus three times that amount of information that is supersonic and of seriously questionable benefit to bother with.*
A great number of folks can, or at least claim to, hear an appreciable improvement in quality with over the top 192KHz recording Vs 44.1, despite there being no apparent justification for there to be a difference. The reason why those recordings sound better is not because of the sample rate, but because the AD converters can have a much more gentle cutoff slope at 192KHz, and as a result have fewer ripple effects defecting back into the audible band. Not because of an inherent limitation of 44.1KHz making it unable to capture the part you can hear in a faithful manner.
Sampling rate snake oil is a topic for another thread though.
Coming back to the original question.. Yes, turning down the level in the preamp DESTROYS dynamic range. The next question should be, does it matter?*
Major downsides to 32 bit linear audio are the increasd resource consumption, including disc space and compute, the need to dither back down to sensible bit depths for everyone else and you're dealing with a market that all but rejected high quality audio in favor of convenience. I'm looking at you Apple...
Even more importantly, THERE'S ZERO AUDIBLE BENEFIT for playback.*
In return what does 32 bit linear get you? Unless you're the bionic man there's no point having anything greater than 16 bit linear audio anywhere near consumer space. Certainly there are applications in the Pro Sphere. Examples include Pro Tools HD having a 48 bit mix bus for mixing multiple 24 bit streams with headroom, double or even triple precision plug-in calculations, but examples like that are exceptions and not the norm for good reason. They also need to be well understood to use correctly, and avoid causing more harm than good.
More bits and higher sampling rates are more marketing and an excuse to have sloppy design in your converters than being required to improve audio fidelity.
Here's a controversial, yet absolutely defensible position:-*
There are many more things that will yield audible results than wasting resources on capturing signal at resolutions that provide no benefit.
If a human can't perceive it, there's no point wasting resources capturing it and playing it back. Movies wouldn't look better if they captured the ultraviolet and/or the infrared spectrums as well as visible light you see in the cinema, so why do people insist that it is necessary to do the equivalent when dealing with audio.
Unless I'm missing something?
Nyquist theorem is very explicit when it comes to the sampling rates required to capture full-bandwidth audio. 44.1KHz contains enough data to reconstruct the entire audible range, 192k contains exactly the same data plus three times that amount of information that is supersonic and of seriously questionable benefit to bother with.*
A great number of folks can, or at least claim to, hear an appreciable improvement in quality with over the top 192KHz recording Vs 44.1, despite there being no apparent justification for there to be a difference. The reason why those recordings sound better is not because of the sample rate, but because the AD converters can have a much more gentle cutoff slope at 192KHz, and as a result have fewer ripple effects defecting back into the audible band. Not because of an inherent limitation of 44.1KHz making it unable to capture the part you can hear in a faithful manner.
Sampling rate snake oil is a topic for another thread though.
Coming back to the original question.. Yes, turning down the level in the preamp DESTROYS dynamic range. The next question should be, does it matter?*
Wow, I hadn't noticed all these replies. Thank you all so much. Lots of very useful info here.
Greg.
Greg.
I agree with the 44.1 VS 192kHz position. But in the 44.1 VS 48 kHz I would go with 48. Same for 96khz... Why?
Easier analog filtering, storage space is dirth cheap, flac compression is easy for the present day DSP's/CPU's...
Easier analog filtering, storage space is dirth cheap, flac compression is easy for the present day DSP's/CPU's...
Sample rate is one of the most fundamental decisions when starting a project. It all boils down to what your destination format (or primary destination format) is going to be.
For me my tracks come out on vinyl or 16 bit 44.1 LPCM download so 44.1 is a no brainer as I get far more available dsp put of my system than at higher sampling rates.
Also, despite all the digital voodoo, my primary mix bus is a large format analog console and my AD/DA converters of choice are the 20 bit converter on the back of my Fairlight MFX3. I have the option to run higher sample rates, and I have some really good 24 bit converters but I love the sound I get with this particular combination of gear.
The analog specs of this stuff far exceeds what vinyl can reproduce, but on paper I should stick to mixing "In The Box" for the best quality results. Reality is that there is a lot of warmth and "finished record sound" that all the distortions, non-linarities and noise introduced by the analog gear imparts simply by being in the signal path.
If someone asked me to produce a 5.1 soundtrack it would be all in the box at 96K since the video guys like multiples of 48KHz and the console doesn't handle more than stereo without getting so convoluted that it's impractical to use.
Moral of the story is that if it sounds good then run with it, but don't let specs and "bigger numbers are better" thinking guide you. You your ears. 🙂
Finally I'm currently at a loss as what to do next, since my console and a large chunk of my gear were recently destroyed by a lightning strike. I'm now going to have to re-evaluate both my process and what's available to replace what was trashed.
Part of that research is what led me here, so thanks for welcoming me into the community!
Rob
For me my tracks come out on vinyl or 16 bit 44.1 LPCM download so 44.1 is a no brainer as I get far more available dsp put of my system than at higher sampling rates.
Also, despite all the digital voodoo, my primary mix bus is a large format analog console and my AD/DA converters of choice are the 20 bit converter on the back of my Fairlight MFX3. I have the option to run higher sample rates, and I have some really good 24 bit converters but I love the sound I get with this particular combination of gear.
The analog specs of this stuff far exceeds what vinyl can reproduce, but on paper I should stick to mixing "In The Box" for the best quality results. Reality is that there is a lot of warmth and "finished record sound" that all the distortions, non-linarities and noise introduced by the analog gear imparts simply by being in the signal path.
If someone asked me to produce a 5.1 soundtrack it would be all in the box at 96K since the video guys like multiples of 48KHz and the console doesn't handle more than stereo without getting so convoluted that it's impractical to use.
Moral of the story is that if it sounds good then run with it, but don't let specs and "bigger numbers are better" thinking guide you. You your ears. 🙂
Finally I'm currently at a loss as what to do next, since my console and a large chunk of my gear were recently destroyed by a lightning strike. I'm now going to have to re-evaluate both my process and what's available to replace what was trashed.
Part of that research is what led me here, so thanks for welcoming me into the community!
Rob
A couple of last points
On 44.1 Vs 48 for me. Both are reasonable choices, 44.1 is the one I use to avoid sample rate conversion when I track back in from the console into my master session. I could of course run at any arbitary sample rate with the analog stage in there, but that final recording needs to be delivered at 44.1 to go to CD so that's why I run with it.
The Fairights converters just sound wonderful to my ears. They're very musical and have a distinct character to them that nothing else I've heard can mimic. They're certainly not "pure" or "crystal clear", but I use them because I prefer them over other gear with "better specs" (numbers wise at least!).
For those paying attention to my dithering comments, of course I dither from 24bit to 20bit prior to hitting the converters. Truncation artifacts caused by simply lopping off least significant bits blindly are a great way to mess up your sound.
On 44.1 Vs 48 for me. Both are reasonable choices, 44.1 is the one I use to avoid sample rate conversion when I track back in from the console into my master session. I could of course run at any arbitary sample rate with the analog stage in there, but that final recording needs to be delivered at 44.1 to go to CD so that's why I run with it.
The Fairights converters just sound wonderful to my ears. They're very musical and have a distinct character to them that nothing else I've heard can mimic. They're certainly not "pure" or "crystal clear", but I use them because I prefer them over other gear with "better specs" (numbers wise at least!).
For those paying attention to my dithering comments, of course I dither from 24bit to 20bit prior to hitting the converters. Truncation artifacts caused by simply lopping off least significant bits blindly are a great way to mess up your sound.
A quick bit-depth illustration
Just to illustrate what was mentioned about the difference between 16 bit and 24 bit.
-a 16 bit number can represent 65,536 different (voltage) levels, 0-65,535
-a 24 bit number can represent 16,777,216.
-an 8 bit number can represent 256.
Therefore, 24 bits can actually give 8 bit 'resolution' to each level 'step' of 16 bits.
-MAJOR difference.
Just to illustrate what was mentioned about the difference between 16 bit and 24 bit.
-a 16 bit number can represent 65,536 different (voltage) levels, 0-65,535
-a 24 bit number can represent 16,777,216.
-an 8 bit number can represent 256.
Therefore, 24 bits can actually give 8 bit 'resolution' to each level 'step' of 16 bits.
-MAJOR difference.
Now lets look at it with respect to reality. Most music these days has 10 db of dynamics if your lucky and even old recordings and new clasic/jazz wont have much more than 20 or 30 db of DR (dynamic range). So 16 bits giving you 96 db DR is more than enough for any music. ( LPs "the poor mans high resolution" only have 60 to 70 db of DR ).
What is the dynamic range of your room? Whats the difference between the loudest level you listen to music and the noise floor of your room. Or: when you got your stereo turned up to the loudest you listen to it and the music ends, do you have to turn it up more to hear the backround noise (hiss) of your audio system if you have to turn it up then your systems DR is more than your room can handle anyways. This is usually the case with 16 bit DR.
24 bit has a DR of 144 db! Dose anything else in the audio chain even come close. (20 bits will exceed the DR of any of your electronics). So why bother? I will tell you why. To reduce rounding errors when processing, thats it. If you are not doing any processing in the digital domain, you dont need 24 bits. The catch is, volume control is processing, so if you are decreaseing level before the DAC higher bit depth is benificial.
sorry but parts of the above post are kinda meaningless, even if a recording has only 30-40db (or whatever) of DNR, you still gain a better representation of those particular 30-40db used by the recording by having more bits to describe them. lossless audio isnt adaptive, it describes a set frequency range, with a finite number of samples; regardless of the information it is being used to describe. using a 10 bit system to describe 30db of info will result in even less dynamic range being portrayed, even though the bandwidth in its entirety is more than enough for the job.
hehe in fact here is an area that MP3 would maybe even be superior, because it uses an adaptive algorithm to assign the music information to the levels available in the range, rather than just arbitrarily assigning the samples to the level closest to the input value
totally talking from the hip there, in that last bit, but in my muddled head at 6am after a long night, it totally made sense to me 😉
32bit digi volume control is the best solution I have heard to the problem of adjusting gain and it is what I use wherever possible (the sabre does a particularly good job here), sure if you turn the 'knob' all the way down you will still lose some info, but very little and not at the levels I listen at; as I have tuned my system to not have too much more gain than I need; availing me of pretty much all the DNR I have on tap, all the time.
hehe in fact here is an area that MP3 would maybe even be superior, because it uses an adaptive algorithm to assign the music information to the levels available in the range, rather than just arbitrarily assigning the samples to the level closest to the input value
totally talking from the hip there, in that last bit, but in my muddled head at 6am after a long night, it totally made sense to me 😉
32bit digi volume control is the best solution I have heard to the problem of adjusting gain and it is what I use wherever possible (the sabre does a particularly good job here), sure if you turn the 'knob' all the way down you will still lose some info, but very little and not at the levels I listen at; as I have tuned my system to not have too much more gain than I need; availing me of pretty much all the DNR I have on tap, all the time.
yep, dual sabre 9012 comes pretty close to that with the right IV, but sure I catch your meaning there. numbers in audio arent my sole pursuit, but I wont pretend they dont mean anything to me.
- Status
- Not open for further replies.