DJTT Blog Post on Bitrates

Home :: General Discussion :: DJTT Blog Post on BitratesReply
DJTT Blog Post on Bitrates
Posted on: 26.09.2012 by Jerica Salava
If you missed it here: http://www.djranking s.com/2012/09/2...-and-bitrates/

It's a nice read until the part where they put down 16 BIT 44.1KHz WAV files.

Really? Now this is unacceptable?

I guess I should re-format my hard drives with hundreds of classic house vinyl rips, because they are all 16 BIT 44.1KHz garbage.

What a joke.

>
Antonetta Wikel
29.09.2012
Originally Posted by ksandvik
AIFFs actually also support metadata tagging. As for WAV format, the chunking could be used but so far no de facto standard has appeared. You could put metadata to WAVs in iTunes but it it stored in the database itself.

Anyway, for WAVs just put most important info into the file name: <artist>-<track>-<bpm>-<style>.wav. Job done.
If the DJ industry would just allow the use of Broadcast WAV then we'd have all the metadata tags we'd need. We've been using BWAV in the Post Production Industry for years. What the heck is taking the DJ software companies so long to get on board?
Nancey Inderlied
29.09.2012
Originally Posted by ksandvik
AIFFs actually also support metadata tagging. As for WAV format, the chunking could be used but so far no de facto standard has appeared. You could put metadata to WAVs in iTunes but it it stored in the database itself.

Anyway, for WAVs just put most important info into the file name: <artist>-<track>-<bpm>-<style>.wav. Job done.
Also, IIRC, outboard music managers like Rekordbox don't use conventional tags, but exports their own metadata compilation.


Originally Posted by djproben
I believe on slower computers it can be distorted because it's more processer intensive. The hardware makes a difference; some sound cards aren't quite as graceful about different rates as others.

This http://people.xiph.org/~xiphmont/demo/neil-young.html is a good article on why anything above 16/44 is overkill and could actually distort the sound by bringing in artifacts. Producing is one thing but for recording vinyl or listening to music there's no point. I might sometimes record in 20/44 ALACs when archiving rare vinyl but I can't imagine I would hear the difference in an ABX.
Also to my knowledge, this stems from shitty SRCs/SRC algorithms. People knock DJMs for being expensive, but their entire current install line rockets right up to 96khz no problem. And since it's an output, it actually can make a difference, e.g. the AA filter.
Danae Dumler
29.09.2012
Originally Posted by ekwipt
I've read that people trying to up-sample on Traktor on the with 16/44 to 24/48 causes bad sound, can anyone chime in on that?
I believe on slower computers it can be distorted because it's more processer intensive. The hardware makes a difference; some sound cards aren't quite as graceful about different rates as others.

This http://people.xiph.org/~xiphmont/demo/neil-young.html is a good article on why anything above 16/44 is overkill and could actually distort the sound by bringing in artifacts. Producing is one thing but for recording vinyl or listening to music there's no point. I might sometimes record in 20/44 ALACs when archiving rare vinyl but I can't imagine I would hear the difference in an ABX.

I only use 16/44 files in Traktor, I've been a big supporter of no MP3s in my DJ sets, mostly for what the article talks about, future proofing, MP3s shouldn't be used anymore.
MP3s are still the only format with decent tag support across a wide range of programs. 320 MP3s are fine as far as I'm concerned sound-wise but if I were to start my music collection all over again I would go lossless. I'm not re-ripping my CDs and vinyl, and I'm certainly not paying dance music stores a premium for lossless. I don't believe MP3s are going anywhere, and their sound quality at 320 CBR is well above anything you need for the dancefloor. For archiving other kinds of music I go lossless as often as possible (assuming I am recording off the vinyl or CD) but I see no point in buying WAVs that are hard to deal with tag-wise for over-compressed dance music that may only see a few months of action anyway.

I'm not necessarily advocating this approach btw... my music collection is a total mess of different formats poorly organized etc...
Jerica Salava
29.09.2012
Originally Posted by djproben
It hurts when you drop it on your foot?
hurts so good.

>
Danae Dumler
29.09.2012
Originally Posted by loverocket
I really don't believe you could ever say anything negative about a Technics 1200 without sounding like an idiot to me.
It hurts when you drop it on your foot?
Jerica Salava
29.09.2012
Originally Posted by Tommi Bass
Ripp Cd at 16 bit 44.100Hz (Cus thats what it is)

Ripp Vinyl at 24 Bit 48hz or higher and use a much better turntable than the Technics.
Are you really putting down a Technics 1200 ?

I know that you want to get out meters and talk technical junk but I really don't believe you could ever say anything negative about a Technics 1200 without sounding like an idiot to me.

I would like a Technics 1200 in my casket please.

Respect.

>
Antonetta Wikel
29.09.2012
Originally Posted by djproben
Just a nitpick, if you're talking about the Oohashi tests, they were flawed, and follow up research has pretty much debunked that claim. See http://www.aes.org/e-lib/browse.cfm?elib=10005 and http://www.nhk.or.jp/strl/publica/labnote/lab486.html for instance.
I wasn't but thanks for the links. Always interesting to see the results of tests like that. No, my reference was to the harmonics of tones that still fall within the range of human hearing.
Danae Dumler
29.09.2012
Originally Posted by SirReal
I'm not just talking about harmonics outside of human hearing, although there's been some tests that have proven they too can have an effect on how we perceive the audio.
Just a nitpick, if you're talking about the Oohashi tests, they were flawed, and follow up research has pretty much debunked that claim. See http://www.aes.org/e-lib/browse.cfm?elib=10005 and http://www.nhk.or.jp/strl/publica/labnote/lab486.html for instance.
Celine Surico
29.09.2012
Originally Posted by Tommi Bass
Ripp Cd at 16 bit 44.100Hz (Cus thats what it is)

Ripp Vinyl at 24 Bit 48hz or higher and use a much better turntable than the Technics.

I used a Linn Sondek LP 12... and custom DAC.


Producing dance music (Electronic) 24 bit 48hz anything over this is overkill.

I'm a mastering engineer of quality dance music and the above is fine..... so long as it has no been limited or compressed on the final mix bus or dithered and must have at least -3db headroom if you send to mastering dudes like me.
Yes, totally agree. If I would record a good singer, let's say Corinne Bailey Rae, in a stop-notch studio with the best possible mic setups and acoustics and a beefy computer rig and agood ADC/DAC converters, I would use 96kHz.

For most if not all electronic dance productions, 96kHz is not needed at all, 41.1/48k is fine. But 24 is a must, you could even hear yourself the diff between 16 and 24 bit concerning the dynamic range. That is then of course squeezed out in modern EDM with all the compressors placed in the channel strips and final mix, using loops that are already compressed and 16-bit samples... Ouch.

Or. Striving for perfection does not make sense when the original material and the processing will squeezed out the fine parts of the audio stream, anyway. The only result is a slower computer due to all the additional CPU and I/O processing.
Antonetta Wikel
29.09.2012
Originally Posted by ksandvik
AIFFs actually also support metadata tagging. As for WAV format, the chunking could be used but so far no de facto standard has appeared. You could put metadata to WAVs in iTunes but it it stored in the database itself.

Anyway, for WAVs just put most important info into the file name: <artist>-<track>-<bpm>-<style>.wav. Job done.
If the DJ industry would just allow the use of Broadcast WAV then we'd have all the metadata tags we'd need. We've been using BWAV in the Post Production Industry for years. What the heck is taking the DJ software companies so long to get on board?
Nancey Inderlied
29.09.2012
Originally Posted by ksandvik
AIFFs actually also support metadata tagging. As for WAV format, the chunking could be used but so far no de facto standard has appeared. You could put metadata to WAVs in iTunes but it it stored in the database itself.

Anyway, for WAVs just put most important info into the file name: <artist>-<track>-<bpm>-<style>.wav. Job done.
Also, IIRC, outboard music managers like Rekordbox don't use conventional tags, but exports their own metadata compilation.


Originally Posted by djproben
I believe on slower computers it can be distorted because it's more processer intensive. The hardware makes a difference; some sound cards aren't quite as graceful about different rates as others.

This http://people.xiph.org/~xiphmont/demo/neil-young.html is a good article on why anything above 16/44 is overkill and could actually distort the sound by bringing in artifacts. Producing is one thing but for recording vinyl or listening to music there's no point. I might sometimes record in 20/44 ALACs when archiving rare vinyl but I can't imagine I would hear the difference in an ABX.
Also to my knowledge, this stems from shitty SRCs/SRC algorithms. People knock DJMs for being expensive, but their entire current install line rockets right up to 96khz no problem. And since it's an output, it actually can make a difference, e.g. the AA filter.
Celine Surico
29.09.2012
AIFFs actually also support metadata tagging. As for WAV format, the chunking could be used but so far no de facto standard has appeared. You could put metadata to WAVs in iTunes but it it stored in the database itself.

Anyway, for WAVs just put most important info into the file name: <artist>-<track>-<bpm>-<style>.wav. Job done.
Danae Dumler
29.09.2012
Originally Posted by ekwipt
I've read that people trying to up-sample on Traktor on the with 16/44 to 24/48 causes bad sound, can anyone chime in on that?
I believe on slower computers it can be distorted because it's more processer intensive. The hardware makes a difference; some sound cards aren't quite as graceful about different rates as others.

This http://people.xiph.org/~xiphmont/demo/neil-young.html is a good article on why anything above 16/44 is overkill and could actually distort the sound by bringing in artifacts. Producing is one thing but for recording vinyl or listening to music there's no point. I might sometimes record in 20/44 ALACs when archiving rare vinyl but I can't imagine I would hear the difference in an ABX.

I only use 16/44 files in Traktor, I've been a big supporter of no MP3s in my DJ sets, mostly for what the article talks about, future proofing, MP3s shouldn't be used anymore.
MP3s are still the only format with decent tag support across a wide range of programs. 320 MP3s are fine as far as I'm concerned sound-wise but if I were to start my music collection all over again I would go lossless. I'm not re-ripping my CDs and vinyl, and I'm certainly not paying dance music stores a premium for lossless. I don't believe MP3s are going anywhere, and their sound quality at 320 CBR is well above anything you need for the dancefloor. For archiving other kinds of music I go lossless as often as possible (assuming I am recording off the vinyl or CD) but I see no point in buying WAVs that are hard to deal with tag-wise for over-compressed dance music that may only see a few months of action anyway.

I'm not necessarily advocating this approach btw... my music collection is a total mess of different formats poorly organized etc...
Rochel Gleese
29.09.2012
I've read that people trying to up-sample on Traktor on the with 16/44 to 24/48 causes bad sound, can anyone chime in on that?

I only use 16/44 files in Traktor, I've been a big supporter of no MP3s in my DJ sets, mostly for what the article talks about, future proofing, MP3s shouldn't be used anymore.
Jerica Salava
29.09.2012
Originally Posted by djproben
It hurts when you drop it on your foot?
hurts so good.

>
Danae Dumler
29.09.2012
Originally Posted by loverocket
I really don't believe you could ever say anything negative about a Technics 1200 without sounding like an idiot to me.
It hurts when you drop it on your foot?
Jerica Salava
29.09.2012
Originally Posted by Tommi Bass
Ripp Cd at 16 bit 44.100Hz (Cus thats what it is)

Ripp Vinyl at 24 Bit 48hz or higher and use a much better turntable than the Technics.
Are you really putting down a Technics 1200 ?

I know that you want to get out meters and talk technical junk but I really don't believe you could ever say anything negative about a Technics 1200 without sounding like an idiot to me.

I would like a Technics 1200 in my casket please.

Respect.

>
Antonetta Wikel
29.09.2012
Originally Posted by djproben
Just a nitpick, if you're talking about the Oohashi tests, they were flawed, and follow up research has pretty much debunked that claim. See http://www.aes.org/e-lib/browse.cfm?elib=10005 and http://www.nhk.or.jp/strl/publica/labnote/lab486.html for instance.
I wasn't but thanks for the links. Always interesting to see the results of tests like that. No, my reference was to the harmonics of tones that still fall within the range of human hearing.
Danae Dumler
29.09.2012
Originally Posted by SirReal
I'm not just talking about harmonics outside of human hearing, although there's been some tests that have proven they too can have an effect on how we perceive the audio.
Just a nitpick, if you're talking about the Oohashi tests, they were flawed, and follow up research has pretty much debunked that claim. See http://www.aes.org/e-lib/browse.cfm?elib=10005 and http://www.nhk.or.jp/strl/publica/labnote/lab486.html for instance.
Celine Surico
29.09.2012
Originally Posted by Tommi Bass
Ripp Cd at 16 bit 44.100Hz (Cus thats what it is)

Ripp Vinyl at 24 Bit 48hz or higher and use a much better turntable than the Technics.

I used a Linn Sondek LP 12... and custom DAC.


Producing dance music (Electronic) 24 bit 48hz anything over this is overkill.

I'm a mastering engineer of quality dance music and the above is fine..... so long as it has no been limited or compressed on the final mix bus or dithered and must have at least -3db headroom if you send to mastering dudes like me.
Yes, totally agree. If I would record a good singer, let's say Corinne Bailey Rae, in a stop-notch studio with the best possible mic setups and acoustics and a beefy computer rig and agood ADC/DAC converters, I would use 96kHz.

For most if not all electronic dance productions, 96kHz is not needed at all, 41.1/48k is fine. But 24 is a must, you could even hear yourself the diff between 16 and 24 bit concerning the dynamic range. That is then of course squeezed out in modern EDM with all the compressors placed in the channel strips and final mix, using loops that are already compressed and 16-bit samples... Ouch.

Or. Striving for perfection does not make sense when the original material and the processing will squeezed out the fine parts of the audio stream, anyway. The only result is a slower computer due to all the additional CPU and I/O processing.
Germaine Bernadin
28.09.2012
Ripp Cd at 16 bit 44.100Hz (Cus thats what it is)

Ripp Vinyl at 24 Bit 48hz or higher and use a much better turntable than the Technics.

I used a Linn Sondek LP 12... and custom DAC.


Producing dance music (Electronic) 24 bit 48hz anything over this is overkill.

I'm a mastering engineer of quality dance music and the above is fine..... so long as it has no been limited or compressed on the final mix bus or dithered and must have at least -3db headroom if you send to mastering dudes like me.
Margie Pavell
28.09.2012
Originally Posted by loverocket
haha! CDJ's suck !
okay...now you just need to stop...
Jerica Salava
28.09.2012
Originally Posted by Shishdisma
somewhat limited by CDJ's 44.1khz output.

haha! CDJ's suck !
Jolynn Schroyer
27.09.2012
Another pointless thread.
Nancey Inderlied
27.09.2012
Originally Posted by SirReal
I'm not just talking about harmonics outside of human hearing, although there's been some tests that have proven they too can have an effect on how we perceive the audio. What about the harmonics of the lower frequencies that are still within the range of human hearing? Anyway, like I said, I'm not here to argue. You can believe what you want but I'd suggest looking/listening beyond whatever you read on the internet and applying your own personal experience to the subject. I did and my experience differs greatly from yours and the testing I've personally done and results I've read from audiophiles that I trust and respect does also.
I recognize that harmonics of inaudible frequencies can bleed into the audible spectrum, but unless you have monitors that can replicate those inaudible frequencies, the harmonics typically won't be generated at all. Note that I also gave concessions to recording of acoustic or analogue instruments with high end mics, as the harmonics will be present in the actual recording, and from there, in the raw audio, but for recording a soft synth or sampleware (the majority of electronic production), the internal sample rate of Ableton typically becomes a waste of drive space and processor power.

Also note that I'm not refuting the concept of oversampling for the purposes of aliasing, but thats more applicable to output, and somewhat limited by CDJ's 44.1khz output.
Antonetta Wikel
27.09.2012
I'm not just talking about harmonics outside of human hearing, although there's been some tests that have proven they too can have an effect on how we perceive the audio. What about the harmonics of the lower frequencies that are still within the range of human hearing? Anyway, like I said, I'm not here to argue. You can believe what you want but I'd suggest looking/listening beyond whatever you read on the internet and applying your own personal experience to the subject. I did and my experience differs greatly from yours and the testing I've personally done and results I've read from audiophiles that I trust and respect does also.
Nancey Inderlied
27.09.2012
Originally Posted by SirReal
I'd be very curious to read where you've got your information from, or is this just personal opinion? Anyway, we can agree to disagree. I'm not going to argue with you as it seems your mind is made up, I just don't want people who read this to take what you're saying as truth and would beg them to do their own research. I've certainly done mine and have 17 years experience in the audio industry and have run double blind audio tests for award winning mixers that refute your statements. I'm not into a territorial pissing match, I just want anybody who cares about the sonic quality of their own tracks to be aware.
Im not really sure what "quality" you're even talking about. 44.1khz reproduces all audible frequencies. 96khz just holds inaudible frequencies, for the purposes of hearing their harmonics bleeding into the mix. Unless you have ribbon tweeters, or some other means of reproducing frequencies above 20khz, the extra frequencies sampled by a 96khz rate above a 44.1 rate cannot be physically reproduced, thus, no frequencies are actually lost.
Antonetta Wikel
27.09.2012
I'd be very curious to read where you've got your information from, or is this just personal opinion? Anyway, we can agree to disagree. I'm not going to argue with you as it seems your mind is made up, I just don't want people who read this to take what you're saying as truth and would beg them to do their own research. I've certainly done mine and have 17 years experience in the audio industry and have run double blind audio tests for award winning mixers that refute your statements. I'm not into a territorial pissing match, I just want anybody who cares about the sonic quality of their own tracks to be aware.
Johnetta Olewine
27.09.2012
Shitdisma just admit it. You got schooled.
Nancey Inderlied
27.09.2012
Originally Posted by SirReal
How can you make a statement so patently false? I believe your understanding of sample rate is flawed. You might be extrapolating that at 44.1 16bit it more than covers the range of human hearing from 20hz to 20khz but there is absolutely no doubt that sample rates higher than 44.1 are sampling at a higher rate and therefore more truly reproducing the original. You're essentially mixing down mulitiple tracks into a final 2 tracks. When distilling anything down you want to start with the most premium grade possible, that is, if you care about the quality of the finished product. If all you're doing is sampling loops of peoples tracks and mixing them together to make music, then I'd agree with your above statement but if you're making original music and recording sounds, you should start with the highest sample & bit rate available to you.
Just because the information is there, doesnt mean we can hear it, unless youre mastering on ribbon monitors, the difference between 44.1 and 96 are gone. The "quality" thats given by a 96khz rate simply cannot be reproduced by normal speakers, and thus the harmonics that bleed in cannot be heard on anything but ribbons, let alone the post-master recording. You can talk about "starting with the highest quality" all you want but you'd be better off inventing a dark matter recorder for all the good it'll do.
Antonetta Wikel
27.09.2012
Originally Posted by Shishdisma
Except for the fact that sample rate doesn't affect audible quality beyond 44.1khz...
How can you make a statement so patently false? I believe your understanding of sample rate is flawed. You might be extrapolating that at 44.1 16bit it more than covers the range of human hearing from 20hz to 20khz but there is absolutely no doubt that sample rates higher than 44.1 are sampling at a higher rate and therefore more truly reproducing the original. You're essentially mixing down mulitiple tracks into a final 2 tracks. When distilling anything down you want to start with the most premium grade possible, that is, if you care about the quality of the finished product. If all you're doing is sampling loops of peoples tracks and mixing them together to make music, then I'd agree with your above statement but if you're making original music and recording sounds, you should start with the highest sample & bit rate available to you.
Nancey Inderlied
27.09.2012
Originally Posted by SirReal
Totally disagree. When producing and mixing multiple tracks for music you should start with the highest quality recordings you can afford to make. Once you've mixed down to your stems and give the mastering person the best quality you can, they do their magic and then you can downsample after your final stereo master is achieved. You'll get considerably better sonic results.
Except for the fact that sample rate doesn't affect audible quality beyond 44.1khz...
Antonetta Wikel
27.09.2012
Originally Posted by ksandvik
The only justification I could believe of for EDM producers using 96kHz is that some mastering plugins might have better downsampling quality when operating with such huge frequency domains. But to hear that result you need really good studio reference monitors. In real life I doubt anyone could hear the differences.
Totally disagree. When producing and mixing multiple tracks for music you should start with the highest quality recordings you can afford to make. Once you've mixed down to your stems and give the mastering person the best quality you can, they do their magic and then you can downsample after your final stereo master is achieved. You'll get considerably better sonic results.
Celine Surico
27.09.2012
The only justification I could believe of for EDM producers using 96kHz is that some mastering plugins might have better downsampling quality when operating with such huge frequency domains. But to hear that result you need really good studio reference monitors. In real life I doubt anyone could hear the differences.
Nancey Inderlied
27.09.2012
Originally Posted by weedsmkn
you should always be in 24 bit when possible and dither down if needed, also 96/48khz to 44.1 is not the big deal its the bits. that being said you will not be able to tell the difference until you are on very high end system in an acoustically tuned room. So dont worry loverocket, your rips should be good to go for quite some time. From now on 24 bit is the way to go as CD quality audio (16bit 44.1khz) is no longer industry standard.
"Industry standard" doesnt mean anything when the lowest common denominator for quality audio is still 16/44.1. Theres nothing to "transition into" its just a negligible quality level.
Jerica Salava
27.09.2012
i believe 16 BIT 44.1 KHz sounds warmer.


>


>


Just kidding.
Cassaundra Lindesmith
27.09.2012
you should always be in 24 bit when possible and dither down if needed, also 96/48khz to 44.1 is not the big deal its the bits. that being said you will not be able to tell the difference until you are on very high end system in an acoustically tuned room. So dont worry loverocket, your rips should be good to go for quite some time. From now on 24 bit is the way to go as CD quality audio (16bit 44.1khz) is no longer industry standard.
Nancey Inderlied
27.09.2012
I've never really understood the obsession with 48/96khz 24 bit production values. 24 bit vs 16 bit is a bit of a jump, yes, but in practice, its a bit negligible. Running in 24 bit is a safe bet, since CDJs output in 24 bit, but 16 is by no means useless. What blows my mind is people who sample in 48/96khz and act like they need to. 48khz is a holdover from sampling DAT tapes. 96khz is used for high end condensers, since acoustic instruments can produce so many partials and harmonics; unless youre listening on ribbon monitors, you can't hear the difference, and even then, it's only harmonics.

CDJs output in 44.1khz anyway, so.... yeahh.....
Jerica Salava
27.09.2012
Originally Posted by Xonetacular
what's misleading about it?
For the 16 BIT image the graph paper would have to be 65,536 squares high not 16.

Each 16 BIT sample (of 44,100 samples per second) can store one of 65,536 levels.

that's pretty fine for my ear.

After dithering and filtering my ears are fooled.

>
Celine Surico
27.09.2012
Maybe on the 44.1Kzh versus 96khz 24 bit case, you get the same amount of data, 24 bits is twenty four bits, but just a wider range of values. The graph makes it look like you get more data.

<< Back to General DiscussionReply

Copyright 2012-2023
DJRANKINGS.ORG n.g.o.
Chuo-ku, Osaka, Japan

Created by Ajaxel CMS

Terms & Privacy