Discussion on sample rate converters and its relationship to mastering?

dav nagle

Hello everyone I am getting an album mastered at 96khz and I dont know what the mastering engineer is talking about when he gets deep into a/d converters an so forth. Can anyone please enlighten me to the quality and importance of this process and how it interacts with the final record please? I just would like people to share their taught and links? What equipment should best be used to convert 96khz? What can go wrong in the process?

Anima

The greater the quality of the convertors i.e. analog to digital and digital to analog, the greater the quality of the final sound. Every time you do a conversion you'll always loose some of the information. If you record a guitar or vocal on a low quality ADC, its always going to sound terrible no matter how its mastered. Same goes for a DAC, if the quality is low, whatever it outputs will be garbage.

Two things you have to look at when dealing with converters is the sample rate and the bit-depth. The sample rate refers to how often it takes a sample of the signal or a snap-shot of it. The higher the frequency, the more frequencies and transients it will pick up. Which means your high end will be brighter and clearer. All you really need to know about bit-depth is that the greater it is, the more dynamic range there will be. Which means both quiet and loud sounds will sound great.

dav nagle

Firstly thank-you for the reply and great information. Is it possible for a tearing sound or distortion to occur with bad converters? Example: a 96khz audio file converted to CD 44.1khz for instance?

brettzy

Anima wrote: »

The sample rate refers to how often it takes a sample of the signal or a snap-shot of it.

This is a fairly common misconception. The sample rate refers to the frequencies recorded and the frequency at which the anti aliasing filter is set to. If you record at 44.1kHz then the anti aliasing filter is set close to 22kHz which is pretty much the range of human hearing and the reason why that sample rate was choosen for CD. At 96kHz the anti aliasing filter can be set higher (48kHz i think) and so the recordings can contain frequencies up to 48kHz.

brettzy

dav nagle wrote: »

Firstly thank-you for the reply and great information. Is it possible for a tearing sound or distortion to occur with bad converters? Example: a 96khz audio file converted to CD 44.1khz for instance?

No, not really. But just to help you understand the process here's a little insight into the process.
The mastering studio will most likely master one of 3 ways.

1. They'll use good DA converter to convert a digital 96khz 24bit mix to audio, run it through some outboard gear (maybe even record it to tape), get the master sounding great and then use a good AD converter to "record" the audio back onto his system. Then within the computer they'll resample the 96kHz 24bit master to 44.1kHz 16bit ready for CD using a dithering plugin to take away the linear effect of quantized digital audio.

2. This way is exactly the same except instead of recording it back onto the same system at 96khz 24 bit they'll record it onto another system at 44.1kHz 16 bit with dithering which eliminates the need of resampling within the computer.

3. The won't use any AD DA converters and just master in the box with plugins and then resample with dithering in the box.

The margin of error is very small these days so don't worry!!

frobisher

dav nagle wrote: »

Hello everyone I am getting an album mastered at 96khz and I dont know what the mastering engineer is talking about when he gets deep into a/d converters an so forth. Can anyone please enlighten me to the quality and importance of this process and how it interacts with the final record please? I just would like people to share their taught and links? What equipment should best be used to convert 96khz? What can go wrong in the process?

The mastering engineer's job is not to bamobozle you with the alogorithms of A/D/A conversion. The way I see it is that they cost too much to spend their time talking about science- unless it's something you ask about. Their job is to master your tunes.

What is your deliverable to him? Is he ripping a CD or playing from a deck and then putting it from the digital domain to the analouge than back to digital? Is he taking from tape? What gear is he using?

brettzy

I just read my last post,

I'm such a feckin nerd!!!!!!

ogy

This is a fairly common misconception

its not a misconception. the sample rate is how many samples of audio are recorded digitally per second, i.e. 44100 samples per second at 44.1khz. the connection to frequency is that your supposed to sample at a rate of twice the highest frequency you wish to capture (the nyquist theorem). an anti-aliasing filter is used so that frequencies above 20khz don't get picked up and cause confusion to the sampled audio (aliasing).

the reason its 44.1khz or 48khz instead of just 40khz is for two main reasons:
- brick wall anti-aliasing filters can't be made, they always have a slope so the extra 4.1khz/8khz gives them a bit of headroom.
- both sample rates divide nicely into the frame rate for american and european film (approx 30 frames per second and 25 frames per second respectively).

dav nagle

Oh and its an electronica album recorded on Reason 4 . I brought the master fader down to 70% on all tracks to leave room for mastering and made sure that there was no clipping. Like I said the mix I sent is clean . Maybe my mix sounds clean on my genelecs but on other speakers its not??

ogy

if it sounded fine at the mixing stage theres no need for it to come back from mastering sounding rubbish, get back on to whoever mastered it, if you paid them your entitled to get it redone.

frobisher

If you're hearing a ripping sound, something's wrong.

dav nagle

if it sounded fine at the mixing stage theres no need for it to come back from mastering sounding rubbish, get back on to whoever mastered it, if you paid them your entitled to get it redone.




I was thinking the very same thing. I was disappointed to hear the distortion on the piano. I got home slapped my master copy on, got to the 9th song and BAM (distortion and clipping that was not there before hand!) Like taking a slap in the face from a guy with a wet fish!

dav nagle

If you're hearing a ripping sound, something's wrong.


Can you throw a few suggestions of what could possibly be going wrong please Frobisher?


I am BAFFLED!

The mastering engineer asked me to get a second opinion so here I am asking you guys for suggestions. I will be showing it to Paul Brewer and Studio Rat for greater examination.

dav nagle

OH thanks Bretzzy for the reply dude! I ll have to read your post a few times to take it all in, in a nerdy way (ITS DEEP!)

Niall - Dahlia

dav nagle wrote:

Hello everyone I am getting an album mastered at 96khz and I dont know what the mastering engineer is talking about when he gets deep into a/d converters an so forth.

Like frob said, if this guy was talking tech to bamboozle you I'd have brought my mixes elsewhere. Doesn't surprise me that you ended up with a bad master if that was the case. The mastering engineer should be able to explain exactly what he's going in terms that are relevant to the artist.

Have you tried the masters on different monitoring systems/hi-fi's/car stereo? If so is the problem still there?

Post a lossless sample of the clippings piano!

dav nagle

Niall - Dahlia wrote: »

Like frob said, if this guy was talking tech to bamboozle you I'd have brought my mixes elsewhere. Doesn't surprise me that you ended up with a bad master if that was the case. The mastering engineer should be able to explain exactly what he's going in terms that are relevant to the artist.

Have you tried the masters on different monitoring systems/hi-fi's/car stereo? If so is the problem still there?

Post a lossless sample of the clippings piano!

How and where will I post the sample? I can post you the original mix and the master mix so you can compare and contrast if you like? Send me your email?

cornbb

dav nagle wrote: »

How and where will I post the sample? I can post you the original mix and the master mix so you can compare and contrast if you like? Send me your email?

Might be an idea to sign up on www.box.net and post it up there. Then you can link to it from here so we can all have a gander

I think there might be a size limit per file (maybe 10mb) so if you want to post a lossless sample (i.e. WAV or AIFF) you might need to edit it down to just the part you suspect is distorted.

dav nagle

http://www.box.net/shared/fji7l2f40w



I listened to the master on a cheap c.d. player and it sounded bad, I listened to the master on a family hi-fi and it sounded bad at a higher level. The master sounds at its worst (clipping wise) when played in my itunes player? Why I don't have the foggiest. When I play the master through peak pro the tearing is not so bad?? Surely no matter what I play it on it should sound clean enough? I.E no clipping or tearing. Anyway have a listen and if your not hearing much play the file through I tunes . Thanks..

brettzy

ogy wrote: »

its not a misconception. the sample rate is how many samples of audio are recorded digitally per second, i.e. 44100 samples per second at 44.1khz. the connection to frequency is that your supposed to sample at a rate of twice the highest frequency you wish to capture (the nyquist theorem). an anti-aliasing filter is used so that frequencies above 20khz don't get picked up and cause confusion to the sampled audio (aliasing).

the reason its 44.1khz or 48khz instead of just 40khz is for two main reasons:
- brick wall anti-aliasing filters can't be made, they always have a slope so the extra 4.1khz/8khz gives them a bit of headroom.
- both sample rates divide nicely into the frame rate for american and european film (approx 30 frames per second and 25 frames per second respectively).

Thanks for that ogy, I read quickly through a article in the last sound on sound mag and misunderstood a fundamental. So last night after I read your post I did some searches on the whole process and found some great insights. I'm glad you put me straight!!

brettzy

Anima, sorry for doubting you mate!!

ogy

Thanks for that ogy

no worries mate!

had a listen to the two samples on crappy laptop speakers, and through decent headphones and had a bit of a look at it in soundforge. its not clipping but it is limited quite hard so the attack of the chords is a quite harsh, im no mastering engineer but i don't think the limiter should have had to work so hard. maybe ask him to redo the tracks and not to smash them so much.

i don't think its anything to do with bitrate/samplerate conversion or anything, think its just limited so much that its brought out some of the harsh sounds of the piano samples which you may not have noticed at the mixing stage cos they were fairly low, this combined with the limiter slamming down on the chords is probably at fault for your grievances:).

cornbb

I don't think I can hear any distortion, if I do its very vert faint. Its certainly using the entire dynamic range but I don't think its clipping. I opened it up in Audacity to have a look at the waveform (don't have Soundforge here at work). There are one or two square-ish peaks around 8-9 seconds in alright. While its not distorted as such, you've lost the nice soft dynamics of the original, maybe thats what you were going for though. I'm no expert on mastering so maybe someone else here can have their say...

brettzy

dav nagle wrote: »

http://www.box.net/shared/fji7l2f40w



I listened to the master on a cheap c.d. player and it sounded bad, I listened to the master on a family hi-fi and it sounded bad at a higher level. The master sounds at its worst (clipping wise) when played in my itunes player? Why I don't have the foggiest. When I play the master through peak pro the tearing is not so bad?? Surely no matter what I play it on it should sound clean enough? I.E no clipping or tearing. Anyway have a listen and if your not hearing much play the file through I tunes . Thanks..

Are those audio samples at the same time in the song?? There seems to be an extra note on the piano in the mastered version compared to the premaster. Can you take a cut from the premaster at the exact same place so we can determine the exact reason for the noise.

You are right though, there is a clipping noise in the mastered version.

cornbb

ogy wrote: »

its not clipping but it is limited quite hard so the attack of the chords is a quite harsh, im no mastering engineer but i don't think the limiter should have had to work so hard. maybe ask him to redo the tracks and not to smash them so much.

Ah yes, thats what I meant to say

brettzy

Lads, have a listen to the chord at 8.82 sec, there is a definite clip (or noise) in the right channel.

cornbb

brettzy wrote: »

Lads, have a listen to the chord at 8.82 sec, there is a definite clip (or noise) in the right channel.

Yes, you're right. The final mixdown isn't clipping but the clip noise is there alright.

ogy

its not a clip, the highest peak on the track is -0.21 dB, its slammed to bits alright but it doesn't clip:)

at the part your talking about in the right channel it hits its peak value about 8 times in 0.1 of a second, which is gonna sound harsh.

dav nagle

Ok I have read all the feedback and I will upload the exact position now . PLEASE will you guys listen to the master on I-TUNES. The tearing is massive compared to any else I have listened to.


Cheers guys.

ogy

Yes, you're right. The final mixdown isn't clipping but the clip noise is there alright.

on closer inspection your dead right, theres a few peaks in a row so there is clip noise (sorry if editing this post caused any confusion:))

PLEASE will you guys listen to the master on I-TUNES

edit>preferences>playback: is "sound enhancer" or "sound check" ticked?
the clipping is very obvious when these are turned on. in fairness the idea of mastering is that it should transfer to different systems without any problems, but these itunes features are best avoided!

so in summary:):
- its limited too hard, ask the engineer to remaster it without smashing the life out of it
- don't use itunes "enhancements"

cornbb

I tried it in iTunes with the equaliser on and that made it sound pretty bad too. Fuppin iTunes!

ogy's right though, it should be mastered so that using an equaliser/Apple Magic Sound iWonderfulater™ shouldn't make it sound terrible.

dav nagle

http://www.box.net/shared/dznfygxcso

http://www.box.net/shared/25nwhdmw4w

Ogy you were on the ball !! Well done.

I am going to be honest and admit that its not my itunes player ( I had sound enhancer on (Massive woops!) However maybe the limiting is a bit harsh in the master so this was not a waste of anyones time.

I have sent these links above for you..

Please note that I appreciate the time you all have put into resolving this issue. It is great to get to the bottom of the problem and I feel confident that the issue will be resolved . I will release the album soon enough and if anyone who has helped me would like me to give them a present let me know.

Peace out

ogy wrote: »

on closer inspection your dead right, theres a few peaks in a row so there is clip noise (sorry if editing this post caused any confusion:))



edit>preferences>playback: is "sound enhancer" or "sound check" ticked?
the clipping is very obvious when these are turned on. in fairness the idea of mastering is that it should transfer to different systems without any problems, but these itunes features are best avoided!

so in summary:):
- its limited too hard, ask the engineer to remaster it without smashing the life out of it
- don't use itunes "enhancements"