Login / Register
 
Upsample to 96k for MFIT?
New Reply
Subscribe
I.R.Baboon
Thread Starter
#1
14th June 2013
Old 14th June 2013
  #1
Lives for gear
 
I.R.Baboon's Avatar
 
Joined: Nov 2006
Posts: 510

Thread Starter
I.R.Baboon is offline
Upsample to 96k for MFIT?

Hi!

I have a label asking for a 24-96 master for MFIT.

The original mixes were 24-44, so it's all rather pointless, but ok.

The artist doesn't want the extra expense of the true 24-96 master.

So he suggested i just re-sample the 24-44 up to 96k (and I'll also drop the volume a dB or so).

But my question is, does anyone think iTunes has some kind of automatic system to detect "fake" 24-96 masters?

And just to re-iterate, the artist does't want the extra expense of a true 96k master as the original master had loads of edits and complicated fades so it'll take me hours to make a "true" 96k master.

Many thanks!.......IR
Soundmaster Alex
Verified Member
#2
14th June 2013
Old 14th June 2013
  #2
Gear maniac
 
Soundmaster Alex's Avatar
 
Joined: Oct 2009
Posts: 215

Verified Member
Soundmaster Alex is offline
they are just reading too much into Apple's suggestion that higher sample rates are better. I've only ever submitted 24/44.1 for MFiT releases (exported straight from 24/44.1 projects) and had no problems at all.

And we (along with probably every other place) don't advocate upsampling anyway. I seem to remember someone telling me that Simon Heyworth, who masters at 96k in general afaik, will actually do the conversion down to 44.1 himself rather than supply 96k files for MFiT.

Anyway my short answer is no, and it seems like your clients understand MFiT about as well as ours...
__________________
Alex Wordsworth

www.emasters.co.uk
I.R.Baboon
Thread Starter
#3
14th June 2013
Old 14th June 2013
  #3
Lives for gear
 
I.R.Baboon's Avatar
 
Joined: Nov 2006
Posts: 510

Thread Starter
I.R.Baboon is offline
Thanks Alex.

Although i think it's good that there seem to be some things going on in terms of trying to improve all round quality, it's a shame some people are getting hung up on bit and sample rate whereas it's actually just the amount of limiting / distortion which is really the problem, and the bit and sample rate is a virtual irrelevance.
#4
14th June 2013
Old 14th June 2013
  #4
Lives for gear
 
wado1942's Avatar
 
Joined: Jan 2009
Location: Boise, Idaho
Posts: 2,697

wado1942 is offline
Yeah, it really needs to be common knowledge that the encoder/decoders need about 1dB headroom minimum to avoid internal clipping and Sound Check puts everything at about -16 LUFS. I'm all for 24/96K as the standard but level is the biggest influence. I've never mastered anything for iTunes at 96K, mostly 24/44.1, but you better believe I respect the suggestion for levels. If the guy is insisting on 96KHz, you can bump up the rate at the beginning of the mastering process and leave it as 96K through the end. That could at least potentially allow for slightly cleaner sound.
__________________
Stephen Baldassarre
www.gcmstudio.com
#5
14th June 2013
Old 14th June 2013
  #5
Gear addict
 
karumba's Avatar
 
Joined: Apr 2010
Location: Germany
Posts: 311

karumba is offline
the MFiT guidelines say, that no "upward-SRC" should be done. only if the sample rate of the original project is >96khz you should downsample to 96khz. in other words: if your project is 24/44, don't SRC.
__________________
Jan Ohlhorst
www.finemastering.de
Quote
2
I.R.Baboon
Thread Starter
#6
14th June 2013
Old 14th June 2013
  #6
Lives for gear
 
I.R.Baboon's Avatar
 
Joined: Nov 2006
Posts: 510

Thread Starter
I.R.Baboon is offline
yes, but if analog gear was used, then you have a choice to capture at 96k or 44.1 or whatever without any SRC.

And that's the issue here, the label is insisting on a 2496 MFIT master, which will require me to re-process the whole album through my analog gear at 96k for minimal (IMHO) benefit, extra cost to the client yadadadada
#7
14th June 2013
Old 14th June 2013
  #7
Lives for gear
 
wado1942's Avatar
 
Joined: Jan 2009
Location: Boise, Idaho
Posts: 2,697

wado1942 is offline
If the client is OK with the cost of the revision, then do it. You have no idea how much material shown on HDTV stations is from SD sources, which is essentially the same thing. I'll maintain that there's benefit to mastering at 96K even if the source is 44.1K but for the label, it's a numbers game. Give them their number and collect your check.
I.R.Baboon
Thread Starter
#8
14th June 2013
Old 14th June 2013
  #8
Lives for gear
 
I.R.Baboon's Avatar
 
Joined: Nov 2006
Posts: 510

Thread Starter
I.R.Baboon is offline
Client isn't ok with the extra cost, as stated in the first post.
dcollins
Verified Member
#9
14th June 2013
Old 14th June 2013
  #9
Lives for gear
 
dcollins's Avatar
 
Joined: Aug 2003
Location: Hollywood CA
Posts: 3,170

Verified Member
dcollins is offline
Quote:
Originally Posted by I.R.Baboon View Post
But my question is, does anyone think iTunes has some kind of automatic system to detect "fake" 24-96 masters?
No.

And if it's a good up-sampling, it's not going to make much difference anyway. Who knows, it might even be better.

There are plenty of fake 96k PCM downloads that are merely up-sampled 44k for sale today and no one seems to care.


DC
__________________
Dave Collins Mastering
www.collinsaudio.com
+1 323 467 5570
Quote
4
24-96 Mastering
Verified Member
#10
16th June 2013
Old 16th June 2013
  #10
Lives for gear
 
24-96 Mastering's Avatar
 
Joined: May 2008
Location: Karlsruhe, Germany
Posts: 2,836

Verified Member
24-96 Mastering is offline
Quote:
Originally Posted by I.R.Baboon View Post
But my question is, does anyone think iTunes has some kind of automatic system to detect "fake" 24-96 masters?

And just to re-iterate, the artist does't want the extra expense of a true 96k master as the original master had loads of edits and complicated fades so it'll take me hours to make a "true" 96k master.

Many thanks!.......IR
Quote:
Originally Posted by dcollins View Post
No.

And if it's a good up-sampling, it's not going to make much difference anyway. Who knows, it might even be better.

There are plenty of fake 96k PCM downloads that are merely up-sampled 44k for sale today and no one seems to care.


DC
I had an experience where a specialist audiophile online store for music downloads rejected a 24/96 master because the original source recordings had been 24/44. So they actually checked each incoming recording for spectrum used, I guess.

In this instance though, the analog mastering session had been captured in 24/96, so the master was initially delivered as 24/96 not to fake anything, but rather to maintain the master's bandwidth and avoid an unnecessary SRC stage at the end of it.

But even after I explained, the store people still insisted that I downsample the master to 24/44 before delivery, because they'd rather match their customers' expectations of what SR terms mean, and avoid customer confusion, than avoid unnecessary SRC. Makes good business sense, I suppose; but seems a bit strange coming from the audiophile community.
__________________
.
Robin Schmidt @ 24-96 Mastering
www.24-96.comfacebook
#11
17th June 2013
Old 17th June 2013
  #11
Lives for gear
 
neilwilkes's Avatar
 
Joined: Sep 2005
Location: London, England
Posts: 1,202
My Recordings/Credits

neilwilkes is offline
Quote:
Originally Posted by I.R.Baboon View Post
Hi!

I have a label asking for a 24-96 master for MFIT.

The original mixes were 24-44, so it's all rather pointless, but ok.

The artist doesn't want the extra expense of the true 24-96 master.

So he suggested i just re-sample the 24-44 up to 96k (and I'll also drop the volume a dB or so).

But my question is, does anyone think iTunes has some kind of automatic system to detect "fake" 24-96 masters?

And just to re-iterate, the artist does't want the extra expense of a true 96k master as the original master had loads of edits and complicated fades so it'll take me hours to make a "true" 96k master.

Many thanks!.......IR
MFIT states quite clearly that upsampling is not acceptable, which is right & proper. If the label wants 24/96 then the label will have to pay you for doing it - they need to have it explained to them that upsampling is actually forbidden by MFIT quite categorically, and upsampling or padding a 44.1 is not doing the cause of genuine High Resolution any favours at all.
__________________
Opus Productions Ltd - surround sound & high resolution audio
DAWplus E5-2687W, 32GB RAM, W7 Pro 64-bit, RME RayDat 1, UAD-2 Octo & 2x UAD-2 Quad, Adam Audio A7 x5.
Cubase (latest) & Nuendo (latest)
Waltz Mastering
Verified Member
#12
17th June 2013
Old 17th June 2013
  #12
Lives for gear
 
Waltz Mastering's Avatar
 
Joined: Mar 2008
Location: 3rd Stone From The Sun
Posts: 3,268

Verified Member
Waltz Mastering is offline
Quote:
Originally Posted by neilwilkes View Post
MFIT states quite clearly that upsampling is not acceptable
I can see why people are confused.. In Apples first round of lit, it did give the impression that they did want you to up-sample or deliver only at 24/96. That's the impression I got. It wasn't until later that they revised it, or cleared it up if I remember correctly. I guess Baboon could send them a copy of the mfit pdf page 3.

http://www.apple.com/itunes/mastered-for-itunes/
Paragraph
Provide High Resolution Masters

"Don’t upsample files to a higher resolution than their original format."
__________________

www.waltzmastering.com

#13
18th June 2013
Old 18th June 2013
  #13
Gear addict
 
karumba's Avatar
 
Joined: Apr 2010
Location: Germany
Posts: 311

karumba is offline
and with original format, the recording is meant, not the analog master. you could also see a 96khz analog master from a 44khz original source as some kind of a SRC. i just master with the original sample rate so i'm prepared for all directions.
Waltz Mastering
Verified Member
#14
18th June 2013
Old 18th June 2013
  #14
Lives for gear
 
Waltz Mastering's Avatar
 
Joined: Mar 2008
Location: 3rd Stone From The Sun
Posts: 3,268

Verified Member
Waltz Mastering is offline
There's always blue laws and loop holes.
Quite a few ME's workflow by default is to up-sample and work at higher rates.
or a 2 daw pitch catch situ.
#15
18th June 2013
Old 18th June 2013
  #15
Gear addict
 
zenmastering's Avatar
 
Joined: Jun 2005
Location: Left Coast of Canada
Posts: 327

zenmastering is offline
I've had the same issue...

The audiophilia nervosa types that hang out on the internet seemingly all know how to use a spectrum analyzer nowadays. So, HD-Tracks, et al, have their hands full with customers complaining about such things and without the knowledge to understand where the 'upsampled in mastering' scenario might make sense.

There's another hi-res download service in Europe that uses the SRC in PT on so-called 'fake 2fs' submissions unless you relent and provide your own down-sampled audio to them...I've a hunch this is the one you're referring to, Robin.

Graemme


Quote:
Originally Posted by 24-96 Mastering View Post
But even after I explained, the store people still insisted that I downsample the master to 24/44 before delivery, because they'd rather match their customers' expectations of what SR terms mean, and avoid customer confusion, than avoid unnecessary SRC. Makes good business sense, I suppose; but seems a bit strange coming from the audiophile community.
#16
18th June 2013
Old 18th June 2013
  #16
Gear addict
 
Joelistics's Avatar
 
Joined: Nov 2010
Location: Copenhagen, Denmark
Posts: 367

Joelistics is offline
Quote:
Originally Posted by karumba View Post
and with original format, the recording is meant, not the analog master. you could also see a 96khz analog master from a 44khz original source as some kind of a SRC. i just master with the original sample rate so i'm prepared for all directions.
This is incorrect according to my correspondent at apple. Capturing the analog chain at a different (i.e. higher) sampling to the mixes' native is not considered SRC and is totally fine.
Quote
2
#17
18th June 2013
Old 18th June 2013
  #17
Gear addict
 
karumba's Avatar
 
Joined: Apr 2010
Location: Germany
Posts: 311

karumba is offline
Quote:
Originally Posted by Joelistics View Post
This is incorrect according to my correspondent at apple. Capturing the analog chain at a different (i.e. higher) sampling to the mixes' native is not considered SRC and is totally fine.
i had the certification process 3 weeks ago. this is the official statement from my apple representative:

Quote:
What is the delivery specification for Mastered for iTunes masters?

The final deliverable is a high resolution PCM digital audio file. 96 kHz/24 bit is the preferred or optimal resolution, but please do not up-sample to achieve this. The appropriate resolution will usually be the native resolution of the project. For instance, if a project was recorded or digitized at 48 kHz that would be the recommended sample rate for the master. If the project was recorded at 88.2 kHz, then deliver at that resolution. All Mastered for iTunes masters must be at least 24 bit.
maybe i'm wrong, but in my understanding with "native resolution of the project" the resolution of the original recording, i.e. the resolution the project comes in for the mastering, is meant. at least my workflow (no upsampling at all, downsampling only to 96khz) has been approved by the apple representative.
#18
19th June 2013
Old 19th June 2013
  #18
Gear addict
 
karumba's Avatar
 
Joined: Apr 2010
Location: Germany
Posts: 311

karumba is offline
Quote:
Originally Posted by Miiika View Post
Mastering is part of "the project", don't you think? If it's recorded at 44.1 kHz but upsampled to 96 kKz for master processing, the "native resolution of the project" now 96 kHz or whatever sample rate the project comes out of mastering. Doesn't make any sense to only think of the recording as "the project".
of course, it is part of the project, but my guess is, that apples guideline is written from the view "prio mastering", since they are talking about "if a project was recorded or digitized at 48 kHz that would be the recommended sample rate for the master".

i've sent an email to my representative at apple to clarify that, but he is on holiday right now & will be back on 22th. i'll post his response here.
#19
19th June 2013
Old 19th June 2013
  #19
Lives for gear
 
neilwilkes's Avatar
 
Joined: Sep 2005
Location: London, England
Posts: 1,202
My Recordings/Credits

neilwilkes is offline
I agree with Miiika here - the recording may well have been at 44.1/48, and I have been sent recordings like this (multitracks) where I have done 2 versions:
1 - As supplied.
2 - Make backup copy of entire project, upsample to 96K and re-run the mix process using same plugs, same automation, same everything - and they sound different. Better is subjective & depends on the project in question but they most certainly sound different - audibly. The 96K version has more detail & air, and I am minded of the difference between 15IPS and 30IPS tape speeds.
So this is now a genuine 96K project and a genuine 96K mix even though the original recording was at a lower sample rate. The same applies to mastering (although great care must be taken when mastering digital mixes with analogue kit, as you are at the mercy of the DAC>ADC part of the chain.

What MFIT is saying, as far as I read it, is that there is absolutely no point at all upsampling the stereo master file. This makes sense, although it could (and has been) argued that upsampling can remove problems caused by poor filter design & eliminate pre-ringing issues too (See Bob Katz' "Mastering Audio, 2nd edition, pp 253) so who really knows?
Ben F
Verified Member
#20
19th June 2013
Old 19th June 2013
  #20
Lives for gear
 
Ben F's Avatar
 
Joined: Feb 2006
Location: Sydney, Australia
Posts: 3,182

Verified Member
Ben F is offline
I wish apple would just accept CD quality uncompressed audio, and have it available for download. All this hoops and roundabouts just to deliver a lossy format!
__________________
Studios 301
Quote
1
#21
21st June 2013
Old 21st June 2013
  #21
Gear addict
 
karumba's Avatar
 
Joined: Apr 2010
Location: Germany
Posts: 311

karumba is offline
here is the answer from apple:

Quote:
The requirements for MFiT are to start with minimum 24 bit sources at >= 44.1 kHz. Our recommendation is to maintain the native sample rate of the project, but there is no hard and fast rule.

For 96 kHz masters we would prefer that the source have been 96 kHz -- the idea being that one cannot capture resolution that wasn't recorded, but I understand engineers may have a preferred sample rate (especially as many converters run best at a particular rate.)

As you know, the delivery requirement for MFiT is 24 bit >= 44.1 kHz. We don't have explicit rules about "valid" 96 kHz masters.
so i was both right & wrong. they prefer the sample rate of the original recording, but they are fine when the analog mastering is done with a sample rate higher than the sample rate of the original recording.
New Reply Submit Thread to Facebook Facebook  Submit Thread to Twitter Twitter  Submit Thread to LinkedIn LinkedIn  Submit Thread to Google+ Google+ 
 
Topic:
Thread Tools
Search this Thread
Search this Thread:

Advanced Search
Forum Jump

SEO by vBSEO ©2011, Crawlability, Inc.