IPB

Welcome Guest ( Log In | Register )

 
Reply to this topicStart new topic
Got me some 24-bit 88.2KHz FLACs. Need to get them onto an iPod, Should I convert to 44.1KHz or 48KHz?
neothe0ne
post May 10 2010, 11:15
Post #1





Group: Members
Posts: 295
Joined: 25-September 05
Member No.: 24684



Bought me some 24-bit 88.2KHz FLAC's from the Final Fantasy Distant Worlds II setlist. I need to put them on my iPod, but iTunes says the 88.2KHz AAC's can't be copied. Should I convert to 44.1KHz or 48KHz?
Go to the top of the page
+Quote Post
pdq
post May 10 2010, 11:24
Post #2





Group: Members
Posts: 3407
Joined: 1-September 05
From: SE Pennsylvania
Member No.: 24233



44.1
Go to the top of the page
+Quote Post
a13xth0rn3
post May 10 2010, 13:23
Post #3





Group: Members
Posts: 17
Joined: 27-November 09
From: UK
Member No.: 75358



convert to 24/48 apple lossless. will work fine.
Go to the top of the page
+Quote Post
db1989
post May 10 2010, 13:28
Post #4





Group: Super Moderator
Posts: 5275
Joined: 23-June 06
Member No.: 32180



No, pdq is right. 44.1 kHz is preferable (though I doubt anyone could hear a difference either way), because it's an integer division of 88.2 kHz. Also, OP appears to want to convert to AAC; I presume s/he is keeping the originals, in which case ALAC will simply waste space and sound no different.
Go to the top of the page
+Quote Post
Notat
post May 10 2010, 17:09
Post #5





Group: Members
Posts: 581
Joined: 17-August 09
Member No.: 72373



48 kHz will give you more bandwidth so why not choose that? The sample rate conversion from 88.2 to 48 kHz is more computationally intensive but it is not, in principle, any less accurate.
Go to the top of the page
+Quote Post
pdq
post May 10 2010, 17:44
Post #6





Group: Members
Posts: 3407
Joined: 1-September 05
From: SE Pennsylvania
Member No.: 24233



QUOTE (Notat @ May 10 2010, 12:09) *
48 kHz will give you more bandwidth so why not choose that? The sample rate conversion from 88.2 to 48 kHz is more computationally intensive but it is not, in principle, any less accurate.

I suspect that the filter that is applied for 88.4 -> 48 is has a much wider impulse response than the one for 88.4 -> 44.1, so I believe that it is less accurate. Whether that is significant or not I don't know.

As for the bandwidth, you are talking about frequencies that are beyond audibility anyway.
Go to the top of the page
+Quote Post
Notat
post May 10 2010, 18:55
Post #7





Group: Members
Posts: 581
Joined: 17-August 09
Member No.: 72373



Same filter theory applies here as to an ADC. To avoid affecting the 20 kHz passband, the 48 KHz filter does not need to be as steep as the 44.1 kHz filter. A less steep filter has a shorter impulse response. So I'm afraid you have it backwards.

The reason 48 is better than 44.1 kHz is not necessarily because of increased bandwidth but because of this more relaxed transition band for anti-aliasing.

The thing that the 44.1 kHz filter has going for it is that because of the 2:1 sample rate, it works out that many of the filter coefficients work out to be 0 and so the implementation is less compute intensive. If you're in a hurry, it would be the better choice.
Go to the top of the page
+Quote Post
a13xth0rn3
post May 10 2010, 19:33
Post #8





Group: Members
Posts: 17
Joined: 27-November 09
From: UK
Member No.: 75358



QUOTE (dv1989 @ May 10 2010, 12:28) *
Also, OP appears to want to convert to AAC; I presume s/he is keeping the originals, in which case ALAC will simply waste space and sound no different.


How will it waste space? The question was should s/he convert 88.2 FLAC to 44.1 or 48. Converting to ALAC will not be wasting space as the original files are lossless.

This post has been edited by a13xth0rn3: May 10 2010, 19:33
Go to the top of the page
+Quote Post
db1989
post May 10 2010, 22:28
Post #9





Group: Super Moderator
Posts: 5275
Joined: 23-June 06
Member No.: 32180



AAC files are smaller than ALAC files of identical duration. If the OP is keeping the original lossless files, IMO s/he may as well use transparent AAC files for the iPod, since they'll occupy less space.
Go to the top of the page
+Quote Post
Canar
post May 11 2010, 06:18
Post #10





Group: Super Moderator
Posts: 3361
Joined: 26-July 02
From: princegeorge.ca
Member No.: 2796



Many lossy codecs are more highly tuned for 44.1 than for 48, IIRC. I'd recommend 44.1. 48 has more ultrasonics to bloat bitrate without providing any audible benefit anyhow.


--------------------
You cannot ABX the rustling of jimmies.
No mouse? No problem.
Go to the top of the page
+Quote Post
X-Fi6
post May 12 2010, 05:23
Post #11





Group: Members
Posts: 13
Joined: 8-October 09
Member No.: 73798



I've always wondered why people care about audio higher than the Redbook standard (44.1KHz, 16 bits per sample).

Humans can't perceive 22,050Hz and they certainly can't hear when samples are 1/65535th higher or lower than they should be, so you're wasting bandwidth.

This post has been edited by X-Fi6: May 12 2010, 05:24


--------------------
Mixing audio perfectly doesn't take more than onboard.
Go to the top of the page
+Quote Post
Notat
post May 12 2010, 15:35
Post #12





Group: Members
Posts: 581
Joined: 17-August 09
Member No.: 72373



QUOTE (X-Fi6 @ May 11 2010, 22:23) *
I've always wondered why people care about audio higher than the Redbook standard (44.1KHz, 16 bits per sample).

Wonder no longer. Simply scroll back and read my previous response (#7). There have been some credible cases where higher sample rates can be ABX'd. In all the cases I'm aware of, the audible difference has been determined to come from design compromises in the anti-aliasing and reconstruction filters.
Go to the top of the page
+Quote Post
lvqcl
post May 12 2010, 16:33
Post #13





Group: Developer
Posts: 3383
Joined: 2-December 07
Member No.: 49183



IOW, those cases are unrelated to OP's iPod?
Go to the top of the page
+Quote Post
Notat
post May 12 2010, 19:06
Post #14





Group: Members
Posts: 581
Joined: 17-August 09
Member No.: 72373



QUOTE (lvqcl @ May 12 2010, 09:33) *
IOW, those cases are unrelated to OP's iPod?

The OP is going to be doing sample rate conversion. There will be anti-aliasing filtering involved. All else being equal, a 22.05 kHz filter will be more audible in the passband than a 24 kHz filter. I don't think any of this will be of significance if he then goes and pushes it through a lossy coder and listens to it on Apple earbuds.

The myth I was taking the opportunity to debunk here is that sample-rate conversion by an even 2:1 ratio will sound better than conversion through a more obscure ratio. Not true. The even ratio is less computationally intensive but is not any more accurate.

This post has been edited by Notat: May 12 2010, 19:08
Go to the top of the page
+Quote Post
lvqcl
post May 12 2010, 20:08
Post #15





Group: Developer
Posts: 3383
Joined: 2-December 07
Member No.: 49183



QUOTE (Notat @ May 12 2010, 22:06) *
All else being equal, a 22.05 kHz filter will be more audible in the passband than a 24 kHz filter.

Adobe Audition, 96->44.1 conversion: the passband is flat up to 20+ kHz. I don't think that such anti-aliasing filter is audible...


Go to the top of the page
+Quote Post
greynol
post May 12 2010, 20:19
Post #16





Group: Super Moderator
Posts: 10000
Joined: 1-April 04
From: San Francisco
Member No.: 13167



QUOTE (Notat @ May 12 2010, 11:06) *
All else being equal, a 22.05 kHz filter will be more audible in the passband than a 24 kHz filter.

As per TOS #8, if you wish to comment on audible differences in sound quality, the onus is on to demonstrate them via ABX.


--------------------
I should publish a list of forum idiots.
Go to the top of the page
+Quote Post
db1989
post May 12 2010, 21:55
Post #17





Group: Super Moderator
Posts: 5275
Joined: 23-June 06
Member No.: 32180



QUOTE (Notat @ May 12 2010, 19:06) *
The myth I was taking the opportunity to debunk here is that sample-rate conversion by an even 2:1 ratio will sound better than conversion through a more obscure ratio. Not true. The even ratio is less computationally intensive but is not any more accurate.

Since I hinted at that, my mistake. I assumed that no interpolation is required (i.e. that one just ignores every other sample) and so less additional noise, etc. is introduced. This probably betrays my lack of knowledge regarding signal processing. smile.gif
Go to the top of the page
+Quote Post
DVDdoug
post May 12 2010, 22:54
Post #18





Group: Members
Posts: 2596
Joined: 24-August 07
From: Silicon Valley
Member No.: 46454



QUOTE
Since I hinted at that, my mistake. I assumed that no interpolation is required (i.e. that one just ignores every other sample) and so less additional noise, etc. is introduced. This probably betrays my lack of knowledge regarding signal processing.
I'm not a DSP expert, so the following might be wrong... But, here's how I understand it:

You can simply throw-away every other sample, but you have to filter first.

Imagine that you have a 30kHz signal sampled at 88.2kHz. If you throw away half of the samples and reduce the clock to 44.1kHz, you will still have "something" there. But, that "something" cannot be 30kHz... It's now an alias.

Go to the top of the page
+Quote Post
Notat
post May 13 2010, 00:50
Post #19





Group: Members
Posts: 581
Joined: 17-August 09
Member No.: 72373



Greynol is right, I should not talk about audibility. These are not necessarily audible differences There are measurable differences between a steep and relaxed anti-aliasing filter - the former will have more passband ripple or attenuation and phase distortion near the cutoff.

lvqcl, your filter looks good in the passband but I can't see the stop band. You should also plot phase.

DVDdoug you are correct, to do a 2:1 downsampling, you can anti-alias filter and then remove every other sample. Sample-rate conversion is usually done in one combined filtering and decimation pass - it doesn't make sense to do the filter calculations for all the samples if you're only going to end up using half of them.
Go to the top of the page
+Quote Post

Reply to this topicStart new topic
1 User(s) are reading this topic (1 Guests and 0 Anonymous Users)
0 Members:

 



RSS Lo-Fi Version Time is now: 21st September 2014 - 03:48