[elfary]

Dithering always beats truncating.
 
Thus the best option to go from 24 to 16 is computer with good software (Izotope RX Advanced is very good at downsampling and dithering).

 

[utkuozdemir]

After trying a lot of software, I decided to go with dbpoweramp's fixed 16 bit depth convertion using triangular algorithm. Tried lots of apps including Audacity, Adobe Audition, izotope RX, samplitude etc. Some of them does not support batch processing, some does't support flac, alac etc. really pain in the neck. After dithering, I apply resampling to 48khz, so I will be using 16bit/48khz alac files. Thank you all for your answers.

 

[castlevania32]

I have a lot of files in different bitrates/sampling and i would like to convert them all to 16/44.1
I downloaded LAME to make VBR v0 but i can't find how it works

do i need something like audacity ?

 

[utkuozdemir]

 
[size=medium]

you can use dbPoweramp, it has most of the popular encoder-decoders for many formats.
​

[/size]


Quote:

I have a lot of files in different bitrates/sampling and i would like to convert them all to 16/44.1
I downloaded LAME to make VBR v0 but i can't find how it works

do i need something like audacity ?

 

 

[freqazoidiac]

I will chime in with my personal tests and experience.
 
Stereophile thread about another player but discuss the Ipod truncating - http://www.stereophile.com/content/astellampkern-ak100-portable-media-player
 
I'm testing with Ipod classic v5 from the mid 2000's. USB with no mods and using itunes.
 
The author of that article is right - but he just scratched the surface. I did sonograph tests with Spark XL and Mackie U.420 which has the unique ability of
being able to feed back into the 24bit chip set of the Mackie with whatever is plugged in (ipod to usb on mac) or played internal on the computer with the
push of a button.
 
What I saw on the 'scope' is when the 24/48 AIFF raw file from  a 24/96 aiff source from vinyl with high peaks at 32khz originally and all high peaks
enhanced with an eq before I stepped it down to 24/48 to make a 24/48 ALAC file - so there would be no mistake if the Ipod was playing back at the resolution
intended.

I must add - this would not conclude if it was playing back at 24bits, but it will show clearly if it's capable of playing back to 48hkz.
 
What it does in this test, with BOTH the AIFF file and the ALAC at 24/48 each is it clearly shapes the sound and truncates. It pushes all the high frequencies down and while
most peaks were at 22khz and above  on the files before they were loaded on the ipod, now most hover at 19khz. The weird thing is it shows these odd spikes
at the highest voltage peaks and frequencies and it reads at above 48khz! But everything else is pushed down. I don't know if this is some sort of additive algorithm or what
but it's nonsense because it's sharp spikes but THEY don't show up unless I push the volume above unity gain on the Mackie.
 
So I wasted some time preparing 24-44 ALAC files of my vinyl to play on my ipod IMO - I started with 24/48 but the ipod had a hard time with some material at that
rate and it seemed to be a buffer issue .
 
My next test will be to do the same thing with  16/48 files to see if it's doing the same noise shaping or if that's just the process it uses to step down from 24 bits to 16.

Very weird stuff here - and as some other person mentioned in this post - it makes no sense if it was loaded with the chipset that is capable!
 
(Yes I had the EQ off in itunes and no 'sound enhancement' 
 
If this thread is expanded and people need proof I will do screen shots of the sonograph and the clear shaping.

 

[freqazoidiac]

Here is some proof of why there is no REASON to bother with 24 bit files on the Ipod classic 5th gen.
(a continuation of my previous post)
 
File dithered from 24/96 to 16/48 using 'ultrabitmax' in Spark XL running mac 10.6.8.
 
The original source is from Vinyl of 1965 recording from analogue master tape and has original peaks of high frequency  into the mid 30khz recorded 24/96 raw AIFF.
 
I don't feel it's now necessary to repeat it for my previous post to see proof of massive noise shaping and frequency reduction of all files.
 
This shows even when playing back a 16bit/48khz .aiff file on Ipod 5th gen, there is the shaping happening just like the 24/48 files.
 
The first picture below is of the 30 second 16/48 file playing native in Spark  and then repeated the same file once it's loaded into the ipod
and played back through the Mackie u.420 mixer with the firewire loop back system (enables to record pure signal what is being played within the mac system up to 24/96)
while playing through the itunes interface with the usb , NOT the headphone out to RCA.
 
___________below - 30 second file played through spark XL
 

 
 
___________below - 30 second file (same one as above) played through ipod

 
 
JUST LOOK AT THAT ROLL OFF !!!

 

[freqazoidiac]

Onto an earlier Ipod.  My personal favorite in sound has always been the Ipod 3rd Generation (and I love the way it looks and feels)
with Firewire.
 

 
 
 
 
Same test as above starting with 16/48 .aiff 30 sec file
 
 
 
___________below - 30 second file played through spark XL
 

 
 
 
___________below - 30 second file (same one as above) played through ipod
 
 

 
 
 
BINGO - looks very , very close  to the native resolution if not simply the same without any compression or shaping after 20khz!
 
 
Now - I will further show you what the Ipod 3rd gen Firewire is capable of handling.
 
 
 
24/48 .aiff 30 sec file_________________
 

 
 
Basically the same , a bit crispier resolution, but just barely and reaching 24khz - but what I find out next show's the limitation of this Ipod.
 
 
24/64 .aiff 25 sec file playing NATIVE IN SPARK XL _________________
 

 
Notice the easy extension of the high frequencies , some tailing going to 32khz. Again this is the Raw file from the analogue Vinyl recording.
 
 
 
___________below - 24/64 .aiff 25 second file (same one as above) played through ipod
 

 
Now we can see the limitation of output of the audio in this model. There is clear truncating or shaping with no extension
going beyond 24khz per channel.
 
 
 
Now let's look at a 24/48 ALAC version , from the Ipod.
 

 
A very nice resolution, pretty close to the AIFF raw file at 24/48.
 


The scope does not lie - and the 3rd Gen Firewire Ipod is keeping the music faithful.  I'm rather disgusted by what is going on with the 5th gen Ipod!

 

[freqazoidiac]

Part 3 of the previous two posts.

I'm just posting information about the equipment and method used to make this test to determine the real output of the Ipods in discussion.
 
What was key to processing the 24bit digital audio is the Mackie U.420 series (discontinued)
 
It has a unique function Firewire Loop Out - meant to capture any audio that is being played through the computer
in any audio rate up to 24/96.
 
Using Macbook pro 2007 model (FW 400 and FW 800 ports) , TC Electronics SPARK XL (final release) and the Mackie.
I could use other software such as Adobe Audition , but I prefer Spark on the platform. Just as long as it has a sonogram / sonograph spectrum tool.
(In Audition you can view the sonogram right on the waveform window )