Sunday, October 28, 2012

New iPod Touch 5th Generation Audio Quality Review: Part 3

No more subjective listening tests.  Time for some scientific audio data! (sort of) :-P

Let me start by saying I am by no means an expert at data comparison studies or methods.  Nonetheless, I decided to compare the new iPod Touch 5G with the previous iPod Nano 6G.  The iPod Nano 6G contains the same Cirrus audio DAC (digital to analog converter) as the iPod classic that I've used for over 5 years, which is also found in a few other iPod models.  I'm not sure which DAC the new Touch uses, but I've put it up agains the Nano 6G for comparison.

This test focuses on the headphone output jack, as the digital connector should theoretically be identical on each device, because it bypasses the iPod amplifier and sends just the audio data directly to your external equipment.  Let me describe the testing method and equipment used.

For recording the audio files, I have used a monster 3.5mm to 1/4" balanced stereo audio cable.  This connects the iPod headphone output to the 1/4" instrument inputs of my Apogee Duet audio interface.  Both iPods were recorded using the maximum iPod volume level with no audio equalization or settings applied.  Anything that would alter the original audio quality on the iPod is deactivated and the audio files transferred to the devices were in apple lossless format.  This is as pure as the iPod audio gets.

I chose the song "Freedom at Midnight" by David Benoit, because this is one the best recordings I've ever heard.  It contains clarity, depth, bass and separation that are simply of the highest quality.  I also di the same test with "Never Let Her Go" by David Gates, but I will only include the David Benoit results, as they are both comparable.

I used Logic Pro to record the audio at the matching 16-bit 44khz audio quality that the original files are encoded in.  After recording the audio files, I used a combination of Soundtrack Pro and Audacity to analyze the audio data and make comparisons.  The following are my results.

First, the iPod Touch 5G seems to have an ever so slightly louder output level.  Overall, the sound quality of the iPods is practically identical.  First lets look at the waveform of each file (iPod Nano 6G on top half, iPod Touch 5G on bottom half):

As you can see they look pretty much the same.  According to Audacity, the RMS (mean volume) is identical:

The frequency curve is also identical.  If you look carefully though, you can see the iPod Touch 5G in the bottom image has a lightly slightly greater output.  This can be seen by the entire graph being a little higher on the y-axis which is the volume (db):

iPod Nano 6G:

iPod Touch 5G:

Next we'll look at some spectrograph, which shows the amount of energy of specific frequencies over time.  This was configured for a maximum dynamic range of 100db to include the full 96db of CD dynamic range and 22khz maximum frequency to allow the full range of frequencies in the 44khz CD sample rate (see audacity manual).

The first view is the full length of the song (iPod Nano 6G on top half, iPod Touch 5G on bottom half):

There are some extremely minute differences, but for now we'll same they are essentially the same.  Here is a slightly closer view (iPod Nano 6G on top half, iPod Touch 5G on bottom half):

An even closer view still (iPod Nano 6G on top half, iPod Touch 5G on bottom half):

This view reveals something interesting; the two iPod play music at a slightly different speed.  This is a very close zoom level and a very small difference at only 1/10th of a second by the end of the song, but it is there.  Probably too small to amount to anything in the real world.  1/10th of a second over a few minutes is probably not an audible difference at all at only .4%.  Interesting nonetheless...

Lastly the spectrograph analyses from Soundtrack Pro:

iPod Nano 6G:

iPod Touch 5G:

And a closer view:

iPod Nano 6G:

iPod Touch 5G:

At first they might seem identical, as in the all of the above screenshots, however there are subtle differences. Take a very close look at the data from both programs:

Audacity (iPod Nano 6G on top half, iPod Touch 5G on bottom half):

Soundtrack Pro (iPod Nano 6G on top half, iPod Touch 5G on bottom half):

Although very small, there are some differences in the trails of data.  Look at the slight differences in the faint data at the top of the Audacity views.  Also note the difference in darkness of the straight horizontal line about a third up from each half of the Audacity view.  On the Soundtrack Pro view note the right side edge of some of the green spectrum data.  There are slightly different shapes and forms.  These are very close looks at the data, so in reality, I doubt any of this is audible.

If I had to guess, I would theorize that some of the difference seen are coming from the inherent noise in the analog systems and cables.  Every analog system has some small noise, and perhaps this noise is being seen at these extremely close views?  If that is the case, the noise would be random, causing the slight differences here and there.  Or perhaps there is some other interference?  It is even possible there is simply a slight difference in audio processing or amplification.  Perhaps the slight difference in output  gain is the cause?  Honestly, I have no idea.

However, in true listening tests and the relative comparison of data, I believe these difference are most likely not having any effect on the audible quality of the devices.  I have compared 256aac and 320aac files to lossless audio files, and the differences are greater than you see here.

In conclusion, I would call the audio quality of the iPod touch to be in line with the previous iPod Nano 6G, iPod Classic 160GB, iPod Touch 3G, iPod Nano 2G and possibly others.  These are the only iPods I have to compare, and while I haven't done extensive data analyses of all of them, I have done very excruciating listening tests, and they all appear to be the same basic quality.

I don't know which audio DAC is in the new iPod touch, or the old iPods mentioned, but after some research I think the implementation of the DAC in the entire audio system of the iPod is more important the the DAC itself.  That is simply based on second hand information, but the listening tests appear to verify this claim if in fact any of my iPods have different DACs.

If you have the new iPod touch, I'd love to hear your opinion on the audio quality.  Please leave a comment below.  But most of all, go listen to your music and enjoy it.  Isn't that the point of all of this? :-)

New iPod Touch 5th Generation Audio Quality Review: Part 2

Regarding my new iPod Touch 5g and its audio quality... as embarrassing as this is, I was doing comparisons for the last day (too much comparison for my own good) and realized something stupid.  I had recently switched to 256 AAC before my classic died, just to test the quality and save space to fit all my music on the classic.  Plus, 320 requires me to maintain a copy of every lossless song, because 256 is the highest on-the-fly conversion quality itunes allows.  Anyhow, I could detect the difference, but after agonizing over space vs. quality, I determined it was best to have all my music on me (i listen to a lot of different music all the time and never know what i'll want to hear next).  Needless to say, I hadn't done very much actual real music listening (for enjoyment, not comparison) with the 256 yet, but it seemed sufficient enough.

I guess I was wrong.  After comparing the devices, I remembered something unfortunate.  I had put 320aac files on my wife and father's ipods, so the comparison I was doing wasn't fair.  So, I copied 320 files to the new ipod touch and lo and behold, the quality was either identical or so close as to be undetectable easily.  Therefore, the ipod audio quality is in fact very good, so I'm guessing it's the same DAC as the classic?  256aac however is not very good.

I was so sure there was something wrong with the ipod quality that I was about to return it (literally the following day), when I realized my mistake and compared 320 files.  My wife and I both couldn't tell the difference between devices.  My wife even surprised me when she (not an audiophile) could tell the difference between 320 and 256 aac on every test blind.  I tried to fool her, but she could tell everytime and even explained how it sounded different without me ever telling her or her having any background in audio terms.

Needless to say, I'm sad to sacrifice space, but 320 is going back on my iPod.  If only we have ipod touch's big enough for lossless or at least larger 320 libraries... ugh...  I have tested 320 vs. lossless, and with my friend doing a double blind test I can tell the difference 80% of the time.  The irony is that sometime I don't even know 'why' I can tell the difference, but I know 'something' just doesn't sound right, and I can tell which is better.  Like it's missing some small amount of space or depth.  However, lossless just won't cut it on an ipod touch unfortunately.  Not enough gig-age there. ;-P

Thursday, October 25, 2012

New iPod Touch 5th Generation Audio Quality Review!

***This post has an important followup post. Please read the update.***

All right.  I am going to give my review of the new iPod touch 5th generation...

I'll start by saying that I have a very discerning ear, and well... basically, I'm super incredibly ridiculously picky about everything audio. :-P

With that said, the new iPod touch is a frustrating beast.  The audio quality is not the same as four other iPods I have.  I had a classic 160GB which died, but I've compared that to other devices, so I know where that stands in the scope of audio quality.  I have a 16GB nano g6, and also a 4GB nano g2.  Right now I have the two nanos and my new touch.

For reference, I have an apogee duet interface.  I'm familiar with a variety of sound systems from harman kardon to denon with polk to infinity alpha speaker systems.  I also have an array of in ear and studio quality headphones.  I have alesis m1 mk2 biamp studio monitors as well.  The point being that I am always comparing and contrasting different audio systems and very often using the differences of each system to determine mixing decisions with my own recordings.

So, with that background, I've compared the new ipod touch to the older ipods and the duet.  The results are interesting.  I don't want to come out and say the quality is worse, however it is definitely different.  I also don't want to over stress the differences, because to some they may not exist, to others they may be night and day.  However, I noticed there was a difference without even comparing devices.  I was listening to a recording by the band secret garden, which contains violins and strings, and I thought "something isn't right here".  There was a slight lack of clarity, and an almost boominess to certain low/mid frequencies.  My first thought was that I actually had an EQ set accidentally (I never use EQ), but after checking, there are absolutely no audio settings enabled.  The sound is as flat and unaltered as possible on the device.  So, I decided to compare it to the nano 2g first, because that was the closest to my classic, which I'm very familiar with.  The difference was sort of surprising.  The 2g nano is old and still extremely thin and small.  The old nano had more clarity.  I don't want to call it detail, and I'll explain in a second.

I compared a few different songs of different genres from david benoit jazz to david gates soft rock to secret garden new age, etc.  The difference was noticeable in each situation.  The more I listen the more difficult it is to determine how this is actually affecting the 'quality'.  I'm thinking the difference might be more of a frequency equalization difference.

Further comparing showed that although every device sounds a very slight bit different, they all sound extremely close to the duet (reference) while the new touch was the most far away from the duet in sound.  So my first thought is that if all the other devices are similar to the duet, which is a renowned audio interface for it's sound quality, then the touch is the failure here.  However, upon intensive listening, I believe the root "quality" is the same or similar, but the frequencies are of a different response curve.  You may argue that this is in fact a difference in quality (i would) but not in the same way as what I would consider "sound quality" in the truest sense.

For instance, you can have a smaller stereo separation or less depth to the sound or simply a lack of frequencies (can't be reproduced by the device).  As far as I can tell most of these aspects are similar across devices.  The differences in songs varied.  Sometimes the guitars in a song stood out more on the touch and sometimes drums stood out more clearly on the nano.  I believe the reason for this is because the touch has what I'll call a "boost" in the mid/low frequencies and a "reduction" in high frequencies (or possibly the illusion of this because of the added lows).  Therefore certain guitars with mid/low range would seem to stand out more on the touch, and crisp drums would on the nano.  Both can reveal all of the details of said instruments, and the nuances and atmosphere of the song are similar.  The real differences (so far as i can tell) are between the frequency adjustment.

If I had to guess based on experience, I would say the ipod touch 5g has a 1db boost from about 100hz to 350hz and a -.5 to -1db reduction in 15k to 20k both with a smooth rolloff.  At least that is what it "sounds" like.  It can be very hard to tell if there are other factors in play and the differences may be smaller than that.  Sometimes the other devices sound like they exhibit small distortion.  I'm referring to the minimal distortion from the amplification system, not clipping from bad recording or mastering...  This is almost undetectable, but then listening to the touch seems smoother.  This could be a cleaner amplification system, or an illusion from the difference in high eq.  The fact that details sound similar makes me think possibly that the audio is cleaner, but it's impossible for me to tell for sure.

I think the main difference is the overal tonality of the eq.  I'm on the fence as to whether I should return it or if i may end up liking it.  The logical experience part of my brain says it's not good.  Everything else is similar to reference, which is the truest the sound should be.  Therefore the touch is not being true to the source material.  However, being primarily an eq difference, it might not be that bad.  There is an enormous eq difference between every pair of headphones and speakers, while this is in the smallest magnitude different from the other devices.

I don't want to disuade anyone from getting a new ipod touch.  The device is amazing in every single way.  Including the audio quality that is crammed into such a feature packed device so freaking thin.  I may even find that a lot of the differences I hear are exaggerated because I "think" they are.  But I have a lot of experience in ruling out placebo type audio effects.  I've spent a lot of time training my ears 'and' my brain to detect differences including doing double blind testing on occasion.

I'll stop here for now and update if i find anything new or come to any different conclusion.  Keep in mind even though my comparison has been painstaking, relatively speaking it is a new device and i haven't had a 'lot' of time with it.  I'd love to hear other opinions of the 5g for those who have one.  Please leave comments below. And i'm glad to answer any specific questions...