Few questions concerning mobile audio

[LightFantastic]

Apart from the 30 some ought bands of EQ, pretty lights, phase control, time alignment, etc, what makes a receiver superior to another in terms of sound quality only?

Clarion. McIntosh. Nakamichi. All of these manufacturers carry dead heads; that is receivers with no internal amplification. And most people who run these do all their processing using an external processor.

So why drop $1k on a receiver just to extract a signal? Is the DAC really that good? Is the DAC alone really worth a thousand dollars? Do you really need 8V to feed into your amplifiers?

I ask this because I made note that I wanted to use my tablet as my receiver to a few audiophiles, and I was laughed at. I don't understand. All of my amplification and filtering are done at the amplifier. So all we are doing is comparing signals. Why is the signal coming from an aftermarket receiver so much better than the signal coming from my Samsung Galaxy Note 3.

I do understand one is low level and one is high level, but that can be fixed using a converter. I also understand my tablet won't play digitally remastered discs. And one morething that might affect the quality: the signal coming from my tablet is already amplified. NowI have another question? How can a dead head pass a signal without amplification of some degree?

Just curious why aftermarket head units have superior signals to those outputed from most tablet s.

 

[Averagesupernova]

First lesson: Never argue with an audiophile.
-
8 volt outputs compared to 4 volt, and prior to that 2 volt, etc. etc. gives you a better signal to noise ratio. The gains at the amplifier can be turned down farther so any noise picked up between the head unit and the amp will be amplified less. What you refer to as a dead head I suspect is simply a head unit that is not designed to drive speakers. Why should manufacturers build the ability to drive a speaker when most head units are hooked to a separate amplifier using line-level outputs?

 

[meBigGuy]

Audiophiles are fun to argue with if you are in the mood to argue with a wall.

I call a dead-head a pre-amplifier. It contains all the electronics to produce a line-level output which can then be routed to a a power amplifier.

Producing a clean line-level output is difficult. In your tablet, the process of converting a digital signal to a line level signal suffers from quality issues and poor tradeoffs in many areas.

For example, just consider power supply noise. If you take the audio output from your tablet and listen to a silent recording, you will hear the power fluctuations caused by processor addressing loops. (Within a PC it is really bad and you can hear disc drive accesses.)

Then, you have to consider the CODEC chip (contains the DAC) that takes digital data and converts it to analog, and especially the filtering that is applied to prevent aliasing. Many shortcuts in the tablet. A tablet might get a 70 or 80 dB SNR, whereas a quality receiver will be 100+ dB SNR. Sophisticated DACs use oversampled converters and discrete element matching techniques to achieve high performance. Just the filters are more complicated than the tablets whole audio chain.

Then the output amplifier. In a tablet the harmonic distortion and crossover distortion are terrible.

Then, you need to consider rate conversion, that is, how the 44.1 wav file is converted to a 48KHz or higher sample rate. That is yet another science.

So, the answer is YES, there is a huge difference between your tablet and even a terrible receiver. The simplest example would be how much better a $10 USB audio device sounds than the PC internal sound card.

And all of this is not even getting close to the debatable audiophile fantasies.

But, the end result is whether it is good enough for you. Some Tests have shown that many people prefer the distortion of mp3 to clean audio (because they are used to it). And, beats headphones are extremely successfull.

If you are just going to play 128Kbit mp3's, then it probably won't make much difference.

 

[olivermsun]

So, the answer is YES, there is a huge difference between your tablet and even a terrible receiver. The simplest example would be how much better a $10 USB audio device sounds than the PC internal sound card.

There are some small audio devices and iPods that measure very well on the bench. I haven't seen an objective test of an iPad recently but I'd imagine it would perform much the same as an iPod. On the other hand there's no reason why a reasonably-designed PC sound card couldn't be excellent.

I wouldn't use any of the above to drive a power amplifier directly, though.

 

[LightFantastic]

Line level equivalent to low level. Okay.

Is there a proper term for high level?

If a preamplifier just prepares a signal to be amplified, then is it fair to state that a LOC (line output converter) is a preamplifier?

Sampling rate. I'm not sure if I quite understand this, so my logic might be off. If I want to record voice, then I'm going to take an analog signal and convert it to digital so I can store it. What determines the quality of the recording? Does the sampling rate only work on the DAC side or does it come into play on the ADC as well. For instance, let's say I record something with a sampling rate of 100. Wouldn't it be a waste if I used a DAC with a sampling rate at anything above 100?

People state the higher the sampling rate, the better. What if the sampling rate of the DAC exceeds the sampling rate of the actual recording. That would be useless, no?

 

[olivermsun]

Your tablet's headphone output isn't very "high level" — it's just meant to drive headphones, not speakers. You typically don't need a converter for this, you just turn it up to a moderately-high volume where it doesn't clip.

An actual head unit obviously has some advantages besides (hopefully) better D/A and line outs. It's plugged in all the time, doesn't slide, has a display and controls that are easy to use when driving, etc. Newer ones often do fancy tricks like connect with your phone and turn down the music when you answer a call. Also the radio usually connects properly to the car's antennas.

On the other hand, for $1000 it better have a built in GPS, wash my windshield, and drive my car for me when I have to take a short nap. :big grin: (Seriously, unless you have some overriding need to have a premium auto sound system, you get a lot better sound for the money in a nice, quiet, spacious environment like, say, your home.)

Line output converters take the speaker-level outputs, usually from a factory head unit, and bring it down to line level. You only use these if you don't have a line-out.

Sampling rate shouldn't be an issue for your purposes. If you record voice, then yes your sound inputs will take an analog input from the mic and convert it to digital (A/D conversion). The nominal sampling rates for most consumer audio are either 44.1kHz or 48kHz, but for both A/D and D/A conversion there are advantages to using a higher "internal" sampling rate than the storage rate.

 

[LightFantastic]

Okay thanks.

So I can just purchase a 3.5mm to RCA splitter and feed my amplifier (model Phoenix Gold M44) directly. Is a line driver worth it to boost the voltage?

 

[olivermsun]

I'd just try it with the Y-cable first. Once upon a time, we used to use a Sony Discman as a line-level test input for some pretty high end car stereos, and it worked fine. If the line driver becomes necessary, then again I'd be weighing the cost against even a low-end head unit, perhaps with an aux in. (Oh and one other nice thing about head units — they usually have a fader control.)

 

[AlephZero]

People state the higher the sampling rate, the better. What if the sampling rate of the DAC exceeds the sampling rate of the actual recording. That would be useless, no?

That depends on the design of DAC. Actually, the cheap and cheerful DACs in consumer audio products work at a MUCH higher sample rate than the digital input - probably at a few MHz. But that is to compensate for the fact that the analog output is only ever 0 or the maximum, nothing in between.

If that seems like the quality would be horrible - well, saying you have a "1-bit Sigma-Delta DAC" looks sexy in the sales brochure. And even more important, it's cheap.

As a general rule, what "people say" isn't worth much, unless you have some measure of what they actually know. IMO the best is actually learn something about analog and digital signal processing, not hang out with people who think they have "golden ears".

@olivermsun

On the other hand there's no reason why a reasonably-designed PC sound card couldn't be excellent.

Of course. And that's why some PC sound cards cost quite a bit more than $1000 (and they don't even have EQ with 30+ flashing lights - if you want that, you buy it separately).

 

[LightFantastic]

The problem I'm facing is the HVAC controls. They are on the same board as my radio. Dash kit for my vehicle is 550 USD, not including shipping from Japan. Then I still have to purchase a receiver. This is why I resorted to bypassing the OEM head unit all together and just using my tablet. I was just curious on the DAC. Ill give it a whirl.

I'm not sure if I want to drop money on a processor like the MS8. I only have 3 channels. Full ranges up front and an SLS 6 on the rear deck IB. That would be overkill. They need to make a 3 ch dsp!

 

[LightFantastic]

IMO the best is actually learn something about analog and digital signal processing, not hang out with people who think they have "golden ears".

Exactly why I'm on here.

 

[olivermsun]

The problem I'm facing is the HVAC controls. They are on the same board as my radio. Dash kit for my vehicle is 550 USD, not including shipping from Japan.

Now i got it.

Are there aftermarket aux in/line out kits for your vehicle? If there is no line out option but the head unit is otherwise acceptable to you, the line level converter is not a bad option imho.

 

[olivermsun]

If that seems like the quality would be horrible - well, saying you have a "1-bit Sigma-Delta DAC" looks sexy in the sales brochure. And even more important, it's cheap.

And many are capable of excellent performance. So why not?

 

[meBigGuy]

1. Actually, sigma delta converters are used in some of the highest and best high end audio equipment.

2. I've measured ipod, ipad, and iphone audio systems and worked on IC's that beat their specs in all ways. We also build audio amplifier IC's. The iphone, ipad, etc are OK, but not great. I use my ipod touch as a signal generator for rough testing.

3. There is no way any PC produces good quality audio with its motherboard audio system. A sound card in a PC can do better, but you would be lucky to get 15 bits if it is running from the PC power supply. I suppose the very expensive ones do OK (I've never tested any) but an external card on its own supply can do better, cheaper.

4. "But that is to compensate for the fact that the analog output is only ever 0 or the maximum, nothing in between."

While what you say is sort of technically correct, the idea of "compensate" has a negative connotation. Also, a sigma delta is much more than just a high speed 1 bit converter. The 1 bit signal is noise shaped (modulated) so that the decimation filter gets more SNR increase for each halving of the sample rate. The order of the sigma delta modulator is a big deal. The nature of the decimation filter is a big deal.

The reason they can be cheap is because it is easier to build high speed sigma-deltas into integrated circuits than to build classic ADC and DAC circuits based on matched discrete elements. It is not just "because they are 1 bit".

5. Regarding "the higher the sampling rate, the better": That is a tricky one. From a pure DSP sampling theory point of view you can perfectly duplicate any signal by digitizing or converting it at 2X the highest frequency in the signal. But there are practical problems. Going to higher sample rates makes it easier to design the filters required to clean up potential aliasing products. So, basically it is true, but higher sample rate does not guarantee high quality.

 

[LightFantastic]

What is aliasing?

 

[olivermsun]

The Wikipedia article on Aliasing actually seems like a good start, if a bet dense. Good pictures and links to more specific examples. You've probably encountered more than a few examples in your own daily life.

 

[meBigGuy]

This is an example for aliasing in the ADC.

If the sample rate is not high enough (at least 2 samples per cycle for the highest frequency) the the effect is that "new" frequencies appear. It generally sounds like low level noise/distortion in the background of the sounds. High frequency noise or content in the original can "wrap around" and appear as low frequency distortion. To avoid this, the original is filtered (by an anti-alias filter) such that any such high frequencies in the original are suppressed. For example, if the sample rate is 44.1KHz, then any content above 22.05KHz will alias. The practical problem is that if the filter affects anything below 20KHz, then it could be audible. Designing a filter that does not affect 20KHz, but totally eliminates 22.05 KHz is difficult(if not impossible), so tradeoffs have to be made. But, if the original is sampled at 88.2 KHz, then a filter for 20KHz to 88.2KHz is easier.

In the DAC, a similar effect will occur. http://en.wikipedia.org/wiki/Reconstruction_filter kind of explains it, although the comments about "remove the stair-step" are misleading.

It basically amounts to how many different waveforms can you draw that fit the points and are in the frequency band you are interested in.