[dura]

Quote:

the Sony tuning of the 'flat' sound is... well, for want of a simpler word, cheating.

The difference from 20 to 200 hz seems to me about .25 db!

[/QUOTE].
http://www.dapreview.net/content.php?article.133

In the picture in this article I see a bass rollof starting above 200hz and ending at -9db at 20 hz for the ipod at 25 ohms. Now this is engineered!

 

[bangraman]

Quote:

Originally Posted by dura The difference from 20 to 200 hz seems to me about .25 db! http://www.dapreview.net/content.php?article.133 In the picture in this article I see a bass rollof starting above 200hz and ending at -9db at 20 hz for the ipod at 25 ohms. Now this is engineered!

No, that's called bad judgement. They aren't seemingly trying to 'load the dice'. They are definitely shooting themselves in the foot.

 

[bangraman]

Quote:

Originally Posted by sno1man and the ear is much more sensitive to very small volume changes at lower frequencies then at higher ones. The old 1DB to notice a difference I think applies to white noise, not specific frequencies or groups of them.

This came up in discussions with non-Sony people as well a while ago, which came back into my memories when I looked at the FR chart. I think I have very good hearing, but I don't think it's exceptional. The fact that I can definitely hear the difference and isolate it is probably just down to being a (ex) musician and spending long hours with EQ. For most others, I guess it might just translate as 'better'. Sony's audio department should have a lot of psychoacoustic bods they can call upon, and I'm speculating here but maybe they did that when they realised their APU wasn't exactly a world-beater.

 

[DigDub]

when i had the hd3, i did notice that there is a bassline thumping away even though the EQ is flat. even turning the bass on the EQ all the way down did not eliminate this bassy 'background'.

 

[gords]

Quote:

Originally Posted by bangraman The HD5 is definitely bassy though... it'll supply a lot more bass than the iPod, at the noticeable (to me) expense of sound quality.

Quote:

Originally Posted by bangraman Simple really, overall clarity of sound. There's less of it. There's less sense of air and space, and you hear less detail in the sound. Or putting it another way, there's significantly more mud in the sound. It's very easily detectable against the iPod and the Karma, and also detectable to a lesser degree against the iRivers.

How large exactly is this sound quality difference? If it really is "very easily detectable" I'm confused as to how the cnet staff could rank the hd5 and ipod on complete opposite sides of the scale to you.

 

[bangraman]

There are a lot of factors at work here... First of all, most of the phones I use with the iPod are higher-impedance and do counteract the bass falloff rather well. The E5 is something of an exception, as Mavis pointed out accurately before they do present a very low load to the headphone jack for much of their overall response. But the extended bass of the E5 is balanced out by the loss of the bass on the iPod.


The difference in bass delivery with a low impedance load between the iPod and the HD5 is quite major. The iPod's working from a disadvantage with the falloff and it has problems with the EQ anyway, and the HD5 has the response you see in the graph plus a working EQ which doesn't clip noticeably.


This comes back to the original points mentioned in the post. It seems to me that the CNET staff couldn't judge relative sound quality as such, but rather depended almost solely on the difference in sonic flavour between the sources. Therefore the iPod came last because they employed a 32-ohm load to do the testing, and the HD5 came first because it deviated from a flat response that the most of the other players would have had.

 

[Beethovenian]

I'm sorry, I respect Bangraman's knowledge, but I don't see why this would be cheating. Is there any rule that forbids a DAP maker of tweaking the sound to make it more pleasant to the general public? It seems to me that Sony just understands a little better the limitations of its hardware and the expectations of a good part of the consumers. And listening tests will always be subjective anyway, so who can guarantee Sony wouldn't win even if it didn't have the .25db boost? Disagree with Cnet's test is one thing, complain that Sony cheated to win that test or another focus group is another, silly one, in my opinion.

 

[sno1man]

I'm not a musician, just a music fan, and I'm not an audio engineer either, my specialty is computer software and like I had said before i do work in a company that uses modified audio in it's products.

This therefore might not directly apply to music, which is sonically more complex, but no persons hearing is flat frequency response either. We have a test where you listen to two tones and are asked if the second tone is louder, softer, or the same as the first tone.

This is a fairly long test and you listen to over 600 tone pairs. When you plot it out on a graph, my hearing looks like a ride at Cedar Point with pronounced dips and rises. I have dips in the 200 hz range and the 5500 hz rang and peaks in the 10000 hz range and 1500 range. I have exceptional high frequency hearing in that I can hear tones up to 24000 hz, but poor low with very little below 125 hz. That explains why I tend to like a more bassy sound.

It's all very interesting, but what matters most are your ears I guess.

Bangrman, keep doing this stuff, I for one find it very interesting and informative. have you had a chance to try the new RH10 minidisc yet? We use minidiscs units for field recording and I might want to get one of these

 

[Embio]

let me get this right - your saying that the great sound isn't due to good hardware, its due to software engineering?

 

[bangraman]

What I'm saying is hopefully not difficult to understand. Inferior sound quality + euphonic colouring can > better sound quality + no colouring in listening tests.

 

[vranswer]

But if it's > {greater than}, how can it be inferior?

 

[bangraman]

I'd suggest you work it out... and the future implications thereof. Seriously guys, I'm getting a bit tired of explaining fairly obvious things constantly apparently because I've besmirched your buying decisions or your manufacturer preference. I'm not talking about this from an iPod perspective. I'm talking about it from an iRiver, Cowon, Creative and Apple perspective.

 

[halcyon]

As a side note...

Sony has a history of "improving the natural" i.e. going for a sound that is not accurate anymore, but is more pleasant to listen to.

This in fact what they succeeded with ATRAC to some degree:

http://www.sony.net/Products/ATRAC3/tech/lab/

I would argue that their CD3000 / SA5000 headphones continue this tradition.

Both are not pure in the critical mid-range where the human hearing is at it's most sensitive. CD3000 additionally adds a lot of unnatural (euphonic) reverbation that some people like a lot.

Not impeccable acoustic engineering (i.e. purposefully distorting the sound), but trying to make it more like the ideal they themselves hold true (i.e. spicing it up).

Now, there's nothing wrong with that, if it sounds good to people.

Also, it's nice that there's some variation in the market place as how to devices sound.

However, as a baseline design criteria, I don't like that approach myself and I agree with Bangraman that a device should measure well first, then made to sound good.

Euphonic signal messing can be done afterwards in the player's EQ / SRS / other silly settings, if one desires. DSP power is relatively cheap these days.

But at stock settings, the device should measure and sound flat (imho), unless specifically marketed as NOT being flat.

Making things sound good even when they measure sub-optimally (for this level of engineering) will just mask some of the failures of the device.

It may be financially valid and it may be be to some people's liking, but for me it is reversing the priorities into wrong order.

First make it measure well, then make it sound good. Not: first make it sound euphonic to some and damn the measurements.

Just my opinon though, not the holy gospel. So don't be offended if you don't agree

 

[vranswer]

No offense taken (although I'm getting tired of having to explain everything to Bangraman).

 

[SDA]

Quote:

Originally Posted by halcyon Quote: Originally Posted by sda Who's to say that this thing is the cause of whatever audible difference there is? bangraman said it himself, it could well be something else. It's not as if everything else is identical with these players, not by a long shot. But that I didn't write that Go back and read carefully what I wrote.

In a way, you did, though you might not realize it.

To explain what I mean: the statement that psychoacoustics > engineering rules of thumb, in that context, logically appears to mean that any audible difference would outweigh the "rule of thumb" that a .5dB FR deviation couldn't mean anything. Implicit in that statement is that the .5dB FR deviation is what leads to the psychoacoustic difference. I don't know if you MEANT to imply that, but it's there.

If you still don't get what I mean, look at it like this:

- breez says that differences of this scale are not noticeable.
- You point out a case in which they are, then say that psychoacoustic evidence should outweigh the engineering rules of thumb.

The psychoacoustic evidence, in this case, must be the audible difference between the players. Saying that this piece of psychoacoustic evidence outweighs the "scale" rule of thumb assumes that the psychoacoustic evidence shows that such a small difference can, contrary to the engineering rule of thumb, be perceived. In actuality, all the evidence shows is that there IS a difference; it does not directly outweigh the rule of thumb, because there is no direct evidence that the audible difference has anything to do with what the rule of thumb deals with.

If you didn't mean to say that, sorry, but that's at least a reasonable way to interpret it. At the very least, do you see where I was coming from?