[Louis8ball]

Cool, thanks!

 

[julian67]

New(ish) measurements of Clip+, iPod Touch 4G and Altmann Tera-Player : http://www.tera-player.com/Tera-Player_RMAA.html

The Clip+ is measured alongside an iPod Touch 4G and the Tera-Player. The analysis is at least as interesting as the measurements.

The time domain measurements are imo the most interesting as are the author's related comments on the nature of oversampling and non-oversampling DACs and the effect on frequency response curves. This offers some insight into why the very vocal "one graph rules them all (FR) and it must be flat" congregation (it is a congregation) might not be the swamis of sound and science after all.

The iPod comes off very badly indeed and I have seen iPod fans who failed to read the whole article get very upset and claim the test is broken. Actually the author acknowledges that his iPod is abnormal and probably faulty and that the measurements confirm his disappointing listening experience. So any fans of any of the hardware in the test please read right through so as to avoid reaching unsupported conclusions or feeling upset about stuff that ought not to be upsetting. It's a really interesting read even if you don't own all or any of the items being measured.

 

[nicholars]

I would try not to offend apple fans because they get very upset when people inform them that apple gear is indeed overpriced and not actually very good. @_@

 

[julian67]

I would try not to offend apple fans because they get very upset when people inform them that apple gear is indeed overpriced and not actually very good. @_@

That has nothing to do with the linked article. The iPod Touch in the test is almost certainly faulty and this is acknowledged in the article by the author.

I have had a ton of grief from people because I put my head above the parapet here and elsewhere and said my Clip+ has some problems with cheap construction, can be noisy, and wasn't designed by Jesus or manufactured from touchstone by angels. Similarly I failed to find unassailable perfection and superiority over the HD 800 in a Koss headphone which I rated very highly and love to use. I have subsequently spent days being bored senseless, practically stalked, by Koss fanbois. One even tried to get an "offending" post taken down by the admins (the post remains). Apple has no monopoly on idiot fanbois.

Like I said in my post above:

.....please read right through so as to avoid reaching unsupported conclusions.

 

[uchihaitachi]

Quote:

New(ish) measurements of Clip+, iPod Touch 4G and Altmann Tera-Player : http://www.tera-player.com/Tera-Player_RMAA.html

The Clip+ is measured alongside an iPod Touch 4G and the Tera-Player. The analysis is at least as interesting as the measurements.

The time domain measurements are imo the most interesting as are the author's related comments on the nature of oversampling and non-oversampling DACs and the effect on frequency response curves. This offers some insight into why the very vocal "one graph rules them all (FR) and it must be flat" congregation (it is a congregation) might not be the swamis of sound and science after all.

The iPod comes off very badly indeed and I have seen iPod fans who failed to read the whole article get very upset and claim the test is broken. Actually the author acknowledges that his iPod is abnormal and probably faulty and that the measurements confirm his disappointing listening experience. So any fans of any of the hardware in the test please read right through so as to avoid reaching unsupported conclusions or feeling upset about stuff that ought not to be upsetting. It's a really interesting read even if you don't own all or any of the items being measured.

Lol. As ever the clip+ outperforms the mighty Tera player. The 'be all and end all' of all players. However, I think new ipod touch measures extremely well.

 

[Hutnicks]

Quote:

New(ish) measurements of Clip+, iPod Touch 4G and Altmann Tera-Player : http://www.tera-player.com/Tera-Player_RMAA.html

The Clip+ is measured alongside an iPod Touch 4G and the Tera-Player. The analysis is at least as interesting as the measurements.

The time domain measurements are imo the most interesting as are the author's related comments on the nature of oversampling and non-oversampling DACs and the effect on frequency response curves. This offers some insight into why the very vocal "one graph rules them all (FR) and it must be flat" congregation (it is a congregation) might not be the swamis of sound and science after all.

The iPod comes off very badly indeed and I have seen iPod fans who failed to read the whole article get very upset and claim the test is broken. Actually the author acknowledges that his iPod is abnormal and probably faulty and that the measurements confirm his disappointing listening experience. So any fans of any of the hardware in the test please read right through so as to avoid reaching unsupported conclusions or feeling upset about stuff that ought not to be upsetting. It's a really interesting read even if you don't own all or any of the items being measured.

Thanks loads for that.

It really is a great workup (albiet by the inventor himself) Designed using Porta Pros, I hadn't known that.
 
I wish he had the Ak 100 or 120 on hand, I would love to see time domain measurements off of those two.
 
Had he not mentioned it I would have thought those were taken from an iPhone not iPod, no wonder they sound so much better with a simple cap mod. Their filtering must be done with caps rejected by the Chinese military.

 

[uchihaitachi]

Quote:

Thanks loads for that.

It really is a great workup (albiet by the inventor himself) Designed using Porta Pros, I hadn't known that.
 
I wish he had the Ak 100 or 120 on hand, I would love to see time domain measurements off of those two.
 
Had he not mentioned it I would have thought those were taken from an iPhone not iPod, no wonder they sound so much better with a simple cap mod. Their filtering must be done with caps rejected by the Chinese military.

The time domain measurements are there to purposefully mislead. The prettiness & the symmetry of the time domain graphs rarely impact the SQ to any audible degree.
 
For the listener, the Tera will simply sound more dull than the clip+ as there is treble roll off.
At least the Tera is better than the HM-801. Now that is garbage.

 

[julian67]

Lol. As ever the clip+ outperforms the mighty Tera player. The 'be all and end all' of all players. However, I think new ipod touch measures extremely well.

Its not the case at all that the Clip+ outperforms the Tera-Player. That conclusion only makes sense if you narrow your focus to the Frequency Response test, follow that up by not understanding why the FR differs between the different DAC techniques, then ignore the time domain measurements, and then ignore everything that differentiates an oversampling dac from a non-oversampling dac. Apart from that you hit the nail right on the head

The time domain measurements are really important but are not in RMAA so are overlooked by RMAA testers/fans. The ripples you see in those square wave tests show that you can expect ringing artefacts, actually pre-ringing because the artefacts precede the sound. Artefacts like ringing and pre-ringing tend to be audible as well as measurable because they affect transients and we (humans) are sensitive to this.

 

[uchihaitachi]

Quote:

Its not the case at all that the Clip+ outperforms the Tera-Player. That conclusion only makes sense if you narrow your focus to the Frequency Response test, follow that up by not understanding why the FR differs between the different DAC techniques, then ignore the time domain measurements, and then ignore everything that differentiates an oversampling dac from a non-oversampling dac. Apart from that you hit the nail right on the head

The time domain measurements are really important but are not in RMAA so are overlooked by RMAA testers/fans. The ripples you see in those square wave tests show that you can expect ringing artefacts, actually pre-ringing because the artefacts precede the sound. Artefacts like ringing and pre-ringing tend to be audible as well as measurable because they affect transients and we (humans) are sensitive to this.

Sure, the Tera will sound dull and without as much high end detail with that kind of high end rolloff.
 
The Tera just seems to have a second-order filter, which may make nice looking square waves, but has to break way down into the audio band to be rolled off enough by the sampling frequency.  The FR plots don't quite go high enough, but it looks like 12dB/octave breaking at 12KHz or so.  
 
There is a severe lack of good correlation between a square-wave image and actual sound.  There are actually some posted samples in the science of sound forum, just can't find them now.  People look at square waves and how "messed up" they look, but they don't realize what that actually means for audibility, FR, distortion, etc. I would recommend the use of square waves for system diagnostics only, not sound quality evaluation.  You can mess up the look of a square wave without changing how it sounds at all.
 
I am amazed nobody has yet to do a (certain type of) test with a Tera player and a Clip+. Didn't want to use the word, I have been warned too many times of late.
 
 

 

[julian67]

The time domain measurements are there to purposefully mislead. The prettiness & the symmetry of the time domain graphs rarely impact the SQ to any audible degree.

For the listener, the Tera will simply sound more dull than the clip+ as there is treble roll off.
At least the Tera is better than the HM-801. Now that is garbage.

No those results don't suggest it will sound more dull. Did you even read through the article? Or look at where it rolls off? Or how? Or why? Or what that implies?

And time domain matters! Here is an interesting listening test http://www.hydrogenaudio.org/forums/index.php?showtopic=100896&hl= in which it is shown that time domain artefacts matter a lot. One of Hydrogen Audio's imam's even states

It seems that if you're particularly sensitive to pre-echo then just about anything with hard transients won't be transparent with mp3, even at 320kbits.

see http://www.hydrogenaudio.org/forums/index.php?showtopic=100896&view=findpost&p=835214

May I suggest looking again at the article, following the links, and taking more than a minute or two to consider it? If I wanted to provoke a discussion that rests entirely on FR curves I could visit HA or just inject battery acid into my eyeballs. My hope was to prompt people to look beyond the limitations of RMAA and one-graph monomaniacs and maybe even invite responses that don't sound like "four legs good two legs better".

 

[nicholars]

Julian67 do you think you could sound a bit more agressive and patronising there?

 

[Hutnicks]

Quote:

Sure, the Tera will sound dull and without as much high end detail with that kind of high end rolloff.
 
The Tera just seems to have a second-order filter, which may make nice looking square waves, but has to break way down into the audio band to be rolled off enough by the sampling frequency.  The FR plots don't quite go high enough, but it looks like 12dB/octave breaking at 12KHz or so.  
 
There is a severe lack of good correlation between a square-wave image and actual sound.  There are actually some posted samples in the science of sound forum, just can't find them now.  People look at square waves and how "messed up" they look, but they don't realize what that actually means for audibility, FR, distortion, etc. I would recommend the use of square waves for system diagnostics only, not sound quality evaluation.  You can mess up the look of a square wave without changing how it sounds at all.
 
I am amazed nobody has yet to do a (certain type of) test with a Tera player and a Clip+. Didn't want to use the word, I have been warned too many times of late.
 
 

 
You beat me to the whole square wave issue.
 
That said you can use a square wave to diagnose interference you could not otherwise observe. Ringing being a good example.
 
http://www.ap.com/kb/show/187

 

[julian67]

Julian67 do you think you could sound a bit more agressive and patronising?

I am at 90%. For people like you that is enough. (joke)

 

[uchihaitachi]

Quote:

No those results don't suggest it will sound more dull. Did you even read through the article? Or look at where it rolls off? Or how? Or why? Or what that implies?

And time domain matters! Here is an interesting listening test http://www.hydrogenaudio.org/forums/index.php?showtopic=100896&hl= in which it is shown that time domain artefacts matter a lot. One of Hydrogen Audio's imam's even states
see http://www.hydrogenaudio.org/forums/index.php?showtopic=100896&view=findpost&p=835214

May I suggest looking again at the article, following the links, and taking more than a minute or two to consider it? If I wanted to provoke a discussion that rests entirely on FR curves I could visit HA or just inject battery acid into my eyeballs. My hope was to prompt people to look beyond the limitations of RMAA and one-graph monomaniacs and maybe even invite responses that don't sound like "four legs good two legs better".

The regression results in your links are contrary to what you are saying.
 
Read this part: The Tera just seems to have a second-order filter, which may make nice looking square waves, but has to break way down into the audio band to be rolled off enough by the sampling frequency.  The FR plots don't quite go high enough, but it looks like 12dB/octave breaking at 12KHz or so.  
 
Hence no audible difference. 
 
I don't trust RMAA either nor am I a single graph monomaniac. dScopeIII is what I use when it is available.
Also, you seem to enjoy conjuring up rather disturbing scenarios. 

 

[julian67]

Also, you seem to enjoy conjuring up rather disturbing scenarios. 

-insert personal remark here-

Abracadabra.