Pan said:
So why not give us a sample that actually could be detected on the (properly executed) blind test ? That should not be too difficult, given hundred "other" cases.
As far as Bryston, if anyone showed to me that output filter was eating into the working bandwidth, I'd change it too (regardless of if it could be heard or not). It doesn't cost anything.
And yes, I do think anyone can call Bryston and tell them this. Especially if you paid for it.
dfdye said:Or just build a better set of speakers, or even move the plant in your room! I am sure we all agree THOSE have an appreciable difference in terms of sound. probably even more so (?) than changing the amp you are using!
But then again, this isn't a "new" observation, so I guess it doesn't count!
Which begs the question, at that point, why not go for the big bang items like speakers or room treatments??
And your belief is that every person on earth that care about the performance from the amp are totally clueless about speaker technology and room acoustics?
/Peter
cuibono said:
Oh, but the differences are so subtle that a single switch can mask them! Or an op amp! Or inferior cables. Or most speakers. Or any stress...
Actually not. According to the findings of LTS and the people involved a bypass test can often be performed with success even in a relatively medioker test set up. Of course there are limitations but what you say is simply false.
Switches, opamps and cables ARE inserted in the test set up but still most amps and digital can be picked out in controlled blind tests.
/Peter
Bratislav said:
Why don't you set up a proper bypass test yourself and learn something? But sure, if I find out what specific amps that have proven to be non transparent in bypass tests I will report about it. However since most amps DO color the signals it should be easy to find one. I have never had an amp in my rig that did not sound different from the other. Building an amp right now that I believe will be transparent though.
It was not only the output filter that was changed and of course it's a cost for the company to set up listening tests, measure and change the parts in the amp. You don't run a company it seems from your understanding.
I don't think anything was paid until the new unit was shipped and passed the test. You still think that anyone can call one of the leading companies in the field and say.. "hey, your stuff isn't really that good but if you rebuild a unit for me and ship it to me I will consider buying it if it pass the test..!"
I don't think so but of course you are free to believe that.
/Peter
SY said:
Well, to be fair Pan's reports of the Swedish testing *seems* to be worth looking into. I haven't had time to do so yet, though.
I doubt that any of the 'radical subjectivist' folks will voluntarily head down this road since anything that runs the risk of undercutting their 'golden eared' status is to be avoided, but I think there are a couple things that can be done to begin to move forward.
1) AB/X yourself on lossy compression. This is what xiph does for a living, and IMHO any audiophile that approaches this with an open mind will be surprised. I strongly suspect that something like 320k Vorbis will be audibly transparent on a lot of material in any system you care to play it on. [I won't say everything all the time though - I don't doubt that with training some folks will detect it on some material]. the fact that we can throw 75% of the signal away and get so close to the original 'should' shed some light on the magnitude of some of the differences folks tend to take for granted.
2) Amp measurement techniques. I'm not an 'amp guy' so much of this may exist already, but if we want to start making any headway on any of the alleged 'not currently measurable' differences, we have to start somewhere. Seems like this would include
a) a set of agreed-upon dummy loads simulating 'real' loudspeakers
b) differential measurement jig that senses amp output with the load connected and maps is down to line-level for input to a soundcard or other appratus. Differential is critical as bridged amps dont' have a ground reference on either terminal
c) Synchronous averaging measurement software (stereo). Geddes talks about this in another thread, but I believe the idea is that you design the test signal to exactly fit in the FFT bin length so that you can directly average out the thermal noise and see very low-level inherent behavior.
Given a setup like this, it should be 'easy' to get started on what I think would be some interesting experiments. For example, we could 'record' the output of various amps and make the files available for AB/X with foobar or other software. If the differences are masked, we could loop-back the process and run the signal through the amp X times until we got a detectable difference (direct loop-back on the soundcard can be used as a control). None of this would be 'proof' in any auditable sense (the use of a soundcard alone will undoubtedly invalidate it in some folks minds), but I think it would at least help start to move the discussion forward. (or at least give us something concrete to bicker over)
Jan, that's pretty close to the PCM-F1 tests which showed "no detection."
I'm less inclined, given that the example he linked to showed... null results. If there's one which shows detection, level-matched, with no major frequency response differences, then that would be worth looking at. Oh, Pan...?
I've done that and was quite surprised at the results.
I'm less inclined, given that the example he linked to showed... null results. If there's one which shows detection, level-matched, with no major frequency response differences, then that would be worth looking at. Oh, Pan...?
I've done that and was quite surprised at the results.
SY said:
Yes, similar to the Tiefenbrunn disaster (I guess you're referring to that). In this test, the choice is between a high-resolution SACD source, or the same source but now going through a 16bit/44.1kHz CD-'bottleneck'. Not jus a matter of one resistor brand versus another, no, a whole process of sending an SACD signal through a DAC and then back through an ADC, all at 16bit/44.1kHz.
Sounds pretty clear cut, right? I mean, that degradation should just jump out at you!
Jan Didden
The results.
/quote
The test results for the detectability of the 16/44.1 loop
on SACD/DVD-A playback were the same as chance:
49.82%. There were 554 trials and 276 correct answers.
The sole exceptions were for the condition of no signal
and high system gain, when the difference in noise floors
of the two technologies, old and new, was readily audible.
As the tests progressed, we repeatedly sorted the data
for correlations with age, sex, upper frequency hearing
limit, or experience. No such correlations have emerged.
/unquote
Your turn.
Jan Didden
PS I'd love to post the whole test but this is copyrighted material that I downloaded and paid for.
You can find it at the AES website:
Audibility of a CD-Standard A/D/A Loop Inserted
into High-Resolution Audio Playback*
E. BRAD MEYER, AES Member AND DAVID R. MORAN, AES Member
(EBradMeyer@att.net) (drmoran@aol.com)
Boston Audio Society, Lincoln, MA 01773, US
/quote
The test results for the detectability of the 16/44.1 loop
on SACD/DVD-A playback were the same as chance:
49.82%. There were 554 trials and 276 correct answers.
The sole exceptions were for the condition of no signal
and high system gain, when the difference in noise floors
of the two technologies, old and new, was readily audible.
As the tests progressed, we repeatedly sorted the data
for correlations with age, sex, upper frequency hearing
limit, or experience. No such correlations have emerged.
/unquote
Your turn.
Jan Didden
PS I'd love to post the whole test but this is copyrighted material that I downloaded and paid for.
You can find it at the AES website:
Audibility of a CD-Standard A/D/A Loop Inserted
into High-Resolution Audio Playback*
E. BRAD MEYER, AES Member AND DAVID R. MORAN, AES Member
(EBradMeyer@att.net) (drmoran@aol.com)
Boston Audio Society, Lincoln, MA 01773, US
Sorry folks, i´m a bit in a hurry now, but will later on provide some links to articles and papers with several dbt studies. Unfortunately some are only available in german language.
What is surprising in general is the lack of exact reproduction in this field. I don´t know what the reason is but it leads to a quite inconsistent state.
The moran-ABX jannemann provided is an example for that. Theile and Steinke from the german IRT presented a couple of years ago a study which showed that it makes a difference if material is presented with 96kHz sampling rate instead of the normal 44.1kHz.
A master thesis paper (aquivalent to the german diploma i think) from the university in detmold searched for people beeing able to differentiate between DSD and PCM HD-Audio and found at least 4 participants that were able to.
It was done as ABX with support of Emil Berliner studios.
Michael Fremer reported some years ago results from a double blind test he conducted for a discussion with Lipshitz/Clark and others. John Atkinson was also participant.
The big stereophile double blind test with amplifiers showed a negative for differences that were otherwise ´proofed´to be detectable.
Paul Frindle wrote in the mentioned article about his surprising results in ABX; Bruno Putzeys did several times report about his ABXs with various sample rate converters and so on.
The famous F1 ABX SY mentioned is another example. The german magazin stereoplay did a quite big double blind in 1991 that showed that an additional A/D-D/A-Conversion (16Bit/44.1kHz) added a small but detectable difference.
Jakob
What is surprising in general is the lack of exact reproduction in this field. I don´t know what the reason is but it leads to a quite inconsistent state.
The moran-ABX jannemann provided is an example for that. Theile and Steinke from the german IRT presented a couple of years ago a study which showed that it makes a difference if material is presented with 96kHz sampling rate instead of the normal 44.1kHz.
A master thesis paper (aquivalent to the german diploma i think) from the university in detmold searched for people beeing able to differentiate between DSD and PCM HD-Audio and found at least 4 participants that were able to.
It was done as ABX with support of Emil Berliner studios.
Michael Fremer reported some years ago results from a double blind test he conducted for a discussion with Lipshitz/Clark and others. John Atkinson was also participant.
The big stereophile double blind test with amplifiers showed a negative for differences that were otherwise ´proofed´to be detectable.
Paul Frindle wrote in the mentioned article about his surprising results in ABX; Bruno Putzeys did several times report about his ABXs with various sample rate converters and so on.
The famous F1 ABX SY mentioned is another example. The german magazin stereoplay did a quite big double blind in 1991 that showed that an additional A/D-D/A-Conversion (16Bit/44.1kHz) added a small but detectable difference.
Jakob
Andre Visser said:
As is customary, they don't mention brands, which imho is a pity. But this is what they say:
/quote
Most of the tests were done using a pair of highly re-
garded, smooth-measuring full-range loudspeakers in a ru-
ral listening room with an ambient noise floor of about 19
dBA SPL, all electronics on. We also took the
test setup to several other locations: a Boston-area mas-
tering facility with very large four-way studio monitors; a
local university audio facility, again with large high-
powered monitors in a custom-designed listening space
(the subjects for this test were students in the recording
program); and a private high-end listening room equipped
with well-reviewed electrostatic loudspeakers and very ex-
pensive electronics and cables. In all venues we performed
informal tests of the subjects’ upper hearing limits to see
whether there was a correlation between this parameter
and the audibility of differences.
For the CD loop we used a well-regarded professional
CD recorder with real-time monitoring. Levels in both
channels were matched to within 0.1 dB using a very
high-performance adjustable analog gain stage, which was
always in the 16/44.1 signal path. Audio switching was
handled by an ABX CS-5 double-blind comparator.
/unquote
Jan Didden
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.