12,000 people studied and I imagine that is a wide cross section. Also 18
studies, I know that I hear a difference and the DSd/MQA/HQ file is not necessarily the better.
Some were trained. I do not know if they were selected and then trained or they were professionals. Whether they were a random sample of the general population is unclear. Nevertheless ,saying they a are a wide cross section is fair. Especially where the studies are independent.
as i am an extreme case of a serial format comparer, both analog and digital, and one who has pursued without limits the best possible gear to play it on, and all the media too, i have strong opinions about this subject.
the critical issue to me is whether these compares are dealing with native sources. which would require parallel recording processes for multiple format native recordings. unlikely.
much confusion about 'better' and 'preferences' has to do with what is really what? when formats are not 'pure' then results have some slight, but not complete, meaning. and also, does the playback process used favor one format over another?
for instance native redbook files can be awesome sounding. remarkable. yet native high rez is mostly better. but it's more recording process related to why the results are not consistent. you need parallel processes to eliminate the variables.
the result represented by the article is no surprise, as it mirrors my experience, but it's not science and nothing resembling proof. but it could be closer to that if we knew more. 12,000 samples is not trivial.
Oh come on as a scientist you can do better than the old bad recording argument. As a veteran of the digital vs analog wars tha t one has been done to death.
the critical issue to me is whether these compares are dealing with native sources. which would require parallel recording processes for multiple format native recordings. unlikely.
Exactly my take on the matter as well. Indeed, while my anecdotal conclusion is that the higher res file sounds better, in most cases the lack of an absolute leads me to assume the differences are in the pre-production stages: mastering, etc. In many instances I wonder if more care is taken when mastering the higher res sample(!).
much confusion about 'better' and 'preferences' has to do with what is really what? when formats are not 'pure' then results have some slight, but not complete, meaning. and also, does the playback process used favor one format over another?
This is it, in fact. On occasion when comparing format samples I have often asked myself what am I really comparing? The CD vs the 24/96 files, the mastering thereof, the CD transport vs the streamer...
At the end of the day, being disinclined to reach an absolute conclusion, I just listen to what sounds best to me - usually the 24/96 - 192 - dsd - or, (best of all) DXD...
Exactly my take on the matter as well. Indeed, while my anecdotal conclusion is that the higher res file sounds better, in most cases the lack of an absolute leads me to assume the differences are in the pre-production stages: mastering, etc. In many instances I wonder if more care is taken when mastering the higher res sample(!).
This is it, in fact. On occasion when comparing format samples I have often asked myself what am I really comparing? The CD vs the 24/96 files, the mastering thereof, the CD transport vs the streamer...
At the end of the day, being disinclined to reach an absolute conclusion, I just listen to what sounds best to me - usually the 24/96 - 192 - dsd - or, (best of all) DXD...
Agreed in part, but to my knowledge has been not tested. I think the better artists and recordings are often selected for his hi res treatment. But this does not make it an ad hoc excuse for hi res superiority. Indeed not all but some offer that explanation while arguing they hear no difference.
Over the years there were studies published in Japan about which few English-speakers are aware, that look at formats better than Red Book. I'm glad that someone dug out that in-English meta analysis of various research on this issue.
I'm old enough now to doubt I can hear the difference between Red Book and formats with more information. A 19 year old Japanese woman can hear much more than I can.
Of course anecdotal evidence positive or negative for one individual is not the conclusion. I can hear a difference. I need to convince someone of of that so they will keep making and buying it. Herbie Hancock is not going to create one hi res album just for me. Everybody should be concerned to put the brakes on the race to the bottom.
IMO the main problem is that reports of these pseudo tests are never transparent - people do not show data or methods, just their opinions transformed in "results" .
I think we can all agree that recording quality, or mastering quality (good mastering generally can't fix a bad recording) is of primary importance over the format. All things being equal, I'm convinced that higher resolution formats sound better. But this is obviously an issue where you have something that was mastered really well off tape at 16/44.1, but then 30 years later they pull out the now worn-out tapes to do a high res remaster, and the mastering engineer hits all the electronics really hot to sound "current", or cranks up the treble because he lost that part of his hearing. You are better off with the original CD quality version in this case.
With the issue of separate masters being on SACDs, I know it has happened, but my sense is that it was not particularly common because it makes zero sense from an economic point of view for a label to pay for two separate masters, when they can just down convert the DSD layer to redbook. All the ones I have owned sounded about the same in character, and with the PCM layer you could tell that it came from DSD.
I've also been doing comparisons in home and studio environments (where I did the recording myself) for almost 20 years, and it's clear there are real differences. It's also clear that higher end DACs in recent times have narrowed the gap somewhat, and a great CD recording always beats a mediocre high res release. I think DSD, if implemented right, is more of a game changer than 24/96, and if I had my way we would have a lot more music at DSD128. But all can be excellent. Of course, I'm just stating the obvious at this point.
what I hear convinced me the real high rez files are different to those high rez files come from most recording studios.
most digital recording studios never do right job for audiophiles.
an expert should tell us why their output is not good.
I think many recording studios does not use high end equipments for recording and most of them use digital processing/editing that is not good for sound.
I think the best digital media now is Native AAD .
A good point. The way these evaluations have evolved is to ask the participants ton distinguish the samples. To match a to x or to match b to x. The presumption being that if you can't do it to statistical significant degree there is no difference. I have always taken issue with that.
In this case i believe the participants are asked to select their preference or the one that sounds best. We can then reasonably infer there is a positive difference.
Of course there are many anecdotal reports of preference for Redbook
what I hear convinced me the real high rez files are different to those high rez files come from most recording studios.
most digital recording studios never do right job for audiophiles.
an expert should tell us why their output is not good.
I think many recording studios does not use high end equipments for recording and most of them use digital processing/editing that is not good for sound.
I think the best digital media now is Native AAD .
I can tell you the biggest problem is low expectations, and the sense that you only have to do a good job to the point that someone will notice on bluetooth speakers or an iPhone. There is very little incentive towards good recordings.
In my opinion, the main problems are too much processing, using digital plugins instead of analog for the processors, and allowing digital clipping. The other issue is that because digital recording setups are more portable, many recordings are being done in spaces with poor acoustics.
Many of these have good microphones, preamps, and even ADCs. But the rooms aren't always good and they don't think twice about mangling the sound with heavy, and artifact laden, DSP.
I can tell you the biggest problem is low expectations, and the sense that you only have to do a good job to the point that someone will notice on bluetooth speakers or an iPhone. There is very little incentive towards good recordings.
In my opinion, the main problems are too much processing, using digital plugins instead of analog for the processors, and allowing digital clipping. The other issue is that because digital recording setups are more portable, many recordings are being done in spaces with poor acoustics.
Many of these have good microphones, preamps, and even ADCs. But the rooms aren't always good and they don't think twice about mangling the sound with heavy, and artifact laden, DSP.
I remember I had a test and my friend used an ADC and a DAC in series . analog signal from tape machine goes to ADC and then goes to DAC without any extra processing.
1. Analog -> Playback
2. Analog -> Digital -> Analog -> Playback
the sound quality (No.2) was very good in comparison to before converting (No.1). it means digital is not bad and we can hear good sound from digital but it seems what destroy the sound is digital processing.
I remember I had a test and my friend used an ADC and a DAC in series . analog signal from tape machine goes to ADC and then goes to DAC without any extra processing.
1. Analog -> Playback
2. Analog -> Digital -> Analog -> Playback
the sound quality (No.2) was very good in comparison before converting (No.1). it means digital is not bad and we can hear good sound from digital but it seems what destroy the sound is digital processing.
I've heard this in the studio. If the digital is live, i.e. not recorded to a file, it sounds much much closer to the analog. It's when its placed in a PCM file that I heard a degradation. I've not heard DSD in the studio though.
I've heard this in the studio. If the digital is live, i.e. not recorded to a file, it sounds much much closer to the analog. It's when its placed in a PCM file that I heard a degradation. I've not heard DSD in the studio though.
in my test also the digital was live and not recorded to file. I do not know what happen to file recording or any other digital processing but what convinced me is Digital could be as good as analog.
That is very interesting. Of course digitals reason d'etre is is storage. I do not believe digital can be as good as analog . They can however make the difference negligible for sll but the extreme analog playback systems. To be fair analog has it faults.
The key is for proponents to acknowledge its' faults and continue ton work on them. IMO YMMV
The gorilla in the room for this discussion is the fact that you can’t run a digital stream into your head yet. Even if you could it wouldn‘t clarify anything.
We aren’t just listening to the data, we are listening to the process equipment - server, renderer, dac etc. Each manufacturer’s equipment goes about it’s business in a different way.
Some in industry have pointed out that high data rates, regardless of format, introduce more process noise than lower data rates. How does the playback equipment mitigate this extra noise? There are different ways to do this. Some may be a compromise. How precise is the clocking, what is the bandwidth of the cabling, interaction with RFI, EMI. Can the data receiver see the data, the leading edge, through this additive noise?
Where is the Goldilocks zone? Admittedly, redbook is intentionally minimalist because of the era it was developed. But if data transmission is highly efficient (jitter and other noise don’t result in too much lost information) then all should be good. Reality is a little different. There will be losses. Jacking up the sampling rates and increased resolution may provide the redundancy needed for a non-optimal data receiver to sound “better” to a point, but the extra noise will always be there too.
The lowest data rate that gets the job done will likely be the best, but that rate will be system and environment dependent.
