Hi George, I love Greek wines, but only the traditional ones, not the modern Shiraz/Merlot/Cabernet types. The Hatzidakis winery from Santorini is to die for, boy what a wine! unfortunately quite $$.
The Oikonomou from Sitia is also very good!
Shame it is so hard to get them, even in EU..
Thanks Pano, I posted this test in a French forum. Up for some laugh!
Last edited:
Brilliant! What the world doesn't need is yet more sources of the same generic crapola that is now pouring out of South America, South France, Italy, and Australia at the expense of their traditional viticulture.
Pano, the next ones you post will also be randomized? I'm still stinging that I couldn't make the time to give this a serious go.
I gave up on California many years ago. Every now and again, something comes by which is pretty good, but it's a very rare event.
I've had some excellent Southern France wines, but they are invariably NOT made with generic cepages, but rather with local varieties which had fallen out of fashion. My exemplar of what not to do is Mas de Daumas Gassac. Syrah belongs in the Northern Rhone, Chardonnay belongs in Burgundy, Cabernet belongs in Medoc, Merlot belongs in Pomerol. Not Languedoc.
I've had some excellent Southern France wines, but they are invariably NOT made with generic cepages, but rather with local varieties which had fallen out of fashion. My exemplar of what not to do is Mas de Daumas Gassac. Syrah belongs in the Northern Rhone, Chardonnay belongs in Burgundy, Cabernet belongs in Medoc, Merlot belongs in Pomerol. Not Languedoc.
Following up with some investigation into the version B files, flac vs. mp3, some interesting things emerge. The files are different lengths of playing time, so had to be synchronised and the first bit of the mp3 file chopped off. This is the result, top file is FLAC, bottom MP3:
Look pretty similar? Let's take a closer look, zoom into where the vertical line occurs in the FLAC at 19 secs, where it looks fairly busy:
Pretty impressive ... the two versions look to correlate very nicely, nigh well perfect match ... or is it?
So, next step is to invert the MP3 version, and mix in with the FLAC, see what sort of null we get. At first glance it appeared reasonable using simple eyeballing, but what was the difference in the measurements we know and love? Well, to make it easier to visually pick up what's going on I amplified the 'null' file by 20dB, and this is what we now have - top is original FLAC, bottom the difference, the 'null' file with gain of 20dB:
Oh dear! Some nasty spikes in the null file, and plenty of areas where the variation, or 'distortion' is less than 20dB down. Let's zoom into the 23 sec mark and see what's happening a bit more clearly:
This is definitely not nice stuff ... remember that the bottom waveform is the 'distortion', amplified by 20dB. Let's zoom in yet again, at the vertical line:
Unpleasant to digest ... a whole slab of waveform where the difference component is sitting nicely at a consistent 20dB down ...
I've found that by taking a great deal of care that it's possible to get mp3 versions which largely remain at 60dB or better difference, which should be close to inaudible. But, who wants to bet that -20dB differences will be likewise inaudible ...?
Look pretty similar? Let's take a closer look, zoom into where the vertical line occurs in the FLAC at 19 secs, where it looks fairly busy:
Pretty impressive ... the two versions look to correlate very nicely, nigh well perfect match ... or is it?
So, next step is to invert the MP3 version, and mix in with the FLAC, see what sort of null we get. At first glance it appeared reasonable using simple eyeballing, but what was the difference in the measurements we know and love? Well, to make it easier to visually pick up what's going on I amplified the 'null' file by 20dB, and this is what we now have - top is original FLAC, bottom the difference, the 'null' file with gain of 20dB:
Oh dear! Some nasty spikes in the null file, and plenty of areas where the variation, or 'distortion' is less than 20dB down. Let's zoom into the 23 sec mark and see what's happening a bit more clearly:
This is definitely not nice stuff ... remember that the bottom waveform is the 'distortion', amplified by 20dB. Let's zoom in yet again, at the vertical line:
Unpleasant to digest ... a whole slab of waveform where the difference component is sitting nicely at a consistent 20dB down ...
I've found that by taking a great deal of care that it's possible to get mp3 versions which largely remain at 60dB or better difference, which should be close to inaudible. But, who wants to bet that -20dB differences will be likewise inaudible ...?
Last edited:
Actually, I was more amused by the frantic post hoc dancing and handwaving. And delighted/impressed at how many people approached this as intended- a fun exercise that can be quite enlightening.
One doesn't "pass" or "fail", one gets an insight into what's actually important to audibility. And it's a nice demo to show that blind testing isn't all that difficult unless you deliberately want to make it so. Pano is a stud!
Is it just me, or did anyone else read that as SPINELESS charge carrier?
Now I remember playing with such parameters on some encoder, maybe LAME, I forget - ISTR there was some feedback process if you told it to do higher quality (this was a parameter separate from bitrate), it would encode a block then compare the decoded output to the original, doing this several times and tweaking things to get a 'better' approximation each time. It took much longer to encode, but the final output was better, thus trading off encoding speed for quality.
I hope for a fair play with beer and wine competition. Beer aficionados are as respectable members of this society as wine taste followers.
I also expect detailed description of the ales used.
George
to that
Maybe just the vapours of this will do the job ....
Armageddon: The strongest beer in the world | News.com.au
Armageddon: The strongest beer in the world | News.com.au
+1 to spineless
Frank, wrt the nulling I think you would have to mp3 encode the exact same flacs to get a reliable result. can you be 100% sure that when you cut down the mp3 that you have it starting at the exact same sample as the flac? any tiny difference in start time will affect the result.
I think pano said they were 320kb you could try converting with default options and see the result. would also be interesting to see the difference between different encoders, which are set to the same settings!
Tony.
Frank, wrt the nulling I think you would have to mp3 encode the exact same flacs to get a reliable result. can you be 100% sure that when you cut down the mp3 that you have it starting at the exact same sample as the flac? any tiny difference in start time will affect the result.
I think pano said they were 320kb you could try converting with default options and see the result. would also be interesting to see the difference between different encoders, which are set to the same settings!
Tony.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.