Uncompressed / lossless audio files (Full Version)

Foro Flamenco: http://www.foroflamenco.com/
- Discussions: http://www.foroflamenco.com/default.asp?catApp=0
- - Off Topic: http://www.foroflamenco.com/in_forum.asp?forumid=23
- - - Uncompressed / lossless audio files: http://www.foroflamenco.com/fb.asp?m=312226



Message


Ruphus -> Uncompressed / lossless audio files (Apr. 4 2018 11:28:20)

Weeks ago I was searching the net for something technical, when I came over something on Heise, a German tech website. It is from ... uhm ... 18 years ago, so not really news, although to me it was new.

If you are like me, trying to omit mp3 encoding; for instance having ripped your CD collection for a 'lossless' compression like flac format ...

and on the other hand having been surprised by some files who sounded so good in spite of being mp3 ...

c't made a test and invited despisers of mp3 and people with golden ears for auditioning. None of them could reliably tell files above 256 kBit/s apart from uncompressed originals in blind tests.

And in fact mp3s were at times perceived as better / tidied up sounding, which was concluded to be resulting from lesser artifacts that tend to naturally occur with untreated material through effects of some frequency ranges on other sections of the frequency band (usually effects from HF on MF and LF). Hence, less of HF content seemed to subjectively appeal.


Needless to say that the audiophile test persons were not really ardent about their test results, and that they after the fact tried all kind of explanation and argument , which albeit ended up rather funny sounding.

Anyway, I thought it to possibly be interesting to some of you too.

quote:

MP3 vs CD in the c't listening test

Hanover, March 9, 2000 - At the listening test MP3 against CD of the computer magazin c't, twelve music-trained test listeners as well as a sound engineer from a record company with 256 kBit / s estimated encoded MP3 files just as often as CD-quality as the CD itself, reports c ' t in the current issue 6/00.

c't had published in the issue 3/00 the results of a hearing test with various hi-fi professionals. No one could immediately identify any MP3 source above 128 kbps as such, or even discover serious errors in it. The fact that no "expert listener" could distinguish well-coded MP3 files from audio CDs surprised even the c't editors. Due to numerous skeptical reactions to this article, the c't editors decided to invite twelve critical readers to Hanover to hear the test.

Using 17 music recordings of various styles, the test listeners should distinguish audio CDs and MP3 recordings on a high-quality reference system. According to the results of the c't hearing test, there is no music genre that is particularly good or particularly bad for compression - enough care in coding and a high-quality encoder provided.

A data reduction of around 5: 1 is apparently possible with MP3 without audible loss. MP3 with 128 kBit / s, which corresponds to a reduction of 11: 1, the testers were able to distinguish however in the majority of cases of CD and MP3 with 256 kBit / s. On the other hand, some 128 kbit / s recordings were even better judged by the competent audience than the originals of CD, to the surprise of all those present.




Escribano -> RE: Uncompressed / lossless audio files (Apr. 4 2018 12:17:01)

Doesn't surprise me, though some swear vinyl is warmer than digital, valves are better then solid sate etc. Same arguments are beginning to apply to photography, although I know film has a greater exposure latitude at the moment.




Ruphus -> RE: Uncompressed / lossless audio files (Apr. 4 2018 13:38:02)

One can´t be wary enough of subjectivity. Especially in regard of the senses.
Anticipation seems to be overruling just too much of sensibility.

Smokers who swear by their brand, turning out incapable of proving their preference in blind test. Connoisseurs who in blindtest dumfoundedly find out that a 3-buck vine outperforms luxury specimens. Same with Italian housewives who unexpectedly find themselves preferring plain and sterile German mozzarella of cow milk over such from Italy and of precious buffalo milk. (That trend / myth led to unspeakable agony with thousands of buffalo calves left starving / alone on worn out pastures in Italy).

We´ve had everyone and their gang crippling their toes in fashionable boots (and after centuries of weird standard still today wearing trendy footwear that is narrower than the feet), lofty perfumes sold as top of the shizznizz that were merely smelling like vanilla, or folks riding their bikes with the saddle higher than the handlebar etc. pp.


Old hands like you, Simon, may recall the times when digital audio came out. Even though the first CDs were only AAD, the world was raving over that crushing defeat of old analog technology. And how years later only listeners of classical music started complaining about brittleness of that times converter tech, until one by one the mainstream followed. (With studios by then having thrown their analog rigs in the bin.)

And how within past 10 years audio paradigm is turning around one more time, now acknowledging that DDD can be pretty awesome / lacking nothing if handled well.

Same with the controversy around valves and solid state. Where the truth being simply that a good circuit is a good circuit, regardless of valve or transistors.

In most of the examples of trend mankind proving lack of independence and of empirical evaluation and being all too seducible.
- Like me who used to dismiss garlic or hot spicing and who now loves em.

-

Digital photography has by far outstripped original prognosis regarding its development and capability within predictable frame of time.

No one can nail down a used camera while inspecting a good shot.
To me, whose discerning abilities may not match the sharpest eyes out there, but which have been demanding enough for to appreciate FF and fine glass, there is no advantage worth keeping up with the hassles of celluloid. Including the mentioned lead with exposure latitude and possible detail of texture.
No at all missing the times in the dark room with developing film and positives, but being fascinated by instant digital options of capturing and post work.

Digital tech, in my opinion, has turned out allowing for outright fantastic tools. And huge investment spent on outboard audio gear might have not been my best decision, except of the microphones. (Good to have -though since years now boxed and stored anyway, as I´m not having appropriate rooms / insulation-, but meanwhile questionably worth it.)




Escribano -> RE: Uncompressed / lossless audio files (Apr. 4 2018 14:00:56)

In all but special circumstances, digital is fine by me.

Fujifilm X-E1



Images are resized automatically to a maximum width of 800px




Ruphus -> RE: Uncompressed / lossless audio files (Apr. 4 2018 14:21:57)

Nice pic!
-But what about the noise through the snowflakes? hehehe |O)

Seriously though, one of the things I appreciate with digital is the colors.
Most films were not even in the spectrum, even though magnum guys used to prove that there could be had some.

If there comes up something to tweak with digital me is rather reducing values, as sometimes them being so brilliant that it borders on artificial. IOW, plenty of it; and that is really great.




tele -> RE: Uncompressed / lossless audio files (Apr. 17 2018 13:49:06)

If someone can't hear the difference it doesn't mean it's not there

The most difference when comparing 128kbps vs lossless is in the lower frequencies and this can be seen also in graphs:



Images are resized automatically to a maximum width of 800px




Ruphus -> RE: Uncompressed / lossless audio files (Apr. 18 2018 11:51:06)

Around 1983 or so, I was presenting my new CD-Player to a friend. He then claimed that he could hear gradual interruption of the digital sound. In disbelieve me put an album on the turntable which I owned as CD too, gain staged for same volume and had the devices in sync.
Switching back and forth, the guy was indeed able of pointing them out blind. (And the vinyl was not crackling / both record and t-table needle in perfect condition. - I for certain was not able of hearing a difference.)

Still, in above quoted test ears were not able to prove what they claimed to be capable of.




Ricardo -> RE: Uncompressed / lossless audio files (Apr. 18 2018 11:52:20)

None of that matters when in the final stage they compress the hell out of everything to master a louder version. Disgusting




Ruphus -> RE: Uncompressed / lossless audio files (Apr. 18 2018 12:17:31)

Yes. That is why Bob Katz suggested a new mastering standard some roughly ten years or so ago. And I assume that many mastering fascilities will be following it these days.

With braodcasting sources however, it must be considered that stations employ their own boost to crank up things.

Some so stupidly even, that one may fall off the armchair when accidentally hitting their station.




Paul Magnussen -> RE: Uncompressed / lossless audio files (Apr. 18 2018 16:03:32)

quote:

there is no advantage worth keeping up with the hassles of celluloid


I’m just an amateur, but it seems to be hard to find a digital camera with a fast lens.

The lens that came standard with my old Olympus OM-4 was f/1.4, and they made an f/1.2 as well. No digital camera I’ve had comes close to this.

And flamenco photography is mostly low-light photography (unless you’re prepared to annoy everyone with flash).

Any solutions?




Ruphus -> RE: Uncompressed / lossless audio files (Apr. 18 2018 16:35:12)

I try to never use flash. It flattens everything, indeed.
Canon offers the EF 50mm f/1.2L USM and the EF 50mm f/1.4 USM.
(It would be weird if they would not function just as well as with analog capturing.)
Have you tried those?

And from what users say the cheapo EF 50mm f/1.8 II delivers surprisingly well, even though it does not appear like a quality glass with its plastic housing. Some even claim it to outperform some of the prime alternatives.

I was planning to get me one and see what gives, but with the gross price gauging over here I couldn´t overcome myself yet.




kitarist -> Actually camera and lenses stuff (RE: Uncompressed / lossless audio files) (Apr. 18 2018 20:15:57)

quote:

ORIGINAL: Paul Magnussen

And flamenco photography is mostly low-light photography (unless you’re prepared to annoy everyone with flash).

Any solutions?


I had faced the same issue when trying to shoot ballet.

There are companies making lenses that fit any of the big 3-4 camera brands, but are cheaper than the brand equivalent.

I had a Nikon camera body and a pretty limited budget, so I bought a couple high quality Sigma lenses (constant aperture F2.8 zoom lenses with optical stabilization). They were quite a bit cheaper than the Nikon versions, but basically the same quality. Sigma makes lenses fitting Nikons, Canons, and others. Another company like that is Tamron.

Here is a list of "the best for 2018" - see how many Sigmas are there.

https://www.pcmag.com/article2/0,2817,2421983,00.asp

Your camera still has to be pretty good, though. It is hard to shoot dance inside in low light at anything less than iso800 or even iso1000, even with the 2.8 aperture, so the camera has to be able to deliver a nice image at iso1000 or higher without visible artifacts.

The other 'tool' is to ignore the standard exposure based on 18% gray and go with what matches what your eyes see - especially if light is focused on the dancers but the rest of the frame is very dark. So use exposure compensation to "underexpose" (referenced to whole frame 18% gray) which will give you higher shutter speed and the right exposure for the dancer under spotlight conditions.




Escribano -> RE: Uncompressed / lossless audio files (Apr. 18 2018 20:57:15)

quote:

And from what users say the cheapo EF 50mm f/1.8 II delivers surprisingly well, even though it does not appear like a quality glass with its plastic housing


The "nifty fifty" as it is known, is one of the best lens for the price. Razor sharp.




Escribano -> RE: Uncompressed / lossless audio files (Apr. 18 2018 21:14:09)

quote:

And flamenco photography is mostly low-light photography (unless you’re prepared to annoy everyone with flash).

Any solutions?


Fast lens f2 or faster and the highest ISO you can get away with. Don't use auto, set to aperture priority and open the lens wide. Try ISO 3200-6400 on a decent digital camera, but the size of the sensor, shutter type/speed megapixel count are all factors. Lots of pixels on a small sensor means a smaller pixel density, so poorer performance in low light (look at the noise in a smart phone low light image). 16 megapixels on an APS-C sensor is fine, 24 megapixels on a full-frame, 35mm sensor is more than enough.

Low shutter speeds will blur moving subjects, though and without a mirrorless camera or leaf shutter lens, you may struggle below 1/20 sec. handheld.

Here are some of mine in low light.

Pentax ME Super and Pentax SMC 50mm @ f/1.4 1/30 sec. handheld on Neopan 400 film


Fujifilm XE-1 with 18-55mm lens (optically stabilised) @ f3.2 1/7 sec. ISO 3200 digital, handheld at midnight



Images are resized automatically to a maximum width of 800px




Richard Jernigan -> RE: Uncompressed / lossless audio files (Apr. 18 2018 21:23:10)

One reason smaller aperture lenses are prevalent on digital cameras is that nowadays the sensors are so much more sensitive than film.

Furthermore many current lenses feature stabilization mechanisms, called by various names by different manufacturers. These compensate for camera shake when hand held. They are advertised as allowing shutter speeds up to four times slower than non-stabilized lenses.

When i was shooting film, about as far as I was willing to push ASA 400 was one stop, to ASA 800. Still the added grain and washed out color was a signifcant penalty.

This photo of a dancer in Bali, and more than a hundred others on this and other evenings, was taken with the stage lighting, no flash. The stage lighting was not dim, but certainly was no brighter than usual for ballet performances in America and Europe.

The camera was a Nikon D800, ISO 6400 (8 times as fast as ASA 800), 1/400 second, the lens was Nikon's 70-200mm f/2.8 VR II. The "VR" stands for Nikon's version of image stabilization, but it was not not needed, since the shutter speed was fast enough to "freeze" camera shake for the focal length used. Although I was on the front row of chairs, the dancer was at least 15 feet away, maybe more. There were a lot of people seated on the ground in front of me.

Converting the 49 megabyte Nikon raw file to the 726 kilobyte jpeg uploaded here has very noticeably reduced the dynamic range and sharpness of the original.

RNJ



Images are resized automatically to a maximum width of 800px




Richard Jernigan -> RE: Uncompressed / lossless audio files (Apr. 18 2018 21:51:42)

Sony and Phillips jointly developed the CD, wrote the Red Book specification, and licensed the technology to other users. They also were first on the scene with CD players. At first other manufacturers followed Sony and Phillips design practices with their players.

My first CD player was a portable Sony. It was audibly different from a good vinyl rig. After a few years, two design flaws were discovered in the early players.

The output of the digital-to-analog converter (DAC) is a stairstep analog waveform. It is smoothed by a reconstruction filter. The early players used a "brick wall" filter: one whose response was relatively flat up to nearly 20KHz, then fell off quite steeply. The problem with this was that such filters, implemented in the analog domain, inevitably "ring" when presented with transients, introducing distortion.

In later designs this was overcome by oversampling. I won't bore you with the details, but essentially every CD player now on the market employs this approach.

The second design flaw was jitter in the digital clock that sequences the digital data into (and out of) the DAC. This was caused by trying to recover the clock from the data read from the disc. The non-return-to-zero recording mode exacerbated the problem. It turns out that the distortion caused by clock jitter is particularly audible, especially on louder passages.

My main CD player, more than 20 years old, generates a stable clock electronically, and uses a feedback loop to control the rotational speed of the disc mechanism. Clock jitter problem solved, but at the expense of added circuitry--another box the size of the CD player itself, 20 years ago. I have no idea how current inexpensive players generate their clock. I would think the improved quality would have motivated development of cheap integrated circuits to handle the problem.

To me, my CD player sounds better than my vinyl rig: a Thorens turntable, Grace arm and Ortofon high end cartridge. Still, there are people who prefer vinyl and vacuum tubes to CDs and semiconductors.

No recording sounds like the real thing.

RNJ




mrstwinkle -> RE: Uncompressed / lossless audio files (Apr. 18 2018 23:11:24)

Make yourself do a series of blind tests. I recorded a bunch of stuff in various formats many years ago (mp3, ogg, flac, wma ) and to my shame, testing blind, preferred wma reasonably consistently over the original.




Paul Magnussen -> RE: Uncompressed / lossless audio files (Apr. 19 2018 19:29:27)

quote:

Seriously though, one of the things I appreciate with digital is the colors


The photos produced by most digital cameras have always seemed to me to have a slight bluish tinge.

In 2006 my wife and I went on a walking tour of Switzerland with a group of other people, and we agreed at the end to share all our photos — they amounted to several thousand but they did offer a good opportunity for comparison. The one camera that didn’t seem to me to have the blue tinge was by Fuji: a FinePix E550.

Here are four photos of the same subject. Can you pick the Fuji?









Images are resized automatically to a maximum width of 800px




kitarist -> RE: Uncompressed / lossless audio files (Apr. 19 2018 20:04:29)

If I had a gun to my head I would say the first one is the Fuji as it seems to go more towards a red/yellow than the rest (I am just 'eye-ing it'; have not looked at the spectrum)

But doesn't this depend also on what white balance the digi-cams had set for each pic? A proper test should have the exact same WB setting to be sure, no?




Escribano -> RE: Uncompressed / lossless audio files (Apr. 19 2018 20:16:15)

To be honest, they are all a bit overexposed and that was a long time ago in digital camera terms.

Fuji do great saturated colours, but it is a matter of personal taste.

You could spend £2000 and get the Sony RX1R with a fabulous f/2 Zeiss lens. I love mine. No tinge here.



Images are resized automatically to a maximum width of 800px




Escribano -> RE: Uncompressed / lossless audio files (Apr. 19 2018 20:18:07)

quote:

A proper test should have the exact same WB setting to be sure, no?


Lower-end consumer cameras do not have manual white balance calibration. Always take RAW and adjust in post-processing.

Apart from the Fuji X100, which takes exceptional JPEGS.



Images are resized automatically to a maximum width of 800px




Paul Magnussen -> RE: Uncompressed / lossless audio files (Apr. 19 2018 20:35:43)

quote:

If I had a gun to my head I would say the first one is the Fuji


Correct: well spotted. The cameras are:

1) Fuji FinePix E550
2) Canon PowerShot SD550
3) Sony DSC-W7
4) Canon PowerShot A700

Of the four, which colour balance would you choose?




kitarist -> RE: Uncompressed / lossless audio files (Apr. 19 2018 23:52:39)

Maybe the third one from top, but exposed properly.




Richard Jernigan -> RE: Uncompressed / lossless audio files (Apr. 20 2018 2:54:23)

quote:

ORIGINAL: Escribano

Lower-end consumer cameras do not have manual white balance calibration. Always take RAW and adjust in post-processing.



A dozen years ago, on our first trip together to California and Spain, Larisa was shooting a northern California landscape with a little Sony point-and-shoot she had bought used on ebay. She had the flash turned on. Shooting the same scene with my Nikon D300, I commented that she might be wasting battery power.

Without a word she shot the scene again, this time with no flash. Then she handed me the camera and asked whether I could tell the difference between the two shots.

"In the one with the flash, the colors are just a little warmer." She nodded. I added, "Apparently Sony adjusts the white balance when the flash is turned on."

"Apparently. Which one do you like better?" she asked.

"The one with the flash."

RNJ




Ruphus -> RE: Uncompressed / lossless audio files (Apr. 20 2018 10:31:58)

With the pretty advanced bits that you have been discussing, needless to mention (other than for the interested lurker) that digital sensors can´t see colors / that colors are being determined / translated by values of grey for which again each brand has their own individual recipe.

Personally, I specially like what me got to see from Canon and Panasonic (while Nikon tends to come out slightly greenish which may have to do with measure of blue). Greater than life maybe, but beautiful.

Regarding hinge on color spectrum, I suppose that we can all agree on that there has never been a greater shade than with color celluloid. (If I recall that right, Agfa colors were specially flat and off.)

Besides from the fact that the digital option of color management can do pure magic within a blink of an eye. Like when you have shot in colored light like under disco strobes, and on camera or in post with a touch of a button the originals colors can be retrieved so naturally as if the motifs were captured in daylight.

Color perception seems like the most divergent factor of visual subjectivity.
I know that my ability of distinguishing very fine shades is relatively minor. Thus, in dental labs I often used to confirm with colleagues about precise match of dentures, to make sure.

On the other hand my individual sense for complementing and countering colors appears to be quite there, as I found out through checking out about my choices with corresponding applications that can be found on the internet.

On a side note: At the risk of repeating myself; current generations of the young are increasingly loosing distinguishing abilities with sound and vision.
A kind of paradox, as those skills could be trained most easily by means of digital trainers.

The dislike against digital options, that some display, springs off from symptomatical thinking. The truth being that the world of 1s and 0s has opened mind-blowing opportunities, and we are not at its limits yet.

Like always with tools, it only depends on how they will be engaged.

Oh, and one more point:
Keep in mind that we are not seeing the same thing, also within objective terms. Due to lacking calibration.

The vast of users do not see parameters as (hopefully) originally generated on a calibrated work station. The pics they are watching tend to show graphics in deviating ways.
And in lesser educated regions you might experience such even with print companies. (Imagine me ordering large positives and getting back such skewed images that for instance a bright red ["Lichtrot" in German] would return as ruby.)

Not everyone might be willing to spend those ~ 250 bucks for calibration gear, but downloading test cards for free and correcting parameters of graphic cards and monitors will go a long way.

More so when you just bought a TV from a show room. Manufacturers tend to exaggerate values on default in order to trick in uncritical consumers (who will be taken by overdone measures), and sales stuff likes to crank the values in show rooms one more time.

Probably the public tuned the most to hinge, residing in GB and USA whose TV format used to be extremely skewed. (The British one showing very greenish, the US one very reddish.)

Over here I show the balanced and brilliant screen of my old TV device to acquaintances, trying to give them an idea of how their TV-set at home could be presenting like.

Only with an accurately tweaked device can the fantastic standard actually be seen that is often times aired today. Beautifully rich, in many cases looking like moving Rembrandts and Dalis.




Escribano -> RE: Uncompressed / lossless audio files (Apr. 20 2018 10:40:43)

Very true, my Mac retina displays are calibrated as best I can.




Ruphus -> RE: Uncompressed / lossless audio files (Apr. 20 2018 11:05:08)

PS:
Another anecdotal circumstance, which you might recall:
Those who grew up with B&W TV tend to dream in B&W. Only yet, generations who grew up with color TV dream in Kodachrome.

Ain´t it funny? :O)




Paul Magnussen -> RE: Uncompressed / lossless audio files (Apr. 20 2018 16:14:31)

quote:

Those who grew up with B&W TV tend to dream in B&W. Only yet, generations who grew up with color TV dream in Kodachrome.


What did people dream in before there was TV?




Ruphus -> RE: Uncompressed / lossless audio files (Apr. 21 2018 10:26:19)

Maybe in the billions of shades available in nature?
Dreaming must have taken way more GC CPU in those old days.




JasonM -> RE: Uncompressed / lossless audio files (Apr. 21 2018 23:03:31)

I used to be an Audiophool. I can hardly detect any difference between lossless and 356k MP3. The recording itself is by far the most important factor, like Ricardo touched on.

There is a blind test challenge for amplifiers with a 10k dollar prize to the first person who can correctly identify one amp from the other. The Amps must have at least a THD < 2% and be level matched.
http://tom-morrow-land.com/tests/ampchall/index.htm




Page: [1] 2    >   >>

Valid CSS!




Forum Software powered by ASP Playground Advanced Edition 2.0.5
Copyright © 2000 - 2003 ASPPlayground.NET