No kidding, HE AAC (Apple CVBR 48-9kbps) still makes classical music sound terribly compressed.
DSEE can't help that. At least AAC 128kbps or mp3 248kbps is the 'sweet spot' in terms of the amount of data the DSEE engine takes in and analyses. CD audio likely is analysed differently, who knows.
But it’s not an overwhelmingly great difference. It is very difficult to characterize the sound changes but it is rare to have any detrimental outcome on compressed sources. I think DSEE would make a very interesting series of measurements for analysis.
The DC phase linearizer is a function that controls the phase of the low frequency range to bring it closer to an analog amplifier.
In general, analog amplifiers have their phase advanced toward the low frequencies. On the other hand, the phase of the S-Master method is flat up to DC. And it is theoretically ideal to have a flat DC.
However, considering the actual music playback, it is not always good. This is because existing music and movie software is made in an analog amplifier environment. So is the speaker. In other words, much music was created on the assumption that the low-frequency phase of the amplifier is advanced even if it is not intentional.
If the phase of this low frequency range changes, the phase relationship between the fundamental and harmonic overtones of the bass will shift, and the waveform will change. Then, it can be said that the state adjusted by the mixing engineer is different from the volume and tone of the bass.
Analog amplifiers are usually AC amplifiers (alternating current amplifiers). In other words, direct current cannot be amplified. When the low range is cut, the phase advances 45 degrees at that -3dB point, and then advances to 90 degrees toward the low range. The cutoff frequency is usually around 4Hz for well-received amplifiers. But the phase rotation has already begun to advance at 10 times the cutoff frequency, or 40Hz. It is no wonder that 40Hz affects the feeling of low frequencies.
Now, in order to achieve exactly the same amplitude and phase characteristics as an analog amplifier, it is necessary to perform a high-pass filter operation called IIR. IIR is an iterative calculation, so if the calculation accuracy is poor, the sound quality will deteriorate.
The A-curve rotates the phase obediently and has three center frequencies of 2, 4, and 8 hertz.
The B-curve is to delay the phase by a few degrees below 50 Hz and then advance the phase at a lower frequency range in order to enhance the effect. In other words, it fakes to emphasize the effect, but it may be too effective, so you can think that the B curve is an effecter.
Leaving it OFF. This is a characteristic that is flat up to DC and is usually difficult to achieve with analog amplifiers. If you have a recording with a very long echo, such as a church recording, and you haven't messed with it too much with mixing, you may hear a very good sound.
https://www.sony.jp/feature/products/vinyl/?s_pid=jp_top_201810_feature_vinyl_feajrny_waljrny
The Sony vinyl processor allows you to listen to digital audio as if you were playing an analog record. Not an occult effecter! The scientific grounds are solid. It is a physical phenomenon.
One of the keywords is to increase the initial sensitivity of headphones/speakers. The other is that the vibration of the record and the turntable changes the sound.
By analogy, he doesn’t explicitly say this, but
clearly he’s saying that if we don’t use this DSP, our speaker/headphone won’t be ‘always speaking’ and therefore if there are moments of quiet, we’d not hear good sound when it starts speaking again because it stops vibrating? and low volume tones get missed as well because it’s hard to express them clearly? And
tone arm does the bass changes while
surface noise is for emphasizing high/mid freq (by increasing the noise in the bass area) and
record/player vibrations caused by the speaker (a random 38 cm subwoofer) cause vocals to become analog/classical becomes grouped together more (perhaps blurred or mixed?).
Mysteriously vinyl makes sound easier to hear?
A bold one-man show off which seeks to recreate vinyl in a way unlike a plain boring VST effect?
The Sony vinyl processor, DC Phase, DSEE HX AI, and 10 band EQ/spectrum analyser/VU meter together make A55 such a nice introduction to the Sony world, a world where measurements are disregarded yet the sound doesn't exactly go top-notch based on subjective evaluations, but at least S-Master is simpler than many DACs and with DC Phase, it can be like a digital amp or an analog amp, while headphone listening with vinyl-like sound is made possible, and the plethora of compressed audio can become the equivalent of hires even though switching off and on doesn't bring a big difference, and you can 'see sound' via spectrum analyser/VU meter particularly the 'HIGH band' which appears if DSEE AI is ON.
There is very little to be found in academic circles/journals or measurements when it comes to Sony walkmans. And asking Sony engineers doing jobs completely unrelated to walkmans yielded little. However it can be said that DSEE HX AI did get tested by mastering engineers at Sony Music Studios Tokyo a couple times and in the latest testing it has been virtually perfected, although it does not surpass the original hires. The rest of the DSP tends to be explained subjectively and I sought to clear up the mess a little by providing here some of the facts that engineers know when they evaluate the sound.