I’m posting here hoping that most readers will have horn type speakers and also have Audacity, JRiver, Foobar and/or other software with lossless pitch control utilities.
Sadly, many of my favorite vintage pop, jazz and soundtrack recordings from the 1960s were deliberately hit with excessive dynamic range compression applied almost routinely during mastering sessions to prevent mistrackings on cheap vinyl players, to extend storage space on vinyl LPs and CDs and/or to allow higher signal levels to compete in the loudness wars https://en.wikipedia.org/wiki/Loudness_war#History.
Consequently, somewhat depending on which speakers used, they can sound especially flat-literally, since the natural dynamic peaks along with all the levels of all different sounds in the recording were limited and/or squeezed to one level.
I’d like to know how over compressed but otherwise reasonably well engineered commercially released CD tracks sound on high resolution horn speakers when the pitch is reduced by between ~ 2 and 7%, depending on the tempo and/or other aspects of the recording. Granted, aesthetically, the idea of doing this to music seems silly or even perverse, but it’s your subjective impressions of hearing the change through your horn speakers that I’m looking for.
Therefore, please run these tests on 6 or more such over compressed CD recordings-ripped and saved as uncompressed WAV files-at your earliest convenience and describe the before/after sound over your horn speakers.
Please reply or pm me for further assistance. Thanks.
Sadly, many of my favorite vintage pop, jazz and soundtrack recordings from the 1960s were deliberately hit with excessive dynamic range compression applied almost routinely during mastering sessions to prevent mistrackings on cheap vinyl players, to extend storage space on vinyl LPs and CDs and/or to allow higher signal levels to compete in the loudness wars https://en.wikipedia.org/wiki/Loudness_war#History.
Consequently, somewhat depending on which speakers used, they can sound especially flat-literally, since the natural dynamic peaks along with all the levels of all different sounds in the recording were limited and/or squeezed to one level.
I’d like to know how over compressed but otherwise reasonably well engineered commercially released CD tracks sound on high resolution horn speakers when the pitch is reduced by between ~ 2 and 7%, depending on the tempo and/or other aspects of the recording. Granted, aesthetically, the idea of doing this to music seems silly or even perverse, but it’s your subjective impressions of hearing the change through your horn speakers that I’m looking for.
Therefore, please run these tests on 6 or more such over compressed CD recordings-ripped and saved as uncompressed WAV files-at your earliest convenience and describe the before/after sound over your horn speakers.
Please reply or pm me for further assistance. Thanks.
Last edited:
The pitch reduction is not truly lossless. You can lower pitch while not slowing down tempo, but its a lot of resampling calculations. IOW, the audio will not longer be bit perfect, to say the least.
Its also possible some of the what you are hearing has something to do with your dac performance (regardless of how it measures over at that other website).
If you would like me to listen to an example track on my reference system, please feel free to PM.
Its also possible some of the what you are hearing has something to do with your dac performance (regardless of how it measures over at that other website).
If you would like me to listen to an example track on my reference system, please feel free to PM.
Sorry but I don't get the connection here. Compression affects amplitude and you give a good synopsis for the reasons it is used. Pitch on the other hand affects frequency. I don't see how changing pitch will compensate for over use of compression. You end up with a compressed and out of tune recording ?
That will likely cause jitter issues but that's not the main problem. If you reduce the pitch, you're dropping the resolution. Having the original being pushed up in level doesn't mean you gain any resolution, in fact, the push up likely triggered a limiter, robbing it of the top dynamic peaks, pushing it down again just means you're limiting its lower end too! So no, that makes it worse instead of gaining anything, that's not how digital media works.
Yes, that's likely dithering, resample artifacts and compression losses. If the DAC can't process the dynamic or bit-range, you will lose additional information.
I don't see a particular reason for slowing it down either. However, if I can hear one or more of the recordings considered to be over-compressed, then I might have much better idea about the nature of the perceived problem.
On the issue of resampling, its probably not anything so simple to pin down as an effect of jitter. It just has to do with the degree of perfection (or lack thereof) used used to do the sample rate calculations. Usually, the freeware algorithms are not exactly the best possible. Jitter effects are not part of the data itself and would occur (or not) later during actual dac operation. Changing the data itself shouldn't change the jitter of the dac. That has more to do with the physical clocks in the dac and related circuitry such as clock power supplies.
On the issue of resampling, its probably not anything so simple to pin down as an effect of jitter. It just has to do with the degree of perfection (or lack thereof) used used to do the sample rate calculations. Usually, the freeware algorithms are not exactly the best possible. Jitter effects are not part of the data itself and would occur (or not) later during actual dac operation. Changing the data itself shouldn't change the jitter of the dac. That has more to do with the physical clocks in the dac and related circuitry such as clock power supplies.
Last edited:
The digital processing is very precise if the timing isn't an issue (streaming).
That's correct and also refutes the algorithms theory. Unless it's a stream without buffering, the algorithms will not change anything in the timing as they are not determining the output stream in any timing related way, they are neither responsible for the input buffer underflow nor the output buffer timing - unless you run out of processing time or any I/O lag (like USB interference, which often happens), that's up to the digital interface or DAC. That means, you have to search for the timing faults somewhere else and not blame the software.
Not exactly how I'm seeing it. There is a difference between precision and accuracy. For pitch lowering, interpolation and or decimation will be required. Resampling always requires filtering. Filter quality is highly variable depending on number of coefficients and bit-depth, along with other implementation details. The best calculations for audio filters are done with 128-bit or 256-bit math. Some digital filters have a million taps, which can be seen in HQ Player and or in reported implementation details of some Chord dac designs. Its not so simple as some people may assume to get the best quality conversion.
So, anyway, it then follows that the conversion can be done offline and the output saved into a file. Then it can be sent to a dac at some later time. There need be no interactive relationship at all between the conversion algorithm and dac operation. That's sort of the way I would probably sum it up.
So, anyway, it then follows that the conversion can be done offline and the output saved into a file. Then it can be sent to a dac at some later time. There need be no interactive relationship at all between the conversion algorithm and dac operation. That's sort of the way I would probably sum it up.
Last edited:
Thus, if I "pitched down" a WAV file by ~ 4% with https://jriver.com/ (not freeware) or Editing/Restoration Suites like https://www.steinberg.net/spectralayers/new-features/ or https://www.izotope.com/en/products/rx.html? and saved the changes there would be no resolution loss upon playback, yes?
And if so, now I'm hoping that when played over highly resolving speakers (e.g. ESLs, compression drivers with beryllium diaphragms, et al) they may somehow sound better-or at least no worse.
Check messages.
Last edited:
Also, how much difference can DAC design make towards achieving this goal?
https://www.oktoresearch.com/dac8stereo.htm
https://www.schiit.com/products/yggdrasil
https://www.denafrips.com/dac
https://www.kitsunehifi.com/product/holo-audio-may-dac/
As I'm less than ideally capable, please review and offer suggestions.
https://www.oktoresearch.com/dac8stereo.htm
https://www.schiit.com/products/yggdrasil
https://www.denafrips.com/dac
https://www.kitsunehifi.com/product/holo-audio-may-dac/
As I'm less than ideally capable, please review and offer suggestions.
"Best quality conversion"....but what is necessary to get a sufficiently-high quality of conversion? In the recordings being referred to, aren't they analog, with their own high levels of noise, wow and flutter, speed errors, etc?
Why would you believe that this pitch reduction process will serve to "improve" the perceived sound quality? It seems antithetical to do so; adding distortion of this nature to an existing recording would only serve to make it worse. The processed track may be different, but it won't be any better.
There are going to be flaws in original analog recordings. But they are likely analog sounding errors. Not sure if you are familiar with the types of errors that can accumulate in DSP? You can take something that sounds okay as analog and with some mediocre DSP quite easily make it sound worse than what it started out as. Why do you think dac chips have multiple selectable built-in output filters? They all sound different. Do you like them all equally well? Have you ever heard the best filters in something like HQ Player?
I begin to wonder about the FOMO factor when I read things like: "some modulators and filters seem to synergize better...XYZ is a combination that digs very deep and reveals everything in the sound. However, it's not always the most realistic sound and is not the best fit for human voice for example....but it creates a better sense of space....feels a bit slower than...feels sometimes kind of too fast...and sound can be kind of detached...kicks harder...it's sometimes a bit too much... which is also a bit more aggressive".
I am left wondering as to what was being described or alluded to above...and then I move my head 1cm and the loudspeaker system's frequency response at my ears changes dramatically, swamping any of the abovementioned potential effects.
I am left wondering as to what was being described or alluded to above...and then I move my head 1cm and the loudspeaker system's frequency response at my ears changes dramatically, swamping any of the abovementioned potential effects.
Last edited:
If that's happening then there is a interference at fairly high frequencies. That's an issue of your speakers but that's not an issue of compression of the signal etc
In this case I don't think it would help. There are two basic types of audio compression that affect SQ: there is dynamics compression, and there is lossy-compression. After listening to a few of the files in question, they sound extremely lossy-compressed (e.g 32kbps MP3, or maybe something even lower). That means almost of all the dynamics, details, extended FR, etc., are lost and gone forever. Unlikely that pitch shifting would help much if at all.
The only way you can play the exactly same samples (at a lower pitch) is by proportionally lowering the sample rate in real-time. Any other 'pitch-shifting algorithm' would give you totally different samples as opposed to the original.
Time stretching / pitch shifting is available for maybe 30 years and for maybe 15 years it's inaudible in action.
That said i don't see how it could change anything to modified transients. Once it's done it's definitive and nothing can restore them.
I know because i'm one of the guy who smash them when clients ask for ( they always ask for it...).
That said i don't see how it could change anything to modified transients. Once it's done it's definitive and nothing can restore them.
I know because i'm one of the guy who smash them when clients ask for ( they always ask for it...).