That's a concern for archiving and data integrity long-term, but not germane to the discussion at hand. Likewise, it would become manifest in a bit compare, provided a bum CD causes read errors. Two identical bit-wise copies played through the same hardware had darn well sound the same.
Honest question: would you guys expect a track to sound different being streamed from a hdd than a ssd? Or even different brands of hdd or ssds?
Honest question: would you guys expect a track to sound different being streamed from a hdd than a ssd? Or even different brands of hdd or ssds?
formatcd4,
I do not find anything in that to argue with.
Digital either works or it doesn't, no question. Which means, and I assume this is parallel to your point, is that all kinds of things can happen to the media/etc and it will still work until it doesn't.
I have found the effect to be the same whether I am using an old CD or a brand new one. Most of my CDs are in excellent condition and have not been played that often. Referring back to the comment I made about the early MERIDIAN players that counted the errors ON THE MEDIA and it was noted that each time they played the CD it would the number of errors would increase. CDs do degrade and that does not cause problems until it does!
I do not think any of this has anything to do with what I am hearing.
It is not that resolution is better or the tonal balance is better, it is that this grey blanket of noise is not audible. I am sure there is some of it remaining but I am not detecting it, yet. Now just as looking through a clean glass makes it easier to see what is on the other side, there is that effect. I do not think the resolution is enhanced but it is more clearly revealed.
The CDROM is mounted on an old turntable stand - metal tubing filled with lead shot and topped with a slab of that material that is used for lab countertops. There are pieces of balsa on the top and bottom (3/8" thick) and on tope of the whole thing a power transformer from a Mark Levenson ML2 amplifier that I got from their supplier decades ago for a good price since they had dents in them. Only using one, of course. Very heavy. Who knows what effect this has. The drive is quieter with this mass added and best of all it does not move around when the drawer opens and closes. Seems like an obvious thing to do ...
Concerning whether the drive itself can affect the sound quality - there are many who report hearing a difference between spinning drives and SSDs. I guess I followed the herd when I used a computer for music playback and used an SSD for music storage.
When I started on this project I read that Tony Lauck had pondered if an SSD might not be the best for storing ripped files because of all the manipulation of the files within the drive which would be occurring as the rip was transferred. This made sense to me so I use a WD HDD laptop drive for storage. Powered by a hybrid linear supply using a pair A123 26650s as the output capacitor.
I would like to place the music files directly onto an SD card but XP cannot utilize AHCI and the SATA SD card readers I have (I have tried five of them) will not work most of the time with IDE and when something is as random as that I figure I best not do it.
I do not find anything in that to argue with.
Digital either works or it doesn't, no question. Which means, and I assume this is parallel to your point, is that all kinds of things can happen to the media/etc and it will still work until it doesn't.
I have found the effect to be the same whether I am using an old CD or a brand new one. Most of my CDs are in excellent condition and have not been played that often. Referring back to the comment I made about the early MERIDIAN players that counted the errors ON THE MEDIA and it was noted that each time they played the CD it would the number of errors would increase. CDs do degrade and that does not cause problems until it does!
I do not think any of this has anything to do with what I am hearing.
It is not that resolution is better or the tonal balance is better, it is that this grey blanket of noise is not audible. I am sure there is some of it remaining but I am not detecting it, yet. Now just as looking through a clean glass makes it easier to see what is on the other side, there is that effect. I do not think the resolution is enhanced but it is more clearly revealed.
The CDROM is mounted on an old turntable stand - metal tubing filled with lead shot and topped with a slab of that material that is used for lab countertops. There are pieces of balsa on the top and bottom (3/8" thick) and on tope of the whole thing a power transformer from a Mark Levenson ML2 amplifier that I got from their supplier decades ago for a good price since they had dents in them. Only using one, of course. Very heavy. Who knows what effect this has. The drive is quieter with this mass added and best of all it does not move around when the drawer opens and closes. Seems like an obvious thing to do ...
Concerning whether the drive itself can affect the sound quality - there are many who report hearing a difference between spinning drives and SSDs. I guess I followed the herd when I used a computer for music playback and used an SSD for music storage.
When I started on this project I read that Tony Lauck had pondered if an SSD might not be the best for storing ripped files because of all the manipulation of the files within the drive which would be occurring as the rip was transferred. This made sense to me so I use a WD HDD laptop drive for storage. Powered by a hybrid linear supply using a pair A123 26650s as the output capacitor.
I would like to place the music files directly onto an SD card but XP cannot utilize AHCI and the SATA SD card readers I have (I have tried five of them) will not work most of the time with IDE and when something is as random as that I figure I best not do it.
Exactly. Which is why noise, when ripping or copying data, is no significance unless it actually causes (detectable) errors.
Again, let's not confuse playing of music (involving the DAC conversion) with simply copying/reading data. Derfnofred asked about streaming - an operation that only involves copying of data, but no digital-to-analog conversion.
Unfortunately that simple command does not work, because the file metadata (such as the track name) will be different and cause the comparison to fail. You need a comparison that only looks at the waveform data.
Yes, sorry, as Julf expanded, I meant to write "using the same electronics", I.e. solely comparing the physical storage device. I could have even gone further to state the specific memory chips, since the bit stream would pass through memory before input/output.
The point I'm trying to make is, provided the data is making it to the DAC correctly (and the microelectronics industry has gotten very good at that with such pedestrian demands as audio on a dedicated system), the DAC is what matters. As long as a zero is a recorded as a zero, you can be confident it'll remain a zero until the DAC turns it into a voltage/current signal, where the linearity is very good, but not perfect. If you've made concomitant changes to your DAC and downstream electronics, then the effect has at least the possibility of being from those mods.
This all gets back to the fact that it the PCM stream is the same, how you got there doesn't matter.
I really encourage you guys to read up a bit better on how digital data works in order to understand why the hypotheses you're presenting are essentially a non-starter. Yes, bits get corrupted, but it's rare and mechanisms within communication protocols work so hard to mitigate the issues. Or computers would be in a permanent state of crashed and communication would be nonexistent.
The point I'm trying to make is, provided the data is making it to the DAC correctly (and the microelectronics industry has gotten very good at that with such pedestrian demands as audio on a dedicated system), the DAC is what matters. As long as a zero is a recorded as a zero, you can be confident it'll remain a zero until the DAC turns it into a voltage/current signal, where the linearity is very good, but not perfect. If you've made concomitant changes to your DAC and downstream electronics, then the effect has at least the possibility of being from those mods.
This all gets back to the fact that it the PCM stream is the same, how you got there doesn't matter.
I really encourage you guys to read up a bit better on how digital data works in order to understand why the hypotheses you're presenting are essentially a non-starter. Yes, bits get corrupted, but it's rare and mechanisms within communication protocols work so hard to mitigate the issues. Or computers would be in a permanent state of crashed and communication would be nonexistent.
There is the theoretical (but rather unlikely) possibility that this particular minimalistic player is sensitive to file layout on the memory card, fragmentation etc., but even in that case, the differences would not be due to the ripping process or noise, and simply copying the file on the SD card could change the sound. Easy to test, of course....
Could you please share more details about your CD ripping setup.
I guess I do not need to optimise my laptop (i7, 4 Gb RAM, 1 TB SSD drive, built in blu ray reader) hardware at this stage.
What version of dBpoweramp you use, they sell R15 reference and packages at the moment?
Last edited:
Simple answer is yes. SSD would sound different compared to HDD and, for example, different SSDs might also sound different. There are a lot of reports and discussions out there about this topic.
As an example of what could influence the sound quality is media generated vibration. SSD compared to a spinning HDD generates no vibration of its own. It is likely as well that these devices in turn have different sensitivity to external vibration and vibration isolation as well.
To expand on your question - how reproduction of the same music files, all other things beeing equal, stored on SSD would differ from SD cards. From pure mechanical point of view I would expect SSD to be superior as it has higher mass, hence potentially it is less sensitive to external vibration.
NB Vibration in audio is a bitch.
Last edited:
formatcd4,
First off, we should be glad that julf has a new friend on this forum concerning a product neither of them use.
My first guess is that the laptop could not be optimized since it has so much inherent power supply noise within its cramped container. I am sure that this will be disputed but that is the reality of a public forum.
The motherboard is the MSI P33 H81 which was chosen because of the arrangement of the SATA ports. I am using a SATADOM for OS and it is wider than a connector so the P33's arrangement allows the use of the SATADOM without blocking the other ports. The SATADOM plugs directly into the connector; it does not use a cable.
I use the SATA 2 ports for the CDROM and the music storage drive. They are quieter than SATA 3. There is no doubt speed is noisy.
The CPU is the INTEL 4130t - the low power version.
Memory is MUSHKIN 2 gB low latency - it was the lowest latency stick I could find and it is cheap.
For music storage a blue WD 320 gB laptop drive - I have four of these which is more than I will ever need.
CDROM is the LITEON IHAS124-14 - I had read something good about it at dBpoweramp's forum.
I am using the latest version of dBpoweramp though I have removed everything from it other than what is required to rip a REDBOOK CD. At this point if one wants to change something, which I know I never will, it is best to do it within the registry using another computer (connect the OS drive to another computer). The link to regedit has been broken. There are no registry modifications allowed in this OS while it is running.
The heatsink for the CPU is a 100 mm square block of brass which is also a very effective damper for the whole board. CPU temps: 35 degrees C average .
ATX power for the P24 rails - EVGA (relabelled POWERFLOWER) 850G2.
CPU power - OPTIMA battery which also powers the video card which uses the ASROCK riser so the card is away from the board and allows use of the "short" pci-e slot. The ASROCK riser makes it easy to use another power source. This battery is float charged with a 12 volts BELLESON reg.
All drives are powered by a pair of the A123 batteries float charged by a 5 volts regulator board from EBAY, one of the better ones.
No exotic cables - they are shielded.
So, ordinary, but good, components.
I had found the biggest advantage from linear supplies was powering the CPU. It is by far the noisiest thing on the board; separating it from the powering of the other rails is probably the biggest advantage. The video card relieves the CPU of video duty. It is an old card, you want one with minimal onboard memory - mine is 256K - I would like one with 128K but could not find one. I think I paid twenty dollars for the thing many years ago.
I have the video set for lowest quality. There is only one font installed - a system font that is required for XP to work. I have simplified the dBpoweramp "screen" - only one icon remains and the system font is not pretty but who cares? Fonts are kept in memory so getting rid of them keeps junk out of the memory. Same with icons - I went through the system files and removed the vast majority of icons and bitmaps, along with most string tables and cursors - only one cursor remains. You would be amazed how much of this stuff is duplicated within the files. I figure at 34 mB the OS runs from memory even though it is not "told" to - I figure everything that is left is needed so that is where it resides. If anything, calls are minimized to almost zero.
In the BIOS I turn off everything that can be turned off. I use only one core of the CPU and voltages and speeds of memory and CPU are set as low as the BIOS will allow.
When the base OS was installed I turned off all of this stuff so none of it would trigger installation of unneeded things from the start.
If you put something together I can email you the OS for you to try. Of course, I would like to think you have purchased a copy of dBpoweramp since Spoon does good work and deserves to be paid!
First off, we should be glad that julf has a new friend on this forum concerning a product neither of them use.
My first guess is that the laptop could not be optimized since it has so much inherent power supply noise within its cramped container. I am sure that this will be disputed but that is the reality of a public forum.
The motherboard is the MSI P33 H81 which was chosen because of the arrangement of the SATA ports. I am using a SATADOM for OS and it is wider than a connector so the P33's arrangement allows the use of the SATADOM without blocking the other ports. The SATADOM plugs directly into the connector; it does not use a cable.
I use the SATA 2 ports for the CDROM and the music storage drive. They are quieter than SATA 3. There is no doubt speed is noisy.
The CPU is the INTEL 4130t - the low power version.
Memory is MUSHKIN 2 gB low latency - it was the lowest latency stick I could find and it is cheap.
For music storage a blue WD 320 gB laptop drive - I have four of these which is more than I will ever need.
CDROM is the LITEON IHAS124-14 - I had read something good about it at dBpoweramp's forum.
I am using the latest version of dBpoweramp though I have removed everything from it other than what is required to rip a REDBOOK CD. At this point if one wants to change something, which I know I never will, it is best to do it within the registry using another computer (connect the OS drive to another computer). The link to regedit has been broken. There are no registry modifications allowed in this OS while it is running.
The heatsink for the CPU is a 100 mm square block of brass which is also a very effective damper for the whole board. CPU temps: 35 degrees C average .
ATX power for the P24 rails - EVGA (relabelled POWERFLOWER) 850G2.
CPU power - OPTIMA battery which also powers the video card which uses the ASROCK riser so the card is away from the board and allows use of the "short" pci-e slot. The ASROCK riser makes it easy to use another power source. This battery is float charged with a 12 volts BELLESON reg.
All drives are powered by a pair of the A123 batteries float charged by a 5 volts regulator board from EBAY, one of the better ones.
No exotic cables - they are shielded.
So, ordinary, but good, components.
I had found the biggest advantage from linear supplies was powering the CPU. It is by far the noisiest thing on the board; separating it from the powering of the other rails is probably the biggest advantage. The video card relieves the CPU of video duty. It is an old card, you want one with minimal onboard memory - mine is 256K - I would like one with 128K but could not find one. I think I paid twenty dollars for the thing many years ago.
I have the video set for lowest quality. There is only one font installed - a system font that is required for XP to work. I have simplified the dBpoweramp "screen" - only one icon remains and the system font is not pretty but who cares? Fonts are kept in memory so getting rid of them keeps junk out of the memory. Same with icons - I went through the system files and removed the vast majority of icons and bitmaps, along with most string tables and cursors - only one cursor remains. You would be amazed how much of this stuff is duplicated within the files. I figure at 34 mB the OS runs from memory even though it is not "told" to - I figure everything that is left is needed so that is where it resides. If anything, calls are minimized to almost zero.
In the BIOS I turn off everything that can be turned off. I use only one core of the CPU and voltages and speeds of memory and CPU are set as low as the BIOS will allow.
When the base OS was installed I turned off all of this stuff so none of it would trigger installation of unneeded things from the start.
If you put something together I can email you the OS for you to try. Of course, I would like to think you have purchased a copy of dBpoweramp since Spoon does good work and deserves to be paid!
I use a very simple python program, but if you don't want to do any programming, I suggest you use flac or sox to produce a pure-PCM version of the files.
And i am glad we can manage to keep this discussion fact-focused instead of resorting to personal slurs.
Would it be possible for you top respond to my suggestion about the double-blind test setup?
Sigh. There's value in contradicting things that are wrong as to avoid others coming to similar incorrect conclusions and waste their time on something that has minimal to no plausibility. I realize that you're annoyed by this, but at the same time, any real effect you *might* have (and you haven't reliably tested) is not due to what you attribute it to, unless you want to rewrite a large portion of physics. Why the patent refusal to understand the mechanisms that get data from a disk to your storage medium and off again, much less a modicum of research into digital logic? Obviously my background in computer engineering and fabricating/characterizing digital logic elements in new material sets (now being used in some lcds) is useless here...
Shockingly, I have ripped all my CDs, and, comically enough, used multiple computers at once to speed process up. Given my scatterbrain nature I did rip a couple CDs on both my laptop and desktop. Both showed error free ripping, albeit a little slower on my laptop because it's not as good a drive and frequently needed to reread a few spots on most discs. I didn't do a bit compare, so I apologize for not having that info, but given ultimately both rips ended up matching the same checksums, I don't see what I'd gain.
And, no, given the amount of processing needed to take the bits off a hdd or ssd, and the fact that a lot of clever engineers have spent man-centuries (millennia?) ensuring that the physical representations of the data on disk or in a nand/nor cell end up as a sequence of logic high/logic low as transmitted over the communication buses and end up in memory reliably time and again, the likelihood of being able to tell the source of those bits is at homeopathic levels of plausibility: essentially zero. What effect the read/write electronics has on the initial voltage values has been wiped out by the number of gates that data had flowed through by the time it ends up on the input gates of your DAC.
Another test: assuming you have a couple SD cards, can you have a wife/friend at least run a single blind test of the sound of storing the data on different SD cards.
Shockingly, I have ripped all my CDs, and, comically enough, used multiple computers at once to speed process up. Given my scatterbrain nature I did rip a couple CDs on both my laptop and desktop. Both showed error free ripping, albeit a little slower on my laptop because it's not as good a drive and frequently needed to reread a few spots on most discs. I didn't do a bit compare, so I apologize for not having that info, but given ultimately both rips ended up matching the same checksums, I don't see what I'd gain.
And, no, given the amount of processing needed to take the bits off a hdd or ssd, and the fact that a lot of clever engineers have spent man-centuries (millennia?) ensuring that the physical representations of the data on disk or in a nand/nor cell end up as a sequence of logic high/logic low as transmitted over the communication buses and end up in memory reliably time and again, the likelihood of being able to tell the source of those bits is at homeopathic levels of plausibility: essentially zero. What effect the read/write electronics has on the initial voltage values has been wiped out by the number of gates that data had flowed through by the time it ends up on the input gates of your DAC.
Another test: assuming you have a couple SD cards, can you have a wife/friend at least run a single blind test of the sound of storing the data on different SD cards.
Which vibration coming from where? What vibrations do SSD's make? What vibrations are they sensitive to? How would this ultimately end up affecting the bit stream or, at most plausible (though minimally) on the voltage/current out of your DAC?
Vibration comes from internal components such as HDD, fan, transformers, from external environment through the feet and through air when you play music. All these affect internal circuitry, indiviidual components and cables/wires inside your digital device.
Vibration has a very detrimental effect on the SQ and affects all components of your system, both digital and analogue: NAS (or any other storage device), cable connecting your NAS, music server or CD player or any other front end device, cable(s) connecting these to your DAC, DAC itself, cable connecting your DAC to pre or power amp, etc. down to loudspeakers, including the loudspeakers. And do not forget your power supply(s) and power cables.
Last edited:
Forget it. Not from me. I have done enough to discover it myself over many years through listening to different systems privately, attending shows and demo's.
So if you are sincerely interested in the topic in order to rip the potential benefits prepare yourself for a long jorney. There will be no free lunch here and you will need to spend time and effort to make your own discoveries. You will only believe your own ears, that's at least how it works for me.
You do not need to wait, however, to meet your local dealer to test isolation platforms or feet or to attend the next CES show. Just use some suitable materials at hand in your house to isolate your CD player, DAC, or your power distribution device. Put a bag of sand on top of any of these devices and listen. You might be up for a surprise.
NB Apologies if you already know all of that.
So if you are sincerely interested in the topic in order to rip the potential benefits prepare yourself for a long jorney. There will be no free lunch here and you will need to spend time and effort to make your own discoveries. You will only believe your own ears, that's at least how it works for me.
You do not need to wait, however, to meet your local dealer to test isolation platforms or feet or to attend the next CES show. Just use some suitable materials at hand in your house to isolate your CD player, DAC, or your power distribution device. Put a bag of sand on top of any of these devices and listen. You might be up for a surprise.
NB Apologies if you already know all of that.
Last edited:
But never put in the effort to actually verify your assumptions objectively? Fair enough, you are not alone in accepting the easy path of faith instead of facts.
Indeed. It involves questioning your beliefs and assumptions, actually verifying your results, and learng about how stuff works. Not easy. Especially for your ego.
There is a bit of contradiction there. Only believing your own ears takes no effort at all, only credibility. Actually questioning your beliefs, preconceived notions and senses is what requires hard work and intellectual honesty. Doesn't come easy.
Agreed. Especially if you only listen using your ears (as in double blind).