Best FLAC converter software

[sandyk]

29 minutes ago, miguelito said:

I don't understand... Isn't FLAC lossless? In which case, doesn't that mean that the data should be exactly recovered - other than possibly metadata (which can be incorporated in WAV) the files should be bits for bit identical. Is this not the case? Why?

 

Notice I am not a "bits is bits" person. In this particular case you aren't comparing FLAC and WAV as you play, but doing math to get from FLAC to WAV... Unless there is something inherently complex about whatever compression happens in FLAC, this makes no sense to me.

 

BTW... This possibly means that WAV to FLAC is not all that reliable. 

 The perceived audible differences are not normally huge, but both .ape and .flac S/W  perform as intended without conversion errors , and the differences appear to be mainly due to electrical noise in the PC/Server which is capable of further improvement in many cases.

[fas42]

25 minutes ago, miguelito said:

Yes, and I agree. But my question remains: is the WAV data bit for bit identical? It should be since this is done offline.

 

Yes, the music data in the WAV file is identical - bit for bit. So, there should be no considerations in this respect ... but as @sandyk pointed out earlier, I was caught out by some samples he had prepared, some time ago, which to my ears were quite different - but when I did a nulling operation in Audacity they turned out to be 'identical' ... so far I have no explanation for this; it's been put in the ToDo basket, for something to investigate down the track ....

[manueljenkin]

18 hours ago, bogi said:

 

... and for me, audio is only fun. Both listening and the technical things. I like to compare these 2 worlds. I am open to anything new but missing explanation of the WOW effect physical nature. Eventually I can listen to files sandyk is pointing to.

 

Maybe the computer noise of FLAC to WAV file conversion process is somehow "remembered" in storage media - it may affect how "precisely" is the information stored in physical media such as any computer memory. There are always some levels of physical values which allow to distinguish logical 0 and logical 1. So the physical nature of storage media can cause that the same file, when stored twice on the same media, is stored a bit differently. Like 2 CDs burned at different burning speeds.  CD player can then produce different noise pattern when reading one CD and the other, although the bits are the same.

 

So if different noise pattern of FLAC to WAV conversion can influence the physical properties how the WAV file is stored and this way to "store" the noise related information with the file on physical media, that differences could propagate

further as different noise patterns when reading the stored files again.

 

The question is how far can such type of noise be propagated. Until a zip file arrives from Australia to Europe, it travels through tens of routers and in our home possibly through a switch and then WiFi network. The different noise pattern would need to be "reproduced" all the way. Through different physical principles how the bits are stored on the way and all the electrical to optical to wireless (and back) converters all the way. It looks to me very improbable that the initial noise difference would not be filtered out during so many hops of travel.

Wow. This is the first time I'm hearing such an explanation, and atleast theoretically it makes sense.

 

I would like to make one correction to the conclusion though. As long as the things remain digital and non realtime stuff can be completely corrected. It is only the noise levels in the final chain that does the final audio output that would matter the most. This would mean not only that the software to do this conversion giving sq change but rather any copy/move action or file system differences, etc would also be a part of the change since all of them might have their signatures of noise/physical data storage pattern embedded in the data. Same goes with access type, the hits/misses, etc. Maybe it's time for an audiophile system partitioner 😅.

 

It is likely to not be a one size fits all solution since SSD access types and noise profiles (static and dynamic) is likely to be different from hdd, and different from flash memory.

 

It's only theoretical possibility though, I have done zero comparison primarily because I am not chasing that dragon yet.

[AnotherSpin]

5 hours ago, sandyk said:

 If the OP hadn't been banned as a result of him losing his cool ,because of the attacks from members such as yourself and AnotherSpin, he would have been in the position as the OP to request that as this is clearly  a Subjective, NOT Objective thread,  that comments like your would have been removed , and not have been permitted.as they inevitably lead to the closure of threads which is their clear intention.

 This is exactly the kind of disruptive behaviour that saw quite a few members either banned or move to A.S.R.

 

Did you try the app aggressively promoted here by OP? Can you share your impressions about results?

[Vlodzimierz]

39 minutes ago, AnotherSpin said:

 

Did you try the app aggressively promoted here by OP? 

It is impossible to read your message without laughing. And also many others who deal with verbal diarrhea. And this is instead of just trying to listen to music from Wav files and compare with the sound of Flac. And compare how different programs convert to Wav. Instead, all Forum participants were mired in controversy that this could not be. I recognize the thinking style of computer scientists, not audiophiles.
I personally spent a lot of time and money comparing different programs. And my friend and I came to you at the Forum with a ready-made result. We wanted to share our experience. And they met an aggressively boorish model of behavior. You look like fools in the issue under discussion. Disrespectful and lazy fools! It's not interesting with you.
Once again I apologize for the google translator. I understand that he translates very badly. But I hope you understand the main meaning.
P.S. mine did not come to promote the conversion program, just to tell you about what you still did not know! I bought about ten programs for long-term comparison. At your own expense. So it can further promote its behavior model. She's bound to make you dumber.

[sandyk]

42 minutes ago, AnotherSpin said:

 

Did you try the app aggressively promoted here by OP? Can you share your impressions about results?

The results of the only other S/W I tried were posted 23 hours ago on page 2

[manueljenkin]

7 minutes ago, Vlodzimierz said:

It is impossible to read your message without laughing. And also many others who deal with verbal diarrhea. And this is instead of just trying to listen to music from Wav files and compare with the sound of Flac. And compare how different programs convert to Wav. Instead, all Forum participants were mired in controversy that this could not be. I recognize the thinking style of computer scientists, not audiophiles.
I personally spent a lot of time and money comparing different programs. And my friend and I came to you at the Forum with a ready-made result. We wanted to share our experience. And they met an aggressively boorish model of behavior. You look like fools in the issue under discussion. Disrespectful and lazy fools! It's not interesting with you.
Once again I apologize for the google translator. I understand that he translates very badly. But I hope you understand the main meaning.
P.S. mine did not come to promote the conversion program, just to tell you about what you still did not know! I bought about ten programs for long-term comparison. At your own expense. So it can further promote its behavior model. She's bound to make you dumber.

I find audiophilestyle to be the most welcoming form for anyone deep into these gremlins. As such, there are people who do try to push you and engage in drama, my suggestion would be to maintain patience. Things will turn around. A soft and polite approach will take time, but is more likely to work in the long run.

 

I made a thread a while ago comparing sq differences between music players (all bit perfect) over here: 

I got a fair amount of support and yes occasionally we do encounter people trying to be dodgy but you need to learn to not give them what they want - "drama". Once you've fallen for that, they've won. Regarding moderation, the freedom of expression works both ways, and there's only so much Chris can do.

 

I feel sorry for the op getting banned, I hope he comes back (and learns to keep calm in the toughest baits).

[The Computer Audiophile]

6 minutes ago, Vlodzimierz said:

It is impossible to read your message without laughing. And also many others who deal with verbal diarrhea. And this is instead of just trying to listen to music from Wav files and compare with the sound of Flac. And compare how different programs convert to Wav. Instead, all Forum participants were mired in controversy that this could not be. I recognize the thinking style of computer scientists, not audiophiles.
I personally spent a lot of time and money comparing different programs. And my friend and I came to you at the Forum with a ready-made result. We wanted to share our experience. And they met an aggressively boorish model of behavior. You look like fools in the issue under discussion. Disrespectful and lazy fools! It's not interesting with you.
Once again I apologize for the google translator. I understand that he translates very badly. But I hope you understand the main meaning.
P.S. mine did not come to promote the conversion program, just to tell you about what you still did not know! I bought about ten programs for long-term comparison. At your own expense. So it can further promote its behavior model. She's bound to make you dumber.

Bye @Vlodzimierz

[sandyk]

12 minutes ago, Vlodzimierz said:

And this is instead of just trying to listen to music from Wav files and compare with the sound of Flac.

  

 I have been saying for many years here that I always convert .flac files to .wav files for serious listening and always prefer .wav over the original .flac files where they were supplied in that format from Linn Records and HDTracks etc. and then converted to .wav files.

[manueljenkin]

8 hours ago, copy_of_a said:

But only for the super short time it takes the playback software to load the PCM into RAM ... after that initial decoding everything should be equal to, say, WAV. It's really totally a non-issue.

Great job framing your personal approach to look like the end all be all of analysis. I'm sorry your description is incomplete. And I would say this is a form of opinion bullying to belittle the approach of someone else. @The Computer Audiophile imo, these kind of opinion bullying can be moderated.

 

The stored data is present as charges (or in case of optical disk, the equivalent in light waves). The way these are written and the surrounding noise patterns during the operation might decide what the internal structure/distribution of this is inside the storage media. When reading back, it could cause a different noise pattern joining the required signal.

 

Now I'm not saying it can be conclusively said that this noise propagates enough to cause audible changes but there is no proof of the contrary either (also it's near impossible to prove the non existence of something so the benefit of doubt will always be in the believer). Whether one finds it an issue or non issue is personal. If you're speaking for yourself make it clear you're speaking for yourself, don't assume it is for everyone.

[AnotherSpin]

21 minutes ago, sandyk said:

The results of the only other S/W I tried were posted 23 hours ago on page 2

 

So, you have not tried the app promoted by nice couple of our new friends. Given your support and your tacit agreement with their style of communication, it is rather strange, as I dare to believe. 

[sandyk]

19 minutes ago, manueljenkin said:

The stored data is present as charges (or in case of optical disk, the equivalent in light waves). The way these are written and the surrounding noise patterns during the operation might decide what the internal structure/distribution of this is inside the storage media. When reading back, it could cause a different noise pattern joining the required signal.

 This is undoubtedly the main reason behind why BluSpec and SHM formats can sound quite different to normal RBCD , and  due to improved readabilty.

I did purchase 3 sets of Sony Blu Spec comparisons several years back and had no trouble clearly hearing the improvement of the BluSpec format under non sighted conditions. 

 

Incidentally, recently I have been backing up many of my favourite RBCD and High Res albums to 25GB BluRay discs, and perhaps due to the improved precision of the Optical head necessary for this format, they actually sound a little better than the original CD when played with my Oppo 103 into a higher quality DIY DAC 

 

[manueljenkin]

Speaking of sony, half a decade ago, they released an "audiophile" sd card. I guess market didn't like that approach and seems to have backfired. I wish I could try one to see if it had any legibility, and considering its from Sony, I have far less skepticism than I generally would with other manufacturer.

[miguelito]

12 hours ago, fas42 said:

Yes, the music data in the WAV file is identical - bit for bit. So, there should be no considerations in this respect ... but as @sandyk pointed out earlier, I was caught out by some samples he had prepared, some time ago, which to my ears were quite different - but when I did a nulling operation in Audacity they turned out to be 'identical' ... so far I have no explanation for this; it's been put in the ToDo basket, for something to investigate down the track ....

What happens if you do a few copies of those files? This would smear out the possibility that some files sound different because of the work done to retrieve them. But seriously, I am very skeptical and I would have to try this myself.

[AnotherSpin]

50 minutes ago, manueljenkin said:

Speaking of sony, half a decade ago, they released an "audiophile" sd card. I guess market didn't like that approach and seems to have backfired. I wish I could try one to see if it had any legibility, and considering its from Sony, I have far less skepticism than I generally would with other manufacturer.

 

Forum participants experimented some time ago with installing an operating system on the sd cards and playing music from the sd card, or from virtual RAM, etc. I remember that different cards indeed 'sounded' slightly differently in such experiments. Sony products were believed to be "good". However, the difference was not at all dramatic and the inconveniences of running the computer from the sd card were more important. There were also attempts to optimize the operating system using different scripts. I would also put the encoding of the files from the flac to the wav into the same category. If you listen very carefully, you will notice the slightest differences, not improvement necessarily, but all this is within the framework of minor upticks that can be safely ignored by any one who simply wants to listen music and avoid masochistic procedures on the way. Imho, of course.

 

You could try search function to find all these discussions on the forum. Some fresh ideas are not so fresh after all.

 

[Miska]

3 hours ago, AnotherSpin said:

I remember that different cards indeed 'sounded' slightly differently in such experiments. Sony products were believed to be "good". However, the difference was not at all dramatic and the inconveniences of running the computer from the sd card were more important.

 

I'm doing that all the time with HQPlayer OS, but I don't find it any more inconvenient than any other method. Dumping the OS image on card and booting it up. Possibly restoring configuration backup and that's it.

 

But in my case content is on SMB network share or from UPnP or Roon stream, not local.

 

[AnotherSpin]

39 minutes ago, Miska said:

 

I'm doing that all the time with HQPlayer OS, but I don't find it any more inconvenient than any other method. Dumping the OS image on card and booting it up. Possibly restoring configuration backup and that's it.

 

But in my case content is on SMB network share or from UPnP or Roon stream, not local.

 

 

I also boot the NUC from USB stick with your NAA image. But, I meant something else. By the way, I remember that you definitely objected the usage of optimization scripts for Mac OS on the computer with HQPlayer Desktop installed 😀

[fas42]

8 hours ago, miguelito said:

What happens if you do a few copies of those files? This would smear out the possibility that some files sound different because of the work done to retrieve them. But seriously, I am very skeptical and I would have to try this myself.

 

Alex himself notes that doing copies often loses what's going on - personally, there are far, far more important things to sort out in terms of getting best sound from a particular recording; optimising the integrity of the playback chain has the greatest benefit IME. So I'm not really fussed about precisely how the music data is stored.

 

Alex would probably be very happy to send you copies of some of the things he's got - if you can't hear anything in them, then you won't need to bother about it ... 🙂.

[Miska]

7 hours ago, AnotherSpin said:

By the way, I remember that you definitely objected the usage of optimization scripts for Mac OS on the computer with HQPlayer Desktop installed 😀

 

Yes, because it tends to break things... Same as on Windows, stopping services needed by HQPlayer won't do much good. If it doesn't make the application outright fail, it may degrade performance.

 

Usually, by default, macOS behaves quite sensible way.

 

[AnotherSpin]

5 hours ago, Miska said:

 

Yes, because it tends to break things... Same as on Windows, stopping services needed by HQPlayer won't do much good. If it doesn't make the application outright fail, it may degrade performance.

 

Usually, by default, macOS behaves quite sensible way.

 

 

And what do you think of converting flac files to wav before they are sent to HQPlayer? Would it be sensible to embed such converting into HQPlayer?

[bogi]

4 hours ago, AnotherSpin said:

And what do you think of converting flac files to wav before they are sent to HQPlayer? Would it be sensible to embed such converting into HQPlayer?

 

I don't understand what exactly you mean with your question. Every audio player needs to decode the input audio file to get raw PCM (or DSD) content. To decode FLAC, a player has to decompress the sample data to get raw PCM samples. That's what HQPlayer is doing with PCM content and the same must be done by any FLAC to WAV converter. HQPlayer is doing that decompression already. And it is doing it on the fly with no special memory requirements for that process (other thing are memory requirements for DSP). Embedding that in HQPlayer would mean converting the whole input file to WAV before further processing, because something as "one sample WAV" does not make sense. And decoding the complete audio file before actual playback would mean:
1. Decode FLAC, get raw PCM data

2. Encode raw PCM data to WAV

3. Decode WAV to get raw PCM data

4. Pass raw PCM data to player engine.

As you see, points 2 and 3 are redundant. And what is important, FLAC to WAV conversion takes some time and one could want to play 20 or more minutes long track - nothing special for classical music, or for rock music of 70's. That would mean interruption of playback between tracks. And it would mean interruption also in the case of track transition without silence ("The Dark Side Of The Moon", "Wish You Were Here", concert recordings ...).

 

Converting FLAC to WAV in advance, but independently on HQPlayer, before an audio content (a whole album is optimal case) is passed to HQPlayer, has IMO still advantages. I am doing so in an automated fashion. Every PCM track I want to play I am converting to WAV, physically it means also transferring audio data from a magnetic USB disk (my FLACs) to RAM disk (the WAVs to be played). I am doing that after my listening tests I did years ago. The only logical explanation I have to explain the difference is about higher level of noise computer generates when reading file from magnetic disc. And some noise could be added also by FLAC decompression in HQPlayer, if FLAC is played instead of WAV. The WAV file decoding is much easier - nothing is to be decompressed, sample values can be directly read.

 

[AnotherSpin]

2 hours ago, bogi said:

 

I don't understand what exactly you mean with your question. Every audio player needs to decode the input audio file to get raw PCM (or DSD) content. To decode FLAC, a player has to decompress the sample data to get raw PCM samples. That's what HQPlayer is doing with PCM content and the same must be done by any FLAC to WAV converter. HQPlayer is doing that decompression already. And it is doing it on the fly with no special memory requirements for that process (other thing are memory requirements for DSP). Embedding that in HQPlayer would mean converting the whole input file to WAV before further processing, because something as "one sample WAV" does not make sense. And decoding the complete audio file before actual playback would mean:
1. Decode FLAC, get raw PCM data

2. Encode raw PCM data to WAV

3. Decode WAV to get raw PCM data

4. Pass raw PCM data to player engine.

As you see, points 2 and 3 are redundant. And what is important, FLAC to WAV conversion takes some time and one could want to play 20 or more minutes long track - nothing special for classical music, or for rock music of 70's. That would mean interruption of playback between tracks. And it would mean interruption also in the case of track transition without silence ("The Dark Side Of The Moon", "Wish You Were Here", concert recordings ...).

 

Converting FLAC to WAV in advance, but independently on HQPlayer, before an audio content (a whole album is optimal case) is passed to HQPlayer, has IMO still advantages. I am doing so in an automated fashion. Every PCM track I want to play I am converting to WAV, physically it means also transferring audio data from a magnetic USB disk (my FLACs) to RAM disk (the WAVs to be played). I am doing that after my listening tests I did years ago. The only logical explanation I have to explain the difference is about higher level of noise computer generates when reading file from magnetic disc. And some noise could be added also by FLAC decompression in HQPlayer, if FLAC is played instead of WAV. The WAV file decoding is much easier - nothing is to be decompressed, sample values can be directly read.

 

 

Does it mean the whole idea of converting FLAC into WAV in advance with dedicated converter app makes no sense at all, if a sophisticated player, as HQPlayer, is used - if FLAC is put into virtual RAM, etc.?

[bogi]

If FLAC is put into virtual disk (RAM disk), HQPlayer still needs to decompress, so there may be some added noise caused by processing against reading WAV from virtual disk. And the decompression costs some (not really much) computing resources which in some border situation theoretically could be missing for DSP. Theoretically. IMO magnetic disk vs. virtual RAM disk makes more significant difference than FLAC vs WAV in virtual disk. And my experience is that RAM disk implementation sonically also matters. I don't like to write something what others could consider a nonsense, but years ago I compared more RAM disk implementations. At that time Primo RAM disk was suggested by for example Geoff and Eurodriver, who made sonic comparisons of more products too. That time I found free Softperfect RAM disk, compared it with others and I am using it till today, without any further comparisons. So if you want to try such a thing, you know my recommendation.

 

 

I am simply doing copying to RAM disk with FLAC to WAV conversion in one step. I am doing so for whole album at once. I am using foobar2000 converter in command line mode for that. It processes up to 4 files in parallel and is able to use 4 cores for that, so it really runs in parallel and very fast. For typical CD it takes about 10 seconds. For hires FLAC content it takes longer, about half minute, or possibly more for 192k or multichannel. Occassionally I tend to do some basic DSP with PCM content yet before it is passed to HQPlayer (the foobar2000 converter can be set up to do DSP during format conversion). For example, when I am listening on headphones, I can run multichannel to stereo to lower HQPlayer load. If the recording is too bright, I can run a tilt filter (discussed a month ago). In HQPlayer I tend to do stereo to binaural processing (since I am listening on headphones). So for me moving of content to be played into RAM disk is bound together with format conversion and possible DSP.

[AnotherSpin]

1 hour ago, bogi said:

If FLAC is put into virtual disk (RAM disk), HQPlayer still needs to decompress, so there may be some added noise caused by processing against reading WAV from virtual disk. And the decompression costs some (not really much) computing resources which in some border situation theoretically could be missing for DSP. Theoretically. IMO magnetic disk vs. virtual RAM disk makes more significant difference than FLAC vs WAV in virtual disk. And my experience is that RAM disk implementation sonically also matters. I don't like to write something what others could consider a nonsense, but years ago I compared more RAM disk implementations. At that time Primo RAM disk was suggested by for example Geoff and Eurodriver, who made sonic comparisons of more products too. That time I found free Softperfect RAM disk, compared it with others and I am using it till today, without any further comparisons. So if you want to try such a thing, you know my recommendation.

 

 

I am simply doing copying to RAM disk with FLAC to WAV conversion in one step. I am doing so for whole album at once. I am using foobar2000 converter in command line mode for that. It processes up to 4 files in parallel and is able to use 4 cores for that, so it really runs in parallel and very fast. For typical CD it takes about 10 seconds. For hires FLAC content it takes longer, about half minute, or possibly more for 192k or multichannel. Occassionally I tend to do some basic DSP with PCM content yet before it is passed to HQPlayer (the foobar2000 converter can be set up to do DSP during format conversion). For example, when I am listening on headphones, I can run multichannel to stereo to lower HQPlayer load. If the recording is too bright, I can run a tilt filter (discussed a month ago). In HQPlayer I tend to do stereo to binaural processing (since I am listening on headphones). So for me moving of content to be played into RAM disk is bound together with format conversion and possible DSP.

 

I made a comparison between: a) FLAC file; b) FLAC file converted to WAV with XLD; c) FLAC file converted with XLD to WAV and placed in a virtual RAM disc. It seemed to me that there was no difference between variant a and b, or the difference could be ignored, and there was a very small, barely perceptible difference between a and c. Maybe, c sounded a little more dynamic. All the files were played through HQPlayer in mac OS + NUC NAA. Anyway, the difference between the various HQPlayer settings is more noticeable. The case put in rest.

 

Also: it seems to me that when I first experimented with the converting FLAC into WAV ten years or so ago, the difference was more noticeable. Since then, linear power supply units were added at all critical points, better quality cables, and NAA have been added to the audio system. Now there is practically no difference. Probably, it could be concluded that in this case the efforts in the field of hardware proved to be more effective.

[bogi]

... and you are probably not using magnetic USB connected data disc. The difference in my case may be bigger. I am using a notebook without any special treatment for all things I need including my work. It has two OS, one for work and the other for home/hobby activities.

My approach is RAM disk + iFi iSilencer (I can hear its effect). When I am listening in my living room, I am using directly connected DAC (small Topping E30 + L30 combo) with HiFiMan HE-500 or Final E4000.

 

I am playing from the same notebook (WiFi connected) also into Atom Z8350 based Win10 NAA (Ethernet) placed one stock up at my main listening place. Here I have better DAC + headamp combo, DAC with iSilencer too, with few LPS and more careful cable treatment. When the notebook and NAA are wirelessly isolated, notebook noise becomes unimportant. But for HQPlayer it is still "less work" to read WAV and I still have the possibility to do some DSP during FLAC to WAV conversion.