Tweet
Craig and Rick;
Being recently interested in ‘restoring’ mp3 sourced wav files and determining the best way of doing this, I took advantage of your suggestion to use the Virtual Valve Amplifier to add more interest and harmonics to what are otherwise flat sounding mp3 sourced audio files. I have also found this same technique useful in restoring downloads or on-demand streaming audio files/WiFi radio files recorded directly from diverse online sources. These files are also reduced in frequency content and fidelity in all sorts of convoluted ways.
May I first say that you both have earned far beyond an ‘A+’ on the design and development of the VVA. I know this tool has been around for a while but only recently have I taken a very close and serious look at it. Since your suggestion, I have spent many hours experimenting with every possible option of the VVA, and I am in wonder about the amount of time it took you to model out all these vacuum tube response curves. It may be that you have to be of a certain age to appreciate vacuum tubes - and perhaps the younger folks just don’t get it – but nonetheless, tubes still have lived on beyond anyone’s imagination, including my own.
Also the idea and irony of using a vacuum tube simulator to restore ‘modern’ mp3 sourced audio files is quite amusing to me.
I found the online Wikipedia article on the VVA to be very useful. This is much of the same info the DC8 manual has, but adds some additional insights. Perhaps one of you wrote it, so thanks for this as well.
I have noticed the VVA can synthesize frequencies to 20 Khz and further, which is beyond the range the DC8 EQ filters can really handle. The level of these additional frequencies at 20 Khz cannot be actually controlled with the EQ filter [this is easily observable with the Spectrum Analyzer]. But since we cannot hear any of this, it is really a minor issue. This is just an observation and you may want to correct it in some way with a modification to the EQ filters, with an absolute cutoff at 20 Khz.
As you suggested, I have found both the VVA and the Harmonic Overtone Synthesizer useful in filling out the upper frequency ranges of these mp3 sourced audio files. There is the great advantage the VVA has of adding diverse harmonics to the whole file spectrum in a way the Overtone Synthesizer does not do, but the Synthesizer can be set up to only add harmonics to frequencies above a certain frequency threshold – I have found this difference to be distinctly useful at times as with working with very clean source files.
In an early posting, I suggested adding some more manual controls to the Overtone Synthesizer to allow precise control of a few of the lower level harmonics in an individual way, rather than just the 2nd harmonic as it is now. I still believe this would be useful because of the operative frequency control the Synthesizer has [but the VVA does not have]. No other audio tool that I know would have such control with the upper harmonics if this tool was developed.
Secondly, I suggest an additional audio tool to the DC product line – an MP3 Restorer – which uses the VVA and Harmonic Overtone Synthesizer software to automatically restore an mp3 file back to the original level of harmonic content. This tool would be modeled to possibly read in the mp3 file and automatically determine the best frequency component additions, adding them to the resultant wav file. Or the user could manually set what was the original coding rate of the mp3 file [32, 64, 96, 128kbs, etc], and the tool would use this information to determine the best audio enhancements. No other DAW audio tool has such a mp3 file restorer, and this would be a great restoration addition to a future generation of DC8.
As I understand it, encoding a file to mp3 does three main things:
1. Truncates the high frequencies, the severity of which is dependent on the coding rate.
2. Does not encode frequency components at or near the threshold of hearing [guessing -70 to -90db, according to a prescribed hearing curve].
3. Ignores audible but lower level background audio information [guessing -40 to -60db] if the average primary audio signal level is very loud.
Additionally, encoding to mp3 may do this as well:
4. Bass response is reduced
5. 1 to 4 Khz frequency range is slightly emphasized
6. Stereo separation is reduced.
In decoding the mp3 back to wav, this new MP3 Restorer tool would be able to synthesize items 1&2 above using the existing software, but I believe item 3 is something that would be more difficult to recreate. However it may be possible, using a look-ahead function, to temporarily bring up the level of low-level synthesized harmonics during loud passages while at the same time temporarily changing the harmonic diversity to partially simulate the lost content of item 3. Items 4-6 would be reversed via Diamond Cut's own proprietary methods.
All of this is just a thought. But mostly I want to thank you for the work you did on the VVA - this is a real hidden treasure that needs to be positively recognized.
Phillip
Being recently interested in ‘restoring’ mp3 sourced wav files and determining the best way of doing this, I took advantage of your suggestion to use the Virtual Valve Amplifier to add more interest and harmonics to what are otherwise flat sounding mp3 sourced audio files. I have also found this same technique useful in restoring downloads or on-demand streaming audio files/WiFi radio files recorded directly from diverse online sources. These files are also reduced in frequency content and fidelity in all sorts of convoluted ways.
May I first say that you both have earned far beyond an ‘A+’ on the design and development of the VVA. I know this tool has been around for a while but only recently have I taken a very close and serious look at it. Since your suggestion, I have spent many hours experimenting with every possible option of the VVA, and I am in wonder about the amount of time it took you to model out all these vacuum tube response curves. It may be that you have to be of a certain age to appreciate vacuum tubes - and perhaps the younger folks just don’t get it – but nonetheless, tubes still have lived on beyond anyone’s imagination, including my own.
Also the idea and irony of using a vacuum tube simulator to restore ‘modern’ mp3 sourced audio files is quite amusing to me.
I found the online Wikipedia article on the VVA to be very useful. This is much of the same info the DC8 manual has, but adds some additional insights. Perhaps one of you wrote it, so thanks for this as well.
I have noticed the VVA can synthesize frequencies to 20 Khz and further, which is beyond the range the DC8 EQ filters can really handle. The level of these additional frequencies at 20 Khz cannot be actually controlled with the EQ filter [this is easily observable with the Spectrum Analyzer]. But since we cannot hear any of this, it is really a minor issue. This is just an observation and you may want to correct it in some way with a modification to the EQ filters, with an absolute cutoff at 20 Khz.
As you suggested, I have found both the VVA and the Harmonic Overtone Synthesizer useful in filling out the upper frequency ranges of these mp3 sourced audio files. There is the great advantage the VVA has of adding diverse harmonics to the whole file spectrum in a way the Overtone Synthesizer does not do, but the Synthesizer can be set up to only add harmonics to frequencies above a certain frequency threshold – I have found this difference to be distinctly useful at times as with working with very clean source files.
In an early posting, I suggested adding some more manual controls to the Overtone Synthesizer to allow precise control of a few of the lower level harmonics in an individual way, rather than just the 2nd harmonic as it is now. I still believe this would be useful because of the operative frequency control the Synthesizer has [but the VVA does not have]. No other audio tool that I know would have such control with the upper harmonics if this tool was developed.
Secondly, I suggest an additional audio tool to the DC product line – an MP3 Restorer – which uses the VVA and Harmonic Overtone Synthesizer software to automatically restore an mp3 file back to the original level of harmonic content. This tool would be modeled to possibly read in the mp3 file and automatically determine the best frequency component additions, adding them to the resultant wav file. Or the user could manually set what was the original coding rate of the mp3 file [32, 64, 96, 128kbs, etc], and the tool would use this information to determine the best audio enhancements. No other DAW audio tool has such a mp3 file restorer, and this would be a great restoration addition to a future generation of DC8.
As I understand it, encoding a file to mp3 does three main things:
1. Truncates the high frequencies, the severity of which is dependent on the coding rate.
2. Does not encode frequency components at or near the threshold of hearing [guessing -70 to -90db, according to a prescribed hearing curve].
3. Ignores audible but lower level background audio information [guessing -40 to -60db] if the average primary audio signal level is very loud.
Additionally, encoding to mp3 may do this as well:
4. Bass response is reduced
5. 1 to 4 Khz frequency range is slightly emphasized
6. Stereo separation is reduced.
In decoding the mp3 back to wav, this new MP3 Restorer tool would be able to synthesize items 1&2 above using the existing software, but I believe item 3 is something that would be more difficult to recreate. However it may be possible, using a look-ahead function, to temporarily bring up the level of low-level synthesized harmonics during loud passages while at the same time temporarily changing the harmonic diversity to partially simulate the lost content of item 3. Items 4-6 would be reversed via Diamond Cut's own proprietary methods.
All of this is just a thought. But mostly I want to thank you for the work you did on the VVA - this is a real hidden treasure that needs to be positively recognized.
Phillip
Comment