bagpuss
Every so often I try blind testing, because I appreciate music enough to want the most pristine quality possible and perceivable.
The vast majority of my files have been encoded with the latest version of the LAME encoder, at a bite rate of 320kbps (with all other presets optimised for quality over ripping speed).
However I recently ripped a CD at WAV for the sake of comparing it with the same songs that I already ripped & encoded to MP3-320.
Using fairly high quality Pioneer headphones, and the technically average Hercules soundcard.
After testing some of my favourite/most listened to tracks (I choose complex tracks with lots of instrumentation and power) I started noticing real differences, although the differences are hard to 'hear' and pin point under focused listening, after some time I started realising the differences and when you do, the differences become obvious and suddenly stand out.
It's a bit like watching a magic trick, even when watching carefully you still believe that you've witnessed magic, but when you know what to look for you'll see everything (and the illusion becomes uncovered).
Perceived differences:
The WAV sounded more dynamic with clear separation between each and every instrument/sound, the MP3 in comparison sounded far more "pushed together", the sounds were less separated (frequency range), with the saxophone sounding less clean and the claps sounding "rougher" in the MP3 compared to the 'clarity' of the WAV. People usually point to the highs as a place where you can usually pick out the difference, and yes the high hats sounded more filtered and less open, which actually transforms the sound to a degree.
Action:
I've decided that I'm a semi-audiophile because I really DO care for the highest quality audio, and with the HD space I have I've decided to use WAVs for the music that I personally enjoy and appreciate, where as for the chart music that I only acquire for the sake of commercial bookings I'll stick to MP3-320 until an all WAV collection is more realistic (in terms of download speed/HD space).
I just thought I'd share my experience with you in case some of you care for the topic, and I'd also like to hear of other peoples experience when performing comparisons between various lossy and lossless audio formats.
The vast majority of my files have been encoded with the latest version of the LAME encoder, at a bite rate of 320kbps (with all other presets optimised for quality over ripping speed).
However I recently ripped a CD at WAV for the sake of comparing it with the same songs that I already ripped & encoded to MP3-320.
Using fairly high quality Pioneer headphones, and the technically average Hercules soundcard.
After testing some of my favourite/most listened to tracks (I choose complex tracks with lots of instrumentation and power) I started noticing real differences, although the differences are hard to 'hear' and pin point under focused listening, after some time I started realising the differences and when you do, the differences become obvious and suddenly stand out.
It's a bit like watching a magic trick, even when watching carefully you still believe that you've witnessed magic, but when you know what to look for you'll see everything (and the illusion becomes uncovered).
Perceived differences:
The WAV sounded more dynamic with clear separation between each and every instrument/sound, the MP3 in comparison sounded far more "pushed together", the sounds were less separated (frequency range), with the saxophone sounding less clean and the claps sounding "rougher" in the MP3 compared to the 'clarity' of the WAV. People usually point to the highs as a place where you can usually pick out the difference, and yes the high hats sounded more filtered and less open, which actually transforms the sound to a degree.
Action:
I've decided that I'm a semi-audiophile because I really DO care for the highest quality audio, and with the HD space I have I've decided to use WAVs for the music that I personally enjoy and appreciate, where as for the chart music that I only acquire for the sake of commercial bookings I'll stick to MP3-320 until an all WAV collection is more realistic (in terms of download speed/HD space).
I just thought I'd share my experience with you in case some of you care for the topic, and I'd also like to hear of other peoples experience when performing comparisons between various lossy and lossless audio formats.
Posted Sat 05 Jan 08 @ 12:47 pm
Dach
One of the things I really like about Virtual DJ is that it does some level of "cleaning up" the sound, to improve it quite a bit. Sometimes, I don't have a say in the quality of the source audio. For example, some Japanese industrial band who emails me 128k MP3's to save space, and due to the language barrier, I can't explain to them that I need higher quality, or they might be offended if I do. But VDJ does something to make the MP3's sound better. On low-bitrate files, I can hear a huge difference in quality between VDJ's output and, say, Winamp or Windows Media Player.
While highest quality is always best, I can't conceive going back to rip 13,000 songs again just because terabyte hard drives are cheap.
While highest quality is always best, I can't conceive going back to rip 13,000 songs again just because terabyte hard drives are cheap.
Posted Sat 05 Jan 08 @ 1:44 pm
djlex
what do you use to encode mp3?
I have a sencore audio analyzer and few weeks ago I did a test of my own.
first I checked my audio acuity and my good days are gone, I do not hear much over 18khz.
then I did encode a cd using Lame @320 and twiking the quality to max with constant bit rate.
I did left the cd in the tray, started vdj, loaded the cd file in deck 1 and the same song mp3 file in the deck 2, then play them simultaneously using MKII and studio quality senheiser headphones, guess what?
neither I or my audiophile friend couldn't say the differance in a blind folded test, but the sencore showed a litlle quality loss on the mp3 side, the detonation of the sound , the slope of the low to high pasages was slightly different, but no differance in the stereo separation or frequency response.
I have a sencore audio analyzer and few weeks ago I did a test of my own.
first I checked my audio acuity and my good days are gone, I do not hear much over 18khz.
then I did encode a cd using Lame @320 and twiking the quality to max with constant bit rate.
I did left the cd in the tray, started vdj, loaded the cd file in deck 1 and the same song mp3 file in the deck 2, then play them simultaneously using MKII and studio quality senheiser headphones, guess what?
neither I or my audiophile friend couldn't say the differance in a blind folded test, but the sencore showed a litlle quality loss on the mp3 side, the detonation of the sound , the slope of the low to high pasages was slightly different, but no differance in the stereo separation or frequency response.
Posted Sat 05 Jan 08 @ 4:20 pm
Paz75
hear, hear bagpuss... I agree with you m8. All my collection is wav as well, but i actively spin 320k mp3s encoded in the Lame gui. I could also very easily use the wavs in spinning as well because I have 7200rpm drives in raid 0 so the performance is there however i prefer having the drivespace on the lappie for other things. I do however use wavs for making demo mixes.
A few things I'd like to add to your dissertation. It's not only the perception of clarity that you gain with an uncompressed wav, but there are performance factors as well. the fraunhofer encoding algorythmn works very similar to jpeg compression of images. Compressing of imagery deals with analyzing the available number of colors and throwing out the color palette which is not used. For instance, an image with predominantly dark colors will throw out the shading pallette of the lighter shades since they are not in use. Compression such a gif goes even further to do something where it checks colors which exist next to each other. For instance, if the background is black in a picture, the reference to the first X number of rows is one color only up until the first predominant color change. An uncompressed image will refer each pixel and it's color value, whereas gif says 'it's black up until this point' which takes lass space to notate.
Now, in mp3 encoded music, the frequency usage of the track plays a role in the clarity and compression abilities of the track. For instance, an old Jazz track recorded in the 20-40's was done on systems that could not utilize the full frequency spectrum above 22khz. This means that you can easily compress them at 96kpbs and it will sound near identical to the wav equivalent. Take modern Jazz, like Miles Davis who uses lots of reverb on his instruments, including drums, a 128kpbs will sound aliased and you will certainly hear the compression artifacts on the decaying cymbals and hats.
So actually the proper method to compress is to analyze the dynamic range of a track. Incidentally, when you look at the entire frequency spectrum, electronic dance does not use a range as big as classical jazz does, thus you can get away with less to maintain the sound usually.
Now, to the main point I want to make. Using compressed files for playback is one thing, but trying to do operations on them is completely another. By operations, I'm referring to pitching and effects. I dont know the precise algorythm that is used in VDJs sound engine, but in principle it's the same as any other pitching method in plugins and vsts. WHen you speed up or slow down a track, you need to send data quicker or slower through the output. When a track is compressed, you have less to work with when applying this. Similarly, you should all know what happens when you take a small image and make it bigger, it only doubles the available pixels and looks crappy. Resing down in images is the only acceptable way to get a good output.
To make an analogy, consider the music is a pie and the clarity of the music is represented in the number of slices. Say you have an uncompressed wav. This means there are alot of slices in the pie, and so you can give out alot of slices for the music. Slowing down the music means those slices get doubled up and the slices get larger, but you still have some to work with. If the compression rate is low, you have much larger slices of the pie, and pitching down means again, doubling up. If (in analogy only) you have 4 slices to begin with, pitching up leaves you only 3 - 1 slices, you have no ability to maintain clarity of sound.
Now, to complicate things further, mp3 is an older codec where you have other more efficient codecs available. From speaking to a nerdier friend who is an expert on the subject matter, Microsofy seemingly wins (before the advent of the new mp3 pro from fraunhoffer) with the WMA codec. You can apply alot of techniques when using this codec to get a better result with less compression rate. For instance, dual pass encoding works like how Firefox browser loads a large image. You see it pass through quickly once, and a second time, making the resolution finer each time. This will dix any defects you might get in a single pass.
Again, this applies only to listening at the speed it was encoded in. It's important to know that once you DJ with encoded products, you need the extra quality to apply pitching operations and maintain quality. In summary, you will only get this truly with 320 or uncompressed wav, and even then you will notice aliasing in the sound.
The morale of this is that professional djs should maintain a high standard of what they spin with and the recommendation from my side is to use no less than 320kbps encoded mp3s. IF you have the space and the spinrate capacity (your drives are fast) you can use Wav if you want proper quality. To use the argument on lower compressed mp3s and say 'it sounds fine to me' is only your perception and depends much more on how it was listened. If you are in one of the very few clubs in the world that uses the absolute best equipment, and even rarer still, was properly tuned to the accoustics of the room, OR, you are in a properly tuned studio environment with proper nearfield monitors; YOU WILL hear the differences.
Majority of clubs are tuned to loudness and not clarity. The club environment or your DJ headphones are not sufficient to analyze the quality of a recording. Unless you understand the basics of how accoustics and frequency reverberation works, you should just assume that the best way to go is with the largest sized audio file possible.
Some final things to is that the older you get, the less your ear percieves high frequencies. This is a proven scientific fact. Additionally, all humans are different, and all perceptions of the world are different. Some people are born with accute hearing and others not. Therefore, just because you cannot hear something in sound doesnt mean others cannot either. There is also a distinct possibility of having varying degrees of tone-deaf which means the brain is not able to percieve small changes in tonality. At an accute level this means music is very 'blurry' sounding. At small levels, it means you cannot hear the minute aliasing that generally occurs in the decay of cymbals in compressed jazz music.
The final word to summarize is that you should play it safe and use only high quality materials if you expect to call yourself a quality digital jock.
A few things I'd like to add to your dissertation. It's not only the perception of clarity that you gain with an uncompressed wav, but there are performance factors as well. the fraunhofer encoding algorythmn works very similar to jpeg compression of images. Compressing of imagery deals with analyzing the available number of colors and throwing out the color palette which is not used. For instance, an image with predominantly dark colors will throw out the shading pallette of the lighter shades since they are not in use. Compression such a gif goes even further to do something where it checks colors which exist next to each other. For instance, if the background is black in a picture, the reference to the first X number of rows is one color only up until the first predominant color change. An uncompressed image will refer each pixel and it's color value, whereas gif says 'it's black up until this point' which takes lass space to notate.
Now, in mp3 encoded music, the frequency usage of the track plays a role in the clarity and compression abilities of the track. For instance, an old Jazz track recorded in the 20-40's was done on systems that could not utilize the full frequency spectrum above 22khz. This means that you can easily compress them at 96kpbs and it will sound near identical to the wav equivalent. Take modern Jazz, like Miles Davis who uses lots of reverb on his instruments, including drums, a 128kpbs will sound aliased and you will certainly hear the compression artifacts on the decaying cymbals and hats.
So actually the proper method to compress is to analyze the dynamic range of a track. Incidentally, when you look at the entire frequency spectrum, electronic dance does not use a range as big as classical jazz does, thus you can get away with less to maintain the sound usually.
Now, to the main point I want to make. Using compressed files for playback is one thing, but trying to do operations on them is completely another. By operations, I'm referring to pitching and effects. I dont know the precise algorythm that is used in VDJs sound engine, but in principle it's the same as any other pitching method in plugins and vsts. WHen you speed up or slow down a track, you need to send data quicker or slower through the output. When a track is compressed, you have less to work with when applying this. Similarly, you should all know what happens when you take a small image and make it bigger, it only doubles the available pixels and looks crappy. Resing down in images is the only acceptable way to get a good output.
To make an analogy, consider the music is a pie and the clarity of the music is represented in the number of slices. Say you have an uncompressed wav. This means there are alot of slices in the pie, and so you can give out alot of slices for the music. Slowing down the music means those slices get doubled up and the slices get larger, but you still have some to work with. If the compression rate is low, you have much larger slices of the pie, and pitching down means again, doubling up. If (in analogy only) you have 4 slices to begin with, pitching up leaves you only 3 - 1 slices, you have no ability to maintain clarity of sound.
Now, to complicate things further, mp3 is an older codec where you have other more efficient codecs available. From speaking to a nerdier friend who is an expert on the subject matter, Microsofy seemingly wins (before the advent of the new mp3 pro from fraunhoffer) with the WMA codec. You can apply alot of techniques when using this codec to get a better result with less compression rate. For instance, dual pass encoding works like how Firefox browser loads a large image. You see it pass through quickly once, and a second time, making the resolution finer each time. This will dix any defects you might get in a single pass.
Again, this applies only to listening at the speed it was encoded in. It's important to know that once you DJ with encoded products, you need the extra quality to apply pitching operations and maintain quality. In summary, you will only get this truly with 320 or uncompressed wav, and even then you will notice aliasing in the sound.
The morale of this is that professional djs should maintain a high standard of what they spin with and the recommendation from my side is to use no less than 320kbps encoded mp3s. IF you have the space and the spinrate capacity (your drives are fast) you can use Wav if you want proper quality. To use the argument on lower compressed mp3s and say 'it sounds fine to me' is only your perception and depends much more on how it was listened. If you are in one of the very few clubs in the world that uses the absolute best equipment, and even rarer still, was properly tuned to the accoustics of the room, OR, you are in a properly tuned studio environment with proper nearfield monitors; YOU WILL hear the differences.
Majority of clubs are tuned to loudness and not clarity. The club environment or your DJ headphones are not sufficient to analyze the quality of a recording. Unless you understand the basics of how accoustics and frequency reverberation works, you should just assume that the best way to go is with the largest sized audio file possible.
Some final things to is that the older you get, the less your ear percieves high frequencies. This is a proven scientific fact. Additionally, all humans are different, and all perceptions of the world are different. Some people are born with accute hearing and others not. Therefore, just because you cannot hear something in sound doesnt mean others cannot either. There is also a distinct possibility of having varying degrees of tone-deaf which means the brain is not able to percieve small changes in tonality. At an accute level this means music is very 'blurry' sounding. At small levels, it means you cannot hear the minute aliasing that generally occurs in the decay of cymbals in compressed jazz music.
The final word to summarize is that you should play it safe and use only high quality materials if you expect to call yourself a quality digital jock.
Posted Sat 05 Jan 08 @ 9:52 pm
SirWolfy
And that revelation came with just using headphones, duplicate the test with a high-fidelity sound system such as the Bose and you will hear as well as feel the difference in music quality. Lossless audio files is the wave (pun sorta intended) of the future.
I recommend making a master library of wave files (as a backup, etc.) and then converting to another file format such as FLAC or WMA Lossless if you desire to save some HDD real estate. But with the prices of 1TB HDDs plummeting and the sustainable bandwidth/data throughput of SATA, it's pretty much a no-brainer for anyone except those who use a laptop.
I recommend making a master library of wave files (as a backup, etc.) and then converting to another file format such as FLAC or WMA Lossless if you desire to save some HDD real estate. But with the prices of 1TB HDDs plummeting and the sustainable bandwidth/data throughput of SATA, it's pretty much a no-brainer for anyone except those who use a laptop.
Posted Sun 06 Jan 08 @ 12:14 am
djlex
just one mention
in custom industry we have a saying:
has no highs,
has no lows,
must be bose :)
in custom industry we have a saying:
has no highs,
has no lows,
must be bose :)
Posted Sun 06 Jan 08 @ 12:35 am
bagpuss
Thanks for the responses, especially to Paz.
Well I think there is some good news with using WAV files:
1) They Rip faster (no encoding)
2) They load MUCH faster in VDJ (no decompression)
3) WAV is the most "future proof" means we have today in having the highest quality audio tommorow, in some years when the multi-terabite HD is common, the super fast broadband services cheap/common, and high quality audio equipment getting cheaper/better in time, those who've made life easier today by using poor compression or low bite rate MP3's are going to leave themselves in a vunerable position.
4) Once you have WAVs you can do what you like with them, from remixing to copying and encoding to a different format whilst maintaining maximum quality.
And lastly, I think hearing the differences for some requires time and effort (it did with me), and it very much depends on whether you 'want' to hear any differences, if you want an MP3 to sound like WAV it will (in perception), but if you are scanning over the track with the intent of noticing imperfections then you will, the difference is that the 'difference' does exist and can be noticed I feel by most people.
Well I think there is some good news with using WAV files:
1) They Rip faster (no encoding)
2) They load MUCH faster in VDJ (no decompression)
3) WAV is the most "future proof" means we have today in having the highest quality audio tommorow, in some years when the multi-terabite HD is common, the super fast broadband services cheap/common, and high quality audio equipment getting cheaper/better in time, those who've made life easier today by using poor compression or low bite rate MP3's are going to leave themselves in a vunerable position.
4) Once you have WAVs you can do what you like with them, from remixing to copying and encoding to a different format whilst maintaining maximum quality.
And lastly, I think hearing the differences for some requires time and effort (it did with me), and it very much depends on whether you 'want' to hear any differences, if you want an MP3 to sound like WAV it will (in perception), but if you are scanning over the track with the intent of noticing imperfections then you will, the difference is that the 'difference' does exist and can be noticed I feel by most people.
Posted Sun 06 Jan 08 @ 7:00 am
Handyman
Quite simply put, rip your CD's to WAV - 44.1kHz/16 bit stereo files.
If you need more space, convert them to a lossless format, such as .flac.
That way, when your original CD's won't play anymore, you have perfect copies.
If you need more space, convert them to a lossless format, such as .flac.
That way, when your original CD's won't play anymore, you have perfect copies.
Posted Sun 06 Jan 08 @ 1:20 pm
Supacon
Hi Bagpuss, thanks for your post:
I'd like to make a few observations and share some of my opinions.
If anyone here is interested in this kind of subject, the place to go is HydrogenAudio.
I used to spend a lot of time on that forum, and it is frequented by a lot of hardcore audiophiles, and some of the foremost experts in the world on audio encoding and audio fidelity. It is generally a forum for the discussion of audio, audio compression, and computer audio technology.
On Hydrogenaudio, they want to maintain a high standard of scientific accuracy, and it is forbidden by forum rules to make claims as to audio fidelity or performance differences without offering some kind of objective, quantitative evidence of your findings. In the case of comparision of audio codecs, this usually means doing an ABX test.
The reason for this is that psychology plays a very noticeable role in doing a comparison of two things when you know exactly what you are listening to - i.e. Here's an MP3 and here's the source wav - what do you think sounds better?
An easy way of doing a blind ABX test is to get foobar2000 from http://www.foobar2000.org/
When you install it with all the optional components, there is an ABX utility there that you can use. When ABXing, you are presented with an interface that lets you choose between four versions of a file - A & B and X & Y. A is the same as X or Y and B is also one of X or Y, but you don't know which, so this eliminates the psychological problems of doing these tests.
You have to try to identify which sample is X or Y several times, and then when you are done, you are given your results in a nice list.
The reason that I'm going on with all this is that I have rarely ever met an audiophile who can actually consistently ABX differences between 320 Kb/s MP3 and Wav with any type of music using the LAME codec. So I suspect that Bagpuss's observation may be largely psychological.
Granted, there are different implemenations of the MP3 codec and some of them are bad. LAME is known to be among the best implementations, and it is constantly getting tweaked and optimized. Others, like XING, perhaps, are optimized more for encoding speed, and do not fare as well, so it is possible that a very poor encoder would produce results that could be ABXed from source by many people.
Also, about Paz's comment that bigger is better, that is not inherently true. If that were true, you could take a youtube video, demux the audio, and convert the resulting audio into wav and it would sound better than a 128 kb MP3 (which it wouldn't, obviously). Some codecs are able to do an amazing job at making low bitrate files that sounds quite nice - i.e. AAC-HE can achieve 48kb/s files that sound comparable to or better than 96kb/s CBR MP3.
Also, the LAME encoder has a very good VBR mode available than can give you ~192kb/s encodes that sound as good as or sometimes better than a 256kb/s CBR encode. My point is that it's not so much the filesize that determines quality, but it's the encoder and the configuration of it that is most relevant.
Of course, use a lossless format, and do away with all worries of this altogether. As handyman suggested, I actually rip all my CDs losslessly, and enode them into FLAC files. FLAC files tend to take up about 60% of the space a WAV does, and have the added benefit of being taggable, unlike wavs. I have somewhere near 1TB of lossless audio, and I have it all tagged with things like genre, date, BPM, key, artist, and title that allows me to search all my music for anything that might be relevant when I'm looking for a song to play when DJing (BTW, I use foobar2000 for this, from which I can drag directly into VirtualDJ).
When dealing with lossless codecs, there are a few good ones out there, but FLAC is the most popular by far, and although it can take time to encode them, decoding them is pretty fast - on a modern computer the decode time in VDJ or anything else should be negligable.
When doing mobile DJing, I can make a playlist of all my FLAC files and encode them into something smaller so that I can easily fit my collection onto my external hard drive - complete with all my tags. (I tend to like Nero's AAC encoder now, especially since my car stereo supports it - AAC achieves much higher quality at lower bitrates than MP3).
In regards to all of this stuff - one reason why I love VirtualDJ: it can support almost any codec (including flac, aac, and Ogg Vorbis) easily through the use of Windows Media Player general audio decoder. (Although it'd be nice to have native FLAC support built in - flac is open source, after all).
I'd like to make a few observations and share some of my opinions.
If anyone here is interested in this kind of subject, the place to go is HydrogenAudio.
I used to spend a lot of time on that forum, and it is frequented by a lot of hardcore audiophiles, and some of the foremost experts in the world on audio encoding and audio fidelity. It is generally a forum for the discussion of audio, audio compression, and computer audio technology.
On Hydrogenaudio, they want to maintain a high standard of scientific accuracy, and it is forbidden by forum rules to make claims as to audio fidelity or performance differences without offering some kind of objective, quantitative evidence of your findings. In the case of comparision of audio codecs, this usually means doing an ABX test.
The reason for this is that psychology plays a very noticeable role in doing a comparison of two things when you know exactly what you are listening to - i.e. Here's an MP3 and here's the source wav - what do you think sounds better?
An easy way of doing a blind ABX test is to get foobar2000 from http://www.foobar2000.org/
When you install it with all the optional components, there is an ABX utility there that you can use. When ABXing, you are presented with an interface that lets you choose between four versions of a file - A & B and X & Y. A is the same as X or Y and B is also one of X or Y, but you don't know which, so this eliminates the psychological problems of doing these tests.
You have to try to identify which sample is X or Y several times, and then when you are done, you are given your results in a nice list.
The reason that I'm going on with all this is that I have rarely ever met an audiophile who can actually consistently ABX differences between 320 Kb/s MP3 and Wav with any type of music using the LAME codec. So I suspect that Bagpuss's observation may be largely psychological.
Granted, there are different implemenations of the MP3 codec and some of them are bad. LAME is known to be among the best implementations, and it is constantly getting tweaked and optimized. Others, like XING, perhaps, are optimized more for encoding speed, and do not fare as well, so it is possible that a very poor encoder would produce results that could be ABXed from source by many people.
Also, about Paz's comment that bigger is better, that is not inherently true. If that were true, you could take a youtube video, demux the audio, and convert the resulting audio into wav and it would sound better than a 128 kb MP3 (which it wouldn't, obviously). Some codecs are able to do an amazing job at making low bitrate files that sounds quite nice - i.e. AAC-HE can achieve 48kb/s files that sound comparable to or better than 96kb/s CBR MP3.
Also, the LAME encoder has a very good VBR mode available than can give you ~192kb/s encodes that sound as good as or sometimes better than a 256kb/s CBR encode. My point is that it's not so much the filesize that determines quality, but it's the encoder and the configuration of it that is most relevant.
Of course, use a lossless format, and do away with all worries of this altogether. As handyman suggested, I actually rip all my CDs losslessly, and enode them into FLAC files. FLAC files tend to take up about 60% of the space a WAV does, and have the added benefit of being taggable, unlike wavs. I have somewhere near 1TB of lossless audio, and I have it all tagged with things like genre, date, BPM, key, artist, and title that allows me to search all my music for anything that might be relevant when I'm looking for a song to play when DJing (BTW, I use foobar2000 for this, from which I can drag directly into VirtualDJ).
When dealing with lossless codecs, there are a few good ones out there, but FLAC is the most popular by far, and although it can take time to encode them, decoding them is pretty fast - on a modern computer the decode time in VDJ or anything else should be negligable.
When doing mobile DJing, I can make a playlist of all my FLAC files and encode them into something smaller so that I can easily fit my collection onto my external hard drive - complete with all my tags. (I tend to like Nero's AAC encoder now, especially since my car stereo supports it - AAC achieves much higher quality at lower bitrates than MP3).
In regards to all of this stuff - one reason why I love VirtualDJ: it can support almost any codec (including flac, aac, and Ogg Vorbis) easily through the use of Windows Media Player general audio decoder. (Although it'd be nice to have native FLAC support built in - flac is open source, after all).
Posted Sun 06 Jan 08 @ 1:55 pm
Very cool topic , Bagpuss ..........
Ok , so i'm intrigued ..... what were the headphones ?
What are your favorite monitor speakers ? club speakers ??
What are your favorite test tracks for serious equipment tests ??
If I told you my test tracks you'd probably think I'm nuts .... well , if you don't already .... lol
Ok , so i'm intrigued ..... what were the headphones ?
What are your favorite monitor speakers ? club speakers ??
What are your favorite test tracks for serious equipment tests ??
If I told you my test tracks you'd probably think I'm nuts .... well , if you don't already .... lol
Posted Sun 06 Jan 08 @ 3:52 pm
Handyman
Just a note: WAV's can have tags - however, most programs & taggers do not implement them. The tags are stored at the end of the files rather than in the file's header. Programs such as Easy CD-DA Extractor support WAV tags.
Since support for WAV id3 tags is inconsistent (and perhaps nonstandard), I find converting everything to FLAC is the way to go. When I rip, I rip to WAV (~3mins to rip a full CD), then batch convert all my day's rips overnight to FLAC (compression level 8). Easy CD-DA extractor makes this very easy.
Another benefit of FLAC files is that they seem to load faster into VDJ than MP3's of the same files. I have my external drives attached via firewire rather than USB and this also seems to speed the loading up.
Since support for WAV id3 tags is inconsistent (and perhaps nonstandard), I find converting everything to FLAC is the way to go. When I rip, I rip to WAV (~3mins to rip a full CD), then batch convert all my day's rips overnight to FLAC (compression level 8). Easy CD-DA extractor makes this very easy.
Another benefit of FLAC files is that they seem to load faster into VDJ than MP3's of the same files. I have my external drives attached via firewire rather than USB and this also seems to speed the loading up.
Posted Sun 06 Jan 08 @ 4:17 pm
jimmy b
Nice work everybody, and thanks makes good reading.
Seeing more and more of these FLAC files convertion, they're getting quite a following.
Jimmy b
Posted Sun 06 Jan 08 @ 5:04 pm
Supacon
I really think FLAC is the way to go, unless you're really chinsy about your hard drive space...
I think it's reasonable to expect anyone serious about DJing to have a desktop system with a couple of 500GB hard drives or whatever (they are dirt cheap, best bang for the buck right now), and then you can put all your FLACs on there... encode them to smaller files if you need to fit them all in a laptop hard drive or the like.
If anyone would be interested, later I could make a tutorial for how I rip and convert between different audio formats, and how I have foobar configured.
I use foobar for most of my file management stuff, because there are lots of components you can get to handle just about any kind of tagging or file management task. I also use foobar sometimes for searching for music... if VirtualDJ supported FLACs properly with tags, then that wouldn't be necessary so much - but the browser in foobar has come a ways from where it was in 3.4!
I think it's reasonable to expect anyone serious about DJing to have a desktop system with a couple of 500GB hard drives or whatever (they are dirt cheap, best bang for the buck right now), and then you can put all your FLACs on there... encode them to smaller files if you need to fit them all in a laptop hard drive or the like.
If anyone would be interested, later I could make a tutorial for how I rip and convert between different audio formats, and how I have foobar configured.
I use foobar for most of my file management stuff, because there are lots of components you can get to handle just about any kind of tagging or file management task. I also use foobar sometimes for searching for music... if VirtualDJ supported FLACs properly with tags, then that wouldn't be necessary so much - but the browser in foobar has come a ways from where it was in 3.4!
Posted Sun 06 Jan 08 @ 5:29 pm
jimmy b
Can someone point me in the right direction regarding FLAC and good software to use to convert to FLAC,
want to try ity out myself.
Jimmy b
Posted Sun 06 Jan 08 @ 6:05 pm
Paz75
Supacon, you missed my point. I was specifically talking about larger files being better for the usage of vdj because we are doing pitch operations. you need high granularity to get good quality output after the operation. Same reason why producers are doing natural recordings at 96khz and higher where possible instead of 48, or 48 instead of 44khz... It gets even more complicated when you go from a 16 bit header to a 24 bit header which can capture reverbs at higher clarity.
A similar argument goes for film which uses high quality film and essentially throws it all away in the editing room to package for cinema and dvd.
While in essence I completely agree with what you said, your standpoint can only work for audio which is played 'as is'. compression techniques are for commercial listeners as they will playback with no changes on top. DJs obviously cannot do this and need to consider other things.
A similar argument goes for film which uses high quality film and essentially throws it all away in the editing room to package for cinema and dvd.
While in essence I completely agree with what you said, your standpoint can only work for audio which is played 'as is'. compression techniques are for commercial listeners as they will playback with no changes on top. DJs obviously cannot do this and need to consider other things.
Posted Sun 06 Jan 08 @ 8:08 pm
kaoz99
chucknorrisyouwimps wrote :
Very cool topic , Bagpuss ..........
Ok , so i'm intrigued ..... what were the headphones ?
What are your favorite monitor speakers ? club speakers ??
What are your favorite test tracks for serious equipment tests ??
If I told you my test tracks you'd probably think I'm nuts .... well , if you don't already .... lol
Ok , so i'm intrigued ..... what were the headphones ?
What are your favorite monitor speakers ? club speakers ??
What are your favorite test tracks for serious equipment tests ??
If I told you my test tracks you'd probably think I'm nuts .... well , if you don't already .... lol
what were the headphones ? - Sennheiser HD 280 Pro
What are your favorite monitor speakers ? club speakers ?? JBL 300w Powered Speakers for Monitor, Club Speakers dependant on venues choice. Personal Choice would be the top end Mackies
What are your favorite test tracks for serious equipment tests ?? Original Version of Black Box - Ride on Time (and not skippy the bush kangaroo either Terry ) :-P
Cheers
Kaoz
Posted Sun 06 Jan 08 @ 9:17 pm
Paz75
An awesome track for testing is Maggot Brain by Parliament Funkadelic. Its got a lot of line noise in the recording, but its got a huge timbral range and very spacial. You really bring out the colors in any system playing this track. Rumor has it was played by a 16 year old kid for George Clinton
Posted Sun 06 Jan 08 @ 9:49 pm
jimmy b
Don't laugh, but I often find certain types of Classical music is a good bed to test on.
Jimmy b
Posted Sun 06 Jan 08 @ 10:43 pm
Tear Em 'Up
jimmy b wrote :
Don't laugh, but I often find certain types of Classical music is a good bed to test on.
Jimmy b
Don't laugh, but I often find certain types of Classical music is a good bed to test on.
Jimmy b
Second that....lots of different tones and sounds in classical music.
Posted Sun 06 Jan 08 @ 10:50 pm
bagpuss
jimmy b wrote :
Don't laugh, but I often find certain types of Classical music is a good bed to test on.
Jimmy b
Don't laugh, but I often find certain types of Classical music is a good bed to test on.
Jimmy b
I never laugh at real music, I laugh at the crap that frequents the charts if anything.
Posted Mon 07 Jan 08 @ 5:24 am
