wldrns1 Posted November 2, 2014 Share Posted November 2, 2014 (edited) This database refers to CD's only, correct? I see the word vinyl in individual listings like Dire Straits....'Brothers In Arms [VINYL]'. Pretty basic I know but confusing. If not CD's only, then what other formats? I have Googled some of the file types to learn what I can. Maybe my terminology needs work as well! Edited November 2, 2014 by wldrns1 Quote Link to comment Share on other sites More sharing options...
Chris A Posted November 2, 2014 Share Posted November 2, 2014 This database refers to CD's only, correct? Where did you get that idea, if I might ask? The DR Database is for any digital track/album, even ones ripped by people from their vinyl records on their own turntables using ADCs, such as this one. There are a many digital file formats and media (physical media) formats that can be in the DR Database: http://en.wikipedia.org/wiki/Timeline_of_audio_formats Quote Link to comment Share on other sites More sharing options...
Schu Posted November 2, 2014 Share Posted November 2, 2014 I find that site very interesting... I have yet to confirm the review of the LP's I own, but I plan on going thru it somewhat before accepting anything as gospel. But seeing as I have as many as 5-6 versions of some LP's/disc's, it would be nice to be able to pair it down before making a purchase. Quote Link to comment Share on other sites More sharing options...
Ski Bum Posted November 2, 2014 Share Posted November 2, 2014 I've noted that the dynamic range of even some of the compressed recordings doesn't prevent me from getting fairly wide dynamic swings in playback (as measured by REW in room). The disparity is kind of confounding. If there is only 6db of dynamic range on the recording, and I monitor using REW or some other rta measurement device that's providing 20+ db difference between average and peak in actual acoustical measurements, I'm not sure exactly how to interpret the info on the database. Quote Link to comment Share on other sites More sharing options...
mustang guy Posted November 2, 2014 Share Posted November 2, 2014 Dynamic Range Database Quote Link to comment Share on other sites More sharing options...
Chris A Posted November 2, 2014 Share Posted November 2, 2014 (edited) I've noted that the dynamic range of even some of the compressed recordings doesn't prevent me from getting fairly wide dynamic swings in playback (as measured by REW in room). The disparity is kind of confounding. If there is only 6db of dynamic range on the recording, and I monitor using REW or some other rta measurement device that's providing 20+ db difference between average and peak in actual acoustical measurements, I'm not sure exactly how to interpret the info on the database. The Dynamic Range "TT" meter is calculating crest factor (and that's what all the DR Database ratings are reporting--not dynamic range). The difference between crest factor and dynamic range, of course, is the difference between the average (per track) and peak, vs. the difference between the softest and loudest. If you think about it, for what the algorithm is measuring, it was designed to identify just how much the peaks on each track have been squashed down to average level, so the measurement is directly applicable to the amount of damage applied by a multi-band compressor to the finished product. The idea is that if they can squash the peaks down, they can raise the average levels up (don't you feel good now about understanding what their goal really is..? LOUDER mentality only). It's not ever been about measuring "downwards compression" which raises the quietest parts up to average. But in reality, this type of compression is also happening on squashed recordings. Edited November 2, 2014 by Chris A Quote Link to comment Share on other sites More sharing options...
mustang guy Posted November 2, 2014 Share Posted November 2, 2014 IMO, there needs to be another database with many more fields. Another thing is that the source is mostly 'unknown'. Also, the singles are listed as 'Album'. What's with that? IMO, we could together come up with a database which would be sortable on a field called 'recommend', which would be based on objective data of the track. Is there currently an algorithm which can show the true dynamic range of a recording rather than just the crest factor? If we need to invent these things here, so be it. I think it would be a great project. If there is already such an animal, I for one, would love to know how to get that information. Quote Link to comment Share on other sites More sharing options...
Chris A Posted November 2, 2014 Share Posted November 2, 2014 (edited) I remember when I found the DR Database a couple of years ago and was really delighted that anyone would have the foresight to accomplish such a milestone in recordings. Imagine...now we can see just how much our recordings have been compressed from the beginning, from 1991, and how much our "remastered" recordings really equate to "further compressed". And the entire database has been crowdsourced--otherwise, I firmly believe that it would never have happened. As it is I see some spin occurring by some of the record company producers to talk over the top, ignore or downplay, or even to downright lie about their "audiophile quality" masterings. I've dumped my "remastered recordings" to go for the discs by version number and producer (usually produced before 1991) that have the highest DR ratings, and have had a 95+% success rate in regaining and surpassing the original vinyl recording DR ratings, and have now been hearing things on these older popular recordings that I've never heard before - truly thrilled. But you make a good point: it would take little to expand the database. By the way, you can already see the approximate dynamic range of some of the tracks by adding the DR ratings by track to their RMS values for those album DR Rating records having an "i" icon with blue background by clicking on the record and reading the in the comments or log section. The problem that I currently have with the DR Database is the fact that they actually hide that information in their database records - so a few of us have been duplicating the information in the comment field so that everyone can see it--not just the database owners. The reason why there are so many records that haven't listed a source medium (CD, DVD-A, SACD, download, etc.) is due to the recent change in the database this year to include those sources, and the fact that it is an optional field--that a lot of people simply ignore when they post their album DR ratings. Edited November 3, 2014 by Chris A Quote Link to comment Share on other sites More sharing options...
mustang guy Posted November 3, 2014 Share Posted November 3, 2014 Thanks Chris, I didn't know about the ratings comments, etc. I wonder if there is a SQL link they could share for pulling that data from the database. I'm sure having server side scripting had the intent of keeping data safe, but so does locking the database for queries only. Quote Link to comment Share on other sites More sharing options...
mustang guy Posted November 3, 2014 Share Posted November 3, 2014 Hey I have a question. I just checked out the Dire Straights Brother in Arms SACD on the database, of which there are a number and they were all downmixed with foobar to 2.0 from 5.1. My question is this. Does it sound as good as the 5.1, not considering of course that there are fewer channels. Is the sub bass as pronounced? I have listened to the CD, the LP and the SACD, and the SACD was far superior. I would rip the SACD to 2.0 and cut a CD for my truck if I thought it sounded better than the CD. Quote Link to comment Share on other sites More sharing options...
muel Posted November 3, 2014 Share Posted November 3, 2014 (edited) Interesting... my 1985 CD easily sounds better than the SACD and my DR numbers agree. It would be an interesting test to hear how your rip for your truck works out. Is it possible there are different releases of the SACD? I have a hybrid SACD. The CD layer is even more compressed than the SACD layer. Edited November 5, 2014 by muel Quote Link to comment Share on other sites More sharing options...
wldrns1 Posted November 3, 2014 Author Share Posted November 3, 2014 This database refers to CD's only, correct?Where did you get that idea, if I might ask? The DR Database is for any digital track/album, even ones ripped by people from their vinyl records on their own turntables using ADCs, such as this one. There are a many digital file formats and media (physical media) formats that can be in the DR Database: http://en.wikipedia.org/wiki/Timeline_of_audio_formats I have (had) no idea. Seeing different formats that couldn't be vinyl so...figured I better ask. Understand better now. As for wiki, I'm sure there are pro/con opinions regarding their discussion...another topic for others on some other day. Here though, learning as I go. Quote Link to comment Share on other sites More sharing options...
mustang guy Posted November 3, 2014 Share Posted November 3, 2014 Interesting... my 1985 CD sounds easily sounds better than the SACD and my DR numbers agree. It would be an interesting test to hear how your rip for you truck works out. Is it possible there are different releases of the SACD? I have a hybrid SACD. The CD layer is even more compressed than the SACD layer. That is interesting. It makes me want the 1985 publication of the CD! I don't have the 1985 CD. I have the CD tracks on the Hybrid, the 5.1 cuts, and the Vinyl. Quote Link to comment Share on other sites More sharing options...
Chris A Posted November 3, 2014 Share Posted November 3, 2014 (edited) Try this one: http://smile.amazon.com/Brothers-Arms-Dire-Straits/dp/B000002L7G/ from 1989: $4 (US) delivered to your doorstep (used), or $3.35 if you've got Amazon Prime. (Any CD version issued before 1991 is virtually always not compressed relative to the master tape. I'm told that 1991 was the year that multi-band compressors came into widespread use by mastering engineers.) I recently received a Tres Hombres CD from 1990 to replace the 2006 "remastered" version. The 2006 version is the one that they push in your face everywhere you look. The difference was like that between the living and the dead: I couldn't believe what I was hearing for the first time--and I've owned the vinyl version since the mid-70s. An amazing difference. Cheap, too ($4.73 delivered). It has the same DR values as the 1988 version: foobar2000 1.3.3 / Dynamic Range Meter 1.1.1log date: 2014-11-03 07:35:40--------------------------------------------------------------------------------Analyzed: ZZ Top / Tres Hombres--------------------------------------------------------------------------------DR Peak RMS Duration Track--------------------------------------------------------------------------------DR13 -0.40 dB -15.27 dB 2:53 01-Waitin' For The BusDR14 -0.40 dB -15.53 dB 3:32 02-Jesus Just Left ChicagoDR13 -0.40 dB -15.64 dB 3:27 03-Beer Drinkers & Hell RaisersDR12 -0.41 dB -14.78 dB 3:31 04-Master Of SparksDR14 -0.40 dB -16.30 dB 3:18 05-Hot, Blue And RighteousDR13 -0.40 dB -14.27 dB 2:31 06-Move Me On Down The LineDR13 -0.41 dB -14.38 dB 3:10 07-Precious And GraceDR13 -0.40 dB -16.14 dB 3:54 08-La GrangeDR15 -0.40 dB -17.14 dB 4:05 09-ShiekDR13 -0.40 dB -15.36 dB 3:13 10-Have You Heard--------------------------------------------------------------------------------Number of tracks: 10Official DR value: DR13Samplerate: 44100 HzChannels: 2Bits per sample: 16Bitrate: 1411 kbpsCodec: CDDA================================================================================ Here is the 2006 version--notice how the crest factor has been squashed (the DR column) AND the average loudness level (i.e., the RMS column) been increased: loud discs are just that--louder (sometimes by as much as 12 dB louder than earlier versions before 1991): foobar2000 1.3.3 / Dynamic Range Meter 1.1.1log date: 2014-11-03 07:45:19--------------------------------------------------------------------------------Analyzed: ZZ Top / Tres Hombres--------------------------------------------------------------------------------DR Peak RMS Duration Track--------------------------------------------------------------------------------DR10 0.00 dB -11.21 dB 2:53 01-Waitin' For The BusDR9 0.00 dB -10.74 dB 3:30 02-Jesus Just Left ChicagoDR10 0.00 dB -11.22 dB 3:26 03-Beer Drinkers & Hell RaisersDR9 0.00 dB -9.98 dB 3:31 04-Master Of SparksDR11 -0.03 dB -12.27 dB 3:18 05-Hot, Blue And RighteousDR9 -0.02 dB -10.07 dB 2:32 06-Move Me On Down The LineDR9 0.00 dB -9.82 dB 3:10 07-Precious And GraceDR10 0.00 dB -12.30 dB 3:53 08-La GrangeDR12 0.00 dB -13.90 dB 4:07 09-ShiekDR10 -0.03 dB -11.58 dB 3:15 10-Have You Heard?DR9 -0.22 dB -10.09 dB 2:42 11-Waitin' For The Bus (Live)DR9 -0.25 dB -10.39 dB 4:03 12-Jesus Just Left Chicago (Live)DR9 -0.15 dB -10.81 dB 4:44 13-La Grange (Live)--------------------------------------------------------------------------------Number of tracks: 13Official DR value: DR10Samplerate: 44100 HzChannels: 2Bits per sample: 16Bitrate: 1411 kbpsCodec: CDDA================================================================================ The extra 3 to 5 dB of extra crest factor makes all the difference in the world. http://dr.loudness-war.info/album/list/year?album=Tres+Hombres Chris Edited November 3, 2014 by Chris A Quote Link to comment Share on other sites More sharing options...
Schu Posted November 3, 2014 Share Posted November 3, 2014 is there any discussion on how these things are being measured? Quote Link to comment Share on other sites More sharing options...
Chris A Posted November 3, 2014 Share Posted November 3, 2014 Are you looking for discussion here or from other sites? Quote Link to comment Share on other sites More sharing options...
Schu Posted November 3, 2014 Share Posted November 3, 2014 where ever... there is less technical discussion here. I guess I am wanting to know how accurate and quantifiable these measurements are or if there is a +/- error rate? Quote Link to comment Share on other sites More sharing options...
Chris A Posted November 3, 2014 Share Posted November 3, 2014 (edited) The average (RMS) values are pretty accurate due to the number of samples taken per track (all samples, in fact). The peak value captured for each track is the statistical second-highest measured per that track. It is a single value but one that is gathered from all the samples on each track, so if the second-highest peak is corrupted, then actually the third-highest peak will replace it. etc. Noting that a sample represents 1/20050th of a second duration on each CD track, if one peak is missed, or discarded due to parity error, then the immediately succeeding high samples on that cum-distribution queue will statistically be that high, too, or within a very small SPL from the second-highest value. If you plot the values for each track and arrange them in cumulative fashion, you'll always see that familiar "S" shaped curve of a cumulative distribution, which tells you that you'll see many more samples at or near the actual second highest value. Note that all the samples on a digital track represent numbers - so there aren't any errors within the DR Meter application unless the track's samples are severely corrupted. If that's the case, then the player won't play the track and the DR Meter application will report an error and move onto the next track. As far as the validity of the measurement, in my experience, the measurement has been very, very useful and essentially always tells me what I need to know. The only area that I see where it gets into trouble is when noise has deliberately been added to the track. This happens with some mastering engineers that obey their music producers/artists to keep increasing loudness and adding "grit" to the track. Usually, the DR ratings for that track are in the 1-6 range (on the DR scale), so I don't really care about these tracks anyway. There is an issue with using the DR meter on ripped vinyl tracks, but that issue results in tracks being reported with higher DR ratings, not lower--in fact, it typically reads about 3dB too high for ripped vinyl tracks. For lossy music (mp3, etc.) I've found that the DR ratings are typically reported as too high, depending on the parameters set for the LPCM --> mp3 translation application--so I always ignore any mp3 (lossy) tracks being reported, as they can be in excess of 6 dB too high in reported DR ratings. Did that address some of your technical needs on this subject? Chris Edited November 3, 2014 by Chris A Quote Link to comment Share on other sites More sharing options...
mustang guy Posted November 3, 2014 Share Posted November 3, 2014 Very helpful information, thank you. I am going to start looking at press dates on CD's. I wanna get a few... I do have Amazon Prime. Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.