Question reference CD Volume [Archive] - Audio & Video Forums

PDA

View Full Version : Question reference CD Volume



Dawgincontrol
12-08-2005, 08:14 AM
Is there a receiver/preamp/amp or a cd player/DAC that balances the volume between various cd's (similar to what a burn program does with volume maximizing)?

I do playlists through my computer and various cd's have volume levels that drive me crazy. This includes store-bought and burned cd's. I don't care about the price as I am upgrading throughout, but wonder if any maker addresses this. Volume change as I listen to Christmas music this morning is distracting. For instance, the "Chicago" Christmas CD seems louder than others.

Any help would be appreciated!!

Resident Loser
12-08-2005, 01:06 PM
...that's why there's a volume control on your remote...or at least some sort of control on your head unit...

Something like that would be costly to implement and definitely non-audiophile in nature...heck, some of 'em won't use/don't have tone controls on their gear much less something that obtrusive...you basically want auto-gain riding. Most pre/control centers can't even balance the various sources; something the nature of your request has too many software variables to be simple and/or cost effective...it goes beyond that, but that's pretty much it in a nutshell.

jimHJJ(...too many bells and whistles already IMO...)

Woochifer
12-08-2005, 01:18 PM
Onkyo has their Intellivolume feature on their receivers, which purportedly equalizes the levels between the different sources. But, I'm not sure if the feature works at equalizing the level between different CDs.

Generally, the trend right now is to bump up the levels as high as possible at the mastering stage with pop recordings, and compress the recording so that the playback level can stay consistently high without the peaks pushing out beyond the digital zero limit. But, with CDs there are no established standards on how high or low the levels should go. You generally have more consistency with the prevailing levels on DVD soundtracks.

Otherwise, the only recourse you have with playlists at least is to use the volume equalizing features that are built into several of the CD ripping/burning programs out there.

Dawgincontrol
12-10-2005, 08:02 AM
...that's why there's a volume control on your remote...or at least some sort of control on your head unit...

Something like that would be costly to implement and definitely non-audiophile in nature...heck, some of 'em won't use/don't have tone controls on their gear much less something that obtrusive...you basically want auto-gain riding. Most pre/control centers can't even balance the various sources; something the nature of your request has too many software variables to be simple and/or cost effective...it goes beyond that, but that's pretty much it in a nutshell.

jimHJJ(...too many bells and whistles already IMO...)


Changing the volume manually is not what I'm looking for. Going from one source cd to another while playing a music list makes for a wish for balance. It really seems to me that there would be some way to scan each cd and go from memory and make it workable. Or a maximum gain switch an an amp. Or some computer hook-up that remembers the volume of each cd. I don't think it would be cost prohibitive. It is all algorhythms anyway.

Plus, we can dream can't we? Maybe nobody has just ever thought of it.

RJW1138
12-10-2005, 08:21 AM
What you're looking for is a device to normalize all your CD audio, and it's actually not a bad idea. As for as I know though, there's no such product on the market.

The product would pretty much have to be encorporated into a CD player, because as you already know, it's not a real-time process. It has to be able to read through the track or disc ahead of time before starting to play. The product would have to scan through the track, or album (possibly a setting), and determine what the maximum or average level is (again, probably a setting), and then adjust the output gain accordingly (just selecting the single gain amount necessary to bring up or down the disc's average/maximum level). In the end, everything that gets played back has either the same average or maximum level.

I disagree about what Resident Loser said, such a product is not non-audiophile in nature, and could provide the exact same fidelity as without.

If you wanted to do this in real-time without knowing the data that's coming, you're looking at a dynamic range compressor, but that's a different beast.

Anyways, bottom line is, I don't know of one that exists, but never stop dreaming Dawg!

Resident Loser
12-12-2005, 07:53 AM
I disagree about what Resident Loser said, such a product is not non-audiophile in nature, and could provide the exact same fidelity as without.

If you wanted to do this in real-time without knowing the data that's coming, you're looking at a dynamic range compressor, but that's a different beast.

Foist, the entire quote is:

"Something like that would be costly to implement and definitely non-audiophile in nature...heck, some of 'em won't use/don't have tone controls on their gear much less something that obtrusive..."

To some, anything beyond the minimalist mindset is verboten...hence the "tone control" bit, filters, equalizers anything beyond gold-plated inputs/outputs, OFC single-crystal internal wiring, hospital grade/mil spec(or better) electronics is considered "non-audiophile" by some. Someone who has spent the equivalent of the GNP of a developing country on a hand-built SET amp and passive attenuator is not in the least interested in the circuitry required to balance the volume level of discs...

Any sort of compression, single ended processing, companding, auto-anything will have the possibility of sonic artifacts, whether it be directly on the signal as in "pumping" or "swishing" or some sort of digital hash from the controlling circuit.

What would be done for a source that contains a varied dynamic range...for example the "1812 Overture" which ends with repeated firing of canon? You would need a circuit that could cataloge and identify EVERY disc(which would require it to "listen" in real-time and arbitrarily determine the overall volume which depends on dynamic range) otherwise it's a simple compressor...a definite audiophile no-no...and that's only for YOUR source material...Who sets the parameters? The listener? The processor? AGCs and the like have been shunned historically by recordists who favored manual gain riding.

I'm not saying it can't or shouldn't be done...I mean some folks are happy with the "fidelity" of what sounds like digital cr@p to others simply for the sake of convenience...so go figure. It's just not an easy thing to do and with little or no consumer weeping and gnashing of teeth over it or similar "audiophile" outcry, I think it will be a long time coming.

jimHJJ(...market demand drives the bells and whistles...)

RJW1138
12-12-2005, 11:53 AM
You're confusing me RL. I'm not talking about COMPRESSING, I'm just talking about a way of NORMALIZING the output through a fixed gain amount. I even mention a compressor, and say that that's not what I'm talking about, and you even quoted that. You should read my post again.

What I'm getting at is that a gain adjustment at the end of your "CD Player w/Normalization" does not inherently degrade sound quality. All CD players already have amp circuitry after the DAC stage to bring the level up to some standardized level (usually 2v). This could use a variable amount of gain (on a per song or per album basis, not continuously changing!) to adjust that level accordingly, instead of just being a fixed amount, to normalize the song or album up to either the desired average or maximum level.

Bottom line, I don't think you understood my post.

SlumpBuster
12-12-2005, 03:37 PM
This is an old question applied to new technology. The volume control technology exists, although I won't pretend to have ever seen it or know what it is. You probably heard it today. A buddy of mine in college who worked the radio station told me commercial stations use a "black box" to prevent volume spikes and match levels between cd's. But that was a long time ago. Most stations don't even use cd's anymore.

Resident Loser
12-14-2005, 07:41 AM
...in my response, anywho...


...Bottom line, I don't think you understood my post...

I did, but I don't think you understood my responses...isn't that special? It's like an O Henry story, how appropriate for the season!

Let's start with a few questions...How should this be implemented? Do we raise those with low volume, or lower those with higher volume? Or do we pick some arbitrary "average"?

Any such scheme would have to be implemented to the analog output as I believe there are far too many fixed parameters and other considerations involved otherwise. Do keep in mind that most transports and other circuitry in all but the most "esoteric" of components are pretty much "off-the-shelf" and on a minimal number of chips.

OK, so we put our "device" at the tail end of the signal path. How do we "control" it or allow it to control itself?

Can't be done in a "real-time" fashion (or as close to "real-time" as the medium permits)...sampling the opening level of any disc isn't necessarily indicative of what is to follow...Again, and I hesitate to fall back on the classics, but something like "Bolero" starts out sotto voce' and ends with a timpanic crash...or the reverse, "Carmina Burana" comes out like gangbusters and then the chorus begins in allmost hushed, whispering tones.

In order for it not to be some sort of companding process, a complete "reading" of the disc would be required to find a happy medium...I'm not sure if the process could be sped up; how long it would take the error correction to do it's thing, buffers to load and dump, etc., etc. in order to be evaluated? And what about the most basic of elements: transport speed? Each new disc, on first playing, would have to be analyzed by some sort of processor,it's TOC cataloged and the appropriate playback level tagged to it. Does it require new circuitry to do this...and additional memory to file the info...and time for the circuitry to compare the inserted disc(s) to it's database and do what is required?...This will add to the cost and again, without the hue and cry from consumers, will the mfrs. really give a rat's @$$?

Of course, it would be nice if the point was moot and all media was engineered within certain parameters...but I'd venture a guess there's about as much chance of that as anything else we've discussed...and for the same reason.

jimHJJ(...hope my POV is a bit more clear...)

RJW1138
12-14-2005, 11:55 AM
...in my response, anywho...



I did, but I don't think you understood my responses...isn't that special? It's like an O Henry story, how appropriate for the season!

Let's start with a few questions...How should this be implemented? Do we raise those with low volume, or lower those with higher volume? Or do we pick some arbitrary "average"?

Any such scheme would have to be implemented to the analog output as I believe there are far too many fixed parameters and other considerations involved otherwise. Do keep in mind that most transports and other circuitry in all but the most "esoteric" of components are pretty much "off-the-shelf" and on a minimal number of chips.

OK, so we put our "device" at the tail end of the signal path. How do we "control" it or allow it to control itself?

Can't be done in a "real-time" fashion (or as close to "real-time" as the medium permits)...sampling the opening level of any disc isn't necessarily indicative of what is to follow...Again, and I hesitate to fall back on the classics, but something like "Bolero" starts out sotto voce' and ends with a timpanic crash...or the reverse, "Carmina Burana" comes out like gangbusters and then the chorus begins in allmost hushed, whispering tones.

In order for it not to be some sort of companding process, a complete "reading" of the disc would be required to find a happy medium...I'm not sure if the process could be sped up; how long it would take the error correction to do it's thing, buffers to load and dump, etc., etc. in order to be evaluated? And what about the most basic of elements: transport speed? Each new disc, on first playing, would have to be analyzed by some sort of processor,it's TOC cataloged and the appropriate playback level tagged to it. Does it require new circuitry to do this...and additional memory to file the info...and time for the circuitry to compare the inserted disc(s) to it's database and do what is required?...This will add to the cost and again, without the hue and cry from consumers, will the mfrs. really give a rat's @$$?

Of course, it would be nice if the point was moot and all media was engineered within certain parameters...but I'd venture a guess there's about as much chance of that as anything else we've discussed...and for the same reason.

jimHJJ(...hope my POV is a bit more clear...)

OK, good, it looks like you do understand what I was getting at. I wasn't really interested in discussing how such a device would be implemented, because I don't really see much point in doing that. I just wanted to make it clear that ultimately, such a device could use a variable gain amount to normalize all played discs or tracks to a certain average or maximum level, and such a device would not necessarily incur any fidelity loss over what traditional players experience, due to all players needing a gain stage after the DAC stage anyways. I believe I've accomplished this.

And yes, I completely do understand about how songs can have a large dynamic range, like the "1812 Overture", "Bolero" and "Carmina Burana". Of course I do, and I wouldn't be able to conceive of how such a product would work unless I understood this basic principle of audio.

Anyways, just to have fun with this, here's how I figured it would work:

The user would select if they want all tracks to be normalized, or all discs to be normalized (in this case, the relative loudness differences between songs would be preserved, but in the former case it would be lost...the latter may be desired as albums tend to be created with a certain flow and coherence and their relative level differences is important...). If they chose a per track basis, then the user would also determine if they want to normalize to a common average level, or a common maximum level. This would purely be a preference. They would choose average level if they wanted all tracks to sound as loud on average, but this would involve accepting that some would have louder peaks than others. Maybe the user would choose the maximum level option if they were concerned about having these loud peaks. On a per disc level, you could only normalize to a common maximum level, because without knowing the average-to-maximum level ratios of every disc you could ever play, you couldn't do this because you could end up overloading your output if you chose an average level that was too high.

So, on a per track level, if you were normalizing to a common maximum level, the player would just need to scan one track at a time, determine the maximum level, adjust the output gain, and play it. If you were normalizing to a common average level, then it would need to scan every track ahead of time, determine the average for all of them and remember that, then start playing back, adjusting back each track by the appropriate amount as it plays it. Obviously the player's transport would need to be a custom job to be able to pull this off. It would need to be really fast, like a CD-ROM transport, and obviously for scanning the whole disc, it would still take a long time. This is what brings up the possibility of having memory and retaining this data for your discs so that it doesn't always have to waste this time scanning the disc everytime you want to play it. Anyways, this whole implementation is totally up for debate, and could go different ways.

So, given your settings, and after it's read through the track or disc and determined the information that it needs to, it then adjusts the output gain, and plays back the track or disc, and that's pretty much it.

Here's some examples:

Let's say there's a disc with 3 tracks, and the maximum/average levels are
1. 0dB/-15dB
2. -5dB/-30dB
3. -10dB/-40dB

"K" is the amount of gain necessary to drive the output to a certain standard, like the Redbook 2V RMS.

1. The user wants to do maximum level normalization on a per track basis. The player reads the first track and gets a maximum level of 0dB. It plays the track back at gain K. It reads the next track and gets a max level of -5dB. It plays the track back at gain K + 5dB. It reads the 3rd track and gets a max level of -10dB so it plays the track back at a gain of K + 10dB. In the end, all tracks are played back at a peak level of 0dB, but their average levels are inherently different. You get 0dB/-15dB, 0dB/-25dB, 0dB/-30dB.

2. The user wants to do average level normalization on a per track basis. The player reads the average and maximum levels from all the tracks ahead of time and stores the values. Since you have to go with the one that has the greatest max/avg ratio so that your maximum peak level is 0dB, you go with the 3rd track, which is a difference of 30dB. The player plays track 1 at gain K - 15dB, track 2 at gain K, and track 3 at gain K + 10dB. In the end, all tracks are played back with an average level of -30dB, but their peak levels are inherently different. You get -15dB/-30dB, -5dB/-30dB, 0dB/-30dB.

As you can see, no compression has occurred, and the dynamic range of each song is left untouched. All that we did was adjust the gain to accomplish what we wanted: tracks that all end up sounding as loud as each other.

On a per disc level, it can only do maximum level normalization. It reads through the disc and whatever the maximum level of the disc is, it adjusts the output gain up by that much so that the maximum level is 0dB. Simple. In our example, it does nothing, since one of the songs hits 0dB. It plays back at gain K. If you played a second disc whose max level was -6dB, it would play that disc back at gain K + 6dB. These two discs would have the same maximum loudness.

Phew. Ok, that was a pretty much a huge waste of time, but fun nonetheless. If anyone actually read through this whole thing, you need to do something better with your time :-P.

hermanv
12-17-2005, 01:53 PM
I own an Alesis Master Link. It allows volume offsets for each track in a playlist.

1. You record your offending Christmas CD into a "playlist" on the Alesis hard disk.
2. You use the built in peak finder and level meters to determine the gain error per track.
3. You set a gain compensation for that track (0.1 dB increments).
4. When all the songs in a playlist are compensated you push "make CD"
New CD with all tracks compensated but otherwise identical to source disk is burned.

Problem, takes hours to do. Master link recorder is not cheap, otherwise works great.

bloosqr
12-17-2005, 09:42 PM
Is there a receiver/preamp/amp or a cd player/DAC that balances the volume between various cd's (similar to what a burn program does with volume maximizing)?

I do playlists through my computer and various cd's have volume levels that drive me crazy. This includes store-bought and burned cd's. I don't care about the price as I am upgrading throughout, but wonder if any maker addresses this. Volume change as I listen to Christmas music this morning is distracting. For instance, the "Chicago" Christmas CD seems louder than others.

Any help would be appreciated!!

I'm a little confused by the question and/or the responses to the question which make this sound likes this is an esoteric feature. If I am not mistaken you are looking for a volume normalizer right? i.e. basically the software should scan ahead a song find the average db (or variation of the theme) and use that to renormalize the volume to a user defined setting?

Isn't this what itunes Soundcheck feature does? Under Preferences->Playback if you tick the box it "Automatically sets song playback volume to the same level".

XMMS and winamp all have plugins that do volume normalization as well.

Dawgincontrol
12-18-2005, 09:07 AM
I use Music Library software which "runs" my Sony ES jukeboxes from my computer. I have over 600 cd's. With this software, it plays the cd from the cd players with the tracks you choose. The music does not go through the computer, it goes through my receiver. Only the track list comes from my computer. Unfortunately the music industry records their cd's (not computer tracks) at various levels.

I want something that can scan all the disks output ratios and "normalize" the volume. Whether it be a stand-alone DAC, a receiver, or a cd player/transport. There should be something that equalizes the volume of different cd's. Not tracks form a computer. I do understand the hard disk usage. And maybe that's the way to go for playlists. But, I was hoping for something easier and simpler.

bloosqr
12-18-2005, 03:32 PM
I use Music Library software which "runs" my Sony ES jukeboxes from my computer. I have over 600 cd's. With this sodtware ot plays the cd from the cd players with the tracks you choose. The music does not go through the computer, it goes through my receiver. Only the track list comes from my computer. Unfortunately the music industry records their cd's (not computer tracks) at various levels.

I want something that can scan all the disks output ratios and "normalize" the volume. Whether it be a stand-alone DAC, a receiver, or a cd player/transport. There should be something that equalizes the volume of different cd's. Not tracks form a computer. I do understand the hard disk usage. And maybe that's the way to go for playlists. But, I was hoping for something easier and simpler.

I think the *easiest* way to do this is move your music to a hard drive/software based solution so your playlists and the music are in the same place. (As an added bonus you wont have jitter issues :)). I think the reason that this is so common in the hard drive world is that the random access and/or "look ahead" of ones music collection is trivially performed so it is much easier to work out how to do volume normalization. Hard drives are pretty cheap and you could even go lossless so you get cd quality w/ prices these days and you can still use your receiver as your dac/amp by using the digital out of your computer.

That said if your computer can access your record cd collection digitally you could probably do an "on the fly hack" as long as your computer can also set the volume of your receiver, through your jukebox program.

The reason why this is more difficult in hardware is to do it right it needs the lookahead for the entire track rather than riding the gain on the fly, i.e. a compressor.

I think there isn't a market for what you are asking for as it is somewhat dependent on the fact that you have the sony jukeboxes and most people w/ "playlists" commonly have the hard drive solution. The closest facsimile is probably a compressor but that is problematic for all the reasons everyone has pointed out.