"set the same volume level for all songs" in spotify, does not take into account the subtle volume transients of an album that audio engineers have spent hours working on!
Solution - Add an option in settings for "Set The Same Volume Level Based On Albums". This will retain the intended volume differences between songs on an album (similar to the way replaygain album mode functions).
"Set The Same Volume Level Based On Albums"(replaygain album mode) DESIGNED TO RETAIN THE SUBTLE VOLUME DIFFERENCES BETWEEN TRACKS ON AN ALBUM)
This solution is extremely quick for spotify developers to implement - Just take the individual track volume values in an album and you already have this information due to your "set the same volume level for all tracks", then get the average of these track volumes for that album and the result is your per album volume reduction value! Albums will finally sound like albums again.
IN THE FUTURE - When you select "set the same volume for all albums" then it should also read the metadata in your personal music files (when the tag is present). The Name of this tag in my mp3 and flac collection is "REPLAYGAIN_ALBUM_GAIN".
"Set The Same Volume Level For All Tracks" - (replaygain track mode) BEST FOR SHUFFLE MODE, OR LISTENING TO LOTS OF RANDOM SONGS)
Spotify already has "Set the same volume level for all tracks" in settings for songs that are streamed, which is the equivalent to replaygain track mode. This currently only works for streamed songs and it doesn't check to see if the tag "REPLAYGAIN_TRACK_GAIN" is present in your local audio file that you're currently listening to, a very simple fix indeed.
DISABLED MODE = NOTHING TICKED.
When unticking "set the same volume for all albums" and "set the same volume for all tracks", all streamed and local music should not play with any volume adjustment on either streamed or library tracks, it is exactly the same effect as disabling "set the same volume for all tracks" that we currently have. There are many reasons for needing to turn of volume normalization (such as troubleshooting volume problems etc). Replaygain volume adjustments should be fully lossless when your dac is 24 bit.
Note to spotify - By using metadata, this should be totally seamless - E.g. If the exact same album is streamed, then local files should play at precisely the same volume as streamed files.
IF EITHER TAG IS MISSING FROM METADATA - "REPLAYGAIN_TRACK_GAIN" or "REPLAYGAIN_ALBUM_GAIN" then spotify should not make any volume adjustment and it should produce the same result as having album mode or track mode not TICKED.
When installing spotify, the default should still be set to "set same volume for all tracks" like it it just now
Ok this post isn't as popular as I thought it would be (thought we'd have 200 votes already : P )
I don't need any replay gain stuff, simply being able to copy my flac collection to the local files and listen to them is enough.
Ok, I entirely support this feature. And Deeb, the main problem this isn't that much liked is because not a lot of people are familiar with the name ReplayGain, or think this is some sort of normalization. Lots of people just want to listen to music, not understanding that ReplayGain saves them volume controlling effort or eardamage.
Then again, louder is better for a lot of people, so I guess they'd "normalize" their music by turning up the volume to match their last and loudest played track before their current.
synth - Thanks for your comment, I was initially confused that people wanted FLAC yet they're not interested in replaygain but you're right and that people don't underrstand. If replaygain or normalization was everywhere then we wouldn't have **** sounding audio because there would be no loudness war.
I read that the current main reason for the loudness wars is down to the fact that SoundCheck is not enabled by default in Itunes. Volume normalization could kill the loudness war.
The main reason I love replaygain though is because when I'm in the mood to play different tracks rather than whole albums, then I don't need to keep changing volume control every 2 minutes.
When I wrote that comment about this thread not being as popular as I thought it would be, there was only 3 kudos! So we're donig a lot better now.
Good one Deeb. Replaygain is a must for playlist and radio listening. I listen to Spotify playlists over a Squeezebox system, and the variations in volume really detract from the experience. When listening to local files over the Squeezebox this is not a problem as it has very good Replaygain support.
I agree with Synth that there could those Spotify users that perceive a loss of volume (or, incorrectly, quality) as an issue, but others will appreciate listening to tunes withtout having to reach for the volume control. Just ensure it can be enabled/disabled in the client. Make it happen Spotify!
Please leave comments / suggestions below.
Spotify makes the replay volume of all songs similar. Broadly speaking this is a good idea (Although an "album mode" would be a great enhancement)
BUT
When Spotify lifts the level of "quieter", more dynamic music, it regulates the louder moments using a limiter, often with very poor quaity results. There are lots of posts here complaining about this problem.
I'd like to suggest two improvements to this:
1 - Reduce the reference volume level, ie. make everything play at a slightly lower volume.
This will reduce the need to add extra limiting, meaning people will hear more dynamic music the way it was intended.
Apple's Sound Check feature already uses a lower reference level of roughly -16 LUFS, for example. The international standard for broadcast is -23 LUFS
[UPDATE May 2017 - See #3 below]
2 - Instead of limiting, just restrict the volume boost to prevent clipping
The limiter is only needed when boosting very dynamic music, to prevent clipping distortion. Instead, Spotify could prevent boosts beyond a point that would cause clipping, meaning the limiter would be unnecessary. This would completely solve the problem.
Some very dynamic music wouldn't be normalised as effectively, but examples in popular genres would be rare, and most people listening to more dynamic genres wouldn't mind - classical is a prime example.
UPDATE
I've been asked to join an AES technical group to help develop a white paper on streaming loudness, to provide recommendations for loudness normalisation for streaming providers like Spotify.
If any Spotify employees are reading this, please contact me so we can include you in this discussion.
UPDATE #2
The AES guidelines have been published, they recommend a maximum streaming playback level of -16 LUFS:
http://www.aes.org/technical/documents/AESTD1004_1_15_10.pdf
UPDATE #3
It looks as if Spotify have implemented Part 1 of this idea ! The latest version (May 2017) seems to be using a normalisation level of -14 LUFS, which is similar to TIDAL and YouTube. This is great news, thanks to everyone who voted for the idea !
We don't know yet whether they have also stopped using the limiter, though (Part 2) - more testing is needed. But so far, so good !
Yes, please.
Musicians and music lovers will enjoy Spotify more if you follow Ian's recommendation.
Ian is right on. He is the leading proponent of appropriate compression levels and his attitude and comprehensive take on the issues involved are reasonable and spot on.
Supporting the Ian´idea. If iTunes works well, why not do the same.
Long live dynamics on music!
Thank you.
I agree with this. I master music and Spotify is a tool that many artists reference. -11 LUFS is just a little too loud but more importantly the use of additional non-transparent limiting changes the artsist intent IMHO. Please consider a more reasonable level. I do find that -16LUFS works nicely for a lot of music save for Classical / Orchestral which should probably be flagged for non limiting.
Agree wholeheartedly! This change will make listening so much better. I notice the compression quite frequently and it's ver annoying. I don't want my Spotify listening experience to sound like the radio - I want it to be as high-quality as possible
Yes,
AND
Please stop blasting the ads.
I realize I am using the free version and that I can upgrade to avoid this, but just like television is not supposed to play ads at higher volume, music services should not either--especially when many people listen on headphones and the extra volume is unexpected.
Definitely agree. I don't get why they force such high LUFS values. -23 LUFS is perfect: it virtually eliminates the need of a limiter on even the most dynamic tracks and it keeps more than enough headroom to keep the music high quality. Heck, even with -35 LUFS music still sounds better than when you apply that irritatingly bad limiter to the music.
For your information, that is the single most important reason for me not to pay for the service. The sound quality is often worse compared to other services like Deezer and even Bandcamp and Soundcloud.
Same with the ad volume. I don't mind the ads. I do mind their volume, and they are exactly the reason why I don't use spotify that often, even though all 4 bands I play in are on there.
I was seconds away from canceling my premium membership after noticing that even with high quality streaming enabled, in direct A/B comparison many of my favorite albums sounded very noticeably worse from Spotify than off of CD or FLAC/ALAC. Then I discovered this (hidden) "set the same level for all tracks" option, disabled it, and now everything's fine!
Volume normalization is not a bad idea at all, but Spotify appears to be either applying a compressor with a short (intra-track) response time, or normalizing tracks to such a high level that the limiter is constantly kicking in. It really crushes the life out of a lot of good music.
I can accept that this "feature" is turned on by default if most of your users want and like it, but A) it doesn't work very well,
its name is misleading and C) the option to turn it off is hidden away in an "advanced" section, as if it was dangerous to change things like the crossfade time.
I am all for an advanced options section, but in that case you should provide real advanced settings like the reference level and the timescale (whole track, whole album, intra-track) at which the normalization is applied.
Consider that if you do this right and publicize it, you will win large numbers of devoted clients who care about dynamics in their music!
Hey guys, I'm marking this as a 'Good idea' and bringing your feedback to teams internally. If I have an update on this I will post here first, thanks!
This is probably a bit more technical than most feature requests, but unfortunately this is something that virtually everyone gets wrong, so here goes...
There's a "Make all tracks play at the same volume" switch in the settings, and I'm not sure how it works, but it's clearly normalizing to a fairly high level, instead of a quieter reference level. This is a common but bad way of setting levels. Either dynamic limiting is being applied to the stream, reducing dynamics and quality a lot, or the level for the whole song is being scaled to a high volume level. That can't be done without clipping (or artificially limiting) the peaks of a dynamic track. If normalization if applied based on peak levels, dynamic tracks will play quieter on average than ones with heavy dynamic compression.
Anyone who understands digital audio well should see that all of these methods are quite bad.
What should be done, is that the tracks are left as intact as feasable, and a replaygain tag added, at which point the player will reduce (or rarely increase) the level of the track to a -14dB reference level so they can all be played at the same volume with sufficient headroom for peaks in dynamic music to playback without clipping.
Unfortunately, most everything else you'll play on a PC is normalized near 0dBFS with loads of compression, so this would make your music all play back fairly quietly in comparison to youtube, etc.
The solution to this is to add an OPTIONAL "post-gain" which will increase all of the normalized songs by a fixed amount, say +8dB This will restore the level of the tracks to close to what most would play at without RG applied, but of couse will cause limiting/clipping of peaks on dynamic tracks (as is unavoidable when using a reference level without sufficient headroom), but it will satisfy those who would rather have everything equally loud as their other non-RG content like youtube, etc.
Those of us who care more about sound quality will turn up our volume knobs instead, bringing the levels back where we want them, without causing clipping.
The post gain could be applied by default, since the majority of users probably wouldn't notice or care about the clipping, and would be more bothered by the reduction in volume, but the option should definitely be there (in advanced options) to turn it off, for those of us who care about sound quality.
Note also that the volume attenuation/gain should be implemented client-side, and absolutely not applied to the stream, else there will be further quality loss from transcoding.
I agree 100%. Please adjust Spotify's volume limiting threshold so that it no longer distorts on louder music. People like me who care about music want to stay with Spotify rather than go over to Apple.
Please allow us to have an Album Mode for volume adjustment. I'm very tempted to go back to Apple because I very much want to hear the songs play at their relative volume in the way the artist intended. Listening to entire albums is the reason I became a paying Spotify customer. Please keep me faithful by allowing me to listen to albums as the artist intended. Thank you.
This is major. I'd like to see this option.
-23 dB, please. And the option to level-match between albums (but not within them) would be great. This way you could retain an identical average between albums while retaining each individual album's dynamics.
"Hey guys, I'm marking this as a 'Good idea' and bringing your feedback to teams internally. If I have an update on this I will post here first, thanks!"
YES and that was MORE than six month ago.
This isn't just "a good idea". It's the ONLY way and an excellent report from Ian Shepherd and he is obviously a professional. I consider it a bug report with a fix guide. Not an idea.
How dumb can a streaming service company be not to understand how to get good audio quality out to it's customers? This is the core of the product. How about spending some time on the CORE product and enhancing the audio instead of wasting time developing all kinds of stupid, irrelevant social media content for Spotify?? Not all of the customers are teenagers.
How it is possible not to have hired any audio engineers long time ago that could have taken care of the problem is beyond me.
Yes! This is important. Please do so. -23 LUFS would be just wonderful!
I'm with Ian on this. Dude knows a lot about dealing with loudness. I hope Spotify takes his advice 🙂
I used to run the following code to convert tags in album mode to itunes "Sound Check", although Itunes "Sound Check" is now intelligent enough to know if you're listening in shuffle mode (the equivelent to track mode) or listening to an album (while retaining the subtle volume differences between songs as part of the album). See the following video for an example.
http://www.youtube.com/watch?v=ehMo0JWRFC8
For .mp3 files "format value" COMMENT ITUNNORM
[$rg2sc($if2(%REPLAYGAIN_ALBUM_GAIN%,%REPLAYGAIN_TRACK_GAIN%))]
For .m4a files "format value" ITUNNORM
[$rg2sc($if2(%REPLAYGAIN_ALBUM_GAIN%,%REPLAYGAIN_TRACK_GAIN%))]
Spotify could eventually read iTunes tags using similar code.