Edits Check Request

MusicBrainz
1

As most know, I have been asking a lot of questions trying to really nail down the best way to add data to MB… where to put what, how to put it there, etc. Most of my earlier edits were more basic, I am now trying to add more details like relationships and paying greater attention to the release differentiation, even in its finest details.

I have done a fair amounts of new adds and some edits, referring only to CD edits. I wish not to discuss anything digital in this query. I would like to ask some of the editors here, especially those with whom I have communicated with already, to have a look at what I have done. I am not really looking for a review to make sure the edit is ok, but to look at the finer details… like the annotations, edit notes, relationships, etc.

I dug deep in order to get some nice content, trying to grab as many of the more unusual and non common ones I could find. I am fairly sure there is nothing wrong with how I have added, I am simply looking for feedback and ideas from others on things like maybe changing this or that to conform to a greater norm of style. I would like to stay neutral, meaning that I want to add things more traditionally, with attention to details. So should someone see something I am overlooking or an area I am forgetting, please share.

All feedback is appreciated. As I have stated before, I am not doing this for my gains as many do… adding content so they can tag. My intent is strictly to preserve data. Also, I have not yet looked at “works”, so there is no need to mention that. That is known and I need to address separably and learn before doing.

To make sure all have: Editor “thwaller” - MusicBrainz

1 Like
2

took a brief look, (can maybe look farther in a bit), and just wanted to give you a high five on adding all those scans~ :raised_hands: if there’s full scans on a release, all the other details can be added by anyone~

one small thing, when I add a booklet, I try and add the page numbers on each image. as long as they’re in order though, it’s not super critical.

if you’re scanning your releases already, another scan you might consider is the matrix (the center of the back of the CD). there’s several people who love documenting all those CD pressing details~

as said, I can probably look at some more after dinner~

3 Likes
3

Sure, that makes sense. You know, I put them in order using the file names I have, but thinking on it now, those file names do not carry over. I could only imagine an editor saying “per page 53 of the booklet” as an editing note…

I often have my booklets stored as “booklet0”, “booklet1”, etc… where 0 indicates the cover and booklet back and 1+ represent the inside pages, in order. Would that be enough to help, or is it preferred to list the actual page numbers… “booklet 1/8”, “booklet 2/3”, etc? I will go back in and add comments to those images accordingly.

Yes, a few have mentioned this prior. Being honest, I historically never saved this information or any images of such. So as I dig through stuff, those images are not likely to be there. What I can do however is see if I can find them to add. Discogs sometimes has them, so I can check vs just adding my stuff and being done. Some recent edits I have included them, but it is by no means the majority.

Could I ask… how exactly do you use this information? I understand that it can be used to identify the specific pressing of the CD, which then can help identify different versions of a release, but is there more that maybe I do not realize? I ask this because MB does not have any fields to store the printing in that area, I see Discogs places it in the notes/annotation. Is there a need/want for anything further than just adding the image?

EDIT: I missed one part of your post. My images come from more sources than I can even list. Not counting the quick camera phone stuff, my images come from personal scanner, personal HD camera, other sites like Discogs, vendor sites, pirate sites, torrents, cover art scenes, etc. It has already been suggested to me to make a better effort to state the source. While I cannot always do that, I am saying things like, from my personal collection, matched to this release by_____.

I hesitate some here. In the video side of AV, it is now determined that sharing of subtitles is considered piracy by some courts/judges. MB has also been “slapped” for artist images, thus they are no longer there :frowning: I tried recently on a few edits to provide a reference, and it was suggested to me to look twice as they were “suspicious”. I looked and agree, and thus removed the references. I added them in effort to try to add a reference to verify/confirm the release. My point, I want to avoid trying to be too detailed that it causes an unintended problem… I 100% support some sort of proof and justification of all edits, but sometimes it is hard. I also had an edit where I used an email as “proof”, which was met by mixed opinion. I can go further down that rabbit hole as I have some materials provided through email exchanges for marketing efforts. Those I cannot use, obviously, but please see the dilemma on providing sources.

Could I ask this? Say I have a release in my collection. I have no reference aside from that, it is in my collection. I might have an email exchange, a subscription reference, or simply just a folder with audio files and images. To what extent do I provide references? I have been searching online to find them, but I almost wonder that if what I find is not a source like Amazon, Discogs, iTunes, etc… just to leave it out and say the release is the proof, source my collection, as that is the actual truth.

1 Like
4

I typically do something along the lines of “Page 2-3” (referencing pages 2 and 3), but your numbering should be perfectly fine~

honestly, I don’t know… lol

that’s probably a question for the folks in the thread I linked (maybe they answered in the thread? I forget).

I have recently been trying to add matrix numbers to the release annotation, that way it might be searchable in the future.

a link is always the best reference, but if you can’t find one, something like “from my library” or “my own scans” should be sufficient. at the very least, no one’s yelled at me for adding such releases… :joy:

1 Like
5

All good. I only want to understand how data is used, so it can be provided better. I do not use such things, so I have no opinion on the matter. Just as I dork on mastering engineers, I am sure there are others who dork on the matrix. Fair is fair, I only want to understand, maybe I can help.

Thanks for the reply. Your recommendation is noted, now I have a lot of edits to do, LOL!

6

Could I ask this? I have a few TB of cover art that accompanies my actual audio files. Just on that release alone, the files are between 5-10MB each. So, yeah, 45*7.5 = over 330MB of artwork. I wondered at a time if I could upload this artwork to MB, relying on Internet Archive to store it.

My concern is that others could delete the files, or any combination of the aspects of public DB. So, one one side, my content will only last my lifetime, and only for me. If I upload ALL of it, that helps many, and gives me a lot of room for releases without needing to my more drives to add to RAID setup. I have OCD, so I use the inefficient RAID1.

EDIT: Sorry, for those who might not know, RAID1 means that if I have say 2TB of data, I need 4TB to store it. RAID1 is a mirror.

2 Likes
7

  • Upload all the images! Agreed re. expanding their lifespan past yours (or your hard drive), that’s exactly my thinking too :tada:
    Keep the local copies if you’re worried about anyone f’ing with them in the database.

  • I leave edit notes like “from a torrent site” all the time. For me personally edit notes are not always about having to provide a ‘perfect source’, and often just about helping people assess the data quality.
    A note like “in hand” = this is some gooood data. “from soulseek” = I won’t worry so much about tweaking things on this release, good to know
    If your source is kinda crappy a note is even more important then if your source is great imo :stuck_out_tongue:

  • The matrix is great for identifying pressing plants, which can give you some nice information about where a CD is from. That’s the only practical application I personally use them for. Oh, also very very useful when there’s no full scans or something else is missing and I’m trying to match a release to discogs to find out more.

4 Likes
8

The reason to documenting the sources is so as to be able to trace back errors. Always shout loud about personal scans \ photos. These are the most accurate. You know you did them and it all hangs together.

Something from a less that reliable source, just get more vague. I see one AE who says “found on my computer”. I’ll say “found down the back of a sofa”. Or “Unknown source”. Or “came with the files”.
And if I see artwork uploads with no comments then I know I can distrust them as just something Google’d.

Think of the editor looking at this in five years time. Should they trust the artwork?

The trouble with pirate sources is they are not always what they seem to be. Like you, I have music from many places, and sometimes that pirate art is dodgy. The creator of that package just wants pretty artwork for tagging, not accuracy. I have often had releases where the Music and EAC log are from one album, and the artwork is from a totally different release. (Sometimes the EAC logs also don’t match)

If the album is unusual, then I’ll upload the images I have in hand. Especially as I know other people will have that exact same downloaded package. If it is a popular album then I’ll need to double check my art in hand against Discogs or elsewhere.

Going back over some of my oldest art uploads today I see errors I made five years ago. I can correct them as I know I made the error. And it is a little funny to me the details I didn’t notice in my first year of editing. Or not understanding how important manufacturing countries are.

It also makes more sense to not add a Discogs photo of a matrix if it is not your actual matrix. Unless you look at your CD you won’t know what the matrix is.

I don’t want to go off on tangents here - but it is easy to have CDs that say “made in EU” which could be printed in three or four different locations across Europe. (Happy to talk for ages on this subject… :crazy_face: :nerd_face:)

CAA storage - an interesting side of CAA is even if someone deletes your art and replaces it, CAA still keeps a copy. Generally someone should only delete it if they replace it with something better. Personally I save as 80%JPG as it does a lot of compression for very little quality loss. Others stick to 34MB PNGs per image. It is all possible.

I’ll have a dig through your releases later… busy day today. Already two I have seen look great. Full artwork is always excellent to see as, as @UltimateRiff noted, it allows others to add more data at a later time. Those booklets always have so many extra credits in them.

4 Likes
9

Oh, ok. I get it, don’t trust the source unless you can confirm it is correct. One of my recent edits involved using an existing Discogs reference where the artwork did not match the printed content of=n their page. I had never seen that before, and I noted the issue in MB annotations. Understood, it makes sense. At a very least I see that stating it came from Discogs as the reference is useful, as it shows it was not editor confirmed.

Please do. It is only as of the last week or so that I started adding “areas” as relationships. I have already seen the complications… as you mention EU, then there is UK… and when you see a release that states DE, well, yeah. I honestly never paid attention to this, only the contents and actual appearance. Meaning I have duplicate CDs that came in different cases, contained different artwork, content, etc… but never really worried about where it was made, printed, etc.

@aerozol Great points. I dislike the “CD in hand” note, but I do see your point. Notes like “from torrent site” can be complicated. There is the obvious … “I downloaded this copy of this release from torrent”. At this point, you can say that the files you downloaded came together. Is it reliable, unsure. The metadata will be a mess, but I guess the audio portion of the container is usually good. [It is funny to see on scrobbles the numerous amounts of “torrent” metadata that is supplied]. I get a lot of artwork from torrents via data mining, which includes a basic pull of public torrents. Some of those people that share music in that way are so detailed it is crazy.

Interesting we think the same on the artwork storage. Storage is cheap, and I have a lot of it, currently 31TB. But I also do not like wasting space. If I can go a year or two without buying more storage, that is great. As I stated before, I have come to realize that storing such things because of my OCD will only last me until I die. Then whom ever gets my storage will likely delete it all anyway.

Thanks for sharing to all. I hope I am not seeming like a bother. As you can see from my edits, I am trying to populate content, not just the edit of a track list with no durations and such. I ask these questions as I realize there is no right or wrong, we have here a platform that we either adapt to and use, or we don’t. Something is better than nothing. [Hint to the DB designers/maintainers - I would enjoy my life more if I could help improve things].

4 Likes
10

They can be a good source as they also have plenty of eyes checking. If I borrow their art, I will always quote the source. Also if I have checked some “unknown” images against a Discogs copy it is a useful reference to others.

As you notice, Discogs is also not perfect. Nothing is. But it can be trusted to a high level.

I work on the basis of letting people know what I do and don’t trust of the images.

I rarely say “From torrent site”, but I may say “borrowed from a friend”. No need to get complex on legalities. (and never quote the torrent site name\ID) If you are mix and matching artwork this way, it is not really stuff you can upload to MB. MB wants art to be from the exact copy for ID purposes. It is not just after a nice image. Don’t add art you cannot verify. This is where Discogs can be useful as a cross reference to that artwork version.

Release Countries can be puzzling as different guidelines seem to get followed. Often the DE has appeared due to an editor adding “I know this was released in my country” so adds details for their country. I use Music Week to check UK release dates, but I know little about Italy. “Europe” is often use but is way too vague, and would usually be wrong for my 1980s CDs. (different topic).

And don’t forget a Release country is different to a Manufacture country. Very common to get French manufactured CDs Released only in UK. (this is another long tangent… Manufacturing details get very geeky)

11

Absolutely. The same is for EAC logs, the TOC. All that I use are clear on source, as mine are not generated by EAC as I use Linux. I would like to attach such files as “proofs”, but cannot do that at this time. I find it actually better than an image of the CD, as that can come from the same as the TOC, just downloaded. The ones I generate will have a header like:

whipper version 0.10.0 (eac logger 0.5.0)

whipper extraction logfile from 05. April 2022, 01:20

So those I upload I can be sure they are real, as I did them myself. I do not know of many that use or even know of Whipper, much less use functionality that does not come with the default install. The same applies to before Whipper was Whipper, that is a different topic though.

Please continue. I also see a difference that a CD often states “Made In”, where MB does not have a “Made In”. Made can mean a lot of things.

12

You don’t need to attach the text file of EAC or Whipper. The point is these all generate the same TOC from that DiscID that is uploaded. Just like Picard does with a CD. That is proof enough writting that clearly in the edit notes. A TOC is hard to fake.

I am now in the habit of adding my Matrix details and DiscID together in a text block in the annotation. This ties manufacturing details to an exact TOC.

“Made In” conversations are best with examples to hand. Notice a separate thread starting up with an old Tubular Bells CD.

“Made in” is not “Released in”. Very separate.

Here is a mad example for you from my collection, with artwork: https://musicbrainz.org/release/8af759be-1495-4cc4-85a7-47edc21fa348

That CD was Glass Mastered in EMI SWINDON (UK)
But Pressed in Italy (and says Made in Italy on CD)
Rights Soc of BIEM/SIAE (Italian)
Paperwork printed in Italy
But it is sold in UK - so release country UK

Many variation sit near it in the same Release group pressed at UDEN (Holland), EMI SWINDON (UK). Paperwork swapping back and forth over the channel. This is why you see so many different options at the bottom of a Release page.

13

I tend to take “Made In” to equate to “Manufactured In”. I do not consider the release to be a part of that.

Well, yes and no. The EAC TOC is formatted in a certain way that those of us non EAC users need to use or write code to duplicate. What I consider a true TOC is not the EAC rendition. I have a lot of releases in my collection, that I personally ripped from source, that do not have a suitable TOC to submit. I just started to look into a script to fix it, to change the format. This is a another topic, again, LOL!

I had some issues with areas. There was a Holland example which triggered my mind, but also an artist. I wanted to add an area and could not. I wanted to say his area was “Midwest, USA”. I could not do that. If I specify a state, it would be wrong. Saying US is vague. He is Illinois, Ohio, Indiana, etc. Old school jazz in this reference.

14

I am having a poke around some of your Releases and adding random thoughts. Some good quality stuff going on there.

Yeah - “Made In” is the “Manufactured In” of the CD. There is also a “Printed In” for when that is shown different on a Release. Common in European releases.

I am not worrying about how EAC’s file is formatted. Just the DiscID that is uploaded, and therefore how the TOC is formatted in MB on the DiscID page. This is the common point we can compare from. An EAC log, and I assume Whipper, can be converted to that same number produced by Picard. Allowing the TOC page in MB to be a common layout.

I am aware other details are not being uploaded. MB can’t store everything. :slight_smile:

There are some unusual choices as to which areas are included. For people they can usually be nailed to a city, but not to a wider area. I think there is an ISO list they stick too.

The Holland \ Netherlands thing is in another thread. In the context on the CD it means the bigger country, not the areas near the coast. This is usually talking about the UDEN plant. Which is not in “Holland”, its in the Netherlands…

1 Like
15

Wow… ok. Thanks for putting in the time to provide such feedback. I was not expecting anyone to put in that much time, so it is much appreciated.

I have run into this a few times now. It seems MB does not recognize Holland, although the releases that are made (or other) in Holland make it very clear that Holland is the location. I do not know the politics behind all of that, so I have just used Netherlands in pace of Holland.

On what I think is similar, I also see I cannot use things like “Central US”, “Midwest US”, etc. This could be useful since there are artists where the area portion cannot really be a single state, nor the entire country.

I refer to the EAC format as it can be used by the EAC TOC submitter, which seems to be a proven process interacting with MB. I do not prefer the EAC format, but that is not the topic here. I agree the TOC is a great and important aspect of a release. I really question things when there is a release with 10 or 20 Disc IDs. Maybe that is possible, maybe not, I don’t know, but it seems like it should not be the case.

I mention the log file as it tells me if I created the log or if someone else did. If the log file was not created by me, that brings into question the whole thing. The log files have a different purpose for me, I have been playing with creating a digital copy of a CD, then using the resulting files to rebuild the CD. [Whole different topic]. So the TOC and such files are also very important to me. For this however, I do not use the EAC version of the log.

RE Notes you have left… You pointed out issues with Discogs, and I have also noticed the same. What I have found is that:

  1. I cannot rely on the Discogs reference being the proper one in MB.
  2. I cannot rely on Discogs in general, as I have found some releases that contradict themselves, having different data in words vs the attached artwork.

It seems that both the MB release, the MB ↔ Discogs reference relationship, and artwork on the MB release and on the Discogs release need to also be verified that it indeed matches the release they are attached to.

1 Like
16

I wanted to note that I replaced the images you suggested rotation on. I got to looking and it seems that most all of my TIF format images were never rotated in my collection. Those are usually the ones that were taken from digital camera and just never converted.

1 Like
17

No need to worry about the EAC logs, they are good quality, but for a tool you do not use.

10 or 20 discIDs is not due to the TOC submitter. You will find the same 10-20 discIDs on a handful of old Releases in a Release Group. Check the edit histories and you will see they go way back in time of MB when the releases were not separated like they are now, and less checks were done on DiscIDs. (Often not even listed in history) At some point in MB history (called pre-NGS) separate Releases were made. And at that time ALL DiscIDs currently in the database were assigned to ALL Releases. This has left a bit of a mess that needs people to sort out. It requires sometimes being a little harsh :fire: and chopping out DiscIDs that are clearly a long way off from the average, etc.

It is why I often favour making a totally new release if I can. Then I can add just one DiscID and one set of images. It is also why I am adding verified discIDs to the annotation to assist in this clean-up.

In Europe the DiscIDs often follow a common value across the factories, so once you spot the good ones it is easier to highlight them. But there is no perfect rule, so you can’t assume your discID is unique to your Release, and you can’t assume no other discID will match your release (though there should never be 10…).

Re: Discogs, depending on the subject, generally Discogs is pretty good. Just like MB there are many eyes on. But just like MB it is not 100% perfect. Nothing is. And they also have different submission rules. This is why it is so important to state when you have your own real release in hand with matching scans. And checking your own track lengths (Discogs is especially bad at track lengths) If you show the verified data, it allows others to question the bogus data.

I’ll go look at those edit notes now. There were some good examples in there of details missed on covers causing bad discogs links. Something you often find, and needs careful reading of Edit Histories to work out which is the “correct” data to keep. Or use your consistent data to make a new example of the release.

2 Likes
18

I would agree with that. It is only recently that I have found issues, however, this is also the same time that I am looking at edits differently. Sort of moving away from an add that has a track listing and durations with a cover, and going to a more detailed add with as much artwork as possible, relationships, etc.

My interest in those details are seemingly different than many here. For me it is all about the recordings and the mastering / editing. I have examples of some Paul Simon releases that are so different to me I cannot understand how they can even be called the same. In more general terms… There is one copy that is at a very low gain. This makes the CD a real pain to listen to, but the audio itself is amazing. Then, there is a copy with what I just call modern commercial compression. The dynamic range is destroyed, signs of clipping can be seen, although the compression sort of hides the audible degradation. I get why they do it, but that does not make it a good thing.

1 Like
19

I think everyone here has different ways they enjoy their music. MB acts as a central common denominator for many collections. It can never be perfect for all, but the common data it hold is valuable to most.

I find it weird when I see people on Discogs who own 20 shrink wrapped copies of an album, but have never listened to them! It makes them happy…

I have a mind for little details, and like seeing these little things that make the Release different to each other - but the music will be on the Stereo at the same time. What I enjoy is that often digging out some little details on an engineer sends me down a rabbit hole of connections and I stumble over a new band. A short bit of EBay fishing late and I have a heap of new music.

Or I’ll start following the Artist who created the cover artwork (blame that in an early interest in Hipgnosis and Storm Thorgerson)

I understand your ear for different audio. You could add notes to the Annotations about that and help to enlighten others. It is why I add those kinds of unique facts. Help people to see their collections in different ways. Open eyes and ears.

Cataloguing my music through MusicBrainz has led to me learning a lot deeper knowledge into my music. Digging CDs out I forgot I owned. Or combing though hundreds of old Floyd bootlegs I had never even listened to before. MB makes us look at the details.

20

LOL! I struggle with this a lot in my collections. Comic books, records, CDs, DVDs (I do not collect BluRay)… Opening them kills the value, but not opening them kills the enjoyment. What I generally do is open the CD, RIP it, then put the CD away. I have many box set type releases, the ones that come in a large folding hard cardboard case, that will “creak” when opened. I am also one that if one of my CDs gets a scratch, I look to replace it.

While off topic, I would like to share some about me, so maybe others can better understand where I come from. I started with music back in the 5th grade I think. Prior to that, sure I listened, but the radio was my source, and if I wanted to hear something, I need to call the station and make a request, which yes, I did often. This is approx the year 1986. It is that time that I started percussion in the school band. I quickly got bothered with the school instructor, for the lessons needed, and went to private lessons which the school accepted.

For some reason, it seems that in my area, the percussionists were the real band members. This is where I got my relationships with what I will call “real” bands, not local garage bands. Two people in specific were classmates of mine and were guest musicians on a few commercial albums. The Violent Femmes is also interesting as they attended the same university as I did (before me though), as did the classmates I mention. You can see how this progresses.

I went through my phases of sound…having the car with 20000 watts (exaggeration) and home stereo that will shake off the gutters… to today where I have high end equipment and play them at low volumes… almost opposite from the past. I always liked real music. A band that is a real band, not electronically produced sounds. Nothing beats recordings from the 40s with the amazing dynamic ranges they offered.

Today, I am lucky to not have damaged my hearing too bad from being a foolish kid. I have a music collection, in order of volume, of digital, CD, vinyl and wax cylinder (along with an old Edison phonograph). I had cassettes, but honestly the technology is trash and I am not even sure where they are anymore. Nearly all are replaced with CD anyway.

In addition to enjoying music, I also record live performances at times, local cover band type stuff. I do this usually my tapping into the sound board. I have done a few where there is a capture of vocals sep from guitar done in “studio”, but that is more rare for me. I am not a sound engineer, but just good at creating a digital version of music that sounds good.

I learned to listen, which sounds weird, but it is not. For example, there are testing programmes where you are given a sample and asked to identify frequency and/or compression. Once you go though even one iteration of such ear training, the things you identify surprise you. This is partially where I still state that there can be an audible difference between say a 16 and 24 bit Flac file. Given the same source, you should not be able to hear a difference. But given a different source, that is where things fall apart. In many cases, it is even possible to hear a difference between the MP3 and M4A, all parameters equal. I never listen to claims of others, I test myself. There is plenty of software to help with A/B testing, so unless you lie to yourself, the results tell you the real story… can you really hear it or not.

EDIT: I have grown to like Lacinato ABX.

1 Like