highstrung,
Just look up Bob Wills and Hank Williams singles I have edited!
1 Like
That’s great and I love that. I’ve done that myself for a while and it’s really time consuming.
The only problem is: without a short note: “Durations taken from ripped mp3” or something similar, no future editor can see where the durations are taken from.
Compared to the time it takes to rip the vinyl/shellac a short note is really nothing 
7 Likes
Maybe I can expand a bit on the ‘social etiquette’ aspect that MB has.
You are absolutely correct that the onus being on someone to disprove the existence of something is extremely troublesome, generally impossible in practice. But, that’s not really what’s happening here. What other users are asking for isn’t evidence, but an inkling that it could be incorrect.
Why do editors ask for this inkling? Two things imo:
- As previously hinted at, there are are millions of edits without (useful) edit notes in the database. And, even with edit notes, often the edits are still based on trust. The “in hand” edit note is a gold-standard for most physical release additions, and is entirely based on trust. That’s not too say some edits aren’t bad. But on balance the MB database works.
- We have a voting system of ‘don’t vote no if the information is an improvement, even if not perfect’. This might shed some more light on the attitude of live and let live - though not necessarily applicable to track times, the general idea is that consequent edits would move towards specificity, rather than blocking edits that are an improvement. We can argue until the cows come home if it’s useful for ‘close’ vinyl track times to exist or not, that’s not what I want to get into, more that the general attitude of a lot of voters is shaped by this system.
Overall I share the reaction of a lot of others - if we follow this precedent and remove everything that has no edit note, the database would become worse, not better.
IvanDobsky is more diligent than most so his background check is gold standard. The minimum for me would be leaving a note/the opportunity for the original editor to reply, as jesus did.
Or, if you presented some other inkling that it could be wrong, like ‘i know there are multiple recordings of this and this could be any of them’, the edit would also be more likely to go through without raising eyebrows.
Anyway, I’d be surprised if you didn’t get pushback on this edit and hopefully that helps shed some light on why .
6 Likes
If you are confident that the data is egregiously wrong and needs to be fixed, you could also consider changing the data quality from Normal to Low. This makes it obvious to all users that the release needs to be fixed. Not sure vinyl times would fit this criteria in most cases though.
1 Like
Good examples there, such as this edit. As @indy133 notes, it would be better with an edit note, but the database would not be improved by removing those timings.
Having edited here for a few years now, I have even come across a few of my own early edits which I wish I had documented better. Good edit notes are a benefit to future editors, which may possibly include yourself!
3 Likes
Yes, and that was a right decision. It’s bad for data quality when incorrect durations are added, but it’s worse when data is removed by guessing. You should add an annotation explaining that it’s unverified instead.
However, track lengths for LPs are generally a problem. I have to admit, I don’t measure the exact durations either and mostly leave it at the printed track lengths. These are mostly net playing times and can at least be looked up later. The lengths of the ripped files are completely subjective and could even lead to false assumptions. But these net lengths don’t contain the silence surrounding the song (like CD TOCs do). And often they are not really exact.
That’s why I always judge track length information on LPs with caution and take that into account when it comes to merging recordings.
EDIT: (Now I forgot to add what I actually wanted to say)
There should be guidelines for vinyl like:
- add printed track lengths if available and if roughly accurate (+/- 3 seconds)
- add ripped file lengths, if there are no printed track lengths (if these are somehow reproduceable)
- measured track lengths (if possible)
- leave them empty (no guessing)
- In any case: Note the method used!
3 Likes
I see the inconvenience of keeping unverified times from this 1 week editor, as explained, as it gives too much fictitious credit to something that is completely unknown.
But I really don’t see what is the added value of keeping these times that we know we will never have any confirmation of (editor was only active between Monday 2015-08-24 to the following Sunday).
3 Likes
… of course, there are cases when I would vote yes, but I don’t think, it would be this particular case. I would write annotations.
Well I’m not quite sure about that. When I compare what’s better: unverifiable data versus no data, I’d prefer the truth. There’s nothing bad about not knowing the duration of a release. And when you talk about guessing, then it’s just what we do when not removing it.
I wonder what Wikipedia would look like if they took that approach (I know the comparison is a bit off, but just try to think about it
2 Likes
No, Wikipedia is a good example. First, a passage of the article is marked: citation needed. It is not hastily removed.
2 Likes
Ok here we go:
Edit #90288351 - Add release annotation.
But it’s a little bit waste of time compared to removing the times.
2 Likes
Ok, you need two sources to add data. Otherwise it will be removed. I remember a case (I don’t exactly know if this is a legend) when a famous author corrected an interpretation of his work on his own wiki page. Wiki removed the correction with the justification: Sorry we can’t except the correction without a second source.
1 Like
… not a second source - a secondary source. It is not allowed to use primary sources to avoid promotional articles. The author must not write his own article.
3 Likes
Well countered 
(is this the correct English term?) but the key word still stays “source”. Wiki needs sources, you can’t add any informaton without them.
1 Like
Yes, and we need sources too. But we keep data as long as it’s not falsified. (with exceptions of course 
)
If I were an author wanting to correct my article, I would log in under an alias and do my corrections silently.¹ Nothing would happen and maybe it would be flagged sometime…
¹) adding information, not removing it
1 Like
I actually agree in principle - but the precedent of just removing everything that has no edit note is not the way to go.
Also key is that this information wasn’t actually unverifiable, but no attempt was made to verify it. That the onus is on the next editor to do this is 100% inefficient and annoying, but as said that is the situation MB find itself in as a database.
For my end-user case I actually couldn’t care very much if the track times for vinyl are a few seconds off. They will vary on your player and how you cut the gap anyway. It’s useful and interesting for me to see the approximate length of a track. If it’s possibly very wrong (e.g. there are very distinct recordings that could have been used), that’s a different matter for me.
Now, doing a full 180 and speaking to my own bugbear… Can we finally get rid of pre-ngs discIDs, because they mean that the DB has mass unverifiable track times 
2 Likes
Discogs says:
“Tracks 2 & 7 taken from the “Marilyn Dreams” 7” & 12" singles."
And specifically notes when another track was a different version:
“Tracks 3 & 5 taken from the rerelease “Nowhere Girl” singles. Nowhere Girl is the popular extended & remixed 12” version that was widely played on the radio."
It seems likely to me that these track times are correct, based on the specificity in that annotation.
3 Likes
One way to solve it… 
(no - really… I have actually gone to EBay before to buy something just to fill in some missing data)
https://www.ebay.co.uk/sch/i.html?_nkw=marilyn+dreams+b-movie
Actually, I think the sellers have spotted this conversation as it is £10-£30. (Oh - and to save your effort… I have already checked all the images in the sales for written track times)
Or another solution I have used before. Find it on YouTube.
A side - 3:42 here => Marilyn Dreams by B Movie - YouTube (Actually fades out around 3:36)
Finding the B side is MUCH harder with a name like “B-Movie - Film Music Part 1”
The 12" is 6:05 => B Movie - Marilyn Dreams (12 Inch Mix, 1981) - YouTube
5 Likes
I wouldn’t remove durations unless they’re copied from another release and are not clearly derived from the sleeve or from a vinyl ‘needledrop’.
These can be very useful, especially if there is some dispute over length, arising from a misprint or similar. Some people source the lengths from third-party websites and making the needledrop can be useful for confirming if that data is correct. It is just as likely for third-party sites to have mistakes as the sleeves of the records. (And having a good source like digital files derived from the needldrop is also useful for reaching out to those third parties to help them correct any typos etc.)
I think this is a good example of someone not stating their source. I wouldn’t vote such an edit down, but I probably wouldn’t vote for it and would ask in the edit note what the source of it is. Probably abstain and see if you can find any information elsewhere about it.
For popular singles and albums there’s often a youtube stream of it which can be used to compare the durations.