Search

Discussion

What Is SHS: A Researched Source or Compiler?

Tar Heel

Member
Posts: 5771

Tar Heel @ 2021-02-09 09:53:51 UTC

A recent exchange got me thinking, is SHS an independently researched source for the works, performances, etc. on site or is SHS simply a site that compiles and archives the data found on other sources and sites?


Other than offering a "one stop shop" for cover songs documented on various other sites, what value is added by the efforts of SHS?


The situation that triggered this pondering is a performance that was only sourced from an "official" database, namely no audio available to confirm or documented editor "ear witness" to audio that's no longer available. In other words, our research was merely coping the info from one site to SHS. In a case like this, SHS can't be a source that attests that Artist covered Song but only that Source attests/attested that Artist covered Song. Where is the actual editor confirmation? In that case there's no audio, no images of the release other than the cover art. Where is the value added to elevate the SHS entry from a simple copy/paste to an independent source?


I seem to see relatively regularly entries made solely on the basis of song titles and/or writing credits that turn out to be incorrect, Physical releases have incorrect credits and typos. Do "official" sites simply document what's printed on the releases, etc. or do they confirm?


Frankly, in the vast majority of (if not all) cases, I can't see how SHS can justify adding a performance without either an editor having some audio confirmation or "ear witness" testimony from a trusted contributor. There may be cases where a combination of trusted sources and available images may provide some rare exceptions.


Anyway, just some thoughts....

baggish

Editor
Posts: 3805

baggish @ 2021-02-15 00:54:08 UTC

I have come to realise that SHS is mostly interested in copying from one database to another. The "Song-writing credits must come from PROs" rule shows that this philosophy is ingrained at a quite fundamental level. Song-writing credits must be copied (and they must be copied from a fixed set of sources) and there is no need to research them properly. So it is no surprise if this philosophy is extended to other pieces of information. I do not accept this philosophy and this is precisely why I do not accept the PROs rule.

To address your main point, yes of course editors should somehow confirm what work is being performed.

______
Really wild, General!

Tar Heel

Member
Posts: 5771

Tar Heel @ 2021-02-15 01:21:54 UTC

Wow... First, thanks Bag for the response. I really didn't expect any.


Second, I can appreciate the "legal" vs. "actual" songwriters differences of opinion, but I consider that a different question. We have to default to one or the other, and the "legal" is rarely in dispute at any given point in time. On the other hand, the "actual" can range from "a near certainty" to "some people think". We can easily handle both by defaulting to one and noting the other in the comments.


As for my point and in contrast, I don't recall ever seeing an entry as a cover of "XYZ" with a comment that some people hear a cover of "ABC". i do see entries of a cover of "XYX" with a comment that the credits are wrong on the release, etc.


Listening to a sample of or the whole recording under review does nothing regarding the songwriting credits, but does confirm that the recording is a cover of Work and may confirm the Artist (depending on the editor's familiarity).


To use my efforts on SHS as an example, I do not file reports on or submit performances unless I have heard the entire recording (with very rare exceptions). On occasion, I do file reports with an "under review" qualifier to capture initial findings on site. Given that, not only have I researched the performance to some level but am also providing "ear witness" testimony of the performance.


Perhaps we need to capture "ear witness" testimony on site in some form. A processing editor checks one of four boxes to attest to:

a) Listened to the entire recording;

b) Listened to a clip;

c) Did not confirm via audio; or

d) Relied on contributor's audio confirmation.


The submission form should include the same:

a) Listened to the entire recording;

b) Listened to a clip; or

c) Did not confirm via audio.

Tar Heel

Member
Posts: 5771

Tar Heel @ 2021-02-15 06:50:30 UTC

Song-writing credits must be copied (and they must be copied from a fixed set of sources) and there is no need to research them properly.


Decided to expand a bit on this.... Given my level of SHS activity, including error reports, I still rarely question writing credits. While I will likely now find some "actual" writers that need to be moved to the comments and a needed revision to the "legal", usually I'm most interested in when the credits differ between the PROS and what's printed on the physical releases, liner notes, etc. of the original artist. This is especially true when contemporary reissues print the same "mistake". I used to suggest a comment noting the discrepancy, but the reaction was almost always hostile.


From FAQ:

https://secondhandsongs.com/page/FAQ

SecondHandSongs is the largest and most accurate database of cover songs.

I suppose we are the largest (how is this measured), but if SHS is just a compiler of info from other sites without independent confirmation, how can we judge "accurate"? Currently, I have little doubt as to "most accurate" regarding the details of the releases, but the performances themselves?

walt

Editor
Posts: 5775

walt @ 2021-02-17 09:23:59 UTC

To address your main point, yes of course editors should somehow confirm what work is being performed.


Luckily, we have TouTube, Spotify, etc to cover most cases. The remainder showing proper credits on credible releases also qualify IMO.

Tar Heel

Member
Posts: 5771

Tar Heel @ 2021-02-17 10:22:07 UTC

The remainder showing proper credits on credible releases also qualify IMO.


Curious, what's a "credible release"?

How does one know the credits are proper without audio confirmation?


Performances can be titled in one language but sung in another. Instrumental vs. vocal. Medleys not indicated by title. Wrong credits. Wrong work. Not a cover. I have run across all of these site.

Tar Heel

Member
Posts: 5771

Tar Heel @ 2021-03-12 04:40:09 UTC

With additional SHS activity while paying more specific attention to this question, I would currently say that most entries are made based entirely on song titles, credits, etc. without any audio confirmation (not even a sample).


While filtering out all non-cover related performances from external sources and compiling the cover related performances on a single site is certainly a valuable service, SHS doesn't really attest to A covering B and originally releasing on C.


The question then becomes, how should we publicly qualify the veracity of our information to users?

Tar Heel

Member
Posts: 5771

Tar Heel @ 2021-03-15 10:35:28 UTC

Had another exchange during which an editor took some offensive at my observation of him/her adding based on song title and undisclosed assumptions. Interestingly this editor effectively admitted this is exactly what was done and likely regularly does so.


After more consideration, this may be a more material matter that I had originally thought. Likely will start another related thread elsewhere....

walt

Editor
Posts: 5775

walt @ 2021-04-02 13:29:42 UTC

VV, this "ear witness" thing isn't a bad idea in itself. But if it would be accepted and implemented, we would be over one million performances behind... Smile


How well intended this all is, I don't think none of the editors want any extra work and speaking for myself, no extra coloring on our songlists, please.


The idea of a disclaimer is more Management level, I guess.

Tar Heel

Member
Posts: 5771

Tar Heel @ 2021-04-02 13:57:46 UTC

VV, this "ear witness" thing isn't a bad idea in itself. But if it would be accepted and implemented, we would be over one million performances behind... Smile


I considered the legacy entries, but how is this different than many more recent enhancements? I recently changed ~50% of the performances on a tribute album to instrumental and filed a report for an editor to review the rest. They looked like legacy entries made prior to the instrumental flag. Ear witness for existing entries can be provided retroactively by editors and CCs going forward.


How well intended this all is, I don't think none of the editors want any extra work and speaking for myself, no extra coloring on our songlists, please.


This is the root of my disconnect. I don't consider audio confirmation extra work but rather an essential and expected part of research. Such confirmation has always been assumed, hence my recent shock that this seems to be the exception rather than the rule. In hindsight, this explain a lot:


I kept being asked to submit "leads", but I won't submit a performance that I can't confirm via acceptable audio. What Management considered acceptable for entry wasn't acceptable for me.


Recently I provided releases on iTunes, etc. with audio clips during a discussion with an editor, who kept asking for "real evidence". He/she apparently wanted a Wiki article on the beast at issue, not the actual beast. I remained confused until this epiphany.

Pommel

Certified Contributor II
Posts: 70

Pommel @ 2021-04-02 15:02:20 UTC

A recent exchange got me thinking, is SHS an independently researched source for the works, performances, etc. on site or is SHS simply a site that compiles and archives the data found on other sources and sites?

Honestly: It must be both! First compile information to secure it over time and prevent it gets buried in history without ever being able to recover it. Secondly refine the information with research as much as possible. I don't se any value in not adding information that cannot for whatever reason backed beyond a certain limit (meaning: somehow the existence of a cover should be able to be proven) at the point of being added. Otherwise you just loose information that will never be added later on as the link to the initial source might be lost. Details always can be added once they are available. Hence this database should neither be one or the other alone.

Tar Heel

Member
Posts: 5771

Tar Heel @ 2021-04-03 01:30:08 UTC

Honestly: It must be both!


SHS can just be a compiler from other sources or it can be both a compiler and a confirmer. Recent experience suggests just a compiler.


First compile information to secure it over time and prevent it gets buried in history without ever being able to recover it. Secondly refine the information with research as much as possible. I don't se any value in not adding information that cannot for whatever reason backed beyond a certain limit (meaning: somehow the existence of a cover should be able to be proven) at the point of being added. Otherwise you just loose information that will never be added later on as the link to the initial source might be lost. Details always can be added once they are available.


I call a yet to be sufficiently confirmed performance a "lead". For years I have been suggesting various ways to best document these on site outside of a formal submission. They have come to nothing, so I ignore most and resort to imperfect tools to document the ones for which I offer notice. I have also suggested allowing editors to add an open placeholder entry to document the existence or possible existence of a missing cover, but again I don't recall any responses.


To make matters worse, every now and then worries over "backlog" rise to the level where submissions that were born (effectively and intentionally) in "Needs Info" status are rejected to clear them out, resulting in all that documented research being lost.

Pommel

Certified Contributor II
Posts: 70

Pommel @ 2021-04-03 17:08:47 UTC

Honestly: It must be both!


SHS can just be a compiler from other sources or it can be both a compiler and a confirmer. Recent experience suggests just a compiler.

Not sure why you have that view when I explained above that it definitely can be both by compiling information of other databases/sources and by adding more research it can definitely become a verifiable and researched source as well.

Tar Heel

Member
Posts: 5771

Tar Heel @ 2021-04-03 19:28:56 UTC

Honestly: It must be both!


SHS can just be a compiler from other sources or it can be both a compiler and a confirmer. Recent experience suggests just a compiler.

Not sure why you have that view when I explained above that it definitely can be both by compiling information of other databases/sources and by adding more research it can definitely become a verifiable and researched source as well.


I said it can either be a) a compiler or b) both a compiler and confirmer. A compiler can exist without confirming. A confirmer can (I suppose in theory) exist without compiling, but generally not effectively.


We can get track lists from one source, images from another, etc., but that's still just compiling. Just because Discogs has the same release details as 45cat, how do we know the info didn't come from the same person or one site just got from the other? Certainly the same info on multiple sites increases confidence in the information, but it's not true confirmation.


Even as just a compiler SHS adds value per our goal:

a) We sort out just the originals with covers and the covers and centralize these performances;

b) We centralize the details of qualifying releases, artists, etc.


We are not a primary source for songwriting credits, as we only list works that have an original and cover, and derivative works, etc.


We are not a primary source for discographies, as we only list releases that include an original that has been covered or a cover, and then mostly only releases that were the first to originally include the qualifying performances.


We are not a database for artist biographies, as we only list artists that....


Frankly, in my view we often lose sight of the primary goals. In reality, writing credits, release details, and artist bios are simply practical ways to differentiate works, releases and artists with similar names. We really only need enough details to serve this site purpose.


When I use the term "confirmation", I'm mostly referring to confirming that the performance is what SHS says it is regarding the work covered, the language being sung, whether instrumental or vocal, whether a straight cover, medley, mash, etc. SHS having an original release wrong is relatively minor, saying performance A is a cover of work B when it's not is far more material. Absent interviewing the artists, producers, etc. involved, we can only confirm most things via external sources or our own music collections, but the "is what we say it is" can be confirmed by listening to the freaking performance.

David King

Editor
Posts: 1484

David King @ 2021-04-11 00:07:18 UTC

When I first got started, I assumed that some kind of audio confirmation was necessary. Ideally, that means listening to a whole Youtube clip. However, depending on the source, most of the time I simply skip through the track, to weed out quality and other issues (sometimes a Youtube clip features a whole album). I save the "whole listening experience" thing for when there is a material question about the cover. "Is this really a cover of the original in question?". It would take too much time, otherwise.


1. Is this an actual cover of the original in question? I've been tripped up a couple of times.

2. Is the language correct? I do dig further with Spanish, Portuguese, and Chinese, to see what the dialect is. Is it Continental, or Latin American, for instance.

Tar Heel

Member
Posts: 5771

Tar Heel @ 2021-08-02 23:46:00 UTC

A processed submission that recently caught my eye as it flowed through the "Participate" queue reminded me of this thread. The performance is a lead that I have returned to from time to time, including within the last few weeks, but since I found no audio and it wasn't particularly rare or interesting, I just moved on each time.


Now I see that it was submitted by a (at least now) "Trainee" and processed by a highly active Editor, without any apparent audio confirmation, i.e. based solely on title, credits and likely release date.


This is simply compilation rather than confirmation. Okay, there's no real doubt that SHS is a compiler of data from external sources. This is fine and provides value, but I again ask if some disclaimer to site users is in order.

SlimD

Retired Editor
Posts: 1377

SlimD @ 2021-08-03 05:28:55 UTC

Our website compiles data from external sources and our website also produces results of internal research. I see no problem with this and I see no claim that we need to disclaim.

Tar Heel

Member
Posts: 5771

Tar Heel @ 2021-08-03 06:02:08 UTC

Our website compiles data from external sources and our website also produces results of internal research. I see no problem with this and I see no claim that we need to disclaim.


To clarify to what SHS attests. It's now rather clear and indisputable that SHS attests that some external source(s) claim/suggest that A artist covered B work which was originally released on C release. Without audio confirmation SHS does not (and cannot) attest that A artist actually covered B work....


The distinction is important.