Hold Me Up (studio) Options

[dangum]

I have found what appears to be a studio version of Hold Me Up.

It will be uploaded in the next few hours.

[ThrowgnCpr]

Ah, good to know.

In your experience are you thinking that the soundcloud clip posted earlier has differences that are a result of frequency lost in mp3 conversion, combined with a distorting tape transfer? I never would have expected that. Would the reverb on the snare be wiped away so much? Also the symbols are more to the left in the new recording. The symbols stay dead center in the older mp3. I am not sure how these methods would get rid of sounds coming out of the left like that, and remove the reverb entirely.

But maybe the media five tracks are not the finished version. The media five hold me up is still slower than the zack and miri version. But maybe the tape transfer slowed the track down. At least we know Live are doing anniversary reissues now. Throwing Copper 25th anniversary reissue in 2019/2020, we may finally get the track.

I am still fairly certain that all the versions we have heard of Hold Me Up have been of the same take/recording. And the differences we are hearing are the result of a combination of factors including source format and generation, transfer process, and finally, delivery format compression.

First, let's consider what the recording was transferred from. Media Five probably received a first-generation copy of a 1/4" reel tape of the mix. That would have been the highest quality source format to transfer. Now, who knows how many times that tape had been played and how it was stored. They certainly weren't the original owners. My guess is that it was given to them recently. Over time, it could have stretched from excessive playback, or become brittle from poor storage conditions. A stretched tape could account for some of the duration discrepancies.

Most songs "leak" from studio sessions on a cassette tape. Cassettes of mixes were often run off for band members and studio staff as a way to assess progress. Sometimes copies were made from friends, or staff made copies without the expressed permission of the studio/band. Those tapes are copied, passed around, and around and around. Previous MP3s may have been sourced from multi-generation cassette tapes, which would have resulted in a big drop in fidelity. If you've ever made copies of cassette or VHS tapes, you know how the quality can drop, and [music] information gets lost.

The transfer process can play a role too. Assuming Media Five has high end archival equipment, I doubt that played much of a role in degrading the new transfer, but it probably factors in to a small extent, especially if we compare it to whatever hardware was used to transfer the old MP3s.

The stereo separation would have been affected by the source format and transfer process. That a lot of the sounds are dead center in the older version is no surprise. That could have resulted in the downgrade from 1/4" reel to cassette, it could have happened in the digitization process, and it could have happened in the delivery format compression. The latter plays a big role in many of the differences we've been debating.

MP3 debuted to the public as a consumer format in the early-90s. It is a lossy compression codec based on perceptual coding or psychoacoustic models. Essentially, the algorithm discards some data and reduces audio information that is least crucial to perception. It drops things like harmonics, reduces the frequency range, and truncates and simplifies sounds. Things like cymbal quality and reverb are the first to go. The early generation MP3 coding algorithms were pretty crude compared to the algorithms in use today. I think the most recent versions are less than 10-years old. Anyway, lossy encoding in general is a pretty brute way of reducing the file sizes, but those early generation MP3 algorithms were pretty crummy. They really reduced the information in the high and low frequencies, which greatly affected the quality of the cymbal sounds and in general made the songs sound muddy and glitchy. The latest MP3 algorithms when used well can actually do a pretty good job.

Lossless compression on the other hand seeks to reduce the file size not be removing data, but by simplifying how that data is stored. Think of a FLAC file as a playable .zip archive. None of the music data is changed, but its just stored differently.

In summary, I believe a combination of all of those things above account for the differences between the older MP3s and this most recent version. I'd go as far as to say that the short clip we hear on Zack and Miri is even the same exact recording. So much of the performance sounds identical to me. The difference is that the Zack and Miri version probably had a final mix from the individual source files (each instrument) and a proper master. And given the tonal quality of this recording, I've no doubt that it was recorded during the same final sessions for the Throwing Copper album. That alone could also explain why we haven't received it yet. The label would have owned it, which was Radioactive--an imprint of MCA that doesn't exist anymore. I don't know much about who owns all the rights to Radioactive's recordings now, but I would guess that it's complicated, and unreleased recordings on paid recording sessions for albums is never clear cut