The print audio has the correct frequency balance for the soundtrack since it has been recorded through a Dolby A decoder. Most of the laserdisc versions are extremely bright and are too shrill in comparison. However, the print audio is less dynamic than any of the laserdiscs, since stereo optical on 35mm film has very little headroom.
Soundtracks made with Dolby A noise reduction (which includes all the mixes for the Star Wars films except the mono versions) sound much too bright in their undecoded state. It is apparent that none of the early laserdiscs have been Dolby A decoded, hence their excessive high frequency emphasis. The RotJ soundtrack from the Special Widescreen Edition does have the correct frequency balance, so it is likely that it has been Dolby A decoded.
The best possible version of the 35mm stereo mix can be produced by using the early US laserdisc (since it has the most dynamic range) and adjusting its EQ to match the print/SWE versions. That way you get the best of both worlds; the greater dynamics and the correct frequency balance. This is exactly what I did for the 35mm soundtrack on Despecialized v2.5, but I haven’t yet put out one like this that is synced to 4K83.
I will release a track like that at some point soon, but for the moment, I would recommend using the SWE version, because it sounds extremely similar to the print audio, but with better quality.
Do any of the SW '77 stereo mix laserdisc audio tracks have the correct Dolby A-decoded frequency balance as far as you know, or do they all sound similarly over-bright? (As in, any of the pre-JSC remix audio tracks.)
There is only one digital source (technically two) for the '77 stereo mix, the 1986 Japanese Pan & Scan LD release with catalogue number SF098-1103 (as well as the re-release from 1991 with #PILF-1236). I speculated back in August that it may not have had Dolby A decoding applied and worked with a software engineer that I met through the stevehoffman.tv forum to decode it properly using a proprietary Dolby A decoder he was developing. The results were shared over at tswt forums, and my opinion is that it did not exactly work. Overall it was a bit too muffled-sounding. To me the original version was preferable.
My results indicate that it’s probably not as simple as saying that the LD’s are missing Dolby A decoding. While it may be true, there are other possibilities. These mixes were digitized at a time before a DAW would allow for easy manipulation of the audio track, as such, the harshness may be due to the more rudimentary ADC’s being used at the time and a limited ability to deal with tape noise effectively (and SW seems to have a lot of tape noise), or simply the EQ preference of the person doing the transfer. Even if Dolby A decoding is indeed missing, it’s possible (likely?) that the audio was eq’ed to make it more pleasing to the ear. Any sort of eq manipulation would make subsequent Dolby A decoding nearly impossible to perfect.
I was asked by SkyDude to do an improved version of of the 1977 audio for his SkyMaster edition, and I was able to achieve much better results by simply correcting the eq and using spectral noise reduction. I haven’t shared that version, as I was going to let that debut when SkyMaster is released.