With AI still in its infancy, how long before RX or SL may be able to remap an overcompressed mono or stereo recording and effectively restore its dynamics?
It may well be getting old fast. There is only that much you can train it upon.
Everything will still depend on your audio recording skills.
We have two tutorials in work which compares them both quite in-depth. One is almost ready for realese on our youtube channel.
The best is to demo them on specific tasks that you need them to do for you.
Clicks pops etc RX is a quick fix and may have better algorithms, SL you can come close by tweaking the FFT.
For spectral repair and ARA you can depend on SLP, it is clearly superior.
There may be projects where you may need both. We almost end up using both on all our film and documentary projects before deciding for ADR.
Did a comparision on some of the processes between RX and SL.
@Rajiv_Mudgal looking forward to seeing your comparision too!
Assuming you’re referring to old recordings (from the 1970’s-1990): I have been working on restoration and believe-it-or-not I’ve gotten some good results. One issue is a lot of older recordings tend to have peak audio in the mid tone range compared to today recordings (where it’s a lot bassier and has a lot higher harmonic peaks especially in the tremble range).
Restoration can definitely be done, I’ve remastered small segments of classic films(1970’s from what’s on youtube) and upmixed them to sound like a modern day blockbuster(a film like The Avengers where the punches and kicks are hard hitting and sound crispy and punchy).
Here’s a few things I’ve learned. In order to truly remaster audio(especially a film from decades ago), you need all the sources separated. For example, when they remastered the star wars film, it seems like they remastered the entire audio track by Equing and compressing the entire audio track, and what ends up happening is that everything is mixed the same way in this monotone 1 dimensional aspect… I also noticed the same thing with the classic legendary film “Scarface” where it was remastered, but because it was remastered from a monotone perspective(it seems like everything was remastered on the same audio track with EQ and some compression), it sounds outdated. It is definitely doable to remaster a film like “Scarface” to where the audio sounds modern, but in order to do that you need each source separate. If you watch and listen to the classic film “The Matrix”, the gunshots and the bullets sound punchy and crispy, whereas the remastered version of “Scarface” loses a lot of that punchyness and crispyness in the gunshot sounds because it was mixed in a monotone-istic way. In order to truly remaster a film like “Scarface” then all the sources(like gunshot sounds) needs to be separated.
The problem with truly remastering a film like Starwars and Scarface is that it takes a long time to do that. If someone(or some hollywood company) has the budget then those classic films can definitely be upmixed to sound modern.
Excellent comparision
[quote=“Joey_Kapish, post:24, topic:129936”] Assuming you’re referring to old recordings (from the 1970’s-1990): I have been working on restoration and believe-it-or-not I’ve gotten some good results. One issue is a lot of older recordings tend to have peak audio in the mid tone range compared to today recordings (where it’s a lot bassier and has a lot higher harmonic peaks especially in the tremble range).
Restoration can definitely be done, I’ve remastered small segments of classic films(1970’s from what’s on youtube) and upmixed them to sound like a modern day blockbuster(a film like The Avengers where the punches and kicks are hard hitting and sound crispy and punchy).
Here’s a few things I’ve learned. In order to truly remaster audio(especially a film from decades ago), you need all the sources separated. For example, when they remastered the star wars film, it seems like they remastered the entire audio track by Equing and compressing the entire audio track, and what ends up happening is that everything is mixed the same way in this monotone 1 dimensional aspect… I also noticed the same thing with the classic legendary film “Scarface” where it was remastered, but because it was remastered from a monotone perspective(it seems like everything was remastered on the same audio track with EQ and some compression), it sounds outdated. It is definitely doable to remaster a film like “Scarface” to where the audio sounds modern, but in order to do that you need each source separate.
If you watch and listen to the classic film “The Matrix”, the gunshots and the bullets sound punchy and crispy, whereas the remastered version of “Scarface” loses a lot of that punchyness and crispyness in the gunshot sounds because it was mixed in a monotone-istic way. In order to truly remaster a film like “Scarface” then all the sources(like gunshot sounds) needs to be separated.
The problem with truly remastering a film like Starwars and Scarface is that it takes a long time to do that. If someone(or some hollywood company) has the budget then those classic films can definitely be upmixed to sound modern. [/quote]
Thanks for this excellent, detailed reply. Unfortunately, your solution won’t work for my problem 1960s recordings because the original mono 2, or 3 track analog tape stems are either lost, trashed or had never existed. A perfect example is composer Dominic Frontiere’s lengthy “Slumbering Giant” suite. It was used to score parts of this episode of producer Leslie Stevens’ Stoney Burke TV series. https://www.youtube.com/watch?v=cKhwKbk5vXU
And used again the following year by music editors John Elizalde and Roger Farris to score this episode of Stevens’ Outer Limits TV series.
Happily, it was one of the best of Frontiere’s many superb suites and cues from his TV music library (which he was much earlier commissioned by TV producer Quinn Martin to create for later use in Martin’s forthcoming TV shows) to be released as a 3-CD set. https://www.discogs.com/release/3519252-Dominic-Frontiere-The-Outer-Limits-Original-Television-Soundtrack
Sadly, while there’s very low distortion and noise on the tracks from that CD set boxed set, the dynamics were originally hit with substantial compression. All of my favorite 60s soundtrack recordings on CD are likewise so compressed, as is much of my favorite pop and jazz music from that and later decades.
This is exactly why probably all I can do is to hope that some day AI-driven apps or plug-ins will be able to very precisely remap-AND thereby regenerate-a “virtual” clone of a recording, minus any deliberate, and perhaps well-intended, but excessively applied compression, which usually killed the original recording’s dynamic range.