Trying to shore-up my backup strategy and am looking for a “best practice” for archiving projects for long-term storage. What’s the method to your madness? 
Things to account for:
Go!
I struggle to get a coherent strategy
I’ve got a thunderbolt black magic drive dock. There’s 4 slots for drives so currently I’ve got 2 ssd drives with samples on, a hard drive dedicated to time machine backup and another hard drive for storing old projects. When the hard drive is full, I take it out and get a new one.
I do also have a spreadsheet with all my production music tracks and albums on where I specify where the tracks are stored because I have uploaded a lot of them to cloud storage.
With regards to older projects not working etc. I’ve got an older mac pro which has a number of different systems on it going back to tiger. I think having older systems that you can launch is great but it’s harder to do with newer macs because they don’t have PCI slots. I dont fancy running the system via thunderbolt dock because thunderbolt cables can be pulled out pretty easily and that could lead to problems.
I hear you! Though it seems you’ve got more of a “backup” strategy than true archival. (If you lose your old Mac Pro for whatever reason, you may lose the ability to open the old projects, no?)
I’ve tried some hard searches on this topic and have yet to find a solid strategy bound in reality. Many seem to involve bouncing out tracks as wav files. But do people with 100+ track projects really bounce out all individual tracks as dry wave files, and then with separate effects tracks for each?
Rendering all the Tracks as individual audio file (and maybe stems too) is your best bet against tech obsolescence. But as you note that can be a lot of work. Render in Place will do a bunch of selected Tracks in one go, so start it up and go get coffee (or adult beverage of your choice).
When deciding what to render or not consider the risk of what might get lost. For example loosing access to a VSTi could result in loosing something you have no way to recreate - good candidate to render.
But does it really matter if a reverb plugin goes south? It truly might, but more likely you can find (or already have) a functioning reverb that will work just as well. Is it worth it to archive that reverb as audio? Depends on how likely are you to ever need it, and if you do how much effort would it take to recreate from scratch.
Keeping old versions of Cubase installed (or available to be installed) also gives you a potential path to access old plugins that don’t work in C10.
Also don’t discount the potential upside of trying to recreate something but ending up with something better.
Cautionary Tale:
The are a bunch of us who have shelves full of magnetic tape reels which are unplayable. The adhesive on the tape is deteriorating and tends to stick to the recorder’s transport. This can be temporarily mitigated by “baking” the tape and transferring the audio before the goo returns. But this is specialized since the tape needs to be held at a temp higher than room temp, but lower than the lowest oven setting. When these recordings were made no one had any idea this issue would crop up.
Thanks for wise tips and insight, raino!
I agree that rendering VSTi is a must! (Have been bitten by that one before!)
Is it worth it to archive that reverb as audio? Depends on how likely are you to ever need it, and if you do how much effort would it take to recreate from scratch.
Yeah - I’m thinking archiving effects will take a lot of effort and drive space. However, I can imagine going back to reconstruct a mix could be a potential nightmare! It’s one thing to open up an old mix for nostalgia’s sake… it’s another when an opportunity arises where you get a request for a special stem or alternate mix on short notice! 
In the abstract it is all about risk. In each person/project’s specific circumstances what risk do you create or avoid by deciding to archive something. Then how much does it cost (time, money, effort, lost opportunities, etc.) to mitigate that risk and what is the cost if the risk actually occurred. If it costs less to mitigate than what you stand to loose then archive; if it costs more then don’t.
" But do people with 100+ track projects really bounce out all individual tracks as dry wave files, and then with separate effects tracks for each?"
Yep that’s what you do!!
Not only that but I keep archives up to date in format as much as possible as tech changes right back to when I stated paid studio work in 1983 on 16 track analogue tape and at least a couple of times a year I`m asked to revisit the archive from what is now nearly 40 years ago as the master tapes that they left with then are now deteriorated.
Its lovely to see the faces light up when I can help!!!
I regard it as the boring part of being pro.
Hippo
I can imagine.
I’m totally convinced that my inaccessible tapes contain wondrous sampling opportunities.
Probably best to never find out otherwise. 
Yikes! That’s what I was afraid of! 
Lol …Does playing back 100+ individually rendered effects tracks really sound the same as the original mixdown/master track?
I have a system that more or less works for me, although I will admit it has evolved quite a bit over the years due to some mishaps and misfires, but basically it involves categorizing my projects based on the probability that I will need to edit them again, including the level of editing that might be required in the future. It’s essentially a risk assessment. Your own risk assessment will of course be different than mine.
Categorize project step 1 – amount of detail needed to archive: group/bus level, track level, or fx level? (I’ve found I can get away with different level of detail needed, depending on the client, situation, contract, etc.) Group/bus level is the easiest obviously… all the way to FX level being the most nightmarish to archive.
Categorize project step 2 – timeframe needed: how much editing is likely in 2 years, 5 years, and 10+ years? I’ve found those timeframes to line up with compatibility/ease of restoration for me re: OS issues, major DAW changes, major plugin changes, etc…
Based on the categorization and risk assessment process, I do entirely different things with them on a sliding scale of effort:
For simplest archiving, I will simply bounce out logical groups/stems/buses and maybe a couple of critical tracks like edited lead vocals, for example. That’s it. Cross fingers I don’t need more than that. This logically scales all the way to meticulous, surgical exporting of tracks, including wet/dry, and critical fx in case I need them. This takes a long time and is a pain in the !#$@#%@#.
In ALL cases, I of course keep the actual project and source/raw files AND I try to keep backups of almost ALL of my most important plugins in case I have to try recreate a DAW environment of some time in the past. I still have old backups of almost every OS installer, DAW, plugin and library I have used going back 15+ years. I figure if I have to, I can go on EBAY and find someone willing to sell me some ancient old computer if I absolutely need to. Cross fingers that never happens.
I have learned to be more selective with which plugin developers to trust longer-term. I have made some poor bets in the past and have been burned. Heck, I’ve been burned by DAW developers too. So I’ve tried to keep my DAW much more simple these days… I’ve consolidated around fewer developers I like, and hope they stay around for a long time. I have accepted the idea that “less is better.”
Lastly, I have a pretty robust backup routine… I use the 3-2-1 backup methodology for just about everything.
It’s not a perfect system, but I treat my projects/content like a business (because it is a business) and it takes time and money to do it. Any time I have slacked off (and I have indeed slacked off a handful of times), I have invariably been burned. YMMV
I have the recording drive, backup drive, backup for the backup drive and burn DVD’s at the end. I label the sample rate in the folder name, keep all wave flies bounced at zero. If we recorded 10 lanes of vocals and did a comp track then all 10 get bounced with and without effects. You have to take a lot of notes and put that in a document but it can pay. It’s got to be able opened in any DAW because when you send it out more than likely they won’t have the same type and version of the software. WAV files bounced at zero to the end will open in any software.
One’s already getting dizzy and flooded with today’s sounds and sampling possibilities? So best let the archives tapes rest in piece?
And once sampled anything, there’s also many ways to manipulate it with variaudio and plugins. So don’t feel that you’re missing out if it happens you’re not able to access your old backup tapes? 
This makes sense.
However, let’s say you have 10 tracks all sending to a single reverb aux track, resulting in a total of 20 archived tracks. Now when you go to reconstruct the mix in the future, you’ll have to play the original 10 tracks + 10 individual tracks of soloed effects all together, right? Even if they are bounced at zero, isn’t there now grand opportunity for phase issues and volume build-up resulting from stacking of the 10 individual reverb tracks?
I would be curious if anyone has done some testing to see how closely a reconstructed mix from a large number of bounced files matches matches the original master track.
In software industry it is fairly common than you create virtual machines (like vmware or XEN) where you install the needed software and their configurations. This approach should work for a DAW too.
I use wavpack to compress wav files (40-50%) space saving
Then move to a Nas which is also backed up (using simple robocopy script) to a bitlockered USB drive I store at the office (so I have some form of off-site backup)
It actually depends on how far your client wants to take this? Nowadays you can save anything that has it’s own input. Overhead mics, ambient mics and mics that capture the overhead and ambient mics and maybe this is routed to separate buses that have their own separate busus, etc…?
The question you need to ask yourself and your clients is what you would need to actually recreate the exact state of a given situation and how far they’d be willing to go to make sure this state can be recreated at any given time? Because saving a simple bounce clearly takes less time and maintenance than saving a complete state of maybe a 1000+ separate audio tracks? The backup process remains the same for 1-~ 3000 files. It just takes longer and takes more time and demands more monitoring and disk space? You’ll just have to agree with your clients and charge them accordingly !
Frankly? Don’t use a Mac. Apple is becoming abominable when it comes to hardware compatibility.
Next level… backup everything to a series of USB external HDs which you rotate and occasionally power up to make sure they hold data. That’s the cheapest way to store copious amounts of data reliably for the long haul.
Frankly… the OS platform & hardware should have nothing to do with your session archival strategy. FYI, A solid archive will be able to be constructed on any DAW on any platform. Who knows what state your OS or DAW will be in 10 years from now.
Thanks, for chiming in though…
A great thread, btw.
Especially because it raises the awareness that backup is not archive. The challenges are very different as clearly pointed out here.
Hippos contribution is most professional. I wish I would be as consequent as he is!
I have not tried it for phasing. I have not had to reconstruct a mix exactly. Who knows what system you will be running in the future. New software may not open old projects. My guess is WAV files will be around for a while and if you save everything you should be able to load it back. We’re just dust in the wind anyway.