How big of an impact can a 0.951% difference in speed have on the listener?
For a long time, we couldn’t really tell, that’s why we kept on using our modded PAL Mega Drive with region switch mod and swapped oscillator.
And yet… after a couple years of activity, people have started telling me they could hear that something was off in our releases and the best description they could give is that some instruments had wrong pitch.
They were, of course, right, because changing the speed at which a track plays means you end up changing the pitch as well.
At the 16-bit Audiohphile Project we could never tell this difference, but once more and more people started pointing this out, we just couldn’t ignore it anymore and we had to do some through investigation to understand what was wrong.
Mind you, those were times where something like MDFourier didn’t exist so we had to use our ears and listen through different tracks over and over again.
Eventually, it became clear that there was indeed a difference which, once you noticed it, you couldn’t just stop noticing it and you could hear it all over the work we’ve done.
At the time we were releasing games at full steam so we had a pretty substantial library of games which had to be ripped again.
This was really an eye opener for me: up until that moment, my goal was just to share Mega Drive soundtracks recorded from original hardware in the best way possible, but it became evident that what I considered “best” wasn’t good enough for our community and that’s when our project had its first big change in scope: it wasn’t anymore about myself sharing my recordings, it was about listening our community and improving over and over again to meet their (very!) high expectations.
If there was something which our community could hear which was wrong, it was time to step it up.
So I decided to do lots of research and eventually came up with using an early Japanese Mega Drive I , revision VA1 which needed a fix due to SEGA making an error in its output stage.
This wouldn’t have been possible without the help of Ace which wrote an exceptional thread in the Sega-16 forums about all the different revisions of Mega Drives and the fixes needed for each one.
And so our adventure continued and we had to introduce the concept of “Remasters” which were our old releases re-recorded from scratch with the new hardware.
You might think this would be the end of the story (and it could have been indeed) but our project managed to draw the attention of another group of people that I never thought they would use our material for their own work and this would eventually lead us into re-thinking everything we did up until then… again.
(Quick note: rather than making a single, huge post no one will read through, I have decided to make a mini-series of articles about the evolution of videogame music preservation and the role 16bap had in it. Several announcement will be made afterwards to keep everyone updated with the most important things going on in our project. Please stay tuned: huge changes are coming and we’re going to need help.)
On the 19th of October in 2012, the 16-bit Audiophile Project was born.
Its origins and goals were humble: in a world where soundtracks available on Youtube and forums were recorded mostly from (inaccurate) emulators, 16bap wanted to show how different real hardware sounded, and do it using the best recording equipment I could personally afford.
This wouldn’t have been possible without the huge efforts made by Project2612, who provided VGM logs of most of the Mega Drive / Genesis library, and Deadfish Shitware, who not only provided a software to play the VGM files on original hardware, but even modified it for us so that we could achieve more accurate timing.
And so, off we went, ripping and sharing music with the world recorded from authentic hardware.
I was astonished to find mixed responses to our work: there were those who were (and are) delighted with our releases and those that felt the sound was muffled, bass-heavy and lacked sparkle. It turned out that there are people that have never heard how a real Mega Drive sounded and their experience was limited to emulators such as Gens and KEGA which sounded way different from real hardware.
This was interesting to me, because while people each have different tastes and ideas on what kind of sound signature they enjoy, I understood that it was important to archive music in its original form, the way it played via its own circuitry.
I started wondering not only about recording quality, but other important aspects such as achieving perfect timing.
For casual listening purposes, no one would notice a 1% difference in speed (and, as a consequence, pitch), but when an expert ear starts listening to certain tracks critically, that 1% becomes easily detectable.
This became apparent after few years when some people started questioning our work and showing evidence that something was off.
Little we knew that there were people so passionate that went to the length of listening critically to our work and comparing it with their own Mega Drives. Their feedback would lead to one of the turning point in our project, one which would eventually lead us into re-recording from scratch our releases in what became known as “Rematsters”…
We’ve been hard at work with different groups on several things concerning our project (among many other aspects of videogame music preservation) and we’ve all been working towards something very big and important which will be announced in due time.
Meanwhile, I’ve added a disclaimer to our Release page: as we’ve been knowing for quite some time, all the “pre-remaster” releases have inaccuracies in their speed (and, as a consequence, their pitch) due to us using a PAL Mega Drive, despite the 60hz switch and crystal swap.
All the releases affected have been marked as such.
This was long due and we came to a point where we were forced to specify this because while even for the serious listener a 0.951% speed difference will probably go unnoticed, on the other hand our releases have been used for critical work (emulators, FPGA implementation, etc.) where even less than a 1% difference in speed has a huge impact.
We want to sincerely apologize to all the people who’ve relied on our releases and now have to double check their work to make sure that they were using either a Remaster or post-remaster release (all those after 3/11/2014 are very close to cycle accurate, with a discrepancy in speed of 0,00479% – newer ones will be cycle accurate).
We’ve been going through very big changes during those 8 years (yes, it’s been that long) and we’ve accumulated an incredibe amount of knowledge, but there’s always something new to learn and discover and now we’re going back to drawing board, looking at how we rip and record games from the Mega Drive, because we can’t afford to make any more mistakes and must make sure that our work is not only done through the best audio equipment possible, but is also done rigorously, making sure that all ends meet.
This means we’re re-hauling most of our own workflow and will make sure to be as transparent as possible as to what we’re doing and how we’re doing it so that people know exactly what they’re getting.
All releases are now on hold until we’ve untangled the mess and come up with something we know for sure that is completely accurate (or accurate enough that the differences are so small as to be negligible).
Thanks to all of you who’ve been following us and using our work and to all the fantastic people we’ve had the pleasure of working with so far.
We’ll be back with very substantial updates (hopefully) soon.
Katcho is back with another big release: EX-Ranza, known outside of Japan as Ranger-X!
This is quite a big release and with 4 bonus tracks which are unused during normal gameplay.
Also, this marks an important milestone in our project as this is the first release made using the new and improved VGM Player, courtesy of the immensely talented Deadfish Shitware. We’ve already discussed what’s changed from the previous version but the two big things, to put it short are:
1 – The old VGM Player’s speed was off by 0,00479%. Yes, that’s a ridiculously small amount and I’d dare anyone telling the difference, but there it is. Now we have perfect timing down to each cycle.
2 – We’re now able to completely disable video circuitry while recording, thus reducing noise to an imperceptible amount.
Katcho himself reported that he really couldn’t tell a difference from our previous VGM Player, which should tell you a lot about how we’re splitting hairs here, but as people have been using our releases to improve their own work (mainly emulation – including FPGAs) we now take it upon ourselves to provide the best quality and most accurate rips of Mega Drive material.
I hope I’ll have some time in the near future to write a small article about a recent discovery which has put us even closer to our ultimate goal.
A small hint: in a previous article, where we talked about MDFourier, we said how we made a complete shot in the dark by choosing a particular Mega Drive revision along with a fix for its output stage, hoping that it would come close to what a developer kit would sound like. Well, we have now one evidence that we might have done just the right choice.
Thanks to Katcho for another contribution and all of you for following us and stay tuned for more information!
Get ready for another contribution from Richard Mixin: Whip Rush!
With this release we’ve officially run out of contributions for now. I know Richard is working on others but those might take some time before they’re ready.
Meanwhile, I’ve been thinking about another way of contributing to the project so that we can get back archiving music ourselves with a bit of help from our community.
I’ll flesh this idea out in the following days and make a proper post to let you know how you can help out!
If you’re interested in contributing and you feel confident in your audio knowledge and recording equipment, we’re always accepting new contributions!
Just make sure to adhere to our guidelines and you’re good to go!
Albeit with a little delay here goes this week’s new release from our contributor, Richard Mixin!
His release include even 3 bonus tracks which look like to be unused during normal gameplay, it’s great to have them included!
Let me take the opportunity to remember to all of our contributors that while including beta/unused tracks is not required, it is absolutely welcome!
I’m also thinking about ways to improve my own workflow and I may need in the future some collaborators which, similar to our contributors, will help us out in archiving Mega Drive’s soundtracks, but won’t require any technical knowledge about recording and all that scary stuff in our terribleawfulincomprehensible (vastly) improvable guidelines.
So stay tuned, hopefully I’ll soon make my mind up and let you know about other ways to contribute to our goal!