isnt that what LLMs are for? that anyone can "explain" complicated things to strangers in social media?

 

what irritates me more (in the developer video) is that he claims the sound of the 900/950 would mainly be made from its "anti aliasing filter".

from what i know, the sound first of all comes from the variable rate DAC. which of course uses antialiasing.

when somebody talks about his 900/950 emulation i would expect that he explains how he implemented the variable rate feature (most likely by high upsampling or eventually even by using true multirate streaming with in its plug-in)

 

It's to save people time when they have to go lookup the max sample length in the first place to complain about it.

If someone quotes you the exact sample length of a sampler down to the decimal place, they went and looked it up too. This is the sample length of my "ugly duckling" S2800: ~2.97 minutes stereo @ 22.05 kHz, ~1.48 minutes stereo @ 44.1 kHz; and there is zero chance I would be able to tell you that with any specificity about any sampler in my rack or not. Why? Because it doesn't matter. It is not a number you have any reason to commit to memory, especially about a sampler you don't even own.

 

Last edited: Dec 20, 2025 at 11:52 PM

there is that small difference between looking up numbers and specs on one side, and copying pages of texts including arguments from a bot pretending it would be your own on the other side.

btw, could the 900 not be upgraded a bit like the 950? the latter could be extended for about +150% when i remeber right. i never did this though.

 

If it's the same as my S950 then whatever sample rate you choose you have to multiply by 2.5 to get the actual "real" sample rate.

So 16 x 2.5 = 40khz max for the s900

 

The problem i have with blaming AI for that is because it is misguided. It was the same thing before when it was Google. Garbage in Garbage out. It's just faster now.

Anyhow, you could upgrade the sample RAM and that would give you some more sample time, but we aren't talking any considerable amount. The real "upgrading" you can do with them is with Gotek floppy emulator, sdsi2sd, or now the ZuluSCSI. But those do not help with the RAM/sample length situation.

 

I havent heard about memory upgrade boards for 900, only 950 had them but there was a third party board called Marion so the sampler could have a different filter and 16bit resolution for A/D and D/A, but internal operation left 12bit, plus a higher dynamic range

Not only is it misguided, but can also present lie with a straight face even when guided corretly as a child. But that’s not the point here. The point is that any tool should be in the right hands. You need a certain amount of experience and inteligence to use it properly. Even a hammer meant for driving nails can be used to attack people in the wrong hands. So the real problem lies with the person useing these LLMs, not the tools themselves.
Looks like some people should be banned from using them if they haven't reached some intellectual level, it would be a good idea to introduce an exam to gain access to them

 

Last edited: Dec 21, 2025 at 8:06 AM

Either way, today's agenda has been to derail a thread about a plugin. nostalgia + moral superiority + arbitrary technical gripe.

Every time someone makes an emulation, it is the same thing.

 

I don’t see anything here that could derail things yet. So far, we only see intentions to do something with S900 fame. It seems to me that the developer chose the wrong approach from the very beginning, perhaps due to a lack of experience. Making such emulations requires being an engineer and being able to dig deep into the guts of the hardware for now its only a quick thrown together demo in Juce

What really derails a proper discussion is someone who overestimates their abilities and struggles with basic fact-checking.

 

Last edited: Dec 21, 2025 at 8:34 AM

Blast from the past... sorry bud, I dont even have a copy of that anymore. If you find one let me know!

 

Last time I had Akai S900/950 in my hands was ages ago and I can't remember. Does the sampler's OS really show 16 kHz (instead of 40) as maximum sampling rate?

 

yes it does but they call it bandwidth

 

Last edited: Dec 21, 2025 at 11:05 AM

Thank you Aymat!

 

why would anyone blame the chatbot for beeing quoted into a forum by a human?

 

A couple of months ago I watched a long intervew with Dave Rossum of E-mu (co-desinger of the Emulator line). He explained a lot of technical nuances regarding early samplers. I learned that the the 8-bit and 12-bit machines were doing far more tricks to get maximum possible sonic results out of 80's technology than I would expect. (E.g. things like companding /different bit precison for different parts of signal's amplitude range/ or coding only the change /delta/ in the signal in comparison to the previous sample with lower bit depth instead of coding absolute values for every sample - all this to squeeze more seconds of usable audio in that tiny memory.) Add to this different techniques of pitch shifting - from changing DAC clock rate to various techniques of digital resampling.

After seeing this I realized that it is unrealistic to excpect any exact emulation of those early samplers nowadays - devs usually focus on one link of the sampling chain, ignoring others. Too much effort to achieve the full goal. TAL-Sampler *may* be somewhat close, but I didn't gel with its UI.

Personally, I've been waiting for an *exact* VST emulation of AKAI S1100 (of course with updated sample loading/storing) for the past 25 years. Not gonna happen.

Still, I wish the VU900 project to grow and deliver a usable plugin.

 

Last edited: Dec 22, 2025 at 12:46 PM

any recent one? I'd like to hear that interview too
I agree with you. emulating these early digital stuff is a much more complex task than many people think. They had to rely on many seemingly unnoticeable tricks to achieve results that cannot be reproduced simply by making a plugin by ear, as people often try to do. That’s why I said in addition to being a programmer, you need to be a good engineer with a lab of test equipment.

 

Last edited: Dec 22, 2025 at 12:37 PM

I think it was recorded 5-7 years ago. It was a long (certainly longer than an hour) online intervew. The interviewer was much younger than Dave. But I cannot find the link now.

 

if it's and old one don't bother I most likely watched it
here is something you'l like from 1989, I use it as a small encyclopedia on early samplers
https://limewire.com/d/Mhg3c#TJLKHy6xXn

 

While I totally agree with you guys, one can also make the argument that do modeling those tricks actually contribute to the sound in any meaningful way instead of just brute forcing the sound with a 'common' method so to speak? I personally think they probably do to some extent, even if it's quite minor and I'd like to see modeling done as deeply as possible. Because some people actually making music do pick on those differences and it is meaningful to them when using these plug-ins. But I can see also that it might not matter to some folks and understand that.

 

Last edited: Dec 22, 2025 at 1:06 PM

Interview with Emu Systems Legend, Dave Rossum at Synthplex 2022

Secrets of the E-mu SP 1200 with Creator Dave Rossum @superbooth 16