Why the old digital units in some studios?

  • Thread starter Thread starter expatguy
  • Start date Start date
SonicAlbert said:
I keep thinking about picking up an SE50. Just never quite made the jump and did it. At this point I'm kind of maxed out on outboard fx anyway, and have more fx boxes than I have aux sends anyway! So I just have to control the spending urges...

Go on! You know you can't resist! I've seen them going for $150AUD thats about the same price as a 57, or half a dozen mic leads! :D
 
dr.colossus said:
Go on! You know you can't resist! I've seen them going for $150AUD thats about the same price as a 57, or half a dozen mic leads! :D

You are clearly a bad influence on me. :eek:
 
People use them because they sound good! The reverb algorythems (sp?) in some of those boxes are great. If you have 32 bit converters and a 384KHz sample rate, and your algorythems suck...your 32 bit converters will still sound bad. OTH if you have 16 bit converters and superb algorythems, chances are you'll still have superb reverb...unless they really screwed up the analog part of the chain really bad. Chances are though, if they have great algorythems, they probably have a great, or at least good analog piece of the chain.
 
expatguy said:
Why the old digital units in some studios?

Some of the older reverbs just sound “better” to many people. The REV7 and even SPX90 were used on countless albums. The older Lexicon PCM70 (heck, even the PCM 60 which toped out at 10kHz) run circles around many of the latest generation (Made in China/Malaysia) Lexicons. I prefer the simple LXP-1 to any of the MPX-whatever things I’ve heard in the last few years.

The sampling rate and bit depth may be greater on the latest whizz-bang, but it’s all about the algorithms. This is where spec sheets will really screw you up.

One thing digital does very well is reverb, echo, etc. In this role the processor doesn’t treat the whole sound… it just adds to the direct sound at comparatively whisper levels. Consequently, you don’t need a 24-bit, 20-20kHz processor to add spatial treatment to the direct sound. Even running directly through the processor and using the wet/dry mix level (rather than through an effects send) the direct sound is passed (hopefully) unchanged.

The SPX90, which tops out @ 12kHz still passes the direct signal @ 20kHz. Same with the LXP-1 and Alesis MIDIVERB II/III which top out wet @ 15kHz. The frequency response is not really a limitation. Natural reverberation doesn't even reach these frequencies in real life. 12kHz is very bright. 18kHz and up is inaudible for most adults... many are lucky to hear 16kHz.

I for one, jumped off the planned-obsolescence train years ago. Manufacturers introduce new products whether they’re needed or not. Many of the names we trusted so many years ago have changed to the degree that you can’t just run out and grab the latest offering and be assured of the quality.

Sales brochures will say you need hundreds of presets and eight effects running at once, but if you just need a convincing reverb you can’t beat some of the vintage units from the mid-80s to ‘90s.

My first digital effects processor was an SPX90 I bought new in 1986. Some of the bells and whistles, like the parametric EQ, pitch shifter, and compressor… well forget about it. But the reverbs and delay are quite versatile and usable. Though I have a rack full of newer (not the latest) processors I still keep the SPX around.

My newest processor is the Midiverb IV... gosh, I guess it's been a few years since anything really caught my attention.
:)

Ah... here it is... A great article in Sound-on-Sound I ran across since last time we were discussing digital processors -- by Paul White... very good:

http://www.soundonsound.com/sos/1996_articles/aug96/improvedreverb.html
 
Last edited:
Just to confirm one point....

I'd like to confirm one thing here if possible.

When I started this thread I was under the impression that digital effects units would *totally* resample, say, a 96kHz-24 bit signal, resulting in a returned signal of only 44.1kHz-16bit. **overwriting** the original signal (resulting in a "diluted" signal that would then be resampled on aux in to the 96khz setting of the project.)

HOWEVER, this idea was wrong because the original signal is *not* overwritten, so the only important issue is the sound of the effects unit.

Gearmasters, does the Grasshopper now grasp the concept correctly?
 
expatguy, I don't understand how you didn't understand it. What's all this about overwriting, and sampling at 24/96k and then downsampling/dithering to 16/44? It just doesn't make sense.

In a typical effects box the analog signal appearing at the "In" connectors is converted to digital, processed internally, and then converted to analog again before being sent to the outputs. Perhaps some of the earlier units would send the dry signal through an analog signal path, and convert a stream to digital for fx processing. But the basic concept is the signal gets converted to digital, processed, and converted back to analog. If the unit has digital I/O, there may be no AD/DA conversion necessary, but the processing stage would be the same.

Chances are, the internal processing will take place at whatever the sample and frequency rates of the converters are. Not always true, but usually.

It's a lot simpler than what you are making it out to be. At least, as far as we as users need to know to use the tools. I'm sure designing and coding an effects processor box is quite complex.
 
Some of those older units just sound good so my guess is that keep them just for that reason. I still have my Ensoniq DP/4 that I bought brand new years back (ouch!) I also have a still use a Roland DEP-5.
 
SonicAlbert said:
expatguy, I don't understand how you didn't understand it. What's all this about overwriting, and sampling at 24/96k and then downsampling/dithering to 16/44?

Yeah, I'm embarassed in hindsight. In a totally analog situation I would never have made this mistake. But I somehow got the impression that the process worked differently in a digital processing chain. I was seeing it as a totally linear process instead of parallel processing. Looking back, I'm not sure why. Still, I'm sure I'm not the first person to give all those digital numbers a significance they never had. (Never was good at math, anyway :confused: .)

Thanks for all your input.
 
Thanks, Beck, for the detailed post, especially about the SPX90. They are going for a song on the auctions here. May as well get one for a song or two of my own. ;)

Beck said:
Ah... here it is... A great article in Sound-on-Sound I ran across since last time we were discussing digital processors -- by Paul White... very good:

http://www.soundonsound.com/sos/1996_articles/aug96/improvedreverb.html

Thanks for the link. I stumbled across it a year ago, when I wasn't really thinking about reverb all that much. Recently, when I actually *wanted* to find it, I couldn't remember where I'd seen it. Great article. I already followed the advice with my Sony MP5 and got a much better reverb than any of the presets.
 
Back
Top