Even more on the Reason's Sound Quality thingy
A few weeks ago I and an engineer friend made some simple tests on Reason's sound quality issues. We especially wanted to see if there is any sound colorization within Reason. So we started by preparing various loops and one shot samples of any sort. We triggered them with various Reason instruments (NNXT, NN19 and ReDrum) and exported them to wave.
Then we spectrum analyze both the source and result wave files with the quite complicated spectrum analyzer of Izotope Ozone. When we compared the result graphics we saw that there is almost no difference. I mean, there is %99,99 no difference between the graphics. The remaining difference was mostly about +/-0.001db at no specific place in the spectrum. So this means to us that there is no colorization or muddiness or any such problems with the audio engine.
Beside this we realized one interesting thing that might be considered as a problem: There was always a noise around -96db in the silent parts of the exported waves. Even when there is no sound was playing at all. This noise was becoming slightly louder if we added more and more devices (FX or instruments) up to -90db when there was about 50 devices. Theorically there shouldn't be no noise, just plain silence when there is nothing playing. We didn't go more in this, though, because it wasn't that important for us when considered the hardware devices we were using in the studio like Roland JV1080, Korg Triton, etc with always have much more noise in their audio outputs. Perhaps it is related with one of the FX processors in Reason...
Anyway, I'm happy to know that there is no serious problem with Reason in the sound quality department. I think that the usual "Reason does not sound fat/bright/clean" threats are the result of the quite simple FX processors that come with Reason. Ok, there are quite working tricks to make some cool and fat FX's by chaining them. But Reason is not fat "out of the box".
As a result you can always try the following tricks to help your sound:
- Use double reverb in series for your master "send" reverb. Use the first one as the early reflection part by setting the size very short like "-45" and use the second one as the long tail, setting the algoritm to Hall and keeping the size above "0". Keep the dry/wet setting of the first one as "pure wet (+64)" and use the dry/wet setting of the second one for the balance between the early reflections and the long tail. This is very essential trick for realistic reverb effects with debth. Think of the Waves Truverb!
- Use double chorus in parallel with different delay and rate settings for Left and Right channels. This makes a huge difference, really!
- Try setting up some interesting delay effects: Forexample if you want a pingpong delay with some modulation on the delay part: First connect a chorus device to one of the effect send outputs of the mixer. Keep the chorus in "only wet" state and attach a different delay device for each left and right outputs of the chorus. Connect one of the two delay outputs to left fx return and connect the other one to right effect return. Now you have independent delays for the left and right channels even with some modulation affect that is achieved by the "only wet" chorus effect before the delays.
- Try returning master effects to mixer via a mixer channel instead of the fx return inputs in the mixer. This way you can create fx feedbacks or you can add the other three master effects to the wet part of the master effect you connected to the mixer channel. Forexample you can add master reverb to the echoes of the master delay.
- And use compressors and parametric EQs to maintain a brighter, cleaner, more in control sound from any instruments.
Anyway you can use too many tricks that can be achieved in a real hardware studio in Reason by creative chaining of mixer and effects and other devices.
BTW, I'm not yet an owner of Reason but I'm definitely very excited about it! I everyday check this forum and this website for anything new. What I wait for is that, I wanted Reason to be just enough to be able to achieve the unique sound I'm into, alone without no help outside. I want Reason to have more special FX devices like lofi processors, surround panner, amp simulator, etc. Ok, you can always use rewire but this sort of usage is not for me. Because rewire kills the simplicity and completeness keywords that Reason is about.
Anyway, I don't complain and just wait. But I wanted you know that if Reason is not that complete yet, it does not engrave most of the negative feedback we see everyday in this forum. I personally am satisfied that Reason's sound is top quality right now, if not as special as everyone need.
Great suggestions iloyd - thanks! (and screw Macs)!
Rea question on this...
This sounds interesting--but it seems to conflict with the results presented by Robin Sarac and 'anj' (?)
about 4 weeks ago.
As I recall, there were fairly substantial gain differences between he different modules and the the source samples. how did you correct for these, or did you not encounter them.
i notice you didn't test Dr. Rex. In addition to
obvious gain issues, Dr. Rex seems to have some
sort of degraded playback.
Re: Rea question on this...
Gain difference is normal I think if you don't expecially adjust the volume to match the source waveform with the mixer or any of the sampler devices in Reason. And gain difference shouldn't effect the spectrum or cause any colorization.
Anyway, in the test we proceeded with normalizing both the original and exported waveform at the final stage right before analyzing the waveforms. And the harmonic content was %99,999 same. That means almost no coloration occures in sample playback. (probably at most, around +/- 0.01 db)
Yeah, we didn't check for it. It isn't a very important device for me really. But maybe I'll manage a test with it soon.
garbage in=garbage out
I remember a few months ago when I had a project I was working on that needed fresh, new sounds. I have the unique luxury of working walking distance from East West Soundwarehouse and a greater luxury of being formerly employed by them.
While I was there I saw the Creative Essentials refills - 3 CDs worth of sounds. The packaging says something like over 7 GB of sounds!!
I ask you, how do they get 7 Gb of sounds on 3 CDs? Doug Rogers (the president) told me that they use "lossy compression" to do it.
Which brings me to my point - I don't think the components in Reason are the issue, but rather the sounds that you put into them. Again garbage in= garbage out.
Re: garbage in=garbage out
:I ask you, how do they get 7 Gb of sounds on 3 CDs? Doug Rogers (the president) told me that they use "lossy compression" to do it.
ReFills, REX2 files and self-contained Reason documents all use **non-lossy compression**. No degradation. On average the compression results in half the size of the files.
Thanks for doing the analysis in a scientific way and reporting on it.
The noise at 90db has been reported on before. I guess it's the result of quantising errors - digital fuzz.
The gain issue is irrelevant (that's what amplifiers are for) - at least at -6db.
Maybe the props can comment?
You're listening to The Sweet
You tell 'em Marcus. n/t
Re: garbage in=garbage out
yes, but the Creative Essentials refill uses lossy compression. 7 GB in 3 CDS is way less than half (under 2 GB of sounds).
Not that I'm knocking it, I bought it after all. And what a buy - it was about $150 or so, can't remember.
Re: Even more on the Reason's Sound Quality thingy
:Beside this we realized one interesting thing that might be considered as a problem: There was always a noise around -96db in the silent parts of the exported waves. Even when there is no sound was playing at all. This noise was becoming slightly louder if we added more and more devices (FX or instruments) up to -90db when there was about 50 devices. Theorically there shouldn't be no noise, just plain silence when there is nothing playing. We didn't go more in this, though, because it wasn't that important for us when considered the hardware devices we were using in the studio like Roland JV1080, Korg Triton, etc with always have much more noise in their audio outputs. Perhaps it is related with one of the FX processors in Reason...
Which version? I think what you're measuring is dithering noise. Are you using 24 bit samples and/or are you rending to a 16 or 24 bit track?
This all makes slightly differences, but this would only be possible to hear using the best soundcard with the best wordclock.
I have to say that I only trust my ears... Most differences in sound people complain about is caused by differences in volume. Reason's playback is not causing clicks and pops, which means: every sample is converted!
r a a p i e
|All times are GMT +2. The time now is 15:47.|