Lets talk about sound baby

[ethios4]

OK its late and i don't have the tools or time available tonite to fully explore this but... i brought a drum loop into Live, unwarped & at original bit-depth/sample-rate, and i recorded it onto another track at the same bit-depth/sample-rate. No FX on any track, all faders at unity. i took the original file into Sound Forge, along with the newly recorded version. Should be identical, right? When i invert phase on one and mix it with the other, there should be nothing left, right? instead there is a strange muted version left over. To be sure i was doing this right i mixed inverted copies of each file with itself and the result was no signal. I also recorded two copies of the file in Live, inverted one and mixed it with the other, and the result was, again, no signal.
Strange... I tried again with a few other files and the result is the same each time...there is always a remainder after mixing the orginal with the Live-bounced copy.
Am i doing something wrong here?

[Mbazzy]

To some degree , an interesting read, but I'm just trying to be a musician so I'm not sure I [can] follow all this techie nitty gritty ...

Just one thing I didn't seem to notice in the comments : you are all comparing live output vs. output in environments like Soundforge, WaveLab ... and saying these sound better ... but can you confirm 100% sure these other environments are NOT colouring the sound output in some way ... have you considered Live just might have a flatter, uncoloured and thus more realistic soundimage ... like a difference in monitors can immensely change your perspective on a certain mix ...

But maybe I'm just plain stupid

[ PS. I'm already holding my heart when Live 26 becomes a complete brainlinked modular host and we'll start seeing discussions here how someone obtained a certain sound by sampling the subbass resonance of the hypothalamus but cancelling out a part of the frequency spectrum by adding an inverted phase of the cortex frequencies, this coupled to a sidechained compressor sync'd to your heartbeat [ mind the discussions we'll have about jitter of the heartclock !!!! It will be Ableton's fault ... "Please add heartbeat latency compensation " ] .... ]

<SIGH>

[drumroll57]

To be brutally honets this thread lost me on the first paqe.
I mean is the original poster some genetically engineered DOG/BAT hybrid
He's recording vinyl with a frequency bandwidth of approximately 32 khz (get over it batboy it's a fact) !!!!

Here's some Brutal Honesty™ for you:

- Vinyl is a very imperfect way of playing the music, but in a very strange manner, the waveforms are smooth and musical to the ear, especially in the upper frequencies, as well as the bottom end. You may think of its frequency response as limited, but there is no brick-wall nasty-ass digital filter to clip all of the harmonics, which can extend up to 40 or 60 kHz, and are significant not because they are directly heard, but how their waveforms create interference patterns with the audible ones. (think: throwing rocks in a pond)

-Digital has some very pronounced filtering effects, and horrendous stairsteps which are most apparent in those same upper frequencies. The digital format commonly used was developed 25 years ago, and designed for the limitations of that time. It is incredible no one has taken the music companies to court (for false advertising on their 'superior sound' marketing claims) over this.

-The human ear can definitely tell the difference. The bigger the sound system, the more they are magnified and apparent in A/B testing.
Go ask Jeff Mills, Derrick May, Carl Cox, Joe Claussell, Harvey, David Mancuso, and many others why they prefer to play the vinyl? It just sounds much more pleasant, warm and musical in a club.

- Even if the source of that music was 44.1 kHz / 16 bit, the mastering process 'glues' all of the stairsteps and artifacts back together into a seamless waveform on the acetate disc, and your ear definitely picks it up when listening to the vinyl.

I have done much testing on this, and please believe me, not every sound system in clubs is crap, quite the opposite. You may not have gone to the right clubs.

Ask yourself: why does every Digital Camera manufacturer double the pixel resolution of their cameras each year or so? (answer: large format prints would reveal that difference, which is the exact equivalent of a club sound system)

And yes, I also believe that Live does alter the sound a bit. But I am not hung up on it, it is something that I can deal with, as many other variables are involved into a musical and artistic performance in front of a crowd.

Sadly, Purple, although I might agree with you, this is not something that will be significant unless you are playing on an 'All-Class A Sound System'. None exist today outside of some private owners.

LEARN WHAT MATTERS. USE THE LIMITATIONS TO YOUR ADVANTAGE.

DON'T FORGET YOU'RE MAKING MUSIC, NOT WORKING IN AN INSTRUMENTATION LAB.

END OF STORY

[Vercengetorex]

I have replicacted the experiment described by Rappie above and got a rather strange result...
The file I rendered from Live had excatly ONE more audio sample then the original. Very strange indeed.
There was no change in pitch...
There was no change in timbre (or whatever "descriptive" word you would like to use here, as they are all equally non-descript)...
The Live rendering file was functionaly identical to the original, except for the addition of one sample...
I am going to try this again with several different originals of different lengths and compositions. Too weird.

[muthafunka]

Please describe your systems and setups otherwise this isn't all very objective/productive. I just opened then exported to aiff at unity, 16 bit/44.1KHz, Hi quality, no warping etc, nothing, the same 16 bit/44.1 aiff in Peak, Quicktime and Live then set all the files plus the original to start at the same time and spent quite a long time flicking randomly between the 4 versions on monitors (Mackie 824s) and hi-quality Technics headphones. I could NOT hear any difference in the highs, lows or mids or any other so-called 'feel-good factor'. I have yet to try the reverse phase test, but so far, to my (I think fairly reliable) ears on a reasonably high quality system I can hear no discernable difference. If anyone has any reasonably short sections they want to mail as .aiff or .wav for comparison, please message and I'll give you an address, I'll be happy to do the same.

Gear as below plus Mackie 824 monitors, mackie vlz 1642 vlz pro mixer, Technics dj1200 headphones, all very high quality monitor and power cables.

[Vercengetorex]

Here is what I have done:

I brought a 16bit, 44.1kHz AIFF drumloop into Live 3.0.4 with Live's audio resolution settings set the same.

I dropped said file in the arrange view, making sure warp and loop options are both off and that the file was extended to its full length.

I verified that both the track fader and the master fader were set to 0.0dB.

I selected the entire loop in the arrange view, and pressed "command-r".

I ensured the render settings were identical to the files original resolution.

I rendered the file to my desktop as an AIFF, closed Live and brought the file and the original into DSP Quattro (OS X Wave Editor).

After determining the Live file had one more sample than the original I repeated the process with a signifigantly shorter sample of a simple sine wave with the same result.

To verify it was not a problem with DSP Quattro misreading part of the AIFF header, I brought all files in question into Peak 4.1 with the same result. All files after rendering through Live with absolutly NO changes grow one extra sample.

Too strange. Any possible suggestions as to what is going on here would be greatly appreciated (looking in the direction of Dr. Henke)

[Purple]

koranek,

Yes, I am using the exact same file, same hardware path etc.

Regarding the driver, that could be an important question but I'm not sure I know how to check for the exact answer..

On Live what I'm using is the ASIO Hammerfall DSP, on Sound Forge I am not sure what driver is being used. I couldn't find it on the wave tab (audio) in preferences, and I couldn't find any mention of 'driver' or 'asio' in the help section search. How do I check this?

drumroll57,

>The human ear can definitely tell the difference. The bigger the sound system, the more they are magnified and apparent in A/B testing<

Amen! Finally I hear this from somebody else! It feels great to know I'm not the only one crazy enough to think that, I mean - after all - it just makes sense, doesn't it? And more important - I can definately hear it.

Look, everybody. I really didn't mean to create such a stir.. I do mean well, believe me!

I actually expected that people would be more familiar with this problem. Now I'm not sure, some of you seem to agree from your own personal experience that there are differences in sound quality of *unprocessed* sound files, while others say there aren't any. There must be just one absolute answer to this question at least.

I know the digital world is a weird and complex place. Aahhh, the days I worked with just two turntables and a mixer Everything seemed so much simpler (and it was just a week ago! )

But you know what, after reading all your comments, even if I won't find a solution to this problem, I will give Live a try after all, and see how much does the sound difference bothers me on the long run.

I never doubted the beauty of Live as a creative tool, in fact it's been a long time that I've been fantasizing on what I will be able to do with it as a DJ, but as somenody said - there is nothing wrong with trying to optimise your gear to your exact needs or desires, and to check how far can you push the system.

Today I have another producer friend coming over, I'll do some more tests with him and let you know.

[Purple]

muthafanka,

>Hey Purple, as a hint, if you want the best possible sound, lose the Numark, they're not exactly top-end to start with (if you want NICE sound get Allen&Heath etc), also, you're much better ditching the dj mixer altogether and using a hi-fi amp and getting some decent cables and even a decent cartridge if you can stretch to it...<

Well, I did think of buying a separate phono pre-amp. I heard great things about the Graham Slee Gramamp 2, supposed to deliver a sound competetive with units of 4 or even 5 digit price tags, in the price of 150 pounds or something similar.

I might still get it.

But I hate buying something I didn't listen to.

My Numark mixer is actually surprisingly sweet sounding, it sounds much better than other Numark mixers I've heard. And miles better than the usual Pioneer mixer that most clubs have over here. Regarding Allen & Heath I know they are good, Rane to is something I was thinking about, but again, I'd have to listen to those in comparison and it's not easy to get them just for a test.

The cartridge I use is also not that bad - Shure M44, the best IMO from the DJ cartridges.

And I also did buy very good quality rca connectors, cost about 35$ for about 1 meter cable.

I gave all these details before, but again, since you asked me to describe my system: Technics DJ1200, the ShureM44 cartridges, the Numark PPD mixer, the RPM RME, the Gateway M675 laptop, and my Dynaudio BM15A Active Nearfield monitors.

Purple

[drumroll57]

I gave all these details before, but again, since you asked me to describe my system: Technics DJ1200, the ShureM44 cartridges, the Numark PPD mixer, the RPM RME, the Gateway M675 laptop, and my Dynaudio BM15A Active Nearfield monitors.
Purple

Trust me here,

If you are worried about sound quality, I must tell you that there are several steps that would make your setup orders of magnitude better:

-1) Get a moving-coil cartridge, but they are very expensive. (a Koetsu Onyx will go for $6000.00, there are obviously cheaper ones, Van Der Hull, etc...) A great moving-magnet can also be fine. Grado comes to mind, although all of these may sound a little thin compared to a top-of-the-line DJ cartridge. I think the Shure 'White' or Stanton 881 are mighty fine as well, some people SWEAR by the Shure V-15 MkIV (older).

-2) Mate it with a dedicated preamp with output volume control, no mixer whatsoever. Cello is great, but you could buy a car for that price. There are some good Krell or Mark Levinson to be had for $1000.00 or so on eBay. Make sure that it has the 'MC' phono preamp, not 'MM'.

-3) Use an ultra-high quality A/D converter and interface. The Metro Halo ULN-2 comes to mind. A newer Apogee with 'soft-limiting' is also quite good. This interface item is somewhat debatable, as many people keep arguing about sound coloration and prefer one interface over the other. The one you have is probably fine.

It is imperative to get the optimum level INTO the interface, as getting it at less than the maximum '0 dB' does not give you a full 24-bit.That is why the Apogee's analog 'Soft Limit' is so good, it lets you squeeze every last db out of the signal without hitting the dreaded red on your digital meter.

All this costs a horrendous amount of money. Be sure that this is what you want to do. There is, however, no question that those factors above will really change the sound of your transfers, and make them much superior to what you've experienced so far.

The best way to test this would be to find someone with such a rig a compare to what your rig at home yields. (before making any commitments)

You mileage may vary. Professional driver on closed course. Do not attempt this with your own vehicle. Batteries Not Included.

[Guest]

All files after rendering through Live with absolutly NO changes grow one extra sample.

Well, it's one better innit?

[fjolublar]

purist attitudes regarding samplerates and bitdepth are a bit of a contradiction if you like the sound of vinyl recordings, which are by their very nature inferior to digital recordings, in purely scientific terms.

If you wanna capture the "rawness" or similar that seems to be the charm of vinyl, record at highest possible settings, with the most expensive and complicated equipment you can afford, just to compress it back to 16bit aiff or mp3's which such recording eventually end up. kinda betrays the point , doesn't it ?

Anyway . . . experts will tell you the most expensive and complicated precedure is the best , depends if you want to do things or endlessly complicate them.

I that case I reccomend the new protools rig with D/A-A/d 96khz 128bit converters with flashing lights built into them , at a price hobbyist's wont touch. record at highest settting using a 45300rpm Hd and then use Wavelab, Kyma or TDM to dither back to 16bit 44khz.

go figure.

[gaspode]

OK its late and i don't have the tools or time available tonite to fully explore this but... i brought a drum loop into Live, unwarped & at original bit-depth/sample-rate, and i recorded it onto another track at the same bit-depth/sample-rate.

Did you use master in, or did you route the output of the drum into an input in live and record? Just curious as, if you used an output/input signal strategy you'd have to adjust your latency and that'd certainly be far from perfect...

However, I'm assuming that you recorded the file through the master out.

[drumroll57]

the sound of vinyl recordings, which are by their very nature inferior to digital recordings, in purely scientific terms.go figure.

In purely scientific terms, please go and tell that to those that are getting paid every day of the week to use those tools in front of tens of thousands of people.

They have access to any from of technology they want, and there is either a reason for them to keep doing it this way, or they are all certifiably insane.

In purely scientific terms, please use your ears comparing analog and digital recordings of the same song in a large (50,000 Watts and above) listening environment and come back and tell me your results.

Actually, no need to argue this point. To each his own. Is that better?

You win. (but if I remember Purple's message, he explicitely stated that he was interested in doing 24bit-96 kHz playback with Live, so why bring in 16-bit or MP3 into the discussion?...)

[noisetonepause]

For inferior, read 'less accurate'. That's indisputable.

The inaccuracies are desirable, though.

-Paws

[fjolublar]

playing straight from comp at 24khz thru a soundcard that outputs the same bitrates makes a clearer sound, I brought in 16bit and mp3 cos thats the media "non-live" recordings usually end up on.

not much need to argue analog-digital, all comes down to preferance. . . A good analog eq connected to a 16bit 44khz output of a casual laptop can make mp'3 sound as warm as vinyl, in my experience, doesnt have to be much more complicated than that . . . unless one likes complex setups.