dudes... i got it! i effin' got it!

[Splinter]

Thermos:

That's interesting about the DB8. I've heard the summing buss in that was terrible. Can't vouch for it myself, but I do understand what you are saying about DP's summing. It can sound ceramic and have that wonky low end, but at other times I've gotten some really rich, fat sounding mixes. Those kind are definitely harder to get in the digital domain than analogue.

I'd love a D2B, an HD192, and a Lucid. Anyone got 4G's they want to give away.

[Siryne]

I have been following this thread with intrest and would like to ask a related question. Assumeing equipment I already have, would I be able to improve my mixes bybussing out to a Apogee PSX-100 AD/DA converter and back in to a stereo track? If so, what would be my best route? I have an 896HD ( i also have the older Apogee Rosetta but it is only A/D so I am assumeing the high grade converters in the PSX 100 would be better since it is A/D and D/A) and I am useing DP4.5.1 with UAD1 and many other plugs. If it is say a 24/48k project would I route out the AES on the 896 and go AES through the Apogee and then back into the 896 AES? Where/who would to the sample rate conversion and the 24-16 bit depth conversion. I am guessing that the high grade converters and clock of the apogee must play into this but I am a bit confused. Could someone suggest a routeing scheme by which I could use this gear to improve my mixes and avoid some of the final bussing problems mentioned in this thread? I thank you and commend you on a great thread.
Peace!

[FM]

Originally posted by Splinter:
Mike I surely understand all you are saying about the smudged glasses, but why now through the same smudged glasses is everything clearer?

As Jayman said, and I agree, Ken's a great engineer, but it doesn't hurt having $20,000 D/A converters and $30,000 hand built EQ's either. I didn't necessarily love everything he did color-wise to the mixes - that's subjective - but the spaciousness of the mix was definitely improved. I know that can be a product of "opening up" a mix with EQ, but I think it is more a product of what FM has discovered: Great converters add warmth, depth and definition even to an already digitally summed mix.

What FM is hearing has nothing to do with summing (since the audio has already been digitally summed ITB), but is a byproduct of "re-converting" his mix through high-end converters. I would even suggest the psycho-acoustic phenomenon of loudness. If the converters are not calibrated perfectly and are adding any gain to the "re-converted" signal, this too could slant the listener to favor it.

I'm not trying to be the naysayer here and I'm highly interested in this discussion, I just want to look at all the variables and suggest some holes in our thinking when it comes down to splitting hairs and spending thousands of dollars

What do you guys think about my HD192 idea?

boy, do i need some coffee.
up all night, bussing and rebussing.
phew!
lemme see if i can be somewhat coherent here...
the reason i'm quoting Splinter is two-fold, he makes a great point and his point is valid. imho.
yes, the reason i'm hearing a difference in my mix is absolutely because of the way i'm processing the signal and the gear involved.
and yes, some of the "perceived warmth" is due to a slight loss of high-end, not enough to make the mix sound dull mind you, at all. so, something happens, and at least to my ears and the way i want my music to sound, the results are acceptable.
as acceptable as using a 16 channel D/A and high-end outboard mastering compression?
of course not. but one of the points we are exploring here, maybe i'm wrong, is to find ways to get the most out of the gear we already got.
i also tried other experiments.
i bussed some instruments thru my preamp (mindprint DTC), bussed VI's thru it using the same bussing i was using for my mixes.
in those instances i have to say that i had no idea VI's could sound so effin' good! forget freezing boys! when it comes to rendering software created sounds this is the sh_t!
anyway, my "final for now" conclusion.
bouncing the whole mix the way i was doing it produces a desirable result in my case, it taskes a little high-end off but the push and pull of voltages between the D/A and A/D process "glues" the whole thing together. me like.
AND, this is the part i really like, if i only bounce two tracks at a time like that then the sh_t really begins to take shape.
whole different ballgame.
but, it is A LOT more work.
again, to me, the point is not necessarily just to see if this actually works but also to find new ways to use my existing gear... you know what they say about necessity.
i didn't get a huge improvement but i learned a lot and i found new ways to get even more out of my rig.
again, yes it makes a difference. it's gotta be some voltage/conversion thing, i dunno, i need a nap.

but first i'm hitting the diner, there's nothing like the smell of a gigantic mound of home-fries waffting from a brooklyn diner early in the morning.
mixing sure makes you hungry.

peace!

FM

FM... sunny side up, extra bacon on the side.

[mchimes]

Very cool observations guys . . . I have to leave town, but I'm eager to try some experiments when I get back. By then this thread will be up to page 10 with alot more to chew on.

I wonder if there are any decent 4 channel D/As out there. I'm thinking I could do this "Out/In" analog "glue" that FM is talking about while still listening to it all through the same good converter. I think Swissonic made one, but didn't they go under?

I did do a "re-amp" of a stereo track through a Brent Averill 312 card the other day, but I haven't really had time to listen to it yet. And of course it was out the 828s converters.

Talk to you soon,
Mike

[rahlo]

Hey guys,

Been following this thread with MUCH interest, and am absolutely humbled at how little I know about this subject compared to you cats.

Please forgive my ignorance, but in reading b/tw the lines, am I hearing you guys say that the 828 mk2 does not have good converters? Are they any better or worse than, say, a digi 002?

This thread is blowing my mind. I always thought (and software companies like Motu, Ableton, etc., seem to suggest) that their methods of faster than real time bouncing/rendering of all tracks itb yields the same or better results as a real time bounce that records the audio into a stereo track. If I'm correctly hearing what you cats are saying, then this isn't true, is it?

[amplidood]

The process that happens when sending all channels straight to a digital out versus an internal bus is *not* the same. These posts are too long and many for me to be sure, but I think the only variable some of you have been changing is *after* all the tracks have been sent to an internal bus (or Master Fader). The change happens when you send *each track* straight to the digital out and then record that out back in to a stereo track. You won't notice much of a difference just by switching where the final bus goes. What should be avoided is using an aux to sum the entire mix before heading OTB.

[csiaudio]

This past summer Sweetwater Sound had their 'Gearfest' which was pretty cool. Bob Clearmountain was there and I would assume that most, if not all, know of Bob (he is an excellent engineer and masterer)...any who, one thing that he said that really struck a chord and is quite obvious is this - if you have marginal converters (MOTU) your mix will, no matter how hard you try, lose dimension/depth. So basically anything that you record through your MOTU stuff even at higher sampling rates will lack depth. So, I went out an bought an Apogee Rosetta 800 and I have to say it does make a substancial difference! In addition to this clocking is EXTREMELY important - which the Apogee has both nice converters and an excellent clock. The Apogee and the like are not cheap but if you want a definite 'step up' in your sound - get it, or save and get it! You will not be disappointed.

[Jaysplace101]

Hey guys, on the converter issue, my understanding is that if you get great converters, lavry, apogee, mytek, whatever, your mixes will improve right away even if you didn't record through the great converters, right? I often send stuff out to friends in other states to get recorded, and those guys don't have great converters, well the drummer has paris, and I like the sound of those, but the guitar guy uses MOTU converters. But, my understanding is the converters help the mix even if I don't record everything through them. And, I think this is because of a good clock.

Sorry for the tangent.

Amplidood, with my setup I described, a DO go out and back in and record that on a stereo track. But, I'll set up an aux buss just to monitor it. Does that make sense? So, audio patch thru wouldn't even have to be enabled for me to hear what's going down. But, I think what your saying is that what I'm hearing through that aux is being internally summed, so it won't sound as good as what I'm printing on that stereo track???

The really weird thing is, and I wish someone could set up and verify this. If I stem some things out and back in through my lightpipe like I've described earlier, it sounds really good, a definite improvement when I'm just listening to the DP file play. But, then when I record it down to a stereo track, it doesn't sound as good. I don't know why. The descrepency here is only apparent when I have things stemmed out through lightpipe. If I'm just doing everything ITB, the recorded stereo track sounds the same as what I hear when I'm just running the sequence down. Hmmmm. We'll get to the bottom of this someday.

I've decided (for now) I'm going to buy the LYNX aurora, and either the dangerous 2buss or the Audient Sumo for summing. Unfortunately, the LYNX won't be available for another 3 weeks or so. I talked to the guy at LYNX, I think Paul was his name, and he said everything was looking and sounding REALLY good with this box. I'm sure he wouldn't say anything else, but I've read enough reviews on other Lynx stuff to know that it's the real deal. After I get this all hooked up, I'm going to run some mixes and I'll post the before and after. Hopefully there will be a big difference. There better be!

Jayman

ps; sorry to be longwinded

[Splinter]

Originally posted by rahlo:
Please forgive my ignorance, but in reading b/tw the lines, am I hearing you guys say that the 828 mk2 does not have good converters? Are they any better or worse than, say, a digi 002?

Well, "good" and "better" is relative. It depends on how snooty you are Let's put it this way: You could record a hit record with an 828! Does an 828 compare with a converter twice as expensive?... no. But, you wouldn't expect it to. And by the way, as admitted by Lynn Fuston, who did the ADC CD, the differences in most converters are greatly reduced when externally clocked to a low-jitter, high end clock. As far as the 002, the 828 is probably comparable.

Originally posted by rahlo:
This thread is blowing my mind. I always thought (and software companies like Motu, Ableton, etc., seem to suggest) that their methods of faster than real time bouncing/rendering of all tracks itb yields the same or better results as a real time bounce that records the audio into a stereo track. If I'm correctly hearing what you cats are saying, then this isn't true, is it?

Well, this depends on who you ask. As some suggest, those with, of course, superior hearing to the rest of the audio industry, bounce-to-disk is inferior to real-ttime bouncing, and even still bussing to an external output is superior to internal bus. I, for one, believe this is a bunch of self-delusioned nonsense. Now, what FM is doing is entirely different, but in the digital domain, math is math, whether it is summed to a buss, an output, bounced fast, slow, real-time or at warp speed.

I'm gonna shoot this baby down and do a test:

Take a mix with all its tracks and bounce it to disk with the BTD feature in the Audio Menu. Take the same mix starting at the exact same point with all its tracks and bus each track to a stereo track and record it in real-time. Lastly, take all the tracks from the same mix, starting at the exact same point, assign them to a lightpipe out and feed the lightpipe back into your audio interface and record that input to a stereo track.

At this point you have 3 "identical" mixes of the same song all bussed differently, but have all remained in the digital domain.

Line up each stereo mix at the exact same point on 3 separate tracks. On the 2nd and 3rd tracks, instantiate a MOTU Trim plug. Playback the first track (others muted), then with the phase inverted in the second and third track's Trim plug ins, unmute each, one at a time.

You will hear nothing when you unmute the phase inverted tracks if they are identical. If you hear slight artifacts, then you know something different is happening in each of the summing techniques being employed.

I have in fact done this before with the BTD and realtime bounce and it was a complete null. THERE IS ABSOLUTELY NO DIFFERENCE! Sorry to burst your bubble guys, but there is no difference even though your brain says it does. You are deceiving yourselves.

I haven't tried the external out yet, but I'm very certain I will find the same results.

Try the test and prove it to yourselves. Others have and it has vastly improved there quality of life using the BTD. Why would MOTU employ an inferior means of bouncing a mix as their primary mixdown operation if it could be done better by other means?

You can do the same test with Thermos's other preposterous thread stating the work priority determines the audio quality.

These are just bad myths. Back them up with solid, scientific evidence so we can all be "enlightened."

[thermos]

Splinter
you should read my last post on my thread, where I did a simple solid test such as yours. Definitively conclusive, and admitted by motu. I don't know if putting stuff through lightpipe will make it sound any different, but according to me and motu, work priorities will.

[Jaysplace101]

Splinter, that's brilliant to do the test that way. I'll try it on going out of the box just on the stereo buss. You could be right there. That is definitive on BTD, and thank goodness. I read a post on OSXaudio that said the BTD was BAAAAAD, so I stopped doing it. I guess I shouldn't believe other peoples ears.

jayman

[Splinter]

Just ran the test.

YOU ARE ALL FAKING YOURSELVES OUT!!!

All 3 mixes, a BTD, a realtime bounce, and the lightpipe out and in to a stereo track all nulled completely.

There is no difference in the summing AT ALL!

Now I should add that this test will only work if you aren't using time variant plugs - that is any plug that would not identically repeat its performance from one bounce to the next. It will be very obvious, such as the guitar part sitting out by itself or a BGV. Anything that may have panning or effects that are not phase locked to the track. Some less obvious time variant plugs would be compressors or gates without lookahead. My Waves RennComps were not an issue because they lookahead 64 samples and perform identically everytime, but Trackplug without the lookahead turned on, revealed the compressor's attack response inconsistancies. Autotune is also a big free for all... a realtime juggling act. But, I knew that before.

None of these factors negate the test though as they would be variant from bounce to bounce each time regardless of what summing method you used. I just added this observation in case you did the test and got all kinds of wacky out of phase info. Any subtle artifacts would be evident with 99% of the mix cancelling out and slight high or low frequency content left. But that's not what I got. It was a 100% cancellation. Sorry guys. It's all in your head.

One last thing. When I bounce the mix through the lightpipe and played it back out of phase there was all this higgh frequency content - and I mean a lot - left. I started wondering if I was totally wrong in this. So, I zoomed down to the sample level and found that the lightpipe mix was latent by 1 sample. As soon as I bumped it forward, It cancelled completely.

Conclusion: Save your time with all your convoluted routings and realtime bounces (where you don't have to) and use the BTD. It is every bit as good - it is exactly the same - and about 4 times as fast.

I haven't tried the Freeze or Work Priority test, but I suspect they will yield the same results. IT'S ALL MATH BOYS!!!

[FM]

Originally posted by Jayman:


I've decided (for now) I'm going to buy the LYNX aurora, and either the dangerous 2buss or the Audient Sumo for summing. Unfortunately, the LYNX won't be available for another 3 weeks or so. I talked to the guy at LYNX, I think Paul was his name, and he said everything was looking and sounding REALLY good with this box. I'm sure he wouldn't say anything else, but I've read enough reviews on other Lynx stuff to know that it's the real deal. After I get this all hooked up, I'm going to run some mixes and I'll post the before and after. Hopefully there will be a big difference. There better be!

Jayman

ps; sorry to be longwinded

hey Jayman,

why the lynx piece?
matter of fact, why this particular setup?
is it because of price?
is it because the sumo has so many features?
the sumo is still on the yop of my list since i am not planning on getting any additional D/A converters?
and after you get it, needless to say, i would love to hear from you what your experience is like.

thanks!

FM

FM calls Alabama The Crimsom Tide; call him Deacon Blues.

[Timeline]

I haven't tried the Freeze or Work Priority test, but I suspect they will yield the same results.

LET ME KNOW WHEN YOU TEST THIS...

Thanks
G

[Splinter]

Tried the Work Priority Test. SCAM! No difference. I bounced the song at High, then changed the Work priority to Low and rebounced. Played them together out of phase. 100% cancelation. Now I wasn't listening to playback quality from one Priority to the other, but as far as the rendering of audio during a bounce, work priority yields the same results.

But, I doubt Work Priority has anything to do with the audio quality, rather works in the background with OSX to determine how DP will use multithreading and manage background tasks. The manual says the Work Priority sets the OSX thread priority and may affect audio performance (I would take that to mean track dropout, drive performance, and such), not quality.

The freeze test was a little more challenging. I first tried freezing a Stylus RMX loop then did a realtime bounce of the same loop. When flipped out of phase there seemed to be apparent "hits" that popped out, but the majority off the loop cancelled. I figured this was a timing or panning issue so I used a Trim plug after Stylus to make it mono and Quantized the loop 100%. I refroze and rebounced, but still had the same issue... hmm. I decided to try something other than Stylus. There are so many variables with that plug who knows wants going on behind the scenes. Softies are suppose to be sample accurate in DP, as are realtime bounces, so there must be something going on in Stylus I'm unaware of.

I decided to try u-he's Zoyd instead in mono. Because synths use so much panning and variables, for the test, I just kept it simple with a mono synth, no LFOs, no velocity, etc. The freeze and real-time bounce cancelled 100%.. For the purrposes of our test, that being summing/bussing, there is no difference in a freeze vs. realtime bounce.

Now one thing to consider with all of this. As I've noted with the plug ins timing variables, it is completely safe to say that one mix could or would sound "better" (or at least different) than another, due to the inconsistancies of the plug ins from one mix to the next. But, this is not a summing issue. As I said before math is math. 1 + 1 = 2 no matter how you get it there. If you change the numbers, now that's a different story, but then we get into floating point math and error correction and I don't really want to go there. That's where my eyes start to glaze over. Within a closed system, like DP, this really shouldn't be an issue.

Anyway, I hope this is truly enlightening for us all. It is so easy to get all mystical about all this stuff, when we should really just trust that the guys who designed this stuff had at least a clue about what they were doing. Now as to whether we like their taste in audio, that's a different issue, but that's subjective.

Objectively, there is no difference between an ITB bounce, a freeze, a realtime bounce. an OTB (digital) bounce, a Low Work Priority bounce, or High Work Priority bounce.

So disspell the myths... there is NO difference.

<small>[ December 28, 2004, 03:08 PM: Message edited by: Splinter ]</small>