x
anonymous 8 March 2012

Welcome to our new forum, Track Talk.
I'll kick it off.

The piano isn't the best but the player makes up for it :) I have Sage, my 13 year old playing a few bars of Tarantella for us.
It is the exact same performance, just summed differently. Nice fast attacks for this comparison.

What track do you prefer?

(Expired Link Removed)

(Expired Link Removed)

NOTE: Don't read through this thread until you vote.

Comments

djmukilteo Fri, 03/09/2012 - 19:08

I listened to both clips on my ATH-M50 headphones three times each...so take my listening opinion with a grain of salt. I was going to vote the same!
They were so extremely close....but I thought SUM1 was subtly clearer and had just a touch more air to it.
That was only noticeble at certain points in the piece where I was listening to the tails..which I focused on after the second or third time....so pretty esoteric.
I think if you only heard either separtely without trying to make a comparison, you wouldn't notice anything that different....they were both great!

audiokid Fri, 03/09/2012 - 19:54

Don, very impressive response and very accurate! You have good ears. Both of you do.

Okay, I see I needed to clean this up. I switched the same tracks out in the Poll but this time removed reverb and did a null test until I got everything as close to each other as possible. The reverb was effecting this test and one track was a bit louder in volume. I also chose a better spot on the song so it should be easier to stay focused.
I also uploaded them to Dropbox as wave files this time. Soundcloud was doing some weird stuff. I think they sound like my tracks should now.
I believe you can download these now or audition them online.

I have uploaded NEW EXAMPLE from this Post Forward.

djmukilteo Fri, 03/09/2012 - 21:46

OK...that worked!
I guess having a Dropbox account isn't needed...I clicked on the link and they downloaded...quick and easy
So...were those stereo clips?
I imported them into Cubase 6 and split them to separate tracks and panned them L and R.
I guess I could have imported them to one stereo track though...right?
I put them one after the other SUM1 then SUM2 and looped them....
My meters indicated SUM1 quieter than SUM2.
I closed my eyes and lost track as to which one was which
And I liked SUM2 best....
I'm still using the ATH-M50's.
SUM2 was richer sounding, you can hear more of the room...
SUM1 was just a touch lighter, but then I don't know that could be the difference in the loudness.
Again very close....discerning ears I think would prefer SUM2....am I right?...LOL
So what is the difference in what you're doing....different summing equipment?

djmukilteo Fri, 03/09/2012 - 21:54

Oh BTW your son plays very well!
I think for me if you did something that was not as fast and crisp as a test, I would have an easier time hearing more detail.
The reason I say that is while it's a good piece it is hard to focus in on anything it's all attack and little or no sustain which would give your brain some time to register and process the sound and develop a feel for it.
With that type of piano sound, the thing you start hearing most is the tone of that piano.
Also wondered how you mic'd that and what the signal chain was....

audiokid Fri, 03/09/2012 - 21:58

Nicely done Don. Yes, two stereo clips.

Everyone should be able to import the stereo tracks without splitting them but either should still be fine as long as they are hard L&R. I'm new to this too. Trying to keep this as accurate as possible using two different summing system won't null perfect but I think its close enough for this test. I set the levels using a null test. Once they got as close to cancelling I left them there. There was only a slight amount left over.

I suppose its safe to say, set the level on the track louder to where you think its balanced. Do you still have the same opinion?

audiokid Fri, 03/09/2012 - 22:09

djmukilteo, post: 386050 wrote: Oh BTW your son plays very well!
I think for me if you did something that was not as fast and crisp as a test, I would have an easier time hearing more detail.
The reason I say that is while it's a good piece it is hard to focus in on anything it's all attack and little or no sustain which would give your brain some time to register and process the sound and develop a feel for it.
With that type of piano sound, the thing you start hearing most is the tone of that piano.
Also wondered how you mic'd that and what the signal chain was....

Thanks for the compliment, its my daughter, only 13 years old. She was just picked to represent our citiy as a delegate for the 2012 Provincials held in a few months from now. She loves music.

I will do another song later that will be much slower and involves more info. A fast attack piece is important IMO because the attacks are the first to show something important here.

I don't want to say much more until the end of this quiz.

audiokid Fri, 03/09/2012 - 22:17

I used a matched pair of DPA 4011's, SPL [[url=http://[/URL]="http://spl.info/fil…"]Premium [/]="http://spl.info/fil…"]Premium [/]Mic Pre – deadly! hi-end microphone preamplifiers , RME ADI-8 QS converters into Sequoia 11. This was recorded 24/44.1 and summed differently then bounced to wave files. I never normalized them or added eq, nothing. Just summed two ways.

djmukilteo Fri, 03/09/2012 - 22:42

audiokid, post: 386051 wrote: Nicely done Don. Yes, two stereo clips.

Everyone should be able to import the stereo tracks without splitting them but either should still be fine as long as they are hard L&R. I'm new to this too. Trying to keep this as accurate as possible using two different summing system won't null perfect but I think its close enough for this test. I set the levels using a null test. Once they got as close to cancelling I left them there. There was only a slight amount left over.

I suppose its safe to say, set the level on the track louder to where you think its balanced. Do you still have the same opinion?

Well I'm looking at peak hold meters in Cubase all faders at 0db and 2 separate tracks for each wave file panned L & R.
Those are going to a Master stereo bus metered and fader set the same way.
I see a couple db difference in the two waves overall and a slight balance difference L to R.
So I don't know if the test is I like Sum2 because it's a little hotter or not.
If I could attach a HTML or .doc file of my screen save to show you the meters I would....but I don't think I can.
What are you using for metering when your summing these?

djmukilteo Fri, 03/09/2012 - 22:58

audiokid, post: 386052 wrote: Thanks for the compliment, its my daughter, only 13 years old. She was just picked to represent our citiy as a delegate for the 2012 Provincials held in a few months from now. She loves music.

I will do another song later that will be much slower and involves more info. A fast attack piece is important IMO because the attacks are the first to show something important here.

I don't want to say much more until the end of this quiz.

Sorry your daughter...LOL tell her I apologize!
Sounds like you have a pianist there and congrats on the gig!
I think?? I understand what you mean with the fast attack thing.

audiokid Fri, 03/09/2012 - 23:02

Nothing was mastered, never intended left or right to match the same. Shouldn't matter IMO because they are the same track summed differently. I stuck the mic's up, measured them for XY, set the pre's to where they sounded "good enough for jazz" and recorded a stereo track as is. One mic is closer to the wall so it picks up reflections more. Its the nature of the beast. But this isn't about my method as much as it is about how the two identical tracks summed in the end. At least that's how I see it at this step.

I have know idea how this is going to play out. I'm as curious as you. We're doing an experiment.

If I share much more, it may create a bias which I don't want to do.

audiokid Fri, 03/09/2012 - 23:10

I should mention, I'm learning here too, and you ask good questions plus I can tell you are hearing this accurate.

One track is a recording of the other but one is summed different. I shifted and adjusted their levels until they came close to nulling. Then bounced them to a wave file.
Did I do it accurately?

For those feeling inquisitive, I think its safe to try and do your own null test on these, Do whatever you like to see which one you prefer. Play with the levels. I'm very curious which track you like better after you've tried everything possible. They are identical tracks but one was summed different that the other. Nothing more added.

How am I doing?

djmukilteo Fri, 03/09/2012 - 23:26

audiokid, post: 386056 wrote: Nothing was mastered, never intended left or right to match the same. Shouldn't matter IMO because they are the same track summed differently. I stuck the mic's up, measured them for XY, set the pre's to where they sounded "good enough for jazz" and recorded a stereo track as is. One mic is closer to the wall so it picks up reflections more. Its the nature of the beast. But this isn't about my method as much as it is about how the two identical tracks summed in the end. At least that's how I see it at this step.

I have know idea how this is going to play out. I'm as curious as you. We're doing an experiment.

If I share much more, it may create a bias which I don't want to do.

OK...I'm surprised you have like 70 views and it looks like I'm the only one who voted...LOL..which is fine by me...I'm up for more tests and discussion. I do question my hearing cuz I'm old...so this is good.
A musician friend of mine told me he had really good hearing and I told him I didn't think mine were that great anymore and I was losing the high end.
But then I heard some flaw in a recording we were doing and he couldn't hear it....no matter how many times I went back and forth over it....then I tested our hearing and turned out I got around 12k and he lost it at around 10k....plus it's all about how healthy your ears are...

bouldersound Sat, 03/10/2012 - 00:21

Okay, I wasn't going to say anything that might bias a listener, but I also heard Sum1 as marginally clearer compared to Sum2, which seemed richer. That is, Sum1 had a bit more around 4-6kHz and/or Sum2 had a bit more at 200Hz. That was with the Soundcloud versions through HD280s connected to Creative powered computer speakers driven by the stock SoundMax sound card. I listened a bunch of times, finally playing a few seconds of one then the same bit of the other repeatedly to the end of the clip.

Just downloaded the samples and will listen again on a better system when I can. I just got through 11 sessions in 14 days plus the day job so my ears and brain need a break.

Excellent musicianship, by the way.

audiokid Sat, 03/10/2012 - 00:38

bouldersound, post: 386066 wrote: Okay, I wasn't going to say anything that might bias a listener, but I also heard Sum1 as marginally clearer compared to Sum2, which seemed richer. That is, Sum1 had a bit more around 4-6kHz and/or Sum2 had a bit more at 200Hz. That was with the Soundcloud versions through HD280s connected to Creative powered computer speakers driven by the stock SoundMax sound card. I listened a bunch of times, finally playing a few seconds of one then the same bit of the other repeatedly to the end of the clip.

Just downloaded the samples and will listen again on a better system when I can. I just got through 11 sessions in 14 days plus the day job so my ears and brain need a break.

Excellent musicianship, by the way.

Thanks for the kind words as well. I'm very proud of her.

Yes, you heard it the same as me too with the soundcloud clips. Kudo's, we are all hearing the same. I'm confident then, the current tracks are accurate to our ears.

Have a good night!

djmukilteo Sat, 03/10/2012 - 00:55

OK...LOL
Sounds advanced for my pea brain!
If you know Cubase boulder, could you run me through the steps?
I'm assuming with this test, I would have Sum1 wav on a stereo track and Sum2 on a second stereo track and then line them both up and play them together into the same output bus?.
Then process one of the tracks Sum1 or 2 to "invert the polarity"...In Cubase I'm not sure how you do that...
Sorry if this is newb stuff...I've just never done this...

bouldersound Sat, 03/10/2012 - 01:06

djmukilteo, post: 386070 wrote: OK...LOL
Sounds advanced for my pea brain!
If you know Cubase boulder, could you run me through the steps?
I'm assuming with this test, I would have Sum1 wav on a stereo track and Sum2 on a second stereo track and then line them both up and play them together into the same output bus?.
Then process one of the tracks Sum1 or 2 to "invert the polarity"...In Cubase I'm not sure how you do that...
Sorry if this is newb stuff...I've just never done this...

That's right, but I don't know Cubase and all the ones I know do it differently. Oddly enough I just posted this elsewhere for a related question:

me wrote: In PT8 highlight the block, go to Audio Suite, select Other, select Invert, click Process button.

In Audacity click the Track Control Panel to highlight the track, go to Effects, select Invert.

In Reaper double click on the track to show Track Properties, check the box for Invert phase (even though it actually inverts polarity).

In Sony Vegas 6 click the channel's polarity button (which is how all DAWs should do it).

You're on your own with Logic as I haven't used it in ages.

djmukilteo Sat, 03/10/2012 - 01:26

Thanks boulder
OK I got it figured out...I HAD done this before.....
easy squeezy....phase switch at the top of each channel.
I tested cancellation with two Sum1 tracks and with phase switch on dead quiet on the output.
When I try that with Sum1 and Sum2 it has a real phasey sound to it but no cancellation at all?!

audiokid Sat, 03/10/2012 - 01:50

The tracks were nulled best I could before I saved them as a a wave file and uploaded them. They will not null dead quiet because something changed in the summing of 1 & 2. But they were close. I haven't checked them after I uploaded them but you are all welcome to hear what remainder is left from your own tests..

If they were both digital dups, however, they would null perfect. So the question I have is, which track sounds better to you, and you have already answered it. The 3 of us appear to be similar. It will be interesting to hear what the crowd says.

I suppose we could also ask why they didn't null perfect after the summing and why does one sound better than the other? The answer is, one went otb and back in again. Using the null test, I time aligned and balanced the levels best I could and what you hear is the difference between two methods of summing. Does everyone get this same result and the answer is, not likely. They would have to be using the exact same gear as me and this is where it gets interesting.

djmukilteo Sat, 03/10/2012 - 02:02

We're up late on a rainy Friday night!
I suppose the difference in the levels could be the reason but I certainly didn't get much of any sort of null...
Unless I did it wrong...Like I mentioned...I tried it on the two Sum1 and it null perfectly and the only thing I did was delete one of the Sum1's and imported Sum2 and ran it.
Should I try and adjust the levels?

audiokid Sat, 03/10/2012 - 02:41

The process above is what I did before I saved and uploaded them for you all. I have no idea what another null test will do again but I would expect them to line up the same. Its hard to know what happens in the upload process. Your own null test should expose the changes that occurred between the two summing processes I did.
That being said, how you hear them on dropbox is how they sound first generation after I summed them the two ways. That vote is the one I value and would expect to count but its all fun and interesting regardless. You have answered it with flying colours. Nothing wrong with your hearing.

I'll share which is what at the end of this poll.

audiokid Sat, 03/10/2012 - 14:41

Okay, I downloaded my Dropbox files and performed a null test on these files also. Not sure why the start points and level changed from within my DAW to the download, I'll have to look at that too now. More fun...

After the dropbox wave files were imported as 2 stereo tracks, inverted, aligned again, levels and panning adjusted I am surprisingly able to get them to almost completely null but this time with a remainder of approx -50 on peaks with upper mids hanging over. I have a feeling if I could figure out how to do ultra fine adjustment in [[url=http://[/URL]="http://www.samplitu…"]Sequoia[/]="http://www.samplitu…"]Sequoia[/] ( never needed to go this fine) I could get that overhang down another 20 db. I wonder which SUM has the overhang. How do you tell this?

Doing a comparison now, I cannot tell which one is what. The tracks sound more identical. So what I find interesting and important, If there was a discernible amount of difference between the two sums at the end of the day, I would question my hybrid rig. If the hybrid rig added noise and lost the edge, I would have tossed it at this point.

What I'm wondering about right now, why did the DAW make the analog track more digital sounding?

So, I move on to the next step. Does hardware improve my hybrid system and how do I keep it from sounding digital and the end of the day?

x

User login