| Author
|
FL Producer 5 Cubase SX 2 comparison.
|
UnderTow
Started Topics :
9
Posts :
1448
Posted : Dec 21, 2004 19:06
|
Quote:
|
On 2004-12-21 18:51, H2O wrote:
Panning laws are set the same way, -3dB in Cubase and unchecked at FL Producer which is exactly the same.
|
|
Ok.
Quote:
|
On picture you can see that FL Producer follow Cubase's signal with 1.5dots lag. If there was no lag at four samples test, i assume the lag is overdrive made one of sequncers or both un precised. For your question, when i said less precise i meant it has lag compared to Cubase.
|
|
When I first read this I thought the whole signal was late in FL but then I looked at the JPEG. Waow. This isn't good. The wave in FL actually has a lower frequency. It seems like FL is like a lopwass filter on the signal.
This would explain why the high-end sounds less good in FL: It is removing some of it!!!
Quote:
|
High end felt by ear, loading both of samples into one line at SoundForge without space between. We listened to it on three different monitors and got same idea. Why FL Producer is in lack of high end is because inside of sequencer we heard a bit different signal than after export. So we record using Sound Forge directly and export same sequence. They sounded a bit different.
|
|
Yeah looking at the JPEG I can imagine. Thanks for the explanation.
UnderTow |
|
|
ZilDoggo
Started Topics :
4
Posts :
663
Posted : Dec 21, 2004 20:34
|
Quote:
|
On 2004-12-21 18:31, UnderTow wrote:
Quote:
|
On 2004-12-21 17:36, ZilDoggo wrote:
every bit you add will double the ammount of information.
|
|
No. That would mean that a 24 bit file is 256 times bigger than a 16 bit file. This is obviously not true.
Adding one bit to a 16 bit word will add 1/16th amount of information. Adding one bit to a 24 bit word will add 1/24th etc etc.
It will double the dynamic range and the amount of possible values but that is a different thing.
UnderTow
|
|
uuh., if you double the dynamic range then you can fit 2 sounds of half the dynamics in it, right?.,
you forget that we are talking binary here.,
if you add an extra bit to a 16 bit number you have space for 2x a 16 bit number
if you add 2 bits to a 16 bits number you have the space to code 4x a 16 bits number,
3 bits is 8x a 16 bit number
4 =16
5 bits = 32
6 bits = 64x a 16 bits number
7 bits = 128x
8 bits is 256 times a 16 bits number!!!
and 24 is 8 bits more than 16 bits.,
realy, if this was not true then 32 bits floating point would never been enough!!., we would have been in the 1024 bits audio by now! 
every bit doubles the dynamic range (which is equal to an increase of 6db i think)
do the math and you will see.,
greetos,.
aka.,
|
|
|
|
ZilDoggo
Started Topics :
4
Posts :
663
Posted : Dec 21, 2004 20:44
|
Quote:
|
On 2004-12-21 18:31, UnderTow wrote:
Quote:
|
On 2004-12-21 17:36, ZilDoggo wrote:
every bit you add will double the ammount of information.
|
|
No. That would mean that a 24 bit file is 256 times bigger than a 16 bit file. This is obviously not true.
Adding one bit to a 16 bit word will add 1/16th amount of information. Adding one bit to a 24 bit word will add 1/24th etc etc.
It will double the dynamic range and the amount of possible values but that is a different thing.
UnderTow
|
|
actually, after rereading your statement i think i understand what you mean.,
i should not have said 'double the INFORMATION'
it can lead to confusion.,
what i mean is that every bit allows you to code double the ammount of data in that with that row of bits.,
so a row of 24 bits can include the total of 256 16bits rows of data.,
24 bits is enough to store the total of 256 16 bits things.,
|
|
|
|
UnderTow
Started Topics :
9
Posts :
1448
Posted : Dec 21, 2004 21:05
|
Quote:
|
On 2004-12-21 20:44, ZilDoggo wrote:
actually, after rereading your statement i think i understand what you mean.,
i should not have said 'double the INFORMATION'
it can lead to confusion.,
what i mean is that every bit allows you to code double the ammount of data in that with that row of bits.,
|
|
Well data really is just another word for information. When you increase the word length, you increase the precision but you do not increase the amount of audio data. In PCM wave, one binary word, or put differently, one sample, contains one sample point. Never more, never less.
You increase the precision/resolution of the data but you don't increase the amount of audio data. To do that you need to increase the sampling rate.
I might be being difficult here but I think it is important to get the facts and terminology right.
UnderTow |
|
|
|
High Pulse
Darkpsy
Started Topics :
57
Posts :
1187
Posted : Dec 21, 2004 22:11
|
|
Trip-
IsraTrance Team
Started Topics :
101
Posts :
3239
Posted : Dec 22, 2004 17:40
|
I'm happy to see someone is putting an effort into checking these audio engines - but we need to figure out a more consistent test criterias I think.
I think we can all agree that 24bits is the ideal since not every sequencer creates 32bit float files.
master volume should be allways on 0dB.
noise signals samples should be peaking at -12dB/-18dB > i wonder how many you can stack and keep the master output on 0dB, keeping channel volumes intact. (since I know fruity shows volume amounts not in dBs or maybe it's changed by now)
Another test would be to change channel volumes - this is also done differently in several sequencers.
The master sum signal cannot clip but must peak at 0.
Would like to hear other opinions.
Please suggest other analyzing tools for checking, not only the sound forge zoom.
 Crackling universes dive into their own neverending crackle...
AgalactiA |
|
|
|
Trip-
IsraTrance Team
Started Topics :
101
Posts :
3239
Posted : Dec 22, 2004 17:58
|
btw,
to get all the big prices straight:
FLStudio XXL (boxed) = $399
Orion Platinum 5 = $298 ($399 XXL bundle)
Cubase SX 3 = $799 (HALion 3 costs extra $399)
Nuendo 2 = $1,499
Logic 7 = $999
Sonar Producer 4 = $959
Reason 2.5 = $369 (seems no Reason 3 on the online shop yet)
couldn't find Digital Performer non-upgrade price on motu.com
I also have to say, that all these packages are different - it's not ONLY the sequencers but also samplers or effects or whatever.
Crackling universes dive into their own neverending crackle...
AgalactiA |
|
|
|
Trip-
IsraTrance Team
Started Topics :
101
Posts :
3239
Posted : Dec 22, 2004 18:05
|
and by all means,
Once starting to test all these audio software - H2O please open a new thread since this topic is out of place allready.
Crackling universes dive into their own neverending crackle...
AgalactiA |
|
|
|
H2O
IsraTrance Junior Member
Started Topics :
16
Posts :
352
Posted : Dec 22, 2004 18:13
|
Trip, FL Producer cost 150$. XXL boxed means whole plugin set. Empty sequencer like Cubase or Logic with only bit of native effects and synths it cost 150$. So do not mistake ppl.
Thanks to Zildoggo and Psyx have joined to help us, we still seek Sonar, Neundo and Reason users who have two weekly hours to help community.
Thanks, Kik |
|
|
|
Mo-Dul
IsraTrance Junior Member
Started Topics :
15
Posts :
135
Posted : Dec 22, 2004 18:21
|
[quote]
On 2004-12-21 18:51, H2O wrote:
Why FL Producer is in lack of high end is because inside of sequencer we heard a bit different signal than after export. So we record using Sound Forge directly and export same sequence. They sounded a bit different.
[/quote]
So direct recording to soundforge will not
destroy the original (and clean) signal?
|
|
|
|
H2O
IsraTrance Junior Member
Started Topics :
16
Posts :
352
Posted : Dec 22, 2004 18:23
|
Sound Engine test is very complex chain of tests, we have no professional experience to test Sound Engine.
We can test: output noise, signal precise, signal identity, CPU overload, different functions.... But not sound engine which we don't know what is that. |
|
|
|
H2O
IsraTrance Junior Member
Started Topics :
16
Posts :
352
Posted : Dec 22, 2004 18:26
|
Modul, with my sound card it is imposible to record well signals, others have same sound card. We sure that FL Producer's problems are sampler or/and export. So avoid export it good thing to do since many many functions at export menu in FL Producer are fucking nonsense like HQ for plugins or interpolation which is different thing at alll when Image Line use it as something pretty different.
Don't ask us now, we need to finnish tests. |
|
|
|
Trip-
IsraTrance Team
Started Topics :
101
Posts :
3239
Posted : Dec 22, 2004 18:33
|
H2O just want to point out that Logic comes with it's own sampler, a powerful ir reverb, many well known instruments - and many many plugins ...
That's why I wanted to point out the XXL bundled version of each product - though Steinberg doesn't come with one.
Crackling universes dive into their own neverending crackle...
AgalactiA |
|
|
|
thockin
Started Topics :
1
Posts :
114
Posted : Dec 22, 2004 18:50
|
It's frustrating to see people who only half-know what they are talking about spreading misinformation. The test seems flawed to me.
There are a number of things you can test.
Do you want to test the pure mixing of a sequencer? Here's how:
1. Create as many audio tracks as you want in any quality that you want.
2. Load all the audio tracks into your sequencer. Use the same volume for all of them (max or -0dB). Pan them all to dead center. No repitching for any of them, and no effects AT ALL.
3. Trigger them all exactly on the same sample.
4. Turn off dither.
5. Render in the highest quality that both can support. That is probably 32 bit float. If one supports 24 bit int and one supports 32 bit float, that's OK - they are close enough to pretty much not matter.
6. Load the two rendered outputs into SoundForge or something similar.
7. Normalize them both to -0dB.
8. Do a bit-depth conversion if you need to (if one was 24bit and the other 32bit). Avoid this if you can.
9. Now invert one of them and mix it with the other.
10. The result is the difference between the two.
Now, if you want to test the panning of a sequencer, or the dither of a sequencer, you can do that in a second test. Those are all new variables and should be tested independantly.
So for example, when the above test is done, set up the panning on all the tracks. Set it however you want, but set it the same for both sequencers.
If you want to test the attenuation, then set some tracks to less than max volume, but make sure you set them the same in both hosts.
If you want to test the VST handling, then add some VSTs, but make sure they are set exactly the same in both hosts.
The key that people tend to forget is that you have to normalize at the end. Some hosts leave headroom in the end of the chain.
Do *these* tests, then you'll have something almost valid to talk about. |
|
|
|
thockin
Started Topics :
1
Posts :
114
Posted : Dec 22, 2004 18:53
|
Quote:
|
This would explain why the high-end sounds less good in FL: It is removing some of it!!!
Quote:
|
High end felt by ear, loading both of samples into one line at SoundForge without space between. We listened to it on three different monitors and got same idea. Why FL Producer is in lack of high end is because inside of sequencer we heard a bit different signal than after export. So we record using Sound Forge directly and export same sequence. They sounded a bit different.
|
|
Yeah looking at the JPEG I can imagine. Thanks for the explanation.
|
|
If this is true, it's a new bug in FL. It certainly was NOT true in the past, and I suspect an error between the keyboard and chair. What jpg are you looking at? |
|
|
|
