Dec 1, 2010 12:09 PM
Preliminary results of HD500 DSP usage of amps and fx
-
Like (0)
14250 Views
71 Replies
Latest reply:
Mar 8, 2013 1:22 PM by
mneph 
-
Fester2k
Nice work. Very interesting and useful and probably about as accurate as you are going to get without a meter of some sort in the LCD. It must have taken you an hour or two at least to work through this exercise.
Thanks for making the effort. I'm sure a lot of people will appreciate it.
regards
Nick
-
nickmattocks wrote:
Fester2k
Nice work. Very interesting and useful and probably about as accurate as you are going to get without a meter of some sort in the LCD. It must have taken you an hour or two at least to work through this exercise.
Thanks for making the effort. I'm sure a lot of people will appreciate it.
regards
Nick
Thanks - you're probably right about accuracy. What we might find with add'l testing is further gradation among the various amp models.
Cheers,
Fester2k
-
-
-
-
-
silverhead wrote:
I absolutely love that term!! I hope you achieve Line 6 immortality with it! I can just hear the talk several years from now...... "Oh Yeah? Well remember what Fester's DSP Nanny study showed?......"
Thanks! Glad you enjoyed it.
I can see it in my obituary now - "Yesterday we lost our dearly departed brother Fester2k. He accomplished many things on this earth, not the least of which was his coining of the phrase 'DSP Nanny' in connection with a groundbreaking 2010 study of DSP limitations in a piece of Line6 gear. He leaves behind..etc..."
Ah, well - hopefully the testing helps somebody out.
Cheers,
Fester2k
-
-
-
-
Thanks a lot for Your superb work Fester2k!
IMHO I think reaching the DSP limit wil be both a problem and a stimulant. A problem if we want to build rich effects and amps structure, with complex textures doing cool stuffs...but WHO really need those things? I think L6 did an excellent job recreating real breathtaking effects and amps, maybe too much, even for a guitar hero!
So the stimulant may be thinking different and try to find different ways to reach the goal, for example: we forget the external loop...need some more stuff? Use external loop and link whatever You want, et voilà! job done!
I'm really optimist and I'll wait further improvement that surely L6 will do, so start suggesting them what to do:
- optimize the code so the POD HD will use less DSP and we may put more on chains;
- realize a "DSP use" bar or percentage (better both...), that may be recalled with a click showing how much more efx or amp we may squeeze in that patch;
- Implement some "simpler" efx using less DSP nanny...
Music is fantasy...
Anyway, thanks again Fester2k...
-
-
-
-
So I quickly just completed the following test. On that patch with 2 full AC30's, I was able to get the following information on the distortion models:
The numbers in brackets are the number of each model I was able to add to my dual AC30 patch before getting the DSP warning. Based on Fester's findings that the Amp models use about 40% of available DSP each (I specifically picked the AC30 because it was in his "mid-tier" category, so should be an average amp in terms of DSP usage), I just divided 20 (the amount of approximate DSP that should still be available after adding dual AC30's) by that number to get the approximate usage of each distortion model.
Note, the numbers basically mean that for the Tube Drive (for example) I was able to successfully add 5 of that effect before getting the DSP warning, which should mean that it would use somewhere slightly less than 4% for each effect (again, based on the assumption that Fester's earlier findings about Amp DSP usage are accurate, which I have no reason to think they're not
). I think that's probably about as close as we're going to get without an official DSP meter (something I hope to see in a future firmware update).• Tube Drive (5) < 4%
• Screamer (4) < 5%
• Overdrive (5) < 4%
• Classic Dist (5) < 4%
• Heavy Dist (5) < 4%
• Colordrive (5) < 4%
• Buzz Saw (5) < 4%
• Facial Fuzz (5) < 4%
• Jumbo Fuzz (5) < 4%
• Fuzz Pi (5) < 4%
• Jet Fuzz (4) < 5%
• Line 6 Drive (5) < 4%
• Line 6 Distortion (5) < 4%
• Sub Octave Fuzz (4) < 5%
• Octave Fuzz (3) < 6.67%
-
-
-
-
Thanks for the kudos man! Really though, it wasn't all that much work to do, just a lot of knob twisting and note-taking. But it didn't take all that long to do - I just had to break it up into pieces while I had some time available. I doubt it was more work than what you did originally, so thanks again for the base testing you did for me to work from
-
-
ozbadman wrote:
Hi Fester,
Based on everybody's great work here, I ran some tests tonight which I believe give some more accurate numbers on some of these things. Running v1.10 of the firmware.
...
Cheers, and thanks again for kicking off this great effort.
David.
Very cool work, David! How did you get such highly accurate results? Honestly I didn't think getting anything +-0.5% was remotely possible when I started testing these things.
One quick question: I'm assuming there must have been a typo in this section, maybe?
Full Amp:
--------------
(Tier 1):37.5 +- 0.5%
(Tier 2):37.5 +- 0.5%
(Tier 3):37.5 +- 0.5%
Unless all the full amps are within 1% of each other (which is possible), these don't really make sense. But maybe if they are all this close how did you figure out that the Gibtone and Divided by 13 amps are the most expensive? Maybe because it's 3:00 in the morning and my brain isn't working correctly, hahahahaha.
Ok, really cool effort! Thanks for contributing to the process!!!
Cheers,
Fester2k
-
fester2000 wrote:
Very cool work, David! How did you get such highly accurate results? Honestly I didn't think getting anything +-0.5% was remotely possible when I started testing these things.
One quick question: I'm assuming there must have been a typo in this section, maybe?
Full Amp:
--------------
(Tier 1):37.5 +- 0.5%
(Tier 2):37.5 +- 0.5%
(Tier 3):37.5 +- 0.5%
Unless all the full amps are within 1% of each other (which is possible), these don't really make sense. But maybe if they are all this close how did you figure out that the Gibtone and Divided by 13 amps are the most expensive? Maybe because it's 3:00 in the morning and my brain isn't working correctly, hahahahaha.
Ok, really cool effort! Thanks for contributing to the process!!!
Cheers,
Fester2k
Hey Fester,
Well, by trial and error I was able to hone in on some pretty accurate numbers on the 63 Spring and Particle verbs, as well as the Tube and Octo. By varying combinations of these I could further hone in on values of other things. Say > 35, < 40, then adjust some effect combination to get >36 < 39, etc.
Yep, it looks like the Full Amps are all between >= 38 and <= 39 so knowing you had already grouped them, I figured that was accurate enough. I took the middle for all 3 groups and added the error.
To work out the most expensive, I put my combinations right on the edge, then went through all the pres/amps until I triggered a "DSP OVER message", then fine tuned it until this only happened on 1 preamp/amp. I couldn't separate the Gibtone and Divided by 13 fulls, even with just a a noise gate difference, so I put them together as the most expensive.
And what's really cool is that now as I put things together, I can fairly accurately predict when adding even a small effect is going to trigger a DSP error.
Dude, thanks again for your brilliant idea on the general approach. It's had awesome results for all of us.
David.
-
-
-
-
-
-
-
-
-
-
bverdon wrote:
This is all great info but you guys realize that each firmware update will most likely change ALL of these perameters right?
Yes - of course that is entirely possible if Line 6 manage to find ways of optimising the amp models more with every firmware update
I just did a simple test based on the recipe offered by another user yesterday in a post in which he was having severe problems getting dual amps to work and the results he was getting just seemed very wrong, but with the currently available firmware (1.22), and found that ALL full amps can be used with ANY other full amp without hitting any DSP limit with the FX specified.
-
-
jodeli wrote:
Yes - of course that is entirely possible if Line 6 manage to find ways of optimising the amp models more with every firmware update
I think I'll wait for buying, until the development has finished and the final version is available.
In the meantime I remain lucky with the X3.
You might have a very long wait then as no-one knows if and when there will be a final update
I'm pleased your X3 is still working out for you. In some scenarios, so is mine, but I rarely use the X3 Live these days because the HD500 works so well all round
-
-
-
-
-
Your comment made me look! I never paid attention to it or needed too.
-
About
-
Follow
Subscribe
Latest offers, special deals and insider updates.
Line 6 Story
News
Press
Jobs
Support
Contact Us
