BerntR said:
From my email exchanges with him it seems Uli does a good amount of experimenting and looking at new approaches. He decided since a wide scoped autonomous approach was impractical then he would allow any owner of Acourate to send him files for hand tweaking to get the desired effect. A new approach that he was willing to share.
Super user? Do I get the T-shirt to go with the title?
Your words not mine.
I come on here to balance the propaganda of a developer pushing his wares. Hahah... just kidding.
The first correction you posted can be done with Acourate but why bother, such overcorrection sounds awful and isn't even worth bringing up outside the context of technical curiosity. In otherwords its not useful for listening to audio.
The second set of examples shown as step responses are simulations and aren't actual measurements from the listening position.
Seems to me nothing you've presented so far is directly comparable. I'm not even looking for screenshots, all I asked was does Audiolense have the features to allow a user direct and targeted manipulation of filters so that they may influence the time domain as they see fit and in this case, allow a level of close matching between pulses.
It must be an incredibly difficult to answer question since you've span it out over nearly 2 pages now and began twisting it into some petty ******* match.
ShinOBIWAN said:
Super user? Do I get the T-shirt to go with the title?
Well I think you've earned it. But from Uli - not from me
Well - you know - all these sound correction systems we are discussing are grounded on Linear Time-invariant Systems theory. In plain English it means that we trust that the impulse response through a given correction filter is predictable any day of the week.
Measurements have been taken time and again that confirms that the LTI assumption is very well justified. Whenever I have done measurements and compared with the simulated response the differences have been down in the noise floor department.
Dont let this thread become a flame.
You both are on a level, the most other will never reach.
I like to understand what i do, but in case of acourate, there is
no manual from Uli and the website only show the basics, not
the tricks, but my problem is generally, i do my steps in acourate
and i do not know why. The result is better than without DRC,
but i never knew this was the best trimming.
Only for stereo, tomorrow i start with six channels...
Regardless of the danger to bore you, i like to knew more precisely why the M-audio 1010 LT is not suiteable ?
You both are on a level, the most other will never reach.
I like to understand what i do, but in case of acourate, there is
no manual from Uli and the website only show the basics, not
the tricks, but my problem is generally, i do my steps in acourate
and i do not know why. The result is better than without DRC,
but i never knew this was the best trimming.
Only for stereo, tomorrow i start with six channels...
Regardless of the danger to bore you, i like to knew more precisely why the M-audio 1010 LT is not suiteable ?
Sonopanic said:
Are you using the demo version? If so it doesn't create filters you can use but does allow you to familiarise yourself with some of the functionality and see how it would influence your measurement data.
I'll try to continue and finish the DRC section of the Acourate guide over the next few days and cover some of what your asking but a completely exhaustive explanation of every function or operation would be a time consuming task so I've specifically distilled it down to what you need to know to get the job done yet still provide some insight as to why your doing it.
Remember, besides my own attempts to help you'll find Uli very receptive to email and there's no better person to ask for advice on Acourate.
I almost forgot about your question on 1010LT. If you look back at the first part of this thread I remember a discussion about whether it would provide the loopback functionality needed to successfully route audio through console. Although I haven't tried the card myself there was at least one person that couldn't get it to work.
Ideally what I'd like for is to compile a list and post it at the very beginning of this thread so as its clear for everyone which cards or interfaces work with this method.
For now here's the list I have so far:
Note that any channel limitations in loopback are commented upon. If there's no comment, there's no practical limit.
This list isn't complete and there will be others out there that support loopback. If you find one not listed please let me know and I'll update.
RME:
HDSP 9632
HDSP 9652
HDSP AES 32
HDSPe AIO
HDSPe RayDAT
HDSPe AES
Digiface
Multiface 2
Fireface 400
Fireface 800
MOTU:
828 mk2 (loopback of upto 8 channels)
828 mk3 (confirmed loopback of 2 channels but possibly 8)
896 mk3 (confirmed loopback of 2 channels but possibly 8)
UltraLite mk3 (confirmed loopback of 2 channels but possibly 8)
TC Electronic
Studio Konnect 48 (confirmed loopback of 2 channels possibly more due to extensive routing capabilities)
EMU:
EMU 1820m (discontinued but excellent interface that pop up occasionally on ebay for not much money)
EMU 1820 (discontinued but same deal as 1820m)
1616m PCI
1616 PCI
1212m PCI
0404 PCI
Forgot about this one:
Audiotrak:
Prodigy 7.1 Hifi (great value and probably one of the most popular to test the waters)
Audiotrak:
Prodigy 7.1 Hifi (great value and probably one of the most popular to test the waters)
the list is great !I have been warned about the ESI cards because of driver
problems. Anything heard about this ?
The e-mu devices looks pretty good.
For my understanding, when i choose the cheap 1212m
i have not enough real analog in and outs but i can do all
with the SP/Difs. The card/driver have the loopback ability
for minimum six channels ?
If this is true and the user don´t want to go analog, then the
e-mu 1010 card is enough, because all systems are based
on the 1010. Then you can start digital with 129 Euro.
Anyway, there is first a problem, i have with the convolver filter
for MPC. I can choose the convolverfilter under external filter
and load the stereo wav correction file, but the sound did not
change.
Convolver from this site :
http://convolver.sourceforge.net/
The same wav correction file works great in foobar (with the
convolver plugin for foobar, of course)
My thought was, i test the convolver first with stereo in the
mpc and if it works, i can go further to six channels.
The ffdshow has a convolver too, but this won´t work, the
whole track will be played with timejumps and no sound is playing.
Hi, Sonopanic,
While I can't speak for the ESI branded cards, I've do have a Prodigy HiFi 7.1 and it seems to have it's fair share of issues with the drivers. I know it surely must be possible to write decent drivers for this card as there a few clone cards like the Chaintech which apparently work fine.
With the Prodigy, I've always had problems with gaps in the sound with both Directsound and ASIO. This is usually when routing through Directwire, but it happens with straight ASIO output too.
Also, the ASIO driver (on most card driver versions) will often freeze when first accessed by certain software like Console (which is what I bought the card for in the first place).
OK, it's possible that I've got a faulty card, but it was brand new, and the problems occur to varying degrees depending on the driver version. There are plenty of people on the Audiotrak forum with similar problems.
The gaps in the sound are worse on some buffer settings than on others, but the severity of the issue also varies with each driver version. I'm now on Vista, so I'm stuck with the Beta driver, which has most of the above issues.
It's a REAL shame, because this card is great value for Room Correction duties as it's well built and has the excellent Directwire interface. I'm absolutely convinced that the issues are simply down to the driver software...
I've tried every possible version of the driver (on XP Pro). I've tried adjusting PCI latency settings / IRQ sharing / fresh installs. (The glitches happen on the analog speaker outputs as well as the optical output.)
When just using the output directly (NOT via Directwire), it still has audible glitches. This is with both Directsound OR ASIO from programs like Console, VST Host, Cooledit, FL Studio, and most annoyingly during DRC log sweeps! (again, both DS and ASIO).
It's also been tried on three different PC's - one with an Athlon XP 2600+, one with an Athlon 64 3400+, and it's now installed in this PC with Vista Ultimate 32-bit, a Quad Q6600 @ 2.7GHz, and 3GB DDR2-800. I still have the audio glitching troubles with the card at all buffer settings, so I'm currently using the onboard Realtek optical output.
In fact, I'm pretty sure I know what the problem is - The glitch happens at regular intervals (say every 40 seconds), and this interval changes depending on the buffer latency setting in the driver. I personally think there's a bug in the buffer handling of the driver which causes this. Aside from the occasional ASIO driver problem, everything else appears to work fine?
Just my ten cents.
OzOnE.
EDIT: Actually, I would be happy to hear if anyone has had good success with the Prodigy HiFi.
While I can't speak for the ESI branded cards, I've do have a Prodigy HiFi 7.1 and it seems to have it's fair share of issues with the drivers. I know it surely must be possible to write decent drivers for this card as there a few clone cards like the Chaintech which apparently work fine.
With the Prodigy, I've always had problems with gaps in the sound with both Directsound and ASIO. This is usually when routing through Directwire, but it happens with straight ASIO output too.
Also, the ASIO driver (on most card driver versions) will often freeze when first accessed by certain software like Console (which is what I bought the card for in the first place).
OK, it's possible that I've got a faulty card, but it was brand new, and the problems occur to varying degrees depending on the driver version. There are plenty of people on the Audiotrak forum with similar problems.
The gaps in the sound are worse on some buffer settings than on others, but the severity of the issue also varies with each driver version. I'm now on Vista, so I'm stuck with the Beta driver, which has most of the above issues.
It's a REAL shame, because this card is great value for Room Correction duties as it's well built and has the excellent Directwire interface. I'm absolutely convinced that the issues are simply down to the driver software...
I've tried every possible version of the driver (on XP Pro). I've tried adjusting PCI latency settings / IRQ sharing / fresh installs. (The glitches happen on the analog speaker outputs as well as the optical output.)
When just using the output directly (NOT via Directwire), it still has audible glitches. This is with both Directsound OR ASIO from programs like Console, VST Host, Cooledit, FL Studio, and most annoyingly during DRC log sweeps! (again, both DS and ASIO).
It's also been tried on three different PC's - one with an Athlon XP 2600+, one with an Athlon 64 3400+, and it's now installed in this PC with Vista Ultimate 32-bit, a Quad Q6600 @ 2.7GHz, and 3GB DDR2-800. I still have the audio glitching troubles with the card at all buffer settings, so I'm currently using the onboard Realtek optical output.
In fact, I'm pretty sure I know what the problem is - The glitch happens at regular intervals (say every 40 seconds), and this interval changes depending on the buffer latency setting in the driver. I personally think there's a bug in the buffer handling of the driver which causes this. Aside from the occasional ASIO driver problem, everything else appears to work fine?
Just my ten cents.
OzOnE.
EDIT: Actually, I would be happy to hear if anyone has had good success with the Prodigy HiFi.
Ok, I know I might be annoying but....
With all of the above said, I still haven't managed to generate a decent sounding DRC filter with an ECM8000 and UB802 mixer with any sound card - can anyone give any tips on log sweep measurement when using this mic?
ie. how to "properly" set up the gain an level settings on both the mixer and the sound card. Also, the issue of whether to point to mic toward the ceiling or directly at the speaker is still in question, as well as the "best" averaged calibration file to use for each mic orientation???
And, whether to use the "MAIN OUT" or "TAPE OUT" outputs from the mixer, and what volume level to run the sweep at (peak / averaged) etc. I actually had much better luck with my DIY WM-61A mic and preamp when I first dabbled with DRC years ago.
I'm currently listening to my Buffalo DAC and RevC amps, so I'm keen to try DRC again (as you can imagine).
Incidently, for some reason when I last ran "rec_impDS" to do a log sweep, it defaulted to the microphone on my Dell monitor's webcam! I thought I'd let it do the full sweep, and when I listened to the resulting impulse file, it had a MUCH sharper "snap" to it compared to with the ECM8000?
So, I must be doing something majorly wrong with the ECM / mixer as the impulses sound really dull?
Any opinions? (and don't say "stop typing!" )
OzOnE.
With all of the above said, I still haven't managed to generate a decent sounding DRC filter with an ECM8000 and UB802 mixer with any sound card - can anyone give any tips on log sweep measurement when using this mic?
ie. how to "properly" set up the gain an level settings on both the mixer and the sound card. Also, the issue of whether to point to mic toward the ceiling or directly at the speaker is still in question, as well as the "best" averaged calibration file to use for each mic orientation???
And, whether to use the "MAIN OUT" or "TAPE OUT" outputs from the mixer, and what volume level to run the sweep at (peak / averaged) etc. I actually had much better luck with my DIY WM-61A mic and preamp when I first dabbled with DRC years ago.
I'm currently listening to my Buffalo DAC and RevC amps, so I'm keen to try DRC again (as you can imagine).
Incidently, for some reason when I last ran "rec_impDS" to do a log sweep, it defaulted to the microphone on my Dell monitor's webcam! I thought I'd let it do the full sweep, and when I listened to the resulting impulse file, it had a MUCH sharper "snap" to it compared to with the ECM8000?
So, I must be doing something majorly wrong with the ECM / mixer as the impulses sound really dull?
Any opinions? (and don't say "stop typing!"
)OzOnE.
Sonopanic said:
I have been warned about the ESI cards because of driver
problems. Anything heard about this ?
The e-mu devices looks pretty good.
For my understanding, when i choose the cheap 1212m
i have not enough real analog in and outs but i can do all
with the SP/Difs. The card/driver have the loopback ability
for minimum six channels ?
If this is true and the user don´t want to go analog, then the
e-mu 1010 card is enough, because all systems are based
on the 1010. Then you can start digital with 129 Euro.
As always, this is a little more complicated than appears.
You see to be able to give console access to a PCM bitstream containing multichannel audio it must be decoded first - this means its now 6 separate channels for 5.1. After you finish with DRC or whatever else your doing to get it back into PCM bitstream ready to pass over SPDIF and to an AV amp for decoding you'd need something capable of real time encoding of multichannel. Such a thing can be had in the form of a hardware solution called Dolby Digital Live and is found on a few cards out there such as Auzentech and Asus. Sadly these aren't test or confirmed as having loopback.
To avoid this problem most people decode multichannel in the PC playback software and select a soundcard with enough analogue outputs to handle the multichannel. Afterwards they can either use power amps or the multichannel line inputs that virtually all AV amps have.
Acourate has the ability to output mono .dbl(64bit double floats). This is ideally what you'll use with convolver.
ffdshow convolver isn't suitable for use with the filter lengths we use.
OzOnE_2k3 said:Ok, I know I might be annoying but....
Agreed
Are you using Acourate?
I normally point the mic at the speaker but I've tried pointing it in the air and only noticed some minor differences above 10Khz.
As for levels, the input from the mic should be just below clipping at peak SPL so as to avoid clipping but also get the best SNR.
You can achieve this with a combination of output level from your speakers and the mic preamp gain.
Not familiar with the Behringer mixer so unable to say.
The mic in the dell probably had falling high frequency response and this was boosted during DRC because it only corrects what it see and would assume your speakers had a deficiency in the upper range and so boosts it.
The solution here is a calibrated mic.
Assuming other things aren't causing you issues you may also not like a flat FR sound so you could tweak the target curve in acourate to give a non flat response which sounds more in line with your preferences.
Finally since you bypassed this with dell mic, it could be the preamp. Behringer kit isn't often known for the best sound.
Sonopanic said:
For this method, yes.
EDIT: You can also output over digital but not to devices expecting multichannel bitstream ie. AV amp. For example you could output 8 channels over ADAT to a DAC and then onto then onto amps or output 16 channels over AES to a DAC.
To do what you want and to do it in realtime you need a card with hardware encoding of multichannel Dolby and DTS.
Sonopanic said:
BTW Sonopanic/others,
I have an RME HDSP 9632 PCI card and also an EMU 1820m interface that you might be interested in.
I planned to use the EMU in the lounge for DRC but I've had it for some months now and I don't think I'll ever get round to doing anything with it. There's 10 analogue outs and 10 analogue in's including 2 mic pre amps with phantom power for measurements. There's also a bunch of other digital I/O's such as ADAT and SPDIF.
The best value interface for sound quality and number of I/O's that I've tried so far.
More info can be found here:
http://www.emu.com/products/product.asp?product=9871
The RME is 2 channel only but can be expanded to 6 with an addon card.
Both work fine and the condition is perfect.
If anyone is interested or wants more details, mail me.
@Sonopanic,
I don't think any of the other clone card's drivers have the Directwire interface. This is by far the best feature of the Prodigy, but it's fairly useless when combined with the glitching issues I've had.
You could of course use something like Virtual Audio Cable which I believe works fine on Vista now, so I might give that a try again.
It's then possible to loop everything through Console, then back out to the speakers.
I'll have another go at building a decent mic preamp so I can bypass the need for the Behringer mixer. What I really wanted from the start was a simple preamp with phantom power and gain control and nothing else on it.
It's difficult to know when the recording chain is accurate though, and I'm not an expert on this stuff. Is there some way of working out the SNR of a log sweep, or a some sort of qualitative "benchmark" for an impulse capture?
What's the recommended SPL for the log sweep, and do certain SPL ranges work better than others?
OzOnE.
I don't think any of the other clone card's drivers have the Directwire interface. This is by far the best feature of the Prodigy, but it's fairly useless when combined with the glitching issues I've had.
You could of course use something like Virtual Audio Cable which I believe works fine on Vista now, so I might give that a try again.
It's then possible to loop everything through Console, then back out to the speakers.
I'll have another go at building a decent mic preamp so I can bypass the need for the Behringer mixer. What I really wanted from the start was a simple preamp with phantom power and gain control and nothing else on it.
It's difficult to know when the recording chain is accurate though, and I'm not an expert on this stuff. Is there some way of working out the SNR of a log sweep, or a some sort of qualitative "benchmark" for an impulse capture?
What's the recommended SPL for the log sweep, and do certain SPL ranges work better than others?
OzOnE.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.