Subject:ACID Optimization

Posted by: Nationalism

Date:8/11/2007 5:39:32 PM

When I press the spacebar to start/stop whatever's on the timeline, there is a delay in the playback. I have to wait a few seconds before the audio starts/stops. I have 3 gigs of ram, so I figure this shouldn't be much of an issue.

Are there settings that tell ACID to go as fast as it can? That seems like a dumb question, but is this possibly a buffering issue? Of the top of your head, are there any ways to speed up the responsiveness of playback in ACID? If not, that's cool too.
Thanks guys.

Subject:RE: ACID Optimization

Reply by: tascolas

Date:8/11/2007 10:56:20 PM

Hi,check your audio settings m8... ASIO is required to run your card under low latency ...I 'm not familiar with the latest creative sound card but one thing i can say for sure ,its not designed to run pro audio applications . Check if they have any decent ASIO driver for u to install. If not get yourself something like a m-audio audiophile192 pci which is incredibly cheap now and will definetely yield results under win xp.
After lookin at your specs i must comment on the 3gig ram u mentioned .Don't waste your money m8, win xp 32bit can only assign 2 gigs of ram per application so unless u r running Acid along with another application (eg. rewire with Cubase or Reason etc.) its pretty much useless. I ve seen ppl using LaaTIDo ,an app that extends theamount of ram adressing in win xp sp2 but im not sure how stable can it be under Acid. Also assuming you have 3 x1gb ram chips your ram doesn't work dual channel ,right? Hope this was helpful ,cheers
PS: there is buffer setting in audio preferences as well :)

Subject:RE: ACID Optimization

Reply by: Nationalism

Date:8/13/2007 12:55:28 AM

Thanks for the reply. I use windows wave mapper, ASIO can't pick up my configuration for some reason. The ram is dual channel. I have 2 x 1 GB and 2 x 512 MB, totaling out to 3 GB onboard ram. The PC shop which built it for me says XP SP2 actually CAN use up to 3 now. He said that going for 4 GB would be a waste.

As far as the quality--it's pretty decent. I just use the RCA inputs from my synth's 1/4 out, and it sounds relatively clean. Here's an example of my work:

http://media.putfile.com/July-21st-58

At any rate, I was wondering what exactly the buffer settings do. They don't really explain. Thanks again.

-Jack

Subject:RE: ACID Optimization

Reply by: pwppch

Date:8/13/2007 11:37:25 AM

What is your bufffer setting?

Are you using external MIDI or softsynths or both?

What type of media is on your timeline?

What FX plugins are you using?

If you are not using ASIO, why not?

Peter

Subject:RE: ACID Optimization

Reply by: Tomdini

Date:8/13/2007 6:19:33 PM

4 gigs of RAM is not a waste. Whoever told you that is wrong.

WinXP Pro SP2 can address up to 4 gigs of RAM. Using the /3GB switch in your boot.ini, you can allocate up to 3 gigs to your applications (Acid, Kontakt, etc.) and 1 gig for the system. By default, only 2 gigs can be used by any given application, but your tech guy might not be aware of the /3GB switch (used extensively in the audio and post-production world), which gives your applications another gigabyte to work with.

I can load up around 2.8 gigs of RAM with the /3GB switch, but if you only have 3 gigs, you'll get maybe 1.7 or 1.8, as Windows has to suck up some of the memory for itself.

However, even if you have the /3GB switch enabled, if your app isn't large address aware (LAA) you still won't be able to use the additional memory. Do a google search for LaaTiDo - it's a simple little app that hacks the .exe and makes the application LAA. I've used it on Acid and Kontakt with no problems whatsoever.

-Tom

Subject:RE: ACID Optimization

Reply by: RobertCorrell

Date:8/15/2007 1:06:48 AM

Sounds like you have track buffering enabled. Go to the Options->Preferences menu. Choose the Audio Device tab. Look down about halfway and see if "Enable track buffering" is enabled. You can enable/disable it from the checkbox or modify the buffer time from the slider.

Subject:RE: ACID Optimization

Reply by: mobiledjs.com

Date:8/16/2007 12:03:10 AM

I have only experienced this problem on my laptop where I am using a PCMCIA E-MU 1616m (EMU is a higher end brand made by creative).

The time marker would get off beat (out of sync) after I finished beat mapping the first mp3 file and opened another one right after. Sometimes this would happen in the middle of mapping the song as described in another thread. The only way to reset it would be to open the audio settings and reselect the audio drive and apply. I tried setting the track buffering to match the latency on the sound card (9ms) and then tried even a lower setting of (1ms). The settings that appears to have worked were the following:

1) Set my Emu 1616m to 10ms delay (shows up at 9ms)
2) Disable track buffering completely (I think this is the root cause as mentioned above)

I also enabled the following but I don't know if they are related or not.

1) Use slower update to prevent playback clicks during editing
2) Close media files when ACID is not the active application
3) Close audio and MIDI ports when ACID is not the active application
4) Automatically detect and offset for hardware recording latency

Hopefully it will not creep up again. I did not have the problem when using the AC97 internal sound card through ASIO4ALL. I think what happens is the track buffering is introducing latency in a card already has the capability to manage it in real time. The buffer is probably useful for regular sound cards that already have 10ms+ latency.

For my desktop system I use an Audiophile 192 and it works like a charm. I would advise strongly against using a M-Audio 1010LT. I had nothing but problems from the card and ended up returning it. The digital input/outputs introduce annoying noises. This card is also very poor with IRQ sharing. If not given it's own assignment (practically impossible) it probably introduce the crackle and pops.

If you are on a budget I would highly recommend the E-MU 0404. It has on board DSP effects and allows you to create multiple ASIO buses. However, some programs which are used to dedicated use of the sound card will not cooperate unless you use the DX drivers which introduce a lot of latency.

I have heard a lot of mis-information about Creative/E-MU from people who say they cannot make good sound cards. Considering how long they have been around I thought this seemed a little off. My other card is an Audigy ZS. I would highly recommend using the KX drivers and switching the front/rear channels to the HQ.

Subject:RE: ACID Optimization

Reply by: mobiledjs.com

Date:8/16/2007 12:24:56 AM

Actually the problem still seems to exist. I have tried to disable several other features and even tweak the latency but the problem creeps back up. This seems to only happen during the beat mapping process. Changing some of the above settings does seem to minimize the noticeable effects but so far I haven't found the magic bullet.

Subject:RE: ACID Optimization

Reply by: tascolas

Date:8/16/2007 11:24:14 PM

''EMU is a higher end brand made by creative'' &''I have heard a lot of mis-information about Creative/E-MU from people who say they cannot make good sound cards. Considering how long they have been around I thought this seemed a little off.''

At first i didn't wanna quote because all the things i say here i consider them common knowledge but its not as it seems so here goes:
Plz ,respect the historical name of E-MU that launched a ton of sampling (and synth) musician careers in the 80's and was (and still is ) used in thousands of records. Actually since '93 when E-mu was acquired by Creative their Soundblaster core dsp was made by E-mu and NOT the other way around. In fact, the current E-mu line of audio cards has the exact same dsp core chip as the soundblaster series without the EAX and the rest of blah blah the more gaming geared Sounblaster series offer. In '98 ,Creative also acquired Ensoniq ,another legend in synth/sampler/fx market and merged it with E-mu.Bottom line is- in my opinion- it took them only 2-3 years to utterly destroy two of the biggest synth names in the industry and actually vanish them from the face of the synth market. Both Ensoniq & E-mu synths and sampler series were discontinued making the hardware synth world a lot poorer. E-mu was founded in '71 and Ensoniq in '82 . They both had breakthrough and succesful products which is not my place to start mentioning as it would make this thread a monster (lol).
Whats not to blame about Creative man? Its not too long ago when i remember their cards would actually breakdown when u would put a line in minijack in its respective input while the pc was on.Actually most of sounblaster cards had this problem until the Live series came out. Dont u hear yourself m8? The Kx driver ? But this is made from Hanz Petrov & Eugene Gavrilov and these guys are NOT Creative staff or work for them in any way. Its not that i'm saying that Creative dont make good cards,i'm saying they wouldn't make cards -period- if it wasnt for E-mu to assist them.
As for the 1010Lt - i have a 1010 which is the same pci card with an additional outboard rack for almost 7 yrs now and have produced numerous albums and singles on it . Still works flawlessly on 2-3ms latency depending the host app...
PS: Plz do not defend the moguls like Creative who came to dominate the pc market after selling millions of defective or minorly defective (if u want it ) products throughout the 90's. Personally i don't and will not forget and keep in my mind i have several SB's meself as well :-)
PS2:Is it me or i just heard u say u have an Audiophile 192 and works flawlessly? Thats why i said in the first place SB is not designed to run pro audio apps m8. The audiophile192 goes for as less as 75 euros(!) and ridiculus any SB when it comes to stability - latency - sound quality -driver design . Oh , i forgot, a big F word to those pc system builders who supposedly build pro pcs and they install SB
for pro audio - they definetely are ingnorant to the pro audio world.
PS3: Not sure if Depeche Mode use a SB but they still use an Emulator III alright:-))

Subject:RE: ACID Optimization

Reply by: JohnnyRoy

Date:8/17/2007 4:37:49 AM

> I use windows wave mapper, ASIO can't pick up my configuration for some reason.

You need to fix this problem and get the ASIO drivers working. This is the only way to get really low latency. I had a SoundBlaster Audify card once and it could never keep hold of the ASIO drivers. I finally got rid of it and haven't had any problems with my M-Audio FW-410 since. Dump the SoundBlaster and get a pro audio card if you are serious about making music on your PC. As others have stated here, the M-Audio Audiophile 192 is a great little card that's not too expensive and works flawlessly. Anything from M-Audio or Echo should be reasonably priced and work well.

~jr

Subject:RE: ACID Optimization

Reply by: mobiledjs.com

Date:8/18/2007 6:00:44 AM

Tas thanks for the great information about E-MU/Creative.

After getting good studio monitors and a decent sound card I didn't realize how much the EAX actually colors the music. I tried the KX drivers for my Audigy ZS but the 48hz sample rate was always a problem. Most other cards will adjust the sampling rate instead of locking it. I still use the HQ output for simple music playback through my Logitec speakers.

I tried pretty much every configuration including both a 1010LT and 192 in the same computer. The 1010LT SPDIF input was not giving me a clean recording signal. You can find a lot of comments regarding this online if you look around. Price was not an issue, but I ended up going back to get the EMU 404 after liking the 1616M. The routing options on this are fabulous. I use the EMU between Ableton Live & Audition. The SPDIF on this is clean as a whistle. No noise to clean up at all. I don't really use the DSP effects on the card thus far. I use my Audiophile 192 dedicated for Acid Pro. However, I did have to optimize a few things along the way to make things work well.

On my desktop system this is how I have achieved very good latency...

1) Audiophile 192
2) Using ASIO and not allowing other apps to use this card at all. Make sure that windows does not reset your driver to Microsoft Mapper.
3) Used common methods of PC optimization including putting priority on background services. There is a website that describes all the things you should look for. Go through everything with a fine tooth comb.
4) Close non-essential programs and background tasks. (yes even AIM)
5) Defragmentation
6) The most important thing is having your audio files on a separate hard drive than your operating system and the software program. This can be a huge factor in latency issues.
7) Raid mirror for both operating system drive and working files. There are other more risky forms of raid. I opted for the safest form.
8) I do not use external hard drives except for accessing files that are not an active part of my current projects.

I also use several instances of Ozone. Make sure that if you are using this that you set the option for "delay compensation". If you get the Sound Forge 9 software it comes with Izotope mastering software which is pretty good esp the multiband compressor. But I have now ditched both that and wavehammer for Ozone.

The next step of optimization will be upgrading my processor to a dual core.

Subject:RE: ACID Optimization

Reply by: tascolas

Date:8/18/2007 9:09:55 PM

Hi again :)
About the raid thing ... Depending on your mobo's onboard raid controllers there are options for 0 and 1 usually with the more sofisticated mobos offering 0+1 and 5 modes.You do a wiki or google on the raid modes and u ll find that u can have better performance as well as data backup with modes 0+1 or 5 ( or even 6 now).Even if your mobo doesnt support these ,pci raid cards have grown so cheap that a typical 4hd raid card costs about 35-50 euro/dlrs. All these of course suggest u have a lot of identical hds for use in raid.
Oh , and for the dual core - do yourself a favor m8 and get a quad core if u can. I can't express how happy i am 7 months now with my Intel 2.66XE quad-core. Just for example :50mono channels with 50 Waves SSL channel strips(E-channel) @ 25% cpu . :-) Cheers

Subject:RE: ACID Optimization

Reply by: pwppch

Date:8/20/2007 6:31:48 AM

>>I think what happens is the track buffering is introducing latency in a card already has the capability to manage it in real time. The buffer is probably useful for regular sound cards that already have 10ms+ latency.<<

There is no such thing as a card that has the capability to manage latency in real-time. It is entirely dependent on the buffer size and how the card reports this latency. The ONLY thing a card can provide is a direct input monitoring scheme so that input latency can be avoided during monitoring.

The track buffering is not about latency directly. It is about "preparing" the audio that is to be streamed early enough so as not to cause drop outs during real-time playback.

What this means is that the "engine" is primed to handle hardware requests for audio by buffering. It does not directly introduce latency, but permits lower latencies from causing gapping.

Why do we offer this? So that complex projects can service the typical low latency ASIO sample frame buffer sizes that users desire. The only time you really want to work at very low latencies is when you are recording against existing material - i.e. tracking or doing comps. This permits software input monitoring to work effectively.

When you are not recording - or using software input monitoring through FX - a large sample frame size is desireable.

We have had reports where the track buffering has caused problems. In general , it does not, but if you have specific scenarios or cases where turning off Track buffering causes problems, please share as many details as you with us.

WRT the EMU hardware.

Be VERY careful with these cards. The ASIO implemenation essentially "emulates" ASIO ports. While technically this is not a bad thing, the way the EMU drivers and their mixer applet perform this, can introduce problems.

It is highely recommended that you do not change the EMU's routing or sample frame size while connected through ACID. It is possible, but due to how the ASIO emulation is done, this can cause problems.

We are working with EMU to correct the issues we have discovered with the EMU ASIO drivers.


Peter

Subject:RE: ACID Optimization

Reply by: mobiledjs.com

Date:8/26/2007 10:58:55 PM

Found ways to deal with latency issues over the past year but over the past week I finally decided to just build a new system rather than upgrading to X2 939 dual core on my current aging board (Asus A8V Deluxe)

System Specs:
Intel Core 2 Quad 2.4ghz
XP Professional (XP x64 was an idea but decided against)
Asus p5b Deluxe board (loaded with features)
2x1GB Ram - 667mhz DDR2 dual channel (kingston)
Sapphire x1650 512MB PCI-E (room for another card)
Raid-1 main for OS/programs (seagate 7200rpm 16mb 3Gb/s)
Raid-1 miror for data (same type)
Audiphile 192
E-MU 0404
Mackie Big Knob

Several pre-built similar systems are now becoming available for around $750-$1200 built. In fact, the Quad core is in some cases cheaper than some Dual Core prices I have seen. This configuration runs about 20x faster than my previous system. I can run ozone 3 with full visual display without a hiccup or freeze. I was also able to lower the buffer settings on my Audiophile 192 to 256 (from 512) and get no latency at all (probably could push it much lower). Overall the improved quality of the music and time stretching is highly noticeable now in ACID PRO 6. The Acid installations did not hang at all during the original install. Media Manger has no lag whatsoever so keeping it open is not a bother either. I am still using the Emu 404 card between Ableton Live and Adobe Audition. This time both were sharing the card simultaneously using ASIO without any problems whatsoever. The onboard sound card I am using for non-essential stuff like system sounds and winamp. The only thing that did not see much improvement was the rendering time in ACID. This might be due to delay compensation being enabled in VST plugins. I saw a demonstration of Sony Vegas using a Quad Core for a faster rendering time so I thought that it would translate into this software also. Overall I am very pleased with the results and it will be a huge productivity boost.

My laptop will have to wait till prices come down on multicore.

Subject:RE: ACID Optimization

Reply by: pwppch

Date:8/27/2007 8:24:18 AM

>>The only thing that did not see much improvement was the rendering time in ACID. This might be due to delay compensation being enabled in VST plugins.

Why do you believe plug-in delay compensation has anything to do with this?

>>I saw a demonstration of Sony Vegas using a Quad Core for a faster rendering time so I thought that it would translate into this software also.

Vegas and ACID are very different.

The improved rendering times for Vegas typically refer to video rendering, which is HUGE compared to audio.

Can you provide examples of your rendering times?

What type of project complexity?

What format are you rendering to? MP3 ? Wave? Other?

When you render, what type of CPU usage does Task Manager show?

Peter

Subject:RE: ACID Optimization

Reply by: mobiledjs.com

Date:9/16/2007 12:41:12 AM

>Can you provide examples of your rendering times?
Actually the rending time is pretty fast now. I think the problem was that my Intel Raid Array was somehow broken, but I fixed this. The CPU barely ever gets overloaded so I can perform multiple tasks without a hitch. I can even open two instances of Acid on two screens. Course I use two different sound cards on each. This is useful when copying stuff over from another project. I updated the INF files for my processor from Intel for my Core 2 Quad (make sure you know what you are doing). My advice is to set a system restore point every time you install something new on your system so you can maintain the stability and back peddle if necessary.

>what type of CPU usage does Task Manager show?
The Taskmanager does show all 4 cores as separate processors in XP Professional. If your system doesn't then you either have a version of XP Home or it was not installed properly.

>What format are you rendering to? MP3 ? Wave? Other
I usually render to mp3 format 256kps CBR

>Vegas and ACID are very different.
I was considering buying Vegas for Audio Production as well as Video. How would this compare to Acid? It seems to have all the functionality and support for more including multi-core. I was looking at the new Vegas 8 Software. Are beatmapping, acid loops, one-shots, all supported? Someone told me Vegas has a feature to manage songs with tempo varations/drift?

>>The only thing that did not see much improvement was the rendering time in ACID. This might be due to delay compensation being enabled in VST plugins.

Actually I found that in Ozone specifically it's good to "enable" delay compensation for the plug-ins however, unless you have severe lagging on your system do not "Freeze" the track otherwise when you re-open the saved project or the latency is different then it will not update the latency amount. Depending on the VST plug-in you are using there are different ways of fixing latency issues. The easiest way is to disable until render time. I also found that lowering the buffer for my sound card and matching this setting in each instance of Ozone seemed to help also.

Overall the biggest thing that I found the optimize this week was that I realized that on my raid-1 (specifically for project files) I noticed that I had not taken out the jumper on the SATA 300 drives therefore they were running at 1.5gbs (half the speed). This also fixed a weird seek noise that was happening on the seagate drives (7200rpm 16mb buffer). Warning: Make sure your board supports SATA 300 before removing jumpers.

In other news I found that when Native Instruments Battery 3 was re-installed on my computer it caused a weird DLL error with Acid 6.0d upon startup when it tried to load the VST plug-in DLL for Battery. After patching the Battery 3 install with the two latest updates (3.01 & 3.02 I think) this problem went away.