A simple hint to avoid clicks, sync problems and similar issues

Out of sync problems, clicks, stuttering and similar errors are ever lasting titles in the forum, regardless of using AA1.x, AA2.0, or AA3.0. I have fought the battle against it many times and despite thorough testing I get victim again sometimes without warning. The most difficult situation occurs when I am recording live in MT mode. Any such error ruins the work definitely, since no repetition is possible. That is why I reflected about this situation and came up to a simple solution.

Most of us use our PC’s as multipurpose machines. That means beyond AA in any of the existing versions, there are countless programs for word processing, internet access, mail, CD/DVD burning, up to tax declaration, just to mention a few examples. On my machine there are some dozens of such entries. There are also peripherals such as modems, printers, scanners and other exotic devices. These all eat up processor power and RAM, though they may not be running when AA is on. Furthermore, the frequent updating of many of these programs, mostly done automatically via the web, as well as installation of new programs introduce changes in the configuration that may be culprit for impairing the former correct action of AA. In fact, all these interactions result in a rather poor real time behaviour.

To get rid of this complex and opaque situation, the solution is simple and straight forward. Most people disregard the possibility of all OS from W2000 up to be installed more than once. Thus, I propose to create a new partition of appropriate size (6-8 GB) and install an additional XP (for AA3.0) or W2000 (for AA1.0) on it. Use all updates (SP3/SP4), but keep the installation strictly minimum. This also eliminates all the additional problems that are encountered with Vista, and W7. No network, deactivate LAN, COM, LPT, possibly even USB, since these all may be accessed via the normal, multipurpose partition that stays available with any kind of OS. Then simply install the audio drivers, and AA, nothing else, and use this particular installation with the pertinent settings for live recording only. These settings have been described earlier. By this way, there is a stable and minimum environment for AA which is repeatable. The price for it is low: You have to restart the machine for all other work. This is boring, but much less than searching for ways to eliminate the errors in a complex multipurpose environment lacking transparency. 8 GB loss on a modern HD does not count. In some cases you may need to repair the MBR which is easily done by appropriate tools such as EasyBCD (freeware).

And/or run EndItAll or SmartClose to shut down all those other programs running unnecessarily in the background.

Many years ago, I opted for a second computer, dedicated to music applications only, with a KVM switch. 

Best 'upgrade' I ever made.

Most of the routine applications - word processing et al - are easily within the capabilities and processing power of a cheap CPU.  Use the big fast one for audio and the cheap one for everything else.

That's really the best option. For work we try and treat the DAW PCs as a piece of audio hardware like a reel to reel tape machine or a mixing desk. We don't let IT anywhere near them.

We don't let IT anywhere near them.

Excellent move! Most IT departments' staff should be locked in a room and forced to use their own systems to attempt to do some real work - perhaps then they wouldn't hack around with them so much...

Or... you upgrade your machine to the point where you don't get "Out of sync problems, clicks, stuttering and similar errors", which is where I have just about gotten to. Not expressly for AA(still run by my trusty Mia Midi), but for all the stuff I use my machine for (AA, Photo shop, games and various information security stuff I do for my job.) So I now have windows ratings of 7.6 on my processor (Intel Core i7 2.8GHz overclocked to 3.66 GHz) and memory (6GB), 7.2 on my system disk (Corsair 120 GB SSD - I am still running it as IDE) and am only brought down by my graphics card (NVidia Ge Force 8800GTX) 6.9. I have 2.1 terabytes of disks locally and 2,5 terabytes of backup and NAS storage on my Gigabit Lan (Synology server and an Amahi server both Linux based). This of course means that all Internet activity is run by my servers and doesn't affect my PC. I can access my server consoles directly from my PC and do all surfing and downloading from there.

Finally I feel I have a system which is robust enough to take everything I throw at it - I hate having to stop programs or re-boot so I feel it is a worthwhile investment and I have learned a hell of  a lot on the way. Oh and I just got a boxee box from D-Link which streams all my movies and music over wireless LAN to my TV! What could be better?

Now I just have to make some music again 

It seems that the clue of my hint was somewhat misunderstood:

1. I fully endorse ryclarks idea to treat a PC dedicated for recording and it’s specific environment as a piece of audio gear, as well as Steve’s comment about banning IT there. But none of us can live without the web, mail, word processing, photofinishing and many more daily gimmicks. It’s also true that for most of these applications, a machine with slow processing is sufficient and a KVM switch is a good tool to manage 2 machines easily as Graeme mentioned. But as soon as you want to do other work demanding high processing power such as video processing, the slow machine will not do – you need another fast and expensive one.

2. There is no doubt that a super fast machine with multicore will do all the load together without showing any problems. I like to call this the brute force method. However having a different configuration behind it every time you get an update of one of the many programs gives the chance for a failure even of a very fast machine without early warning. There is little help about this to know the fault is not being at the machine but at the supplier of a lousy tested update.

3. I have a very fast multicore machine myself with a lot of peripherals and cabling around. I hate to dismantle these every time I go for field recording just to bring my fast machine along and doing the reverse again after that. I appreciate this power mainly for mixdown and post processing as my working centre. For outdoor work I prefer to carry 2 small lightweight single core powered machines which are not much above the minimum system requirements of AA. By establishing a stable environment as described I can make sure that my concept works, and is repeatable. Despite that I can also do other work such as labeling, burning DVD for backup just using another OS an a different partition of the same machine. I just have to care to clean the disks before a new recording session.

Eventually this is a good compromise between 1. and 2. at reasonable performance/cost.

I agree, I would not use this solution if I had to be mobile. Fortunately I am a stationary type, my roots extend throughout the house!

I based my concept on recoverability - so I take an image backup of my disk weekly, and all important data daily, with a weekly off-site backup of my music and my photographs (the only critical data for me). I am sure for people where music is a business this would be impractical, but for me it covers all important bases. If I lose my system disk, I can restore it - if the worst happens (virus, corrupt image etc.) I can rebuild it and restore all the data.

3. I have a very fast multicore machine myself with a lot of peripherals and cabling around. I hate to dismantle these every time I go for field recording just to bring my fast machine along and doing the reverse again after that. I appreciate this power mainly for mixdown and post processing as my working centre. For outdoor work I prefer to carry 2 small lightweight single core powered machines which are not much above the minimum system requirements of AA. By establishing a stable environment as described I can make sure that my concept works, and is repeatable. Despite that I can also do other work such as labeling, burning DVD for backup just using another OS an a different partition of the same machine. I just have to care to clean the disks before a new recording session.

Hmm... I have a dedicated recording machine too. It has exchangeable HDs which can be read directly on the editing machine, has an OS loaded from ROM, a dedicated display with mostly hard key operation, and for location work is absolutely foolproof. If I'm desperate I can connect it to the house network and download files from it like that, but reading the disks directly is a lot quicker. One huge advantage of it is that the OS didn't come from Microsoft, so nobody is trying to give it viruses or anything like that. It's about the same size as a typical desktop machine, only all the sound hardware required is incorporated into it, so no massive connecting job required when you're on location. But the cost in total is no more than purchasing a PC or laptop, and a sound device of equivalent quality - in fact probably less, because the converters are rather good.

It's made by Alesis and it's called HD24XR. No it doesn't run Audition! Other advantages are that it doesn't burn CDs, so people don't make unreasonable requests during recording sessions. If anybody actually wants work copies to take away, they can have the monitoring feed output recorded to something if they want, but that's all. I've also got a Fostex rackmounted solid state two-track recorder, but that doesn't get as much use.

The only thing I would ever dream of using Audition for in a live session is as a backup. No particularly bad experiences, but it's simply not anywhere as easy to set up on location as the Alesis is. It's the best multitrack machine I've ever had - by a long, long way. For anybody intrigued by urban myths about Tascams - it quietly blows them away, simply because it sounds better and is way more reliable. It has been favourably compared to the Rolls Royce of digital multitracks (Radar) and some say it sounds even better than they do. Not that I'm biassed or anything... 

It happens quite often that a topic ends up in a very different corner from where it was started. I just wanted to show up a way to set up a PC most appropriate for recording without sacrificing all the other possibilities of computing. I never claimed this to be the ideal situation for recording. I have no word against  Steve’s praise of the Alesis HD24XR and I wished to posses one. Then I would immediately forget everything about field recording on a PC. Yet I have to protest vividly about his comparison regarding costs. My recording PC’s are second last generation single core which I bought for an apple and an egg, as we say in german. Spending some additional RAM and a 80 GB 7200 RPM HD, as well as a DVD burner for upgrading did not cost more than 300 SFr. ( ~ $ !) for two recorders. That  is about 10 % of an Alesis !

I do my work both as hobby and as a volunteer and therefore do not get any money from it. Thus I avoid the word business since it is no such and the expense for a 24 track machine is out of budget for some 4  to 6 sessions per year and overdone with respect of the number of tracks. I could imagine to go for an Alesis like machine for less than half the price but feeding only 8 tracks. That would certainly save a lot of troubles.

May be I am on the wrong track discussing in the forum since I am not formally an audiomaster and have no business interests. I only claim to have some technical knowledge I like to share with other people.

Yet I have to protest vividly about his comparison regarding costs. My recording PC’s are second last generation single core which I bought for an apple and an egg, as we say in german. Spending some additional RAM and a 80 GB 7200 RPM HD, as well as a DVD burner for upgrading did not cost more than 300 SFr. ( ~ $ !) for two recorders. That  is about 10 % of an Alesis !

And what high quality 24 track sound device have you included in your calculations? come to that, you haven't included a monitor either. Plus you have all the hassle of connecting this lot up when you get to your location. Alesis still wins! Well, certainly as a 24-track solution:

I could imagine to go for an Alesis like machine for less than half the price but feeding only 8 tracks. That would certainly save a lot of troubles.

You should hunt around a bit - there are quite a few devices like this on the secondhand market that are much cheaper, and would suit you fine. The worst case scenario with them is that you have to change the HDs - but that's about the only thing that will wear out, and they are generally easy to replace. As an example, I know where there are a couple of Fostex D108 Hard disc recorders going for £200 each... that sound pretty cheap to me. And as a location recorder, even though it's 16 bit (with 20 bit A-D converters), that's fine for just about any real location you come across - you'll never need greater bit depth at the capture stage at all. And I'm sure that there are other examples around too.

And what high quality 24 track sound device have you included in your calculations? come to that, you haven't included a monitor either. Plus you have all the hassle of connecting this lot up when you get to your location. Alesis still wins! Well, certainly as a 24-track solution:

I grant one point for the missing A/D converter rack and one more for the hassle, which I admit is stressing some times. For the A/D conversion I had a setback of some 650.- SFr (~ $) for an M-Audio 1010 rack type I bought some years ago. Although it is an old device running under any type of OS, starting with Win95 !,  I think its performance is still up to date and not subject to discontent. The display is not a big expense since I use a small 15” flat screen that was around 100 SFr.- as it has 3 pixel errors that don't bother me for monitoring. But definitely I would miss to follow the history of all the tracks on the screen as is the case with the Alesis. For me, there is some important information in it since for example a trumpet track has a very typical asymmetry.

Compared to all the other material, starting by the mics, stands, cables, and the many small other tools, and devices, I also admit the bare recording box is in fact is the smaller part of the money spent so far.

And as a location recorder, even though it's 16 bit (with 20 bit A-D converters), that's fine for just about any real location you come across - you'll never need greater bit depth at the capture stage at all.

My opinion about field recording with a 16 bit device is very different. In field recording, the conditions often are much less predictable than in a well known environment and usually call for more headroom. Usually I spend some 18 dB for that, thus eating up 3 to 4 bit in resolution. That still preserves the full 16 bit range for the final CD. With 16 bit recording, the loss is considerable or you risk clipping with less headroom unless you have compression on the analog side..

My last recording was a Dixieland Jazz Band at a place I had never been. Upon my estimate, I turned all mic preamps 20 dB lower than I am used to. I still assumed to have my usual headroom stated above. Due to horrible snow fall, the fellows came just in time and started the concert without a single tone of rehearsal. Despite my 20 dB additional margin, the original 18 dB were eaten up completely and I was close to peaking. 16 bit would have been a real disaster !

Eventually I might consider an Alesis, but I found a smaller brother of the HD24XR named HD24 which sells for about 2/3 of the price. As a main difference I found the small version is limited to 48 kHz and has different converter brands although the specs are the same. I could well live with that version as sampling rates above 48 kHz are not significant to me. A less encouraging point is that a found a number of complaints in the internet about HD problems and other issues. I would like to get some comments from the forum based on experience.

But definitely I would miss to follow the history of all the tracks on the screen as is the case with the Alesis. For me, there is some important information in it since for example a trumpet track has a very typical asymmetry.

If it doesn't overload, there's nothing you can do about it at recording time, so why do you need to see it then?

Compared to all the other material, starting by the mics, stands, cables, and the many small other tools, and devices, I also admit the bare recording box is in fact is the smaller part of the money spent so far.

You're not kidding. Just one of a pair of DPA 4006's costs way more than the HD24XR did! In fact, the mic stand I generally use for the Soundfield cost about the same as it.

My opinion about field recording with a 16 bit device is very different. In field recording, the conditions often are much less predictable than in a well known environment and usually call for more headroom. Usually I spend some 18 dB for that, thus eating up 3 to 4 bit in resolution. That still preserves the full 16 bit range for the final CD. With 16 bit recording, the loss is considerable or you risk clipping with less headroom unless you have compression on the analog side..

My last recording was a Dixieland Jazz Band at a place I had never been. Upon my estimate, I turned all mic preamps 20 dB lower than I am used to. I still assumed to have my usual headroom stated above. Due to horrible snow fall, the fellows came just in time and started the concert without a single tone of rehearsal. Despite my 20 dB additional margin, the original 18 dB were eaten up completely and I was close to peaking. 16 bit would have been a real disaster !

You're not seriously telling me that a Dixieland Jazz band, even recorded 18dB down is going to use all of the dynamic range, are you? Because I'd find that rather hard to believe... let's assume that your band has an unbelievable dynamic range of 45dB (more than any Dixieland band I've ever come across), then you have all of the wanted signals captured with 65dB of your available 90dB, which is more than adequate. So when you convert to 32-bit for editing, you have no problems at all.

For years, most pro stereo recordings were made on 16-bit DAT machines, and invariably you'd leave about 12dB of headroom at a minimum. And even doing this, you could record classical music with no real difficulty, and that generally has a far wider dynamic range than Dixieland. So I'm afraid that based on my own experience, and that of a lot of other recordists, I don't buy that argument at all.

Eventually I might consider an Alesis, but I found a smaller brother of the HD24XR named HD24 which sells for about 2/3 of the price. As a main difference I found the small version is limited to 48 kHz and has different converter brands although the specs are the same. I could well live with that version as sampling rates above 48 kHz are not significant to me. A less encouraging point is that a found a number of complaints in the internet about HD problems and other issues. I would like to get some comments from the forum based on experience.

The HD24 has a SNR of 103dB, and the HD24XR has a SNR of 112dB. Also it only has a slightly lower stated overall THD+N figure, other aspects of the AD-DA conversion measure better as well. It really does perform better.

As for the HD problems some people have found, you have to bear in mind a couple of things; Firstly, you only ever hear complaints, you don't ever hear from the vast majority of people who treat drives sensibly and don't have these problems... Secondly, it's also worth noting that you can get caddies for newer drives as well, and these apparently work fine. Also, there's an independent PC toolbox available for the HD24, (HD24tools) and this can rescue just about anything, including the effects of a brownout (don't ask me how I know this!), and that makes the whole thing a lot more secure to use. The ultimate security though comes from running the recorder from a portable UPS, which is what I now do, and with that, the whole rig is pretty much bomb-proof.

In the light of your vote about 16 bit recording with a floor of some 90 dB, it will not matter having a range of 103 dB or 112 dB in the recorder anyway. I also can hardly image that you can distinguish a THD of 0.00x % from 0.00y % since all speakers and headphones for the final conversion are at least one decade worse ! Thus the HD24 should do well for my case !

You're not seriously telling me that a Dixieland Jazz band, even recorded 18dB down is going to use all of the dynamic range, are you? Because I'd find that rather hard to believe... let's assume that your band has an unbelievable dynamic range of 45dB (more than any Dixieland band I've ever come across),

Yes I am. During the last days I occupied myself by mixing down the Jazz session mentioned before. Although I used 4 mics only, the raw mixdown many times well exceeds a range of 50 dB, in few cases even 65 dB. It will be even worse, if I consider a single track only. Although the final mixdown will have at least some manual compression to lower this figure, there is very little space left on the recording side in the field. Why otherwise would everybody go for 24 bit ? You may check my sample, from which, due to the length of the piece, I had to cut some part. You simply have to belive me that I made no level adjust.