Print To Tape - Humble Pie Time!!!!

Grazie wrote on 6/28/2002, 4:22 AM
Well Folks, I've spent even more time on this PTT issue. And Chienworks's "Gut feeling" is still ringing in my ears, So "Humble Pie Time" - saving files and the defaults didn't cut it. I don't know where I saw the drop from 60% to 30% of processor usage, but since I posted the message I have not been able to repeat the success. Oh dear!

ANYWAYS.....

Ahem.... I've interrogated the Sytem Monitor for hours to see if there are any clues to my situation.

I am still getting "drop-out" on freshly rendered clips and at different places on those clips. However, this sometimes happens - it is not a consistent "fault". Sometimes I can PTT for 2,3 or 4 PTTs on the trot, then maybe on the 5th attempt drop-out. Then onwards through 6,7 etc and no problem. There is something Hit 'n Miss about what I'm doing. Because of these successful PTTs am I right in thinking that the timecode is not the issue?

All I want to do is be "secure" in the knowledge that I can rely on PTT happening and, if "drop-out" is case, also be knowledable enough to take preventative measures to ensure PTT being succesful.

Getting back to System Monitor - here's the thing. I am registering Disk Cache Misses at around 900/second. I don't know if this is a large number or it is something I should ignore? Is RAM taking up the slack and therefore being "overrun" with Disk Cache call-outs?

If the PTT process is in fact dependant on making correct write-to-disk then I may have "stumbled" on the reason for my problem. What do you think?

Could someone do a PTT test with System Monitor running and tell me if they are getting the type of DIsk Cache "misses" that I am? If they are, then I need to go back to the drawing board! If it can be proved that my 900/sec misses are the reason then I need to somehow envoke a way of eliminating this.

I can tell you that I am not aware of my on board C: drive taking up any of the slack. That is I am not aware that the C: drive is "handling" any Cache writing - I don't see my red light flickering on my C: drive - yeah I'm not very high tech. My external F: drive is flickering nicely - this is where I've got the VF video files. All my VF programme files on the C: drive.

Next question: Does VF in PTT make use of a Windows Temp file or other Disc Cache writing procedure to ensure the fidelity of the PTT process? Is this "temp" file/directory set-up by VF? If not how do I force this to happen?

Please get back to me. I know I'm nearly there with this wonderful NLE package, I just need to go that last inch.

Grazie

Comments

kcarroll wrote on 6/28/2002, 7:16 AM
Grazie;

As I may have mentioned before, I'm not what you'd call a "Windows Guru", so I stand prepared to accept correction from more qualified people on this forum: Having said that, her goes!

You mentioned that you are using an external hard drive: As an experiment, would you be able to move the file you're trying to PTT over to your "C" drive and try it again?

I ask this because of a problem I had. Early in my search for more disk space, I bought a QUE 60gig 1394 external drive. It installed easily, and worked beautifully: right up to the point where there was other traffic on the 1394 bus. The instant that happened the drive disappeared. Windows no longer saw it. I had to restart Windows in order to find it again.

Needless to say, this drive was not solving my Video Capture problems, so I took it back. I was prepared to forget about 1394 externals, and go SCSI, when one of the salespeople said; "I can't tell you what's going on with the QUE, but why don't you try a MAXTOR? I use one myself for video!"

I decided that I had nothing to loose, so I bought a MAXTOR 80gig 1394 external, and tried that. The problem never re-occured. I have been using that drive for capture, files in process, and PTT operations for the last three months, and have had NO problems.

What I think I learned from this is that there are definate differences between drive controllers, and these differences make them more or less able to deal with other traffic on the 1394 bus.

It just might be such an issue that is causing your difficulty.

kcarroll
Grazie wrote on 6/28/2002, 9:25 AM
I too have a Maxtor [60gb 5% free] - it is the F: drive I speak of. Yes, I have tried putting the file onto the C:. Same result.

However, I do feel there is, what you call a "traffic" problem, occuring. Hmmmmm....

Kcarroll, would you be prepared to do a test for me? It is the one about using System Monitor and see what, if any, cache misses you get when you PTT a small, even one minute would do?

Thanks anyway for your input.

Grazie
soundguy63 wrote on 6/28/2002, 11:11 AM
I too have recently had some PTT drop-outs. However, these would occur at exactly the same point everytime, but not if you just played that section (either from the timeline or from the rendered file). They only occurred when attempting to PTT the entire rendered project (29 minutes). This suggested it was either a file fragmentation issue or some kind of RAM/Cache/Buffer/Read bottleneck that was occuring at the same two places during playback of the entire file (roughly 10 mins in and 20 mins in).
With the deadline approaching, I did the following:
1. Several defrags/bounce to another drive/re-defrag/bounce back to the video drive steps until I could confirm that the file was not fragmented (i'm using Windows2000 Professional which gives you those details).
2. Did PTT successfully with Pinnacle DV7 using the defragmented file because I just didnt have anymore time to risk on another unsuccessful try with VF.

I havent yet had time to try PTT'ing this same program again with VF to see if it will now work successfully with the fully defragged file. One thing I noticed is that the preview monitor isnt active during PTT from a rendered file with DV7. The preview monitor represents alot of load on the computer's resources.
Have you tried turning off the preview monitor in VF when PTT'ing to see if that free's up enough resources to prevent dropouts? I will let you know when i'm able to repeat the test with VF.
Grazie wrote on 6/28/2002, 11:46 AM
Thanks SOundguy63 - SOund Advice!

You have been through it, haven't you? Sooo it's not just me. Yes I do feel there is a bottleneck or as Kcarroll says a "traffic" issue!

No I don't see anything in the Preview window when I'm PTTing. I've got 85-90% system resources when I'm PTTing. Have reduced Startup msconfig to just Systray.

Please do a System Monitor experiment and post your results. Cheers!

Grazie
BillyBoy wrote on 6/28/2002, 10:29 PM
You can almost guarantee dropped frames if you use the same drive as your operating system (typically C) so for best results invest in a seperate drive. Just making a seperate partition is useless. Use a seperate drive on a seperate IDE channel and that should get rid of dropped frames. If that don't do it, most likely something with the camera or interface card (assuming digital firwire) not being up to specs.
Grazie wrote on 6/28/2002, 10:56 PM
Billyboy

I have an external Maxtor drive F:. Is this a partition or a separate drive?

Video drops frames in different places. Is that something wrong with video?

Video gets sent through Hollywood Bridge. Is that something wrong with Camera?

"interface card (assuming digital firwire) not being up to specs." Hint at what spec to go for please?

3 out of 5 attempts work. Hint at what is "tripping" this up, please?
Chienworks wrote on 6/28/2002, 11:34 PM
Grazie: partitions reside on the same physical drive. Your maxtor external is quite obviously NOT the same physical drive as the one inside the computer ;)

I dunno. Looking over your procedure and equipment, i have another gut feeling (sorry!) that your problem is the Dazzle unit. I've seen way too many other people both here and in the Vegas forum talk about ditching the thing for the Canopus or Sony units and seeing all their dropped frames go away.

Wish i had better guts for ya!
BillyBoy wrote on 6/29/2002, 12:14 AM
I have an external Maxtor drive F:. Is this a partition or a separate drive?

Like Chienworks said. ;-)

Video drops frames in different places. Is that something wrong with video?

Probably not.

Video gets sent through Hollywood Bridge. Is that something wrong with Camera?

For either Video Factory or Vegas Video to work as well as they can, both your camera and interface card must be 100% IEEE 1394 DV (firewire) OHCI Compliant. The rub is SoFo doesn't offer any "approved" list of cameras or interfaces cards so you need to rely on what the manufacture says in this regard. I was very skeptial in buying my camera (a Canon ZR45) because it claims to be firewire supportative but no where on their site or in calls I made would anyone confirm it was 100% OHCI compliant. Funny, I saw a little Windows XP certifed sticker on a camera demo in the store and at lease knew Windows should have no problem with and it don't. I was lucky enough to get a 100% OHCI IEEE card.

"interface card (assuming digital firwire) not being up to specs." Hint at what spec to go for please?

Like I said above.

3 out of 5 attempts work. Hint at what is "tripping" this up, please?

Any chance you're overlooking the obvious? Turn off screen savers, anti-virus, and background task, even something has harmless as a newsreader refreshing its keep alive signal, (so don't be online) anything that may distract the OS from the single task at hand. What you can do is download a copy of Windows Power Toys which contains Tweak and set up a mimial configuation to ensure nothing but what absolutely has to be running is running. This is much the same advice for burning a CD. While doing the editing and rendering you can use your computer for other tasks and it won't hurt anything. When you transfer from camera to PC or the other way around that is all you should be doing. You see check to see if your hard drive is in need of a defrag also. If the video you're trying to "print to tape" is badly fragemented the extra hundred or so microseconds for the OS to find all the parts can also result in dropped frames.
Grazie wrote on 6/29/2002, 2:56 AM
Brilliant. Good feedback from both of you. Sensible, timely and coherent.

Cleared up the F: C: partion or separate drive thing. Chienworks, external would also mean to me that I had a separate drive. I stated this previusly that I had an External drive. BillyBoy in asking me the quesion made me doubt my knowledge on the issue. A partition is part of an exsisting drive - yes/no?

Concerning "100% IEEE 1394 DV (firewire) OHCI Compliant" this card has been Dell Qualified as such. WHo am I to disagree?

BillyBoy, in my previous postings I've explained that I've reduced everything down to SYStray, Scanreg and Powerprofile [as per recommendations by VideoGuys Video NLE Tweaks]. This has afforded me 95% of Sys resources. I've disabled virus protection and I don't use anything else. The only other programme I'm using, to monitor the processes, is System Monitor. And yes, before anyone says it, before I loaded this I was still getting the problem.

Okay, here's the rub or observation or solution [I'll also put this in also as a new post "Print To Tape: VF versus Studio7" as I think it is very useful] and do see Soundguy63 above.

Print To Tape: VF versus Studio7

Well, following Soundguy63 also getting the problem of Drop Out using PTT in VF, and having to complete a project on time, sent the project using PTT in Studio 7. I too again fell back on my "back-stop" of doing the same through S7.

Here are my results:
Using System Monitor I have been able to establish that when using Studio7, I have achieved a reduction in processor usage of 50%.

In VF: 60% - 70% processor usage.

In Studio7: 20% - 35% processor usage.

NB: This is what I got when I posted the "PTT Cracked It!" thread. - Sorry folks it was in Studio7 that I achieved this reduction - NOT VF.

Soooo… what does this tell me:

1. My system of External Maxtor drive and Firewire set-up works. It also confirms to me that the rendered pre-PPT clip was capable of being used for PTTing - no breaks/leaks in timecode.
2. Processor Usage [PU] is critical to PTT - Too much PU [60%-70%] = Drop-out; Optimum PU [20%-35%] = No Drop-Out.

Added to which, within S7 you get position control of the transport of the clip, right down to a single frame point. I don't see this as available in VF. This means, should I ever get "Drop-Out" in S7 in the future, I can stop the process, re-cue both PTT process and shuttle tape on VCR to the "Drop-Out" position and recommence. This is both neat and controllable.

I was hoping to get a solution for VF's PTT. Hey "You can't always get what you want!". However this might be the identification of a bug in the VF PTT procedure, which our friends at SF might wish to consider correcting. Stranger things have happened!

Grazie
BillyBoy wrote on 6/29/2002, 8:25 AM
"A partition is part of an exsisting drive - yes/no?"

To get a little more precise a partition in part of a file system. A brand new hard disk out of the box is just a container that can hold "X" number of bytes. To be useable you need to partition the drive which simply means divide it into areas. A hard disk may be just one large partition taking up the entire disk or several smaller ones, depending on your version of Windows and your BIOS.

Fdisk creates partitions. It allocates space. Next you need to format the partition which also assigns a drive letter and determines how large file clusters are going to be. Under older file systems like FAT32 there can only be so many clusters, so if you have a very large drive the clusters can become excessively large and you end up wasting space if you have just one or a couple large partitions.

The difference between the actual bytes in a file and how many bytes a file takes up spanning clusters is called file slack. Windows needs to write files to the hard drive in blocks, filling up one or more clusters. As a drive fills up less contiguous clusters are available and so files are split across scattered clusters all over your disk. When this happens to excess your drive becomes fragemented, slowing access speed down and also putting more strain on your hard drive since it needs to move its read/write heads around a lot more.

"Using System Monitor I have been able to establish that when using Studio7, I have achieved a reduction in processor usage of 50%."

That actually may be a bad thing rather then a good thing. How much of your CPU's computering power a application tries to grab is more due to how well it is written. For example Vegas Video, VF's big brother will grab 100% of available CPU time slices if nothing much else is running. The more heavily used in general the faster you should get done with whatever you're doing. That does not mean you can't use your PC for other tasks. VV will gladly 'give up' some of its CPU access if you start other applications. Some poorly written applications even if you shut them down don't properly 'give back' memory they are no longer using.
Chienworks wrote on 6/29/2002, 8:32 AM
OK, a little partition information, not that this will help the current problem at all, but it may clear up some confusion ;)

- Every (usable) hard drive has at least one partition, else it wouldn't be usable. Often this partition fills the entire drive.
- Every partition is assigned a drive letter under MS-DOS (C:, D:, E: ... etc.)
- Hard drives can have more than one partition.
- Therefore a drive gets a letter for each partition it contains.

The advice to have your media files stored on a separate physical drive from the OS and program drive (rather than just another partition on the same drive) is due to the fact that while a program is running, the Operating System and program must often refer to it's own data files. This causes the drive head to move, often quite a bit. If the media files are on the same drive, the head must then be repositioned back to the media files. These moves take time and can last long enough for the video output stream to get interrupted. If the media files are stored on a separate drive then this situation doesn't occur.

Badly fragmented media drives can also cause this to occur, but it's rare for a media drive to get so badly fragmented for this to be a problem. It is the nature of large media files created in one shot that they are minimally fragmented even on a drive that has had no fragmentation maintenance. Modern 7200RPM ATA100 drives are also fast enough that having a slight head reposition between every few sectors probably won't slow it down enough to interfere with a 3.75MB/s DV data stream.
Grazie wrote on 6/29/2002, 11:26 PM
Brilliant technical teaching here!

- Soooo.... why does VF sometimes fall over at irregular intervals in PTT? VF is meant for WinME - it says so on the box and in several other places too. I've got an external Maxtor drive [F:] which holds my media files and I am going to order another extenal firewire Maxtor [80gb 7200 rpm] in the coming week.

What would you two advise I do?
kcarroll wrote on 6/30/2002, 8:54 AM
This has been a great discussion.

One thing still troubles me:..... Windows ME.

I too own a Dell, and my machine originally came with Windows ME. The whole time I ran ME, I couldn't start a serious project without first first burning incense, and sprinkling Holy Water around my office. Windows ME (or at least my installation) was simply demon possessed.

Since I am not a Windows Pro, I contacted the Consultant that my company uses and asked him about my troubles. His advice was to dump ME, and go to XP or 2000. I paid him to properly load and configure 2000 on my machine, and my problems vanished.

I do not have problems with dropped frames, and I have NO problems producing. (I'm not taking personal credit for this, I'm just making the observation.) Are we really sure that you're not simply a victim of one of Bill Gate's more extravigant practical jokes?

kcarroll
BillyBoy wrote on 6/30/2002, 10:50 AM
This is not meant as sour grapes, however I must tell you my history with the Maxtor 80GB external drive. I had three. Actually I bought one, on my second replacement. The issue isn't the drive itself, rather the cheesy little interface card that connects the drive and the external connectors that extend through the case. Mine failed three times doing nothing more than attaching and removing the firewire and power cables repeatedly.

The failure is easy to spot, if you have one you'll know what I mean. Normally when you power up the drive you get the normal green power indicator, then a red one which indicates the drive is ready to read/write. Well when the interface card fails, no red light and what you seem to have is a "dead" drive. Actually the drive is fine. I overcame this by removing the drive from the case and now I am using it as a regular drive which is all that is inside.

So a caution. If you use the Maxtor as an external drive on your desk and don't fiddle with the connectors going back and forth from one computer to another like it is suppose to do, no problem. If you do go back and forth a lot, there may be a failure in your future. This isn't just me. Over at Cnet forum over 56% of the people (if I remember the number correctly) reported displeasure with this drive. Extremely high negative number and perhaps why Maxtor keeps replacing them and not really asking questions why. They seem to know its going to fail.

As long as this post is negative, I may well comment on Windows ME. I think and many agree that ME is the worst version ever of Windows. If ME works for you without problems then fine leave it be. If however you have problems you'll probably save yourself a lot of stress upgrading to XP which is by far the most stable version of Windows yet. That said, if you have legacy hardware or software or are a game freak then XP probably isn't a good choice because XP will throw a fit if you try to install unsigned drivers. For those not familar with this new term, XP is much more stable because Microsoft has taken the step of certifying drivers which as you probably know is the number reason for problems. Microsoft mantains an extensive Compatibility List which shows what will and what likely won't run well with XP. So if you're thinking about upgrading do check out the list and also you hardware/software sites for updates.

http://www.microsoft.com/windowsxp/compatibility/default.asp

Grazie wrote on 6/30/2002, 1:58 PM
Hmmmm... Thank you for the advice....

No I don't plug and re-plug the Maxtor. I am very careful with it and leave it well alone - apart from defrag and scan disk.

Question: Do you think I should "partition" the new 80gb Maxtor I am getting? I've read that a separate partition used solely for the editing thrashing about would ensure that the cluster searching that goes on would be best placed in a "defined" hd area. What do you think?

I'm still waiting for Soundguy63's feedback on his experience with PTT drop-out.

What I'm still left with is the feeling that if SF says VF runs on WinME then surely it would. Or am I being naive?

Sour grapes - nah! - BB you have hard won expereince and thank for sharing it with me. I promise you your words have not fallen on deaf ears. And no I am not going to ignore what is, after all, a very diplomatic stance you've taken - really much appreciated.

Kcarroll - sounds like a good idea to spend money on a "consultant" - I just want to be assured that I have done the best I can.

Another thing that is really getting up my nose, is that Studio7 will not fall over when I PTT. Hmmm....

Oh yes, my File system is set up to "Write Behind" and it is also set up as a "Standalone PC". I've recently read (thanks BB) that I would be good to set it up as a network computer.

Any further thoughts, or are you all fed up with my "grasping-at-straws" ("Drop the Dead OS System!") and believe that I should get rid and move on.
Chienworks wrote on 6/30/2002, 2:25 PM
Grazie, i think the disparragement against ME is directed at all the shortcomings in ME, not about whether VideoFactory is able to run under it or not. I'm sure that Sonic Foundry followed all of Microsoft's guidelines in creating a Windows ME compliant program. However, what we're saying is that ME itself is just too brain dead to be of any earthly use ;)

Also, partitioning your external drive probably wouldn't be of any benefit. It's already physically separate from your OS drive and that's what counts. Partitioning it would just cause confusion and potentially make more problems when working with larger projects. Leave it be as it is.
Grazie wrote on 6/30/2002, 3:40 PM
Hmmmmm..."itself is just too brain dead to be of any earthly use".

Thanks
BillyBoy wrote on 6/30/2002, 3:52 PM
Yep, my remarks are about ME in general, not how it works with Video Factory. If it works for you, others, fine, I'm also a strong believer in: if it ain't broke, don't try to fix it. LOL! I was happy with my Windows 98SE, which I went back to after trying ME where I couple application I had used just didn't get along with ME very well. The only reason I upgraded to XP was to take advantage of NTFS and didn't see any point it going to Windows 2000 when XP was already out.

I think both Video Factory and Vegas Video are among the most stable Windows application out there which is a tribute to SoFo's team of software engineers.

kcarroll wrote on 6/30/2002, 5:46 PM
I appreciate the warning on the Maxtor Drive connector! Fortunately, I don't do a lot of plugging and unplugging. My main reason for going external was that all four of my internal IDE connections were full.

Since I had heard some strange stories from people who had experimented with a second IDE controller, (for the purpose of having more than four IDE devices) I didn't go that route.

The 1394 external drive seemed like the best possible solution.

kcarroll