Jump to content

Pending new releases and existing issues. RESOLVED


Chris Ward

Recommended Posts

@Chris Ward Chris for some reason your @ are not showing up correctly. I think because you are not “selecting” the member from the drop-down menu?

As to LatencyMon. I’ve run it numerous times, but have no idea how to interpret the results. Maybe someone will post a link to how to read and utilize the findings.  Maybe I’ve just “missed” it?

t

Link to comment
Share on other sites

Latency Monitor :

Green = good

Red = bad. 
 

I’ve never gone beyond that point because since W7 I’ve never seen Red. 
 

I think I mentioned that running the Samsung Magician was a good starting point in my first or second reply. 
 
Reading some of this I see lots of red flags.
Especially plug ins that are known to chew a giant hole in your resources. 

Doesn’t hurt to try a different audio interface. I just watched a review video about interfaces and he put the Steinbergs low on the list due to pour quality pre amps. But would assume they have good drivers. I just tossed a Motu because of the driver. 

Edited by John Vere
Link to comment
Share on other sites

30 minutes ago, John Vere said:

... I just watched a review video about interfaces and he put the Steinbergs low on the list due to pour quality pre amps. And they are a Mac oriented company.  ...

my experience with the Steinberg IO (2 ch and 4 ch units) is completely different - built like tank, solid drivers on Windows, and excellent audio performance. definitely NOT the best for any desired "preamp" magick, but then again IO units (imho) should be about perfect transparency and not some nuanced preamp magick. ? that's what your rack of 1073's are for...

  • Like 1
  • Haha 1
Link to comment
Share on other sites

I always have had respect for Steinberg but the point is if they don’t update the windows driver then it can be a factor in the buffer issues. Definitely my experience. 

Same projects same computer 

Motu M4 has to be at 520 or there’s crackles.

Tascam is 1641 runs at 256

 Scarlett 6i6 at 256 

Zoom L8 at 128

Soundcraft signature 10  520

Behringer xair 16. 256.  

So in my experience drivers are very important and Mac oriented company’s like Motu seem to fare worse. 

Edited by John Vere
  • Like 1
Link to comment
Share on other sites

6 hours ago, DeeringAmps said:

As to LatencyMon. I’ve run it numerous times, but have no idea how to interpret the results. Maybe someone will post a link to how to read and utilize the findings.  Maybe I’ve just “missed” it?

You haven't "missed it" because no one ever gets specific with what to do with the results.  After you run the program for a few minutes you can click  on each of these tabs to see what it's telling you.  Green doesn't necessarily mean it's time to move on.  You can still use it to tweak processes to improve overall DPC latency spiking.

image.png.5c8d6150fded781d1a7da7fd5462b466.png

Stats will give you the headlines.

Processes will give you an idea of what's running in the background.

Drivers will show you all the drivers and which are running the highest ISR, DPC counts and the highest executions in ms.  It'll tell you the company that the driver belongs to, etc.

CPU will tell you which CPU(s) is(are) getting hit the hardest with DPC latency and which driver is causing it.

You can then search those drivers and processes online to see what exactly they do, etc, and if there is anything you can do to alleviate the problem.  Maybe you can just shut off a process...maybe you can't.  Maybe you can make sure that process doesn't run on the same core you're running audio on, etc.  Maybe it helps you figure out what to change in the bios.  Once you try a fix, you can run LatencyMon again and compare where you were to where you are now.  You really can chase down problematic drivers and processes that are causing latency spikes.  Just takes time.

 

  • Thanks 1
Link to comment
Share on other sites

Hi everyone,

I have only JUST noticed a pile of responses/additions to this thread on page 2 and can't think why I haven't spotted them before. So, gentlemen all, please don't think me rude for not acknowledging each one. I DO SO NOW !!!

All messages were helpful/informative and have helped me focus tighter on what I need to investigate in more detail. RAM and disc issues is what I shall have my Techie look at when he comes next Friday 12th. I expect to receive another SSD on Monday. I will task him with mirroring my 'C' Drive to safeguard all my installed apps and associated folders/files. I will then ask him to uninstall everything from my existing drive other than Win 10, CbB and Band in a Box, as well as check the drive for any faults or failings/decline.

He has already read through this thread (even though I didn't tell him it was CbB forum ??!!!) and has also checked possible upgrades for my motherboard. Unfortunately, Intel have stopped making the i9 chip for the one I have and therefore the only upgrade possible is RAM. What seems logical to me is to go this route of checking 'C' Drive and RAM for failings, increase RAM anyway and see what improves (or not), BEFORE splashing out on a new Audio PC for CbB and Band in a Box only.

What I have been doing yesterday and today is play around with the 'duplicate' I made of my current project. This involved freezing ALL tracks, both VSTs and audio, and noted 'Performance' readings in CbB. I then proceeded to 'bounce to track' the VSTs only and noted the readings. I then bounced the audio tracks as well and noted the readings. I then closed the project without saving and reopened so that I was back to my VSTs being 'open' and instead disabled ALL the plugins (some on the VSTs and others on the audio tracks. THIS IS WHERE I STRUCK A HAPPY PLACE. Audio Processing shot down to 0.5% (max 3.03%), Engine Load down to 2.9% (max 6.8%) and, of course, zero Late Buffers.

What has completely surprised me is that the relatively few plugins (11x) take a greater hit on my rig than the 'open' VSTs !!!  There was very little difference between ALL VSTs and audio tracks frozen then bounced to track and then deleting the originals compared to leaving the VSTs 'open' and disabling the plugins. (This could well be pointing to my 32Gb of RAM as being the culprit.

If any of you are interested in seeing the details of my findings, let me know and I will upload them. They may be helpful to others who are having similar issues and can act as a system comparison.

So, blowing out all the crap from inside the box; re-seating my RAM; checking RAM and drive for failings; increasing RAM to 64 or 128Gb; double checking my CbB, VSTs and samples are where they should be; downloading and running Samsung Magician; plus all the other tips you have so kindly suggested will all be undertaken. James Hook has just commented on LatencyMon as I'm writing this - thanks James.

Sometime after next Friday when new RAM is added as the only upgrade possible on this machine, I will post an update. I have to say guys, you are an absolutely terrific bunch of helpers. A thousand thanks to you all !!!

All the very best.

 

Link to comment
Share on other sites

5 hours ago, John Vere said:

I just tossed a Motu because of the driver.

I hope you don't mean you dumpstered it. There are always folks who can use old stuff like that. People with older systems and such.

(You know how I am about keeping legacy stuff running)

Link to comment
Share on other sites

7 hours ago, DeeringAmps said:

As to LatencyMon. I’ve run it numerous times, but have no idea how to interpret the results.

As John said, on the front page it will give you its opinion of whether your system is in good shape.

If you want to see what's limiting your performance, the best place to look is the Drivers page. That will show you what driver(s) are chewing up system latency.

That's how I spotted my bad NIC driver, I saw that NDIS.SYS was clobbering my system. NDIS.SYS is a driver for network operations, so I started looking into it.

  • Thanks 1
Link to comment
Share on other sites

I ran LatencyMon for 4hrs 15mins this evening and for the first time, having run the software over 20x times in the last 6x days or so with green light, I got a 'fail'. Is this because I have been playing this damn song over and over and over again ??  Is my SSD being accessed at the same memory zone each time ??

I'm leaving it alone now until the drive is safely mirrored next Friday.

I've been clearing out unecessary projects as well and came accross one of the songs I did last year without glitches.

It's got EZD, AmpleGuitar, ManyBass VSTs 'open' (still in midi) with plugins on them and 12 x audio tracks with 2 - 4 plugs on each, plus my Master Bus with 4x plugs and 2x send Busses with 4x plugins between them. That's 69x plugins in all and the damn song plays cleanly 5x times over with Audio Processing maxing @ 56.51% and Engine Load maxing at 50.8%. Yet here I am on the same computer playing my latest project and having to 'bake' my 4x VSTs to bounced tracks as well as my 5x audio tracks, also bounced in order to bake them, before I can get zero Late Buffers.

This gets weirder and weirder.

Link to comment
Share on other sites

2 hours ago, Starship Krupa said:

I hope you don't mean you dumpstered it. There are always folks who can use old stuff like that. People with older systems and such.

(You know how I am about keeping legacy stuff running)

I just bought it last year. It’s not in the dumpster and I will sell it to someone with a Mac

?

Chris it sounds like possibly the difference might be as many of us have said Is certain plug ins will kill your system. Possibly the project you had no problems with does not contain problem child plug ins. Problem solved.  

Edited by John Vere
Link to comment
Share on other sites

Hi John,

Thanks for your latest. That mention in the past did not go past me and I am double checking right now what new plugins I may have bought since last year. Not many is my reckoning. Nectar 3 upgraded to Nectar 4; Waves Silk Vocal; Soundtoys Superplate. As I only have 11x plugins in my current project, that investigation won't take long.

Link to comment
Share on other sites

All 3 you mentioned are well known as being CPU intensive.  
How can this thread be 3 pages long and this was not obvious?
Either we are not being very helpful or you are not focusing on troubleshooting methodically but as I say.”  It’s always a plug in! “. 

Edited by John Vere
Link to comment
Share on other sites

Update on the plugins matter :

I have removed all the plugins on this current project (as I've done before last week) and whilst the Audio Processing and Engine Load figures dropped noticeably, there was a spike at one point on the Engine Load which shot the percentage figure to 227.2% Max, resulting in a Late Buffer. This has been the case over and again these last 2x weeks. And bear in mind that the old project I played through last night had 60+ plugins as live inserts plus 'live' VSTs  and the Audio Processing up to 35% and Engine Load up to 50% with no Late Buffers. In other words, the load on Processing and Engine does not necessarily correlate to my experiencing a Late Buffer or not. Those figures can be high as on last year's project and give no Late Buffers, whilst being low on this current project can give me, on average, 7x Late Buffers.

I'm beginning to wonder if it has anything to do with Memory Paging, but don't really know anything about it. How can it be attributable to one specific audio project. I shall ask the Engineer next Friday. I did download and then run Samsung Magician last night and the result on my 'C' Drive was "GOOD".

Last night when I got a 'FAIL' on LatencyMon for the first time, it was after running it for 4hrs plus. All previous 'PASSES' were achieved after running for 30mins - 1Hr. Now, I have mentioned earlier in this thread that I can stop the engine and with my cursor resting in the 'Performance' window, watch the Late Buffers count continue to rise, sometimes by a single unit every 30mins and occasionally shoot up at the rate of 20-25 per second. The more these experiences run through my head, the more I'm inclined to think this is an issue with computer memory, whether physical RAM or 'virtual' Paging memory.  I can't wait to see what we find in a few days time.

NOTE: LatencyMon has been running for 52mins now and I am showing all clear again. Is it the case that the longer it is run, the more likely it will move into the red on everyone's PC ??

Cheers.

Link to comment
Share on other sites

Hi folks. This may be the last post in this thread and my apologies if it's going to be another long one.

Even as John Vere was coming back with certainty that it's always a plugin that is to blame, I was already deleting all the plugins inserted in the VSTs and audio channels (even though I had already done this some days back on his earlier advice - see earlier posts). With them all gone, and since the start of this thread, my Steinberg proprietry driver set at 2048 Samples, I started playing back. My 4x VSTs in this project were 'live' (being played through). ZERO Late Buffers and max Engine Load around 45 - 50%. I then started to reduce my Sample rate in order to be methodical. Amazingly, I got it all the way to the lowest available setting of 32 Samples and at Steinberg's "Stable" condition. Each time with ZERO Late Buffers. Fantastic - but no plugins. I then decided to push things further and set the ASIO driver to Steinberg's "Standard" setting and still at 32 Samples. Still ZERO Late Buffers - with no plugins. I pushed again and set the driver to Steinberg's "Low Latency" setting and still ZERO Late Buffers, although the hit on Engine Load was now creeping up to 90% (but roundtrip latency to the lowest possible with the UR44 of 5Msec). I played the project through 6x times to make sure it was real.

At this point, I thought to close the project and NOT save, in order to get back to the earlier state with plugins inserted. This was so as to remind myself what plugins they were. Once I had noted them down, I was intending to replace them with as many of the plugins I had used on the last year's project I had mentioned, which had 60+ plugins and 2x VSTs all in an 'open' state (being played through). You will remember me questioning how that could be (no Late Buffers encountered), when on the same system in the current problem project, nothing but Late Buffers.

John Vere's kind and diplomatic response 2x posts ago, urging me to methodically investigate my plugins, because it would only take 1x rogue plugin to cause the problem, was fully noted. And so I closed the project to return to its earlier state. When I went to re-open it, CbB got as far as saying "nearly there", when up came the window of death telling me that "CbB has stopped working and will try and find a solution". It didn't. So I thought to remain calm and reboot the PC. I tried to open the project again and a 3rd time and got the same message !!!

I then opened my 'copy' of this project, removed all plugins (which is where I got to before) and played through, expecting to get zero Late Buffers as before. To my surprise. the Engine Load spiked to a max of 9719.5% - yes that's right 9719.5%. On the next 4x plays, it decided to return zero Late Buffers and Maxing the Engine Load to around 90%.

I have long thought that my problem has had something to do with this specific project being corrupted from the beginning. Is this proof of it, or are we to say that it's just unfortunate coincidence ??

The only reason I haven't smashed up my house is because I had the forsight to make a copy when I thought to open a new 'test' project and singly copy and paste the tracks and insert new instances of the same VSTs and plugins to make that check. I have now made yet another copy to remain with 2x.

So now my thoughts are not only on what was causing my incessant Late Buffer strikes, but also on what caused the crash and what is stopping me from re-opening that original project. I WILL LEAVE THIS THREAD OPEN FOR 24 - 48HRS in case any of you helpful souls have some guidance on the latter. Otherwise, my initial problem is up in the air for now. Between now and the end of the week when the Engineer calls round, I will stop adding to the lengthy 3x pages and concentrate instead on continuing to do what I was planning to do on my copy project - start loading it up with the same 60x plugins I used in last years trouble free project to prove if John Vere's (and others) suspicions are correct. Only then can I offer a conclusion or you good folks, a solution to this thread, which as I said previously, all threads must have to be of any use to others.

I guess my middle name is Bad Luck ??   Cheers.

Link to comment
Share on other sites

27 minutes ago, Byron Dickens said:

60+ plugins!?!?!?

i've seen some of the pro tools folks have videos with 12 or more PER TRACK and 40+ tracks + another 20+ busses sometimes with 5-10 plugins... so something like 400+ plugins on a project... of course if they had recorded it properly things would likely be easier...

one had crazy layers like EQ, compressor, pitch, EQ, compressor, saturation, tape effect, de-esser, EQ, final compressor, slap echo, then reverb... on each of 8 vocal tracks... but i guess that's what it takes sometimes to get platinum records... lol

Edited by Glenn Stanton
Link to comment
Share on other sites

Hey Byron -

The recording I refer to was my first ever - about 15yrs ago (or even more) and as a beginner I was just throwing the kitchen sink into every track just to see what it would do. I'm different now. Thanks.

And Glenn is quite right - I too have seen many a Pro's video sending tracks through umpteen plugs for reasons that MY ears don't always hear. But that's not me criticising them - they know far, far better than me and I'm sure that their ears are also far better tuned.

I did say I wouldn't add more to this thread - but whilst here to acknowledge the above posts, that mix I did from yesteryear was me playing it through at 2048 (max) Samples. An hour ago when I felt like playing it through again, I had left my ASIO driver at 32 Samples and boy, did it rack up the Late Buffers. So at 2048 I will start adding the plugins I used back then and see how many my current project can take. When it starts to glitch, I can make comparisons and then start replacing with the plugins I started with on this current project, to find the culprit. That would be my conclusion.

The solution would most likely be better as another thread - what to expect on a given system/rig and how to go about tracking and mixing to overcome the limitations. That would be very interesting because I'm fairly sure there would be a whole lot of different ways coming from each of you.

Cheers

Link to comment
Share on other sites

1 hour ago, Chris Ward said:

The solution would most likely be better as another thread

@Chris Ward I've been following this thread since the start. Sorry I don't have the solution, but I'd like to suggest that if your problem is fixed, you should show the answer right here, so that future readers will have ready access to all the context of the previous three (or more?) pages. On that glorious day you can edit your original post and add the word "solved" to the title.

Also, if you want to alert another forum user that you're saying something to them, type the @ symbol and then start typing the user name. A menu of names will drop down. Find the user you're trying to alert and select that name from the list.

Good luck!

  • Great Idea 1
Link to comment
Share on other sites

Please sign in to comment

You will be able to leave a comment after signing in



Sign In Now
×
×
  • Create New...